Silicon ChipNovember 1999 - Silicon Chip Online SILICON CHIP
  1. Outer Front Cover
  2. Contents
  3. Publisher's Letter: DC power in the home; it could be made to work
  4. Feature: USB: Hassle-Free Connections To Your PC by Peter Smith
  5. Project: A Speed Alarm For Cars; Pt.1 by John Clarke
  6. Book Store
  7. Serviceman's Log: Price isn't everything by The TV Serviceman
  8. Project: Multi-Colour LED Christmas Tree by Les Grant
  9. Feature: Electric Lighting; Pt.15 by Julian Edgar
  10. Product Showcase
  11. Feature: Setting Up An Email Server by Peter Smith
  12. Project: Build An Intercom Station Expander by Paul Hoad
  13. Vintage Radio: The case of the disappearing TV sets by Rodney Champness
  14. Project: Foldback Loudspeaker System For Musicians by John Clarke
  15. Project: Railpower Model Train Controller; Pt.2 by John Clarke & Leo Simpson
  16. Order Form
  17. Notes & Errata
  18. Market Centre
  19. Advertising Index
  20. Outer Back Cover

This is only a preview of the November 1999 issue of Silicon Chip.

You can view 34 of the 96 pages in the full issue, including the advertisments.

For full access, purchase the issue for $10.00 or subscribe for access to the latest issues.

Items relevant to "A Speed Alarm For Cars; Pt.1":
  • PIC16F84(A)-04/P programmed for the Speed Alarm for Cars [SPEED254.HEX] (Programmed Microcontroller, AUD $10.00)
  • PIC16F84 firmware and source code for the Speed Alarm for Cars [SPEED254.HEX] (Software, Free)
  • Speed Alarm for Cars PCB patterns (PDF download) [05310991/2] (Free)
  • Speed Alarm for Cars panel artwork (PDF download) (Free)
Articles in this series:
  • A Speed Alarm For Cars; Pt.1 (November 1999)
  • A Speed Alarm For Cars; Pt.1 (November 1999)
  • Speed Alarm For Cars, Pt.2 (December 1999)
  • Speed Alarm For Cars, Pt.2 (December 1999)
Articles in this series:
  • Understanding Electric Lighting; Pt.1 (November 1997)
  • Understanding Electric Lighting; Pt.1 (November 1997)
  • Understanding Electric Lighting; Pt.2 (December 1997)
  • Understanding Electric Lighting; Pt.2 (December 1997)
  • Understanding Electric Lighting; Pt.3 (January 1998)
  • Understanding Electric Lighting; Pt.3 (January 1998)
  • Understanding Electric Lighting; Pt.4 (February 1998)
  • Understanding Electric Lighting; Pt.4 (February 1998)
  • Understanding Electric Lighting; Pt.5 (March 1998)
  • Understanding Electric Lighting; Pt.5 (March 1998)
  • Understanding Electric Lighting; Pt.6 (April 1998)
  • Understanding Electric Lighting; Pt.6 (April 1998)
  • Understanding Electric Lighting; Pt.7 (June 1998)
  • Understanding Electric Lighting; Pt.7 (June 1998)
  • Understanding Electric Lighting; Pt.8 (July 1998)
  • Understanding Electric Lighting; Pt.8 (July 1998)
  • Electric Lighting; Pt.9 (November 1998)
  • Electric Lighting; Pt.9 (November 1998)
  • Electric Lighting; Pt.10 (January 1999)
  • Electric Lighting; Pt.10 (January 1999)
  • Electric Lighting; Pt.11 (February 1999)
  • Electric Lighting; Pt.11 (February 1999)
  • Electric Lighting; Pt.12 (March 1999)
  • Electric Lighting; Pt.12 (March 1999)
  • Electric Lighting; Pt.13 (April 1999)
  • Electric Lighting; Pt.13 (April 1999)
  • Electric Lighting, Pt.14 (August 1999)
  • Electric Lighting, Pt.14 (August 1999)
  • Electric Lighting; Pt.15 (November 1999)
  • Electric Lighting; Pt.15 (November 1999)
  • Electric Lighting; Pt.16 (December 1999)
  • Electric Lighting; Pt.16 (December 1999)
Items relevant to "Railpower Model Train Controller; Pt.2":
  • Railpower PCB pattern (PDF download) [09308991] (Free)
  • Railpower panel artwork (PDF download) (Free)
Articles in this series:
  • Build The Railpower; Pt.1 (October 1999)
  • Build The Railpower; Pt.1 (October 1999)
  • Railpower Model Train Controller; Pt.2 (November 1999)
  • Railpower Model Train Controller; Pt.2 (November 1999)
  • Railpower Model Train Controller; Pt.3 (December 1999)
  • Railpower Model Train Controller; Pt.3 (December 1999)

Purchase a printed copy of this issue for $10.00.

November 1999  1 2  Silicon Chip Contents Vol.12, No.11; November 1999 FEATURES 4 USB: Hassle-Free Connections To Your PC USB makes it easy to add peripherals to your PC but what hardware and software do you need? – by Peter Smith 39 Electric Lighting; Pt.15 How neon signs are made – by Julian Edgar 56 Setting Up An Email Server USB: Hassle-Free Connections To Your PC – Page 4. Want to set up your own email server and cut telephone costs? We show you how – by Peter Smith PROJECTS TO BUILD 16 A Speed Alarm For Cars; Pt.1 Is your licence looking a bit dodgey? This compact speed alarm will help you stick to the speed limits – by John Clarke 31 Multi-Colour LED Christmas Tree A low-cost microcontroller chip produces multiple colours and multiple patterns – by Les Grant Build A Speed Alarm For Your Car – Page 16. 62 Build An Intercom Station Expander It’s called the “Addacom” and it lets you add four extra stations to any existing 2-way intercom – by Paul Hoad 72 Foldback Loudspeaker System For Musicians It uses readily available drivers and can be built using basic hand and power tools – by John Clarke 80 Railpower Model Train Controller; Pt.2 Final article describes the IR remote control circuit and gives the full construction details – by John Clarke SPECIAL COLUMNS Multi-Colour LED Christmas Tree – Page 31. 26 Serviceman’s Log Price isn’t everything – by the TV Serviceman 68 Vintage Radio The case of the disappearing TV sets – by Rodney Champness DEPARTMENTS 2 22 44 53 55 Publisher’s Letter Circuit Notebook Mailbag Product Showcase Electronics Showcase 89 90 93 94 96 Subscriptions Form Ask Silicon Chip Notes & Errata Market Centre Advertising Index Foldback Loudspeaker System For Musicians– Page 72. November 1999  1 PUBLISHER’S LETTER www.siliconchip.com.au Publisher & Editor-in-Chief Leo Simpson, B.Bus., FAICD Production Manager Greg Swain, B.Sc.(Hons.) Technical Staff John Clarke, B.E.(Elec.) Peter Smith Ross Tester Rick Walters Reader Services Ann Jenkinson Advertising Enquiries Rick Winkler Phone (02) 9979 5644 Fax (02) 9979 6503 Mobile: 0414 34 6669 Regular Contributors Brendan Akhurst Rodney Champness Garry Cratt, VK2YBX Julian Edgar, Dip.T.(Sec.), B.Ed Mike Sheriff, B.Sc, VK2YFK Philip Watson, MIREE, VK2ZPW Bob Young SILICON CHIP is published 12 times a year by Silicon Chip Publications Pty Ltd. A.C.N. 003 205 490. All material copyright ©. No part of this publication may be reproduced without the written consent of the publisher. Printing: Macquarie Print, Dubbo, NSW. Distribution: Network Distribution Company. Subscription rates: $69.50 per year in Australia. For overseas rates, see the subscription page in this issue. Editorial & advertising offices: Unit 8, 101 Darley St, Mona Vale, NSW 2103. Postal address: PO Box 139, Collaroy Beach, NSW 2097. Phone (02) 9979 5644. Fax (02) 9979 6503. E-mail: silchip<at>siliconchip.com.au ISSN 1030-2662 * Recommended and maximum price only. 2  Silicon Chip DC power in the home; it could be made to work Last month’s Publisher’s Letter about the possibility of DC power in the home triggered off quite a response from readers. Some raised the obvious safety issues of the difficulty of safely switching high voltage DC and the possibility that a DC shock can be more dangerous than AC. These drawbacks must be admitted. Others though, saw the potential in the idea and went on to expand the concept. My feeling is that most people have such a reliable 240V AC supply that they would never contemplate ever having any other system; it works, why fix it? For those that do have an unreliable mains supply, and I include myself in that category, the occasional inconvenience might be extremely irritating but it would not justify the in­vestment and time necessary to eliminate it. And whether a DC power system would be the way to go would probably be a moot point. However, for those who are in remote locations far away from any mains power supply, a combination high and low voltage DC system based on solar arrays could be made to work. As one of our readers points out in the Mailbag pages this month, quite a few appliances could be made to work on DC. But would it be safe? Now that the problem of high voltage DC switching has been highlighted, could be it be overcome? The answer is yes. But it would not be necessary to have large mechanical power switches to turn the appliances, lights or whatever, on and off. The logical approach would be to have electronic switching which could cope with high voltage DC and AC. This would be pretty straightfor­ward, when you think about it. After all, many appliances these days do not rely on me­chanical on/off switches; they use electronic switching. Virtual­ly any appliance which comes with a remote control uses electron­ic switching. The same point applies to microwave ovens, many washing machines and dishwashers. An electronic switch based on a power Mosfet or IGBT (insulated gate bipolar transistor) could be made to handle the switching job for AC and DC. So a compact, reliable and rugged power switch is not an insurmountable prob­lem. Nor is the problem of automatic degaussing for TVs and computer monitors running from DC insoluble - there has to be an electronic solution. So as I remarked last month, there is no reason why most appliances could not be made to run on 250V DC. Will it ever happen? Probably not. To be realistic, if you were faced with providing power in a remote location, the most practical approach would probably be to power as many appliances as possible at 12V DC and for those that cannot be run from low voltage DC, use a 12V DC to 240VAC inverter which would only run when an appliance was switched on. Many of the bigger inverters already have an auto-sensing feature and it is very worthwhile because it stops the inefficiency of running inverters continuously. Leo Simpson            ­—˜—   –‹‹Ž­š‚ ‹      ——Ž  › •œ„› •  – •‚       ™ ž– Ÿ‚   –‹Ž—— *Full details at www.tol.com.au          ††  ­ ­          ­   € €‚       ” €  € € ‘›Œ€›Œ ˜ Œ¡¢ Œ¢ €  ƒ ‚ €  ƒ€€   ‚   „„   †‡‡  ŽŽ‡€‚ ‡  •‚ € € ˆ ƒ‚     €  ƒ „      š  œ‚  ƒ   Œ ‘”Ž‹†   ”  ‚   €€     „„‡      †‡  € ­      ƒ    ‰  „†‰       †‡ € Ž‹  ‚ £’ €      Ž ˆ€€ € Œ€ƒ   £’ ƒ    „‡†     ‡ š €   €œ €  „ކ††  ­ އ   Š‹ˆŒ €£    „‰‰   ‡‡ Š  ŒŠ   ‚     „‰†‰   ‡‡ š€‹œ‚  ¤ ‹”      €   ¥ €  ¥ Œ      „‰‰  ‡‡          „‰‰   Ž  ŒŽ‹Ž †‹  ™Œ¦Œ ƒ€†£       ††  ƒ‚  †‡†‡ €€    €  Ž ˆ   €  †‰  ƒ‚  †‡†‡   •  –   †„  ƒ‚  †‡ˆ‡ † €€   €  Œ ”  €  †‡  ƒ‚  †‡ˆ‡ †Ž ‘’Š “Œ‚ ˆŒ  ” ‹   ŽŽŠ•ŠŠ“”§‚†Š’ˆŠ‚‚  †‰Ž  ƒ‚  ††‡    € • ‚§ ’‚¨¥ˆ‚€  †‰  ƒ‚  ††‡ ‰  ‚ ”   ‚   –   †‰  ƒ‚  ††‰‡ ‚ ”‘‚£’‹  ˆ’ Œ †   € ‚  ‘‹€ƒ  †‰†  ƒ‚  †‡‰‡ †‰ ¤€€  €€  „‰Š  ‹ Š  Œ‹Ž ‚‡  „    އ  ƒŠ ‹ Œ• €’Š Ž„  „  € ‹  ˆ  €‚ ƒ   ‘”Ž­­ €‚   ˆƒ€€ ‚   € €       Š€ ¥ ŽŽ‹   €  €€ €  ƒ €  €  ‡     š€œ ‹     ƒ‚ Ž  ‚ ”        ‚ƒ € „† —˜  ’€™      „Ž  ƒŠ ‹ † †   „Ž  ƒŠ ‹ † ‡   š › €  „Ž  ƒŠ ‹†† Ž   œ € ‚ ‹ š ›ž  „‰‡  ƒŠ ‹††† Ž ‹œ   €Ž      „     ‡  „‰Ž    ‡   ­ Ÿ€  „ €          € ‘€  ‡‰šŸœ­š¡œ­ Žšœ  ˆ€€  Š‚   ‘ ’ €       ‚  ­ ¦‚  €  „‰†  „‰‰  „‰„  „‰„ Š  „   ‚ ‘†„  ’“ ”     •  ‡‡ ‡ ‡ ކ‡ ‚    –­  © §Œ”     ­  €      €   ­ €   ‘    Œ ”    €  €ƒ §Œ”‹ Œ” ’˜ ’ € †¡€  €ƒ €   ‰‚    –­ Œ„ “ ‰‡ ‡    ‘     „‚    –­  ˆƒž €€   ‰ € ƒ  – ‹„ ­  †‡    – ‹„ ­  ‡  „‰„    ކ‡  „Ž    ‰Ž  †– ‹„­ ’Š ‡‡ E & OE All prices include sales tax MICROGRAM 1199 Come and visit our online catalogue & shop at www.mgram.com.au Phone: (02) 4389 8444 Dealer Enquiries Welcome sales<at>mgram.com.au info<at>mgram.com.au Australia-Wide Express Courier (To 3kg) $10 FreeFax 1 800 625 777 We welcome Bankcard Mastercard VISA Amex Unit 1, 14 Bon Mace Close, Berkeley Vale NSW 2261 Vamtest Pty Ltd trading as MicroGram Computers ACN 003 062 100 Fax: (02) 4389 8388 Web site: www.mgram.com.au FreeFax 1 800 625 777 November 1999  3 Connecting a new device to your PC can be a real hassle. But imagine being able to add peripherals without opening the case, adding interface cards or even having to switch the PC off or reboot it. Those are just some of the advantages offered by USB. By PETER SMITH M OST OF US at one time or another have struggled with the software or hardware configuration when installing new peripherals on personal computers. The Plug and Play (PnP) standard introduced with Windows 95 helped PC HOST CONTROLLER & ROOT HUB USB PORTS to ease the pain a little but what do you do when that scanner, printer and Zip drive just won’t work together on the same parallel port? The answer is USB! USB is an acronym for Universal Serial Bus. Developed over the last few years by a group of industry leaders including Intel, Microsoft, Compaq and NEC, USB has finally opened the way for stress-free peripheral connection. The key goal for the USB designers was to create an expansion bus that would make adding peripherals as easy as plugging a connector into a socket. A multitude of USB-ready devices are already available, including scanners, cameras, Zip drives, modems, keyboards, mice, ethernet (network) adapters and joysticks. So let’s have a MONITOR & 2-PORT HUB KEYBOARD & 2-PORT HUB UPSTREAM PORT 4-PORT HUB SPEAKERS MOUSE JOYSTICK DOWNSTREAM PORTS The USB ports on this computer are located on the rear panel, just below the two PS/2 ports. 4  Silicon Chip Fig.1: most modern PCs come with two USB ports and many USB peripherals include inbuilt hubs so that other devices can easily be added to the bus. Alternatively, you can use a dedicated hub to add extra ports. CLIENT SOFTWARE USB DRIVER SOFTWARE HOST CONTROLLER DRIVER SCOPE OF OHCI/UHCI SPECIFICATION HOST CONTROLLER HARDWARE USB DEVICE Fig.2: the OHCI and UHCI standards affect the host controller hardware and its software driver. Motherboards that use Intel chipsets are UHCI compliant while many add-on controllers are OHCI compliant – see text. look at some of the advantages of this new system and how it all fits together. USB basics Adding a new device to your PC in the past usually meant opening up the case, installing an interface card and worrying about such things as IRQs and I/O addresses, etc. That no longer applies with USB – you just plug the device into a USB connector and it works. Up to 127 devices can be connected to the USB bus – more than most of us will ever need. Connections to the bus can be made without switching off the power (known as “hot-plugging”), making it possible to add peripherals and to move them between machines with a minimum of fuss. USB has two bus speeds to optimise efficiency: (1) a “low-speed” mode that operates at 1.5 megabits per second (1.5Mb/s); and (2) a “full-speed” mode that operates at 12 megabits per second (12Mb/s). The low-speed mode is intended for interactive peripherals such as game pads and mice, whereas the full-speed mode provides lots of throughput for printers, scanners and video cameras, etc. High-speed devices like fast disk drives and fast (100Mb/s) networking controllers are not suitable for use with USB. Table 1 compares the fullspeed USB transfer rate with the rates of other interfaces. As can be seen, it’s a lot faster than a standard serial or parallel port. The majority of IBM-compatible PCs designed from 1998 onwards have The USB standard has two connectors: (1) an upstream Series “A” connector as shown at right and (2) a downstream Series “B” connector as shown at left. The different connector styles make it impossible (at least, in theory) to connect two downstream ports or two upstream ports together. USB ports as standard equipment. But that doesn’t mean that you can’t have USB if your PC is older than this. Many older machines can easily be upgraded to support USB but more on that later. Adding ports Most PCs provide only two USB ports but expansion is a simple matter of adding a low-cost hub. USB hubs provide one port for connection to the PC side of the bus (called the upstream port) and typically between four and seven output ports for connecting peripherals or even another hub (these are called downstream ports) – see Fig.1. Some peripherals also come with hubs built in, allowing devices to be daisy-chained. Both the signal and power connections are provided by a single USB cable, eliminating the need for separate power sources for low power peripherals. Devices that draw power directly from the bus are said to be “bus-powered”, whereas devices that have their own power source are called “self-powered”. Self-powered hubs (and this includes the “root” hub that is part of the USB host controller in the PC) can supply a maximum current of 500mA per port. On the other hand, bus-powered hubs can supply only 100mA per port, so this is something to look out for when selecting a hub or a peripheral with an embedded hub. As a matter of interest, power distribution on the bus is under software control and this ensures that the bus is not inadvertently overloaded. USB connectors Because simplicity is an important part of the design, only two types of connectors exist in USB: (1) Series “A” and (2) Series “B”. The Series “A” type is used for all upstream connections (ie, towards the host PC), while the Series “B” connectors are used for the downstream connections (towards peripherals and hubs). Small devices usually have a captive (moulded in) cable with a Series “A” plug on the end. Larger devices, such as monitors, printers and hubs, usually have removable cables, with a Series “A” plug on one end and a Series “B” plug on the other. These are commonly referred to as “A-B” cables. November 1999  5 USB Peripherals From Namlea Data Systems  A self-powered 4-port USB hub: an upstream port (not visible) connects to the PC while the four downstream ports are for the peripherals. (Namlea Data Systems, phone (02) 9429 0800; www.ndsonline.com.au).  The idea behind the different connector types is to prevent accidental “bus loop-back”, by making it impossible to connect two downstream or two upstream ports together (eg, on two different hubs). We should mention here that although it is physically possible to connect two computers together using a non-standard USB cable (ie, with a Series “A” connector on both ends), the results could be catastrophic! To perform this function, go shopping for a USB bridge – it performs the necessary magic and protects your motherboard from possible damage. Typically, cables are available in 2, 3 and 5-metre lengths, with five metres being the maximum allowable length. It is possible to extend this distance by cascading hubs (up to five levels deep) but USB was never intended for long hauls. If you need to cover long distances but still want to use USB, consider using USB-to-Ethernet adaptors. Dual standards This bus-powered USB-Ethernet adapter with LED indicators lets you connect a PC fitted with a USB port to a network. (Namlea Data Systems).   USB to IEEE1284 parallel converter cable. (Namlea Data Systems). 6  Silicon Chip While the Universal Serial Bus itself is now a well-defined standard, two different standards were drafted for the PC interface side of the USB host controller and the software components that communicate with it. In particular, Intel developed the Universal Host Controller Interface (UHCI) standard, while other USB developers, including Microsoft, Compaq and NEC, developed the Open Host Controller Interface (OHCI). Naturally, all motherboards with Table 2: Win 95/98 Releases Version Number Windows Release 4.00.950 Windows 95 retail 4.00.950A 4.00.950C Windows 95 OSR 1 Windows 95 OSR 2.0 or 2.1 Windows 95 OSR 2.5 4.10.1998 Windows 98 4.10.2222A Windows 98 SE 4.00.950B Intel chipsets are UHCI-compliant, as are motherboards with Via Technologies chipsets. However, many add-on USB cards use hardware from other chip manufacturers (such as OPTi, Ali, CMD, National Semiconductor and RealTech) and are OHCI-compliant. A few problems have surfaced recently with a number of peripherals when connected to OHCI-compliant controllers. Windows 98 Second Edition fixes some but apparently not all of these problems. Check the list of supported USB controllers and peripherals in the Windows 98 SE hardware.txt file if you are having intermittent problems or plan to upgrade. Further information on USB support in Windows 98 can be found on the Microsoft web site at http:// support.microsoft.com Operating system support If you want to use USB peripherals on a PC, Windows 98 is the way to go. It includes full support for USB, with its Plug and Play (PnP) system A bus-powered USB to RS232 serial port converter. This could be used for connecting a modem, for example. (Namlea Data Systems). If you don't already have USB ports on your computer, a USB adapter card is the answer. This unit plugs into a spare PCI slot on your motherboard and gives you two USB output ports on the backplane connector. ($49 from Microgram Computers. Phone (02) 4389 8444; web site www.mgram.com.au). perfectly suited to the task. If you plug in a simple device like a mouse or keyboard, it is usually (but not always) immediately recognised and a generic driver automatically installed. Other more complex peripherals may also be recognised but their functions will often be limited until the manufacturer’s USB driver software (supplied with the hardware) is installed. Of course, this only needs to be done once. Unfortunately, Windows 95 provides only partial support and as far as we know, no future updates will be made available to extend this support. Some USB peripheral manufacturers have even dropped support for Windows 95 altogether, so if you’re running Windows 95 and don’t want to upgrade just yet, check out the operating system requirements before you buy USB peripherals. What about Windows NT4? Unfortunately, NT doesn’t have true Plug and Play support and so is unsuitable for USB. On the other hand, IBM provides limited support for USB in their OS/2 Warp 4 operating system. However, you will need to obtain the OS/2 USB Basic driver from IBM’s web site (see Table 3). For Mac users, Apple provides USB support with their Power Mac G4, iBook and iMac systems, with many peripheral drivers pre-loaded. If you have a Mac, check to ensure that the operating system offers full USB support. Finally, USB stacks are now appearing for various flavours of Linux and FreeBSD. Adding USB to your PC By now, some readers will be wondering whether their PC has USB support or not. Alternatively, they may just wish to check that the USB function is working correctly before adding their first USB peripheral. So what are the basic hardware and software requirements for USB? We’ll discuss the requirements below and describe some of the potential pitfalls as they relate to IBM-PC compatible systems running Windows 95/98. Motherboard manufacturers began integrating USB controllers into their first generation Pentium designs. Unfortunately, not all early Pentium motherboards have USB controllers built-in but if you have a free PCI slot, you can purchase an add-on USB controller card that will do the job. These are now widely available and retail for around $40. In addition, PC assemblers do not always install the necessary USB port connectors. If you think that your motherboard has an inbuilt USB controller but there are no external connectors, check the system docuNovember 1999  7 Fig.3: this entry in System Properties (Device Manager tab) is for an Intel UHCI-compliant USB controller. Fig.4: this dialog box is similar to Fig.3 but shows the entry for an OHCI-compliant USB controller. Fig.5: this dialog box lets you check the hardware version of your PCI-to-USB host controller. mentation for details or give the supplier a call. Suitable brackets with port connectors and cables that plug into the motherboard are available but note that the motherboard connector styles can vary between manufacturers. Many first generation Pentium mother­boards utilise the Intel 82371­ SB PCI-to-USB host controller. However, the first revision of this controller may not work reliably in some applications, as standards development was still under way when this chipset was designed. To check if your machine is affected, right-click on the My Computer icon on the desktop and select Properties. Click on the Device Manager tab and expand the Universal Serial Bus controller device – see Fig.3. If an 82371SB PCI-to-USB host controller is listed, double-click on that line to display its properties – see Fig.5. On the General tab, look for the “Hardware version” line. If the version is “000”, then you may have difficulties. Later versions are OK. Motherboards that incorporate Via Technologies VT82C586B or VT82C­596 PCI-to-USB host controllers need a software patch installed to correct a number of problems. Check out the Via Technologies website (see Table 3) for details. Many Pentium II and III machines will already have all the right hardware installed but the USB controller may not be enabled in the BIOS setup. The procedure to check this varies considerably between machines (depending on the BIOS), so refer to your motherboard’s manual. In rare cases, you may also need a flash BIOS update. If you are an experienced user, you can download the latest BIOS for your motherboard from the manufacturer’s website and install it yourself. Be careful here though – if you mess things up, you will be left with a machine that won’t boot until you get the BIOS chip replaced. Depending on the age of the machine, your supplier may also be able to help with BIOS updates. Finally, if all else fails, it’s possible to disable the on-board USB controller in the BIOS setup and install an addon controller card – assuming that you have a free PCI slot. USB TV Tuner: it turns your PC into a TV set and lets you convert live video into AVI files Called the “LifeView”, this external TV tuner simply plugs into a USB port on your PC (no need to turn the power off), making it easy to “hot-swap” from one machine to the next. It supports all TV standards including PAL, NTSC and SECAM in their various formats and features external video inputs (both composite and S-video) so that you can connect a VCR. The video window can be scaled from 80 x 60 up to 640 x 480 pixels using the supplied software. In addition, live video can be captured and saved as AVI files and you can also capture and save still images. A digital camera suitable for 8  Silicon Chip Windows 98 video conferencing or video email is included with the unit, which is also TWAIN-compliant. The unit is available from Vision Beyond 2020. Phone (03) 9558 0333. As mentioned before, Windows 98 automatically detects and installs the correct drivers for most built-in USB controllers. If you’re installing an add-on USB controller card though, you’ll probably need to load the manufacturer’s driver. The appropriate instructions and software will be supplied with the card. A simple utility from Intel called USB Ready gives an indication of USB hardware and software status and is available for free download from http://www.usb.org/data/usbready. November 1999  9 Table 3: Useful USB Websites Fig.6: the USB View utility (supplied with Windows 98) allows you to quickly check the status of the USB hardware connected to your PC. exe Yet another utility called USB View is supplied on the Windows 98 CD – look for it in the \tools\reskit\ diagnose folder. Fig.6 shows a typical output from USB View. In this case, we have a 4-port hub connected to USB Port 2 of the PC. A USB-Ethernet Adapter and a USB Com Port have then been plugged into Ports 1 & 3 of the hub. It’s also worth checking that both the USB host controller and root hub appear in System Properties. This can be found by right-clicking the My Computer icon on the desktop, selecting Properties and then clicking the Device Manager tab – see Figs.3 & 4. A red cross through either the controller or root hub obviously indicates a problem. Sometimes this can be cured by deleting the devices and restarting the machine – Windows 98 will detect the devices again and reinstall the drivers. The first release of Windows 98 apparently has a number of USBrelated problems, many of which are addressed in Service Pack 1. This pack can be down­loaded from http:// windowsupdate.microsoft.com or contact Microsoft to get a copy on CD. Windows 95 We said earlier that Windows 95 does not provide full USB support. If you’d like to give it a shot anyway, you will need to have Windows 95 OSR 2.0, 2.1 or 2.5 installed – earlier versions won’t work. You also need a UHCI-compliant USB controller, as OHCI-compliant controllers are not supported. Note that motherboards with built-in USB controllers using Intel and Via chipsets are UHCI-compliant, while addon USB controller cards are generally OHCI-compliant. You can check which version of Windows 95 you have by right-click10  Silicon Chip USB Impl ementer's Forum http://www.usb.org Intel USB techni cal http://www.intel.com/design/usb Intel USB support http://support.intel.com/support/technologi es/usb Intel chipset support http://support.intel.com/support/chipsets Microsoft Windows 98 http://www.mi crosoft.com/hwdev/busbios/usbwin98.htm Appl e http://www.appl e.com/usb Appl e Macintosh http://www.macintouch.com/imacusb.html Appl e USB Peripheral s http://guide.appl e.com/uscategories/usb.html IBM OS/2 Warp http://servi ce.software.ibm.com/os2ddpak/html /uni versa Vi a Technologi es http://www.viatech.com/dri vers CMD http://www.cmd.com/semiconductor/support/docs/670/usbpatch.cfm FreeBSD http://www.etl a.net/~n_hibma/usb/usb.pl Linux http://www.linux-usb.org Aten Technology http://www.aten-usa.com/ USB Stuff http://www.usbstuff.com USB Workshop http://www.usbworkshop.com A l l U SB http://www.allusb.com ing the My Computer icon on the desktop, then select Properties. The version number will be shown on the General tab. If USB support is already installed, it will be listed in Add/Remove Programs in Control Panel as “USB Supplement to OSR2”. Note that if the supplement has been uninstalled, the Windows version number will change from 4.00.950C to 4.00.950B! Table 2 shows the various Windows 95/98 releases. If you have Windows 95 OSR 2.5, the USB supplement files can be found on the Windows CD in the \other\updates\usb folder. Look for a file called usb.txt, which describes the installation procedure. Windows 95 OSR 2.0 and 2.1 users can download the USB supplement from the web in a single file called usbsup.exe. We couldn’t find this file on the Microsoft website but you can get it from a number of other sites – try the USB Workshop (see Table 3). You simply run usbsup.exe to perform the installation and reboot you machine when it is complete. Note that a minor problem can occur when installing the USB supplement. Windows may pause after it detects the host controller and prompt you for the location of the file uhcd.sys. All you need to do is change the path to c:\windows\system and hit OK. A further complication arises with Windows 95 and hardware support. Motherboards designed after Windows 95 was released incorporate new features that are not detected by the operating system and this includes the USB controller. A disk (or CD) may have been provided with your PC that includes the necessary drivers for Windows 95 and this should have been pre-loaded by your supplier. However, if Windows 95 has since been reinstalled from scratch using the original Microsoft CD, the drivers will also need to be reinstalled. If you don’t have these drivers and you know which chipset is used on your motherboard, you can download them from the manufacturer’s website (see Table 3). Windows 95 users can also run the USB Ready utility to verify the USB hardware and software status on their PC. What, USB 2.0 already? Readers who are already familiar with USB 1.1 may have heard about the new USB 2.0 specification. At time of writing, USB 2.0 was still in the drafting stages but we can tell you that the main difference between 1.1 and 2.0 is the speed. USB 2.0 will run at 360 megabits per second (360Mb/s) or more, while still maintaining backward compatibility with USB 1.1 peripherals. And no, that’s not a SC misprint! R VA EAL $1 LUE AT +$5 2.9 5 ea O r bu P&P g y5 pos et themand tage free Order by phone or fax from SILICON CHIP - or use the handy order form in this issue November 1999  11 SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.dse.com.au SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.dse.com.au SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.dse.com.au SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.dse.com.au Speed Alar Is your driver’s licence looking a bit dodgy? This easy-tobuild speed alarm can help prevent a fine and save you from losing any more demerit points. As a bonus, it can also function as a digital speedometer while still maintaining the speed alert function. Pt.1: By JOHN CLARKE A NYONE WHO DRIVES a car inevitably exceeds the posted speed limit on occasions, either deliberately or due to lack of atten­ tion. But these days, it’s really not a good idea to speed. Apart from the obvious safety considerations, there are lots of speed cameras about and it’s all too easy to cop a heavy fine and maybe even lose your licence. You don’t have to be a speed demon either. On a long trip, your speed can gradually creep up as you become used to the road conditions. It’s also 16  Silicon Chip quite difficult to stick to the speed limit in a 60km/h zone after you have been driving at high speed on the open road – 60km/h seems agonisingly slow after driving at 100km/h. In this situation, a speed alarm can keep you on your toes and ensure that you stick within the posted limit. Another situation where it’s easy to inadvertently exceed the speed limit is if you using a cruise control. Now while cruise controls are a great help when it comes to maintaining a set speed, they do have one inher- ent limitation – the speed of the car can increase beyond the set limit on downhill stretches. Once again, a speed alarm can instantly warn you when you’ve over­stepped the mark. Main features Our new Speed Alarm is quite compact and fits neatly into the smallest available jiffy/zippy box. By contrast, our previous Speed Alarm (described in the December 1997 issue) used a case this size just for the display circuitry. The rest of the circuit was rm All the parts fit on two small PC boards which are housed in a compact plastic case. Note the black cardboard sleeve around the 7-segment displays in the photo at left. This prevents light leakage from the LEDs adjacent to the pushbutton switches from spoiling the appearance of the readout. housed in a separate instrument case and while it was OK for large vehicles, it wasn’t all that easy to squeeze into the average family sedan. So how have we managed to shrink the circuitry so dramati­ cally? The answer is that we have replaced all the discrete control circuitry with a low-cost PIC microprocessor and come up with the necessary software to control the internal “smarts” of this device. The resulting circuitry all fits on two small PC boards which are stacked inside the case. It’s also just as easy to drive as before. As shown, the front panel carries a 3-digit LED display, a LED indicator and three pushbutton switches. Two of these pushbuttons let you set the alarm speed in 5km increments between 0km/h and 155km/h (one switch increases the speed; the other reduces it). As soon as you exceed the preset speed, the indicator LED lights and an internal piezo alarm briefly sounds at 10-second intervals to provide a warning. The third switch selects between three display modes: (1) the alarm speed value; (2) the actual vehicle speed (ie, the unit functions as a digital speedometer); and (3) the alarm off mode. Each press of the switch cycles the unit to the next operating mode – it really is that easy to operate. The alarm off mode is indicated by three dashes (---) on the display. In this mode, the alarm is off and there is no overspeed indication (either audible or visual). The speedometer mode displays the vehicle speed with a resolution of 1km/h over the range from 0-159km/h. If you exceed 159km/h, the display shows 888 to indicate overrange so the circuit is not suitable for use on a racetrack. By the way, the speed alert function continues to operate when the digital speedometer mode is selected. You can even adjust the alarm speed while the unit is in the speedometer mode by pressing the up and down buttons. Each time one of these buttons is pressed, the piezo alarm and the LED both give brief “blips” to let you know that the alarm speed has increased or decreased by 5km/h. If power to the Speed Alarm is interrupted (ie, if the ignition is turned off), the unit “remembers” its current speed alarm and operating mode settings. These settings are then au­ tomatically restored the next time the engine is started. Options OK, so that’s the basic operation of the unit and most drivers will be content with just those features. However, this design uses a microprocessor and that means we can easily include lots of options just by programming them into the soft­ware. And that’s just what we’ve done, to make this unit as versatile as possible. These options are as follows: (1) Disable repeat alarm: this is done by pressing the Up button at the same time as the ignition is turned on. The righthand display will show a dash (-) until the button is released. The Speed Alarm then resumes normal operation but with the repeat alarm feature disabled (ie, the audible alarm now sounds only once when you exNovember 1999  17 Main Features • • • • • • Overspeed indication range of 0-155km/h in 5km/h steps. • • • • • • • • • Speedometer indication from 0-159km/h. Audible and visual alarm indication. Visual alarm stays on during overspeed. Repeat audible alarm sounds every 10 seconds during overspeed. 3-digit LED display. Unit can be switched to display alarm speed or vehicle speed (ie, speedometer mode), or switched off. Overspeed alarm works for both alarm speed and speedometer modes. Audible and visual acknowledgement when a switch is pressed. Repeat alarm and speedometer mode functions can be switched off. Two selectable alarm speed threshold points. Display brightness automatically adjusts to suit ambient light conditions. Illuminated switches for night-time operation. Automatic calibration. All selected settings restored when power switched on by igni­tion. ceed the preset speed limit). The repeat alarm feature now remains disabled even if the ignition is turned off and on again. It is re-enabled by again holding down the Up button as the ignition is turned on. In this case, the display will show an “r” to indicate that the repeat alarm has been reactivated. (2) Disable speedometer mode: this is done by pressing the Mode switch as the ignition is turned on). A dash (-) is indicated on the lefthand display until the button is re­ leased, after which the speedo­meter mode can no longer be select­ed. The Mode switch now simply toggles the unit between off and the speed alarm mode. The speedometer mode is reactivated by again pressing the Mode switch at power up. This time, the lefthand display shows an “S” to indicate that the speedometer option has been enabled. (3). High or low alarm threshold: if the “low” threshold is selected, the alarm sounds as soon as the set speed is reached and stays on until the speed drops by 1.25km/h. Alternatively, if the “high” threshold is selected, the alarm sounds when the speed is 1.25km/h above the set limit and stays on until the speed drops back to this limit. The upper threshold mode is useful if you normally use the speedo-meter setting. It will allow you to travel at the set alarm speed without the alarm sounding. The high or low threshold is selected by pressing the Down button during power up. If the centre display Specifications • • • Overspeed detection accuracy better than 1% above 65km/h. • • • • Hysteresis (alarm on to alarm off speed) 1.25km/h. Speedometer linearity and repeatability to within 1km/h. Speedometer and overspeed detection update time typically 0.5-3 seconds (depends on calibration). Operating current typically less than 300mA. Calibration accuracy typically .002% (depends on oscillator drift). Memory storage endurance typically 10 million times. 18  Silicon Chip shows an “L”, the low threshold is selected. Conversely, if the display shows an “H”, the high threshold is selected. Saving the settings All settings made using the Up, Down and Mode switches are stored in an EEPROM (Electrically Erasable and Programmable Read Only Mem­ ory), so that they are saved when the power is switched off via the ignition. This type of memory can tolerate about 10 million write operations per bit, which means that it will never wear out (at least not in this design). Note that during normal program operation, the Speed Alarm utilises standard RAM which does not suffer from a limited life­time. Both the EEPROM and RAM are included in the PIC microcon­troller, so we don’t have to use separate ICs for this memory. When it is first built, the Speed Alarm contains a set of default values as follows: alarm speed = 60km/h; repeat alarm on; speed­ometer mode enabled; low threshold selected; and calibration = 100Hz per 100km/h. The display will be in the alarm speed mode and so it will show 60km/h. Circuit details Refer now to Fig.1 for the complete circuit details of the Speed Alarm circuit. It’s dominated by IC1, a PIC16F84 microcon­ troller which forms the basis of the circuit. This device takes its inputs from a speed sensor and from the various switches and drives the LED displays and the piezo alarm element. Let’s start with the speed sensor. It consists of a coil which mounts on the chassis, plus two magnets which mount on a drive shaft (or tail shaft). As the magnets spin past, they induce a voltage into the coil and this is detected by comparator stage IC2a. The top of the coil connects to a 2.5V supply, derived from a voltage divider consisting of two 2.2kΩ resistors between the +5V rail and ground. This 2.5V rail is decoupled using a 47µF capacitor and biases pin 3 (the non-inverting input) of IC2a via a 22kΩ resistor. It also biases the pin 2 inverting input of IC2a via the coil and a series 1kΩ resistor. Diodes D3 & D4 clamp the input signal from the coil to ±0.6V, while the 0.1µF capacitor filters the pickup signal. IC2a functions as an inverting com- parator. The output signal from the coil is a 250mV peak-to-peak pulse waveform as shown by the top trace in Fig.2. This is fed to the inverting input (pin 2) of IC2a and each time the input swings negative, the output of IC2a (pin 1) goes high (ie, to about 10V). Note, however, that the output from IC2a is fed to pin 6 (RB0) of IC1 via a 2.2kΩ limiting resistor. This is done to convert the 10V pulse train on pin 1 of IC2a to a +5V pulse train at the RB0 input of IC1. So how does it do this? The answer November 1999  19 ZD1 16V 1W 10 1W 47F 16VW S 1k D3 LED4  LED3  LED2  +5V D4 47F 16VW +5V 22k 0.1 3 2 4 8 +5V 15pF 1M IC2a LM358 47F 16VW OUT +12V 7805 REG1 GND IN 2.2k +5V 0.1 4 14 15 16 6  LED1 10k 5 RB7 RB4 RB3 RB2 RB1 RB6 RB5 13 10 a  g f d e c b a B IC2b 7 E c b a B B B C +5V A K Q1 BC328 3,8 C E E C DISPLAY 1 HDSP5301 Q4 BC338 2 d 1 e 9 f 10 g 4 6 7 680 Q2 BC328 3,8 C E DIMMER 6 5 LDR1 DISPLAY 2 HDSP5301 OUT 10 9 2 1 4 6 7 GND IN 7805 b DISPLAY 3 HDSP5301 c Q3 BC328 680 560 VR1 100k 3,8 C E f g 2 d 1 e e 9 d f 10 g 9 c b a 8 6 7 B CAL 4 680 1k DOWN MODE 7 x 150 1k UP 7 12 11 RA2 1 RA1 18 RA0 17 RA3 RA4 3 2 IC1 PIC16F84 RB0 PIEZO ALARM ELEMENT D1 D2 SPEED ALARM 15pF 3.58MHz XTAL1 1 0.1 0.1 2 x 1N914 Fig.1: the circuit is based on IC1, a PIC16F84P microprocessor. This processes the pulses from the speed sensor on its RB0 (pin 6) input and drives three 7-segment LED displays in multiplex fashion. LDR1, IC2b and Q4 automatically dim the LED displays so that they are not too bright at night. SWITCH LIGHT INDICATORS 680 680 2.2k 2.2k 2x 1N914 680 SPEED SENSOR AND COMPARATOR N L1 L1: 500T 0.18mm ENAMELLED COPPER WIRE ON 8mm DIA FORMER +12V VIA IGNITION SWITCH Fig.2: the output signal from the sensor coil is a 250mV peak-to-peak pulse waveform, as shown by the top trace in this scope shot. The bottom trace shows the processed speed sensor waveform that’s fed to pin 6 of IC1. is that the RB0 input includes internal diodes which clamp the voltage on pin 6 to a maximum of 5.8V. The resulting processed speed sensor waveform into pin 6 of IC1 is shown as the bottom trace in Fig.2. Note how the waveform has been squared up and limited to 5.8V. The 1MΩ positive feedback resistor sets the hysteresis of the Schmitt trigger and prevents false triggering due to noise. Switch inputs The pushbutton switches are all monitored at the RA4 input. The other sides of the Up, Down and Mode switches also connect to the RA0, RA1 & RA2 outputs respectively, while the Cal switch connects to ground. Normally, the RA4 input is pulled high (ie, to +5V) via a 10kΩ resistor. However, when a switch is closed, it initially pulls the RA4 input low. The microcontroller then tests which switch is closed by first taking the RA0, RA1 & RA2 outputs all high. If RA4 is still low, then it must be the Cal switch that is closed. If the Cal switch hasn’t been pressed, the RA0-RA3 outputs are taken low in turn until RA4 also goes low. In this way, the microcontroller quickly determines which switch has been pressed. For example, if RA4 goes low when RA0 is low, then it’s the Mode switch that’s been pressed. The 1kΩ resistors in series with the Mode and Up switches are there to ensure that the RA0, RA1 & RA2 outputs can not be shorted if more 20  Silicon Chip Fig.3: the top trace on this shot shows the RA0 output (2V/div) from the microcontroller, while the lower traces (on 5V/div scales) are for the RA1 and RA2 outputs respectively. These outputs drive transistors Q1-Q3. than one switch is pressed at the same time. While this is not a major problem for the microcontroller outputs for a short time, it can produce strange display results. We haven’t included 1kΩ resistors in series with the Down and Calibrate (Cal) switches, since these are unnecessary. Note that the Calibrate switch is only accessible with a small probe and it is unlikely that this switch will be pressed at the same time as any of the other switches. Pressing the Cal switch places the unit in calibration mode. This switch is accessed through a small hole in the Speed Alarm front panel using a pen or a similar probe. Basically, the unit counts the pulses from the speed sensor over a fixed time period to calculate the road speed. During calibration, this time period is automatically extended until the number of pulses counted equals 8 per 5km/h. This time period becomes the calibration number and is permanently stored in the EEPROM. In practice, this means that if you are travelling at 100km/h, the counter period is long enough to receive 160 pulses from the speed sensor. And because of the way the software oper­ ates, the unit is virtually self-calibrating, as we shall see later on. LED displays The three 7-segment LED displays are driven by IC1 in multiplex fashion. As shown, the individual segments are driven directly from the RB1-RB7 outputs via 150Ω current limiting resistors, while the RA0-RA2 outputs drive the individual dis­plays via switching transistors Q1-Q3. To drive one of the displays the microcontroller must bring the corresponding RA0, RA1 or RA2 line low. When RA0 is brought low, for example, Q1 turns on and applies power to the common anode connection of display 1. Any low outputs on RB1RB7 will thus light the corresponding segment(s) of that display. After this display is lit for a short time, the RA0 output is taken high and display 1 turns off. The RA1 line is then brought low to drive Q2 and display 2. The new 7-segment data on the RB1-RB7 outputs is then presented to this new display, after which RA2 is taken low to drive display 3. Because the displays are switched on and off at 944Hz, they appear to be continuously lit. Fig.3 shows the RA0 output on the top trace (2V/ div), while the lower traces (on 5V/ div scales) show the RA1 and RA2 outputs respectively. Alarm output The alarm output from IC1 appears at RA3 (pin 2) and per­forms two functions. First, it drives the alarm LED to produce a visual alarm output. Second, it provides a modulated 1.4kHz tone to drive the piezo element with a characteristic “beep, beep” sound. In practice, the RA3 output goes high and low at a 1.4kHz rate for about 80ms, then the output stays high for 80ms. The 1.4kHz tone is then pro- duced for another 80ms, after which the output goes low for 10 seconds and the cycle repeats (assum­ing that the repeat alarm feature is enabled). As well as the piezo alarm, the RA3 output also drives the alarm LED (LED1). This means that when the alarm speed is ex­ceeded, LED1 flashes twice (because it is driven by the two 1.4kHz 80ms pulses). The LED then stays lit until the vehicle’s speed drops below the alarm speed. The two parallel diodes in series with the piezo element prevent any low volume tone from being produced due to modulation of the RA3 output as the display is multiplexed. By including the diodes, the modulation must exceed 600mV p-p before any sound is heard from the piezo element. Display brightness IC2b is used to control the display brightness. This op amp is wired as a voltage follower and drives a transistor buffer stage (Q4) which is inside the negative feedback loop. Light dependent resistor LDR1 controls the voltage on the pin 5 input of IC2b according to the ambient light level. The op amp, in turn, controls Q4 and thus the voltage applied to the emitters of the display drivers (Q1-Q4). During daylight hours, the voltage on pin 5 is close to +5V because the LDR has a low resistance in strong light. IC2b controls Q4 so that the voltage on pin 6 is equal to the voltage on pin 5, so Q4’s emitter will also be close to +5V. This voltage is applied to the emitters of Q1-Q3 and to the 560Ω resistor in series with LED1. This lights the displays at full brilliance, so that they can be seen during daylight hours. Conversely, as the light level falls, the resistance of the LDR increases and the voltage on pin 5 of IC2b decreases. In fact, when it’s completely dark, the voltage on pin 5 is deter­ mined by the setting of trimpot VR1. As before, this voltage appears at Q4’s emitter and so the displays are all driven at reduced brightness. In practice, VR1 is adjusted to give the requisite display brightness at night. LEDs2-4 are the switch indicator lights. They shine light through translucent rings fitted to the holes surrounding the switches, so that their positions can be seen at night. Parts List For Speed Alarm 1 display PC board, code 05310991, 78 x 50mm 1 processor PC board, code 15310992, 78 x 50mm 1 plastic utility case, 83 x 54 x 30mm 1 front panel label, 80 x 51mm 1 dark red transparent Perspex or Acrylic window, 50 x 20 x 2.5 1 piezo transducer, 13.5mm OD x 3.5mm (Kingstate KPE-165); use KPE-827 (30mm dia.) or equivalent if a louder external alarm is required 1 3.579545MHz parallel resonant crystal (X1) 1 LDR (Jaycar RD-3480 or equivalent) 3 9.5 x 11.5 x 2mm translucent rings (optional – see text) 4 or 6 button magnets 1 coil former, 15mm OD x 8mm ID x 7mm 1 20m length of 0.18mm enamelled copper wire 1 6mm x 25mm steel bolt, washer and nut 6 PC stakes 1 8-way pin header launcher 2 7-way pin header launchers 1 DIP-16 IC socket with wiper contacts (cut for 1 x 8-way single in-line socket) 1 DIP-14 IC socket with wiper contacts (cut for 2 x 7-way single in-line sockets) 1 small rubber grommet 3 PC-mount click action push-on switches (black) (S1-S3) 1 tactile switch (S4) (Jaycar SP0730 or equiv.) 1 500kΩ horizontal trimpot (VR1) 3 6mm tapped spacers 2 M3 x 6mm countersunk screws 1 M3 x 15mm Nylon screw 1 M3 x 15mm brass screw Clock signals for IC1 are provided by an internal oscilla­tor circuit which operates in conjunction with crystal XTAL1 (3.58MHz) and two 15pF capacitors. The two capacitors are in­clud­ed to provide the correct loading for the crystal and to ensure reliable starting. The crystal frequency is divided down internally to produce separate clock signals for the microcontroller 2 M3 nuts 2 M3 plastic washers 1mm thick (insulating bush and washer with bushing cut off) or 1 x M3 plastic washer 2mm thick 1 400mm length of 0.8mm tinned copper wire 1 2m length of single core shielded cable 1 2m length of red automotive wire 1 2m length of black or green automotive wire (ground wire) Semiconductors 1 PIC16F84P microprocessor programmed with SPEED.HEX program (IC1) 1 LM358 dual op amp (IC2) 1 7805, LM340T5 5V 1A 3-terminal regulator (REG1) 3 BC328 PNP transistors (Q1-Q3) 1 BC338 NPN transistor (Q4) 3 HDSP5301, LTS542A common anode 7-segment LED displays 1 5mm high-intensity red LED (LED1) 3 3mm red LEDs (LED2-4) 4 1N914, 1N4148 diodes (D1-D4) 1 16V 1W zener diode (ZD1) Capacitors 3 47µF 16VW PC electrolytic 4 0.1µF MKT polyester 2 15pF ceramic Resistors (0.25W, 1%) 1 1MΩ 3 1kΩ 1 22kΩ 6 680Ω 1 10kΩ 1 560Ω 3 2.2kΩ 1 10Ω 1W Miscellaneous Automotive connectors, aluminium bracket for sensor, heatshrink tubing, long cable ties, Silicone sealant, super glue, thin black cardboard. operation and for the alarm tone and display multiplexing. The crystal frequency is also used to give a precise time period over which to count the incoming speed signals at RB0. The number of pulses within a set period indicates the speed. That’s all we have space for this month. Next month, we will describe the power supply circuit and give the SC full construction details. November 1999  21 CIRCUIT NOTEBOOK Interesting circuit ideas which we have checked but not built and tested. Contributions from readers are welcome and will be paid for at standard rates. Refinements to the PC Monitor Checker The PC Monitor Checker published in the August 1999 issue can be modified to provide more appropriate scan frequencies for the various monitor types than the original circuit published in the May 1999 issue of Popular Electronics. Another modification produces the pre­ferred sync polarity for VGA & EGA monitors. Table 1 shows the common horizontal and vertical scan rates and sync polarities for the various monitor types. Obtaining the horizontal frequencies from the 1MHz clock signal requires the division ratios from IC2 as shown in Table 2. The circuit of Fig.1 shows the suggested modifications which involve using a 2-pole 5-position switch (eg, DSE Cat P-7508). Note that IC5c is now unused and should have its pin 5 tied high (+5V) or low (0V). These modifications would enable the testing of monitors such as the Philips/Commodore 1084 which use 15,625kHz and 50Hz line and frame rates respectively, and can accept RGB and composite video inputs. To produce a 50Hz vertical scan frequency, the spare gate in IC9 (4012) can be employed. Counter IC6 is Table 1: Scan Frequencies Vertical Horizontal Sync Monitor Frequency Frequency Polarity MDA 50Hz 18.432kHz +H -V CG A 60H z 15.75kHz +H -V EG A 60H z 21.8kHz n/a VG A 60H z 31.5kHz -H -V Table 2: Division Ratios Frequency Divisor IC2 Outputs 15,625Hz 64 Q7 15,750Hz 64 Q7 18,432Hz 54 Q6,Q5,Q3,Q2 21,800Hz 46 Q6,Q4,Q3,Q2 31,500Hz 32 Q6 22  Silicon Chip Fig.1: suggested circuit modification for producing the required horizontal scan frequencies. 3 9 Q1 7 Q4 13 Q8 3 Q14 IC6 Q11 15 14 Q10 12 Q9 R 5 4 3 2 12 11 10 9 EGA CGA VGA 1 60Hz +VE SYNC 13 IC11 IC5f 8 9 IC9a 12 J2/14 - VGA J5/9 - EGA IC5d 13 IC9b J1/9 - MDA/CGA 50Hz MDA COMMODORE 12 13 11 Fig.2: suggested circuit for obtaining a 50Hz vertical scan rate. presently wired to divide its 500kHz input clock signal by a factor of 8329, result­ing in an output of 60.03Hz. By connecting the Q9, Q10, Q11 & Q14 outputs of IC6 to IC9b, the resultant output frequency is 500,000/9984 = 50.08Hz. This modification is shown in Fig.2. The circuit of Fig.3 corrects an error in the labelling of +VSYNC and -VSYNC on the original circuit. Furthermore, the original circuit (and PC board) produced +V and -H sync for both EGA and VGA monitors. These are best driven by -H and -V pulses since this appears to be a standard mode for 640 x 480 resolution at 60Hz and 31.5kHz. -H +V is associated with 640 x 400 resolu­tion at 70Hz and 31.5 kHz. Finally, Fig.4 is a minor wiring addition to allow CGA monitors to be driven from the MDA socket (J1). 11 IC3d Fig.3: this circuit produces -H and -V sync pulses for EGA & VGA monitors. J4/3 - CGA J5/3 - EGA 1 J1/4 - CGA J5/4 - EGA IC8 2 J1/5 - CGA J5/5 - EGA 3 Fig.4: this circuit allows CGA monitors to be driven from the MDA socket (J1). Note: the MGA output should have been labelled “MDA” (monochrome display adap­tor). The CGA (colour graphics adapter) frequencies correspond to those produced by the original circuit (60Hz and 15,750Hz) but those for MDA are significantly different (ie, 50Hz & 18,432Hz). There was also an MCGA variant which was integrated into some PS/2 motherboards but this used 60Hz or 70Hz (vertical) and 31.5kHz (horizontal) scan rates (similar to VGA). Franc Zabkar, Barrack Heights, NSW. ($100) Daytime Lights for Cars: negative line headlight switching +12V Simple Christmas light tester FUSE 87 Q1 47 D 86 G S RLY1 HIGH LOW 85 100k GLOBE SOCKET TO PARKING LIGHTS, 30 INSTRUMENT LIGHTS, ETC. 87a DIP SWITCH 470 D8 9V BATTERY LIGHTS L1 RELAY CIRCUIT FOR OTHER LIGHTS LED MODIFIED DAYTIME LIGHTS FOR CARS CIRCUIT As Christmas lights consist of moulded lamp sockets all connected in series, it is quite difficult to find a faulty bulb if one goes open circuit. In effect you have to painstakingly remove each bulb from its socket and then check if for continuity with a multimeter. This tester does not avoid the need to remove each bulb from its socket but does make the continuity test a breeze. If the LED comes on, the bulb is OK. G. Moore, Seven Hills, NSW. ($15) EXISTING HEADLIGHT CIRCUIT This circuit shows the modifications necessary for vehicles that have negative line switching for their headlights. Note that inductor L1 is now connected directly to ground. Simple lights-on warning HEADLIGHT FUSE BUZZER DSE CAT. L-7009 IGNITION SUPPLY K A This is perhaps the simplest possible headlights-on warning circuit. The red wire of a buzzer is connected to the fuseholder for the lighting circuit. The black wire is connected via the diode to any other fuse which is de-energised when the ignition key is switched to the accessories position. When the lights are on and the ignition is switched off, there will be a 12V potential difference between the fuses and the buzzer will sound. W. Movigliatti, Warabrook, NSW. ($15) The wiring modifications are quite minor, as this diagram shows. Note that a 0.1µF capacitor may need to be connected between the fused side of the positive supply to the lights and chassis, to prevent RF interference The Daytime Lights for Cars circuit published in the August 1999 issue of SILICON CHIP was designed to provide positive line switching to the headlights. However, some cars have their switching in the nega­tive line to the chassis rather than in the positive line from the battery. This modified circuit and wiring allows for negative line switching. Note that the 0.1µF 250VAC capacitor is removed from the PC board and L1 now connects to chassis via the terminal on the side of the case. The drain connection to Q1 now connects to the dip switch as shown. The relay provides the supply for the remaining lights. A 0.1µF capacitor or an automotive suppression capacitor may need to be connected between the fused side of the positive supply to the lights and chassis. This should remove any inter­ference evident in your car radio when the Daytime Lights for Cars circuit is driving the lights at the 80% level. John Clarke, SILICON CHIP. November 1999  23 TECHNICAL LOOK: TEN NEW NEW! TCP/IP EXPLAINED By Philip Miller. Published 1997. $ 90 This concise and practical book offers readers an in-depth understanding of the Internet Protocol suite. It assumes no prior knowledge of TCP/IP, only a basic understanding of LAN access protocols, explaining all the elements and alternatives. It leads the reader through the Internet protocols, combining study questions with reference material. Examples of network designs and implementations are given. 518 pages, in paperback, at $90.00. LOCAL AREA NETWORKS: An Introduction to the Technology NEW! SETTING UP A WEB SERVER A complete reference for anyone setting up a web server. Covers all major platforms, software, links and web techniques. It details each step required to choose, install and configure the hardware and software elements, create an effective site and promote it successfully. The book covers the main web server software applications, how they differ, and which work best in each environment. 273 pages, in paperback, at $65.00. NEW! 65 By Tim Williams. First published 1991 (reprinted 1997). By PK McBride & Nat McBride. Published 1999. $ O R D E R H E R E 29 95                 If you want to create web pages for your business or your own home site, but don't know where to start . . . or if you have some experience of Web page design and now need to master all aspects of HTML form then “HTML4.0 Made Simple” is for you. it uses a combination of tutorial approach, carefully focussed examples and quick reference guides. 198 pages, in paperback, at $29.95. TCP/IP EXPLAINED.............................................$90.00 LOCAL AREA NETWORKS..................................$65.00 HTML 4.0 MADE SIMPLE...................................$29.95 SETTING UP A WEB SERVER.............................$65.00 THE CIRCUIT DESIGNER’S COMPANION...........$59.95 ELECTRIC MOTORS AND DRIVES......................$59.95 UNDERSTANDING TELEPHONE ELECTRONICS....$55.00 AUDIO ELECTRONICS........................................$79.00 GUIDE TO TV & VIDEO TECHNOLOGY...............$55.00 EMC FOR PRODUCT DESIGNERS.......................$95.00 THE ART OF LINEAR ELECTRONICS..................$80.00 INTERNET HOME PAGES MADE SIMPLE...........$24.95 DIGITAL ELECTRONICS .....................................$59.95 ESSENTIAL LINUX..............................................$85.00               ORDER TOTAL: $............. 24  Silicon Chip Includes grounding, printed circuit design and layout, the characteristics of practical active and passive components, cables, linear ICs, logic circuits and their interfaces, power supplies, electromagnetic compatibility, safety and thermal management. Aimed at the practising designer who needs straightforward, easy-to-follow advice. 302 pages, in paperback, at $59.95. $ HTML 4.0 MADE SIMPLE $ 65 $ THE CIRCUIT DESIGNER’S COMPANION NEW! By John E. McNamara. 2nd edition 1996. Intended for those who want to become more familiar with local area networks (LANs) without facing the challenge of a 400-page text. The goals of the book are to give prospective LAN users or purchasers familiarity with the concepts involved and to provide a head start for reading more detailed texts. 191 pages, in paperback, at $65.00. NEW! By Simon Collin. Published 1997. 59 95 ELECTRIC MOTORS AND DRIVES NEW! By Austin Hughes. Second edition published 1993 (reprinted 1997). This book is for non-specialist users of electric motors and drives. The author explores most of the widely-used modern types of motor and drive, including conventional and brushless DC, induction motors (mains and inverter-fed), stepping motors, synchronous motors (mains and converter-fed) and reluctance motors. 339 pages, in paperback, at $59.95. 59 95 $ Your Name_________________________________________________ PLEASE PRINT Address ___________________________________________________ ___________________________________ Postcode_______________ Daytime Phone No. (______) __________________________________ STD  Cheque/Money Order enclosed OR  Charge my credit card –  Bankcard  Visa Card  MasterCard Signature_________________________ Card expiry date______/______ PLUS P&P (if applic): $.............. TOTAL$ AU.................... ALL TITLES SUBJECT TO AVAILABILITY. PRICES VALID FOR MONTH OF MAGAZINE ISSUE ONLY. BOOKSHOP WANT TO SAVE 10%? SILICON CHIP SUBSCRIBERS AUTOMATICALLY QUALIFY FOR A 10% DISCOUNT ON ALL PURCHASES! TITLES AVAILABLE! UNDERSTANDING TELEPHONE ELECTRONICS By Stephen J. Bigelow. Third edition published 1997 by Butterworth-Heinemann. $ 55 (To subscribe, see page 53) A very useful text for anyone wanting to become familiar with the basics of telephone technology. The 10 chapters explore telephone fundamentals, speech signal processing, telephone line interfacing, tone and pulse generation, ringers, digital transmission techniques (modems & fax machines) and much more. Ideal for students. 367 pages, in soft cover at $55.00. AUDIO ELECTRONICS   GUIDE TO TV & VIDEO TECHNOLOGY $ By John Linsley Hood. First published 1993. NEW SECOND EDITION 1998. 80 All you need to get started. Create and design your own Internet home pages that include both text and graphics, using this practical, easy to follow, jargon free guide. This edition has been enhanced and updated and now covers HTML 4.0. 182 pages, in paperback, at $24.95. 79 $ Eugene Trundle has written for many years in Television magazine and his latest book is right up to date on TV and video technology. The book includes both theory and practical servicing information and is ideal for both students and technicians. 382 pages, in paperback, at $55.00. 55 EMC FOR PRODUCT DESIGNERS NEW! P&P Add $A5.00 per book – Orders over $100 P&P free in Australia. NZ: Add $A10 per book, $A15 elsewhere 24 95 $ DIGITAL ELECTRONICS – A PRACTICAL APPROACH By Richard Monk. Published 1998. $ 59 95 With this book you can learn the principles and practice of digital electronics without leaving your desk, through the popular simulation applications, EASY-PC Pro XM and Pulsar. Alternatively, if you want to discover the applications through a thoroughly practical exploration of digital electronics, this is the book for you. A free floppy disk is included, featuring limited function versions of EASY-PC Professional XM and Pulsar. 249 pages, in paperback, at $59.95. ESSENTIAL LINUX By Steve Heath. Published 1997. By Tim Williams. First pub­­lished 1992. Second edition 1996. Widely regarded as the standard text on EMC, this book provides all the information necessary to meet the requirements of the EMC Directive. It includes chapters on standards, measurement techniques and design principles, including layout and grounding, digital and analog circuit design, filtering and shielding and interference sources. The four appendices give a design checklist and include useful tables, data and formulae. 299 pages, in soft cover at $95.00. NEW! By Lilian Hobbs. First published 1996. Second edition 1999. By Eugene Trundle. First pub­­lished 1988. Second edition 1996. $ This practical handbook from one of the world’s most prolific audio designers has been updated and amended to make it the leading practical source of information for those interested in linear electronics and its applications, particularly in the world of audio design. 348 pages, in paperback, at $80.00. DESIGNING INTERNET HOME PAGES MADE SIMPLE By John Linsley Hood. First published 1995. Second edition 1999. This book is for anyone involved in designing, adapting and using analog and digital audio equipment. It covers tape recording, tuners and radio receivers, preamplifiers, voltage amplifiers, audio power amplifiers, compact disc technology and digital audio, test and measurement, loudspeaker crossover systems, power supplies and noise reduction systems. 375 pages in soft cover at $79.00. THE ART OF LINEAR ELECTRONICS NEW! 95 $ Provides all the information and software that is necessary for a PC user to install and use the freeware Linux operating system. It details, setp-by-step, how to obtain and configure the operating system and utilities. It also explains all of the key commands. The text is generously illustrated with screen shots and examples that show how the commands work. Includes a CD-ROM containing Linux version 1.3 and including all the interim updates, basic utilities and compilers with their associated documentation. 257 pages, in paperback, at $85.00. 85 $ NEW! POST TO: SILICON CHIP Publications, PO Box 139, Collaroy NSW, Australia 2097. OR CALL (02) 9979 5644 & quote your credit card details; or FAX TO (02) 9979 6503 N Dovember ecember 1999  25 SERVICEMAN'S LOG Price isn’t everything This month’s lead story concerns a rather expensive, top-of-the-market European TV set. Unfortunately though, its upmarket status didn’t prevent it from being a right proper stinker on the serv­ice bench. Did you know that Loewe is a very popular brand of TV set in Canberra? This is because our Lords and Masters have them installed throughout Parliament House which means that they must be pretty good performers. In fact, Loewe colour TVs have been available in Australia since 1974 and are highly-respected, German-made sets with many advanced features. However, this information really has nothing to do with the Loewe Studio 70 (110C91 chassis) which Mr Canaris (not his real name) reluctantly allowed me to take back to the workshop. He initially complained of the sound dropping in and out, as well as half vertical deflection, but when I switch­ed it on, it was dead. “Oh yes, that too”, he replied – offhandedly dismissing this all too obvious fault when I phoned him back to ask why it hadn’t been mentioned. Fortunately, I was able to obtain a service manual for the set (they are available from a firm called Interdyne in Mel­bourne and cost around $40). That at least was a good starting point and when it arrived, I wasted no time in opening up the set for a preliminary investigation. The reason the set was dead was that the power supply had blown, due mostly to the electrolytic capacitors (this set was now about 10 years old). Getting it going again involved replac­ ing IC611 (TDA4601), the switchmode transistor, four electrolytic capacitors, R613 and the start-up PTC resistor (R622) – see Fig.1. This done, the set came on perfectly. I adjusted variable resistor P633 Fig.1: the power supply circuit in the Loewe Studio 70-110C91. 26  Silicon Chip (4.7kΩ) and set it for 155V at point UB, as shown on the circuit. (Note: some European manufacturers use the desig­nation “UB” to indicate the main HT rail, while some German manufacturers may sometimes just use “U”). Unfortunately, there was some initial confusion regarding this HT rail value. The circuit involved is simple enough, though. Diode D651 rectifies the output from the switchmode transformer secondary at pin 18 and the resulting DC is then filtered by R651 and C651 (47µF) to chassis. The HT is then further filtered by coil L651 and a second 47µF capacitor to chassis to give the 155V HT rail (UB). Although this value is clearly marked on the circuit dia­gram, the parts layout diagram indicated 145V at what appeared to be the same point. However, closer analysis of the layout diagram clarified this; the 145V reading was the voltage at C651, whereas the reading on the circuit indicated the voltage at C652. But this only created further confusion. Why was the read­ing across the first filter capacitor less than that across the second capacitor? The logical explanation is that the waveform across C651 still contains sufficient ripple to upset the reading. It needs L651 and C652 to produce pure DC. Anyway, the set was working for the moment and I put it aside for a soak test, hoping that the sound and vertical height faults would eventually show up. Impatient customer Unfortunately, I hadn’t counted on the impatience of Mr Canaris who was on the phone the very next day. I explained that I had fixed the total failure but hadn’t been able to observe the other faults he had mentioned and that obviously they were inter­mittent. I would need time for them to show up so that I could find and fix them. He was very disappointed and although he didn’t actually say so, I got the impression that he thought I was incompetent. Apparently, I was supposed to wave some sort of magical wand and all the unseen faults in his set would disappear. Anyway, I told him that he would have to wait a few more days and that I would phone him when I had fixed all the faults. Unfortunately, the message didn’t sink in because his wife phoned the next day. Once again, I explained the situa­tion and said that I would ring when the job had been completed. The next day, I switched it on and it came on with only half a scanned picture. I heated it and I froze it and suddenly it worked OK, before I had time to work out which component was responsible. Then Mr Canaris was back on the phone wanting to know when he could pick up the set; his wife had said it would be ready today. Feeling somewhat frustrated with him by now, I told him that she had misunderstood me and that the set was still not ready. The next day the fault was back, so I changed the vertical output IC (I561, TDA8175). The fault didn’t show again until three days later but Mr Canaris was still phoning every day. This was a man who had trouble understanding plain English. Finally, I decided to replace all the electrolytics in the vertical timebase. After that, the set worked well and there were no sound problems but I Items Covered This Month • • • • Loewe Studio 70 (110C91 chassis Sharp VC-A200X VCR Pye radio/cassette/CD player Teac CT-M144 34cm TV set was not convinced that I had solved all the intermittent problems. The manufacturers and agents are generally more familiar with difficult problems, particularly intermittents, than individual servicemen, so I phoned Loewe in Melbourne for advice. As it turned out, they were very helpful. In particular, they suggested I replace IC I441 (APU2471) and fit a special kit to the horizontal output transformer to connect the ferrite core to chassis, which might be flashing over. I had already reworked the entire chassis for dry joints. I acquired and fitted IC I441 plus the special kit as suggested, but then just when things were looking good, the next disaster hit – the set went dead again. In the meantime, the Canaris were still phoning every day and getting very shirty, which didn’t help matters. This time, both the horizontal output transistor T534 (S200AF) and diode D536 BY228 had gone short circuit. Replacing these restored the picture but the horizontal system was over­scanning and there was no east/ west (E/W) pincushion correction. By going into the service mode with the remote control, I found that I could adjust the picture but not enough to correct the problem. I also noticed that the E/W output transistor’s heatsink November 1999  27 was getting very hot. By now, Mr Canaris was no longer phoning me but complaining long and hard to Loewe in Melbourne, who then did their best to help me. The voltages seemed correct everywhere and so did the wave­form on pin 24 of IC I511, except for some horizontal pulses superimposed on the lower part of the parabola. I removed and checked all the transistors and replaced a number of electrolytic capacitors (including C558, C546, C583, C581, C512 and C542) but it was all to no avail. I then replaced C594, C536, C537, C538, C541 and C531 but still no joy. It wasn’t until I noticed some of the old brown goo on L538 (incorrectly marked as R538 on the cir­cuit) that I realised I had a clue. This large coil (1.6mH but marked 14323) measured shorted turns when checked on my shorted turns tester, so I acquired a replacement from Melbourne and fitted it. Success at last – the picture was good and the only thing left was transistor T594 (BD­537B) in the E/W correction circuit, which (I felt) was run­ning far too hot. I replaced it and its other half of the Dar­lington pair, transistor T593 (BC­546B) and rechecked the voltag­es. There was 15V on the collector and 0.65V on the base. In addition, the waveform on T594’s collector was 100% correct. I rechecked all the transistors in the E/W correction as well as all the resistors and everything was correct. 28  Silicon Chip In the end, I felt I had taken all possible steps to solve this problem and the only lame idea I had left was to add more aluminium to the heatsink to get rid of the heat (it was literal­ly too hot to touch). This was done and after soak testing for another 24 hours, I finally agreed to let the set go home. Mr Canaris complained long and hard about the service, the cost, the delay and how I didn’t know what I was doing. For my part, I was thoroughly fed up with him and did little to hide my annoyance. In fact, I doubt very much that I will hear any more about this set. There’s so much bad blood between us that even if it does fail, I’m sure he will take it somewhere else. That’s a pity really, as I would like to know if any more faults subsequently showed up. Some customers really are their own worst enemies. Perverse inanimate objects I’m in a whinging mood at the moment. I don’t usually whinge, at least not in print. But fair dinkum, I’m getting fed up with components that keep giving different measurements. During the last few months, I have been beset by several such frustrating experiences. Perhaps they are due to what an ac­quaintance calls “the perversity of inanimate objects”. His philosophy was that some objects have mind of their own and that if you want them to behave in a certain way, they will do all they can to frustrate you. A questionable philosophy? Well, maybe it is. But have you ever tried to fit a nut to a screw, in an awkward corner of a chassis? Or have you tried to fit a pigtail through a hole in a PC board from the hidden side? Of course, these are relatively simple mechanical situa­tions. It’s when these inanimate objects are part of an electron­ic circuit that the fun really begins. This month, I had a Sharp VCA200X VCR with no display. I didn’t have a circuit but I felt it shouldn’t matter as the circuit is so simple. Apart from disassembling it, it wasn’t difficult to establish that there was -28V on the segment legs of the fluorescent display panel but no filament volts on the ends. Moving along to the switchmode power supply, I measured the voltages on each diode. There were appropriate positive or nega­tive voltages on all diodes with respect to chassis, except for diode D921. As I quickly discovered, this diode rectifies an output from the switchmode transformer and feeds the fluorescent display. The output from D921 is filtered by a 100µF 6.3V capacitor, C921. There was no voltage across this capacitor, the capacitor wasn’t short circuit and it made no difference when I connected another capacitor across it. Diode D921 (FR103) couldn’t really be measured in circuit because of the low impedances everywhere, so I unsoldered one end and found that its forward resistance was too high for my liking. This looked like the culprit but when I removed it completely and measured it out of circuit it measured perfectly. Still, I really didn’t have any other clues so I fitted another diode, a BYV96E, in its place. This immediately restored the 3.5V rail needed to drive the 3V filament and the display with the word “SHARP” came up at full brilliance. So that solved that problem. But it really cheeses me off, having to keep remeasuring components because of the uncertainty that the first reading was correct. So that’s my whinge for the month. I know it’s nobody’s fault and there is nothing I can do about it. And having had my whinge, I feel better already. And now here is a story from a colleague, P. K. I’ll let him tell the story in his own words. The Ghettoblaster This story concerns what is often referred to as a ghettoblaster; in this case, a Pye radio/cassette/CD player with two tape cassettes. The unit had origi­nally come in for service about two weeks previously and I diag­nosed the problem as being in the CD player, which required cleaning and testing. This time it was a cassette problem. I replaced a fuse which had blown, after which there were some signs of life. The radio worked, as did the CD player and the “A” cassette. But when I pressed the play button for the “B” cassette, everything went dead. Closer inspection revealed that a switch associated with the play button had failed. It was a leaf switch and one of the two leaves had broken off at the base and was hanging loose. And in order to understand the implications of this, a brief descrip­tion of the switch’s associated mechanisms should help. The play button for each cassette – the “B” button in this case – activates the mechanical loading functions, moving the tape against the head, closing the pinch roller against the capstan, etc. At the same time a lever – at chassis potential – activates the leaf switch, which is suitably insulated, closing its two contacts. The intact leaf carries the 12V supply, while the broken one connects to the load; the motors, the audio, oscillator and other circuits. Fig.2: this simple circuit was used as an electronic switch to replace the broken mechanical unit in a Pye cassette player. But now, when the button was pressed, the lever contacted the remaining, live 12V leaf, taking it to chassis. The result was inevitable; a blown fuse. It was a simple enough diagnosis but what could be done about it? I contacted Philips but I was advised that a replace­ment was not available. But even if one had been available, it would have been a major job to pull the unit apart to fit it. My next idea was to simply bridge the two leads. This would mean that the “B” cassette drive would function continuously, while ever the set was switched on. This was not as a wild an idea as it sounds; a number of other model cassette players use this arrangement. Well, it was worth a try. And at first I thought that it had worked. In fact it had, to the extent that the “B” tape worked perfectly. But now the “A” tape would not play – the wheels worked but there was no sound. When i disconnected the two switch leads which I had bridged, the “A” tape came good but, of course, the “B” tape was dead again. So I could make one or other cassette work, but not both at the same time. Why? – I don’t know; it would have been too time-consuming to figure it all out. All I knew was that while the 12V rail to the “B” cassette was activated, the “A” cassette would not work. Doubtless, given the time and enough technical backup, one could analyse the device in sufficient detail to work out how it functioned and perhaps find a solution. But, at a practical level, this approach was out of the question. Considering the age of the unit, I was beginning to fear that the customer might be forced to cut his losses and settle for only one cassette player. After all, he could only use one at a time! Then I had another thought. Was it possible to substitute an electronic switch for the faulty mechanical one? In fact, this looked to be relatively simple. I selected a BC327 PNP transistor as the switch, connecting the emitter to the 12V supply rail and the collector to the load. The base was connected to the emitter via a 1kΩ resistor, which would ensure that the transistor was switched off unless other­ wise instructed. Obviously, an “instruction” would be needed to turn the transistor on when the “B” play button was pressed. And this could have been tricky. Fortunately, the remaining switch leaf came into its own. It was no longer connected to anything but still made electrical contact with the lever when the play button was pressed. So the leaf was connected to the base via a 3.3kΩ resistor, applying forward bias to the base and turning the transistor on. Did it work? Yes it did – just like a bought one! And I had another happy customer. Spring crisis When it is a glorious day in spring, one arrives at work feeling euphoric, convinced that nothing could possibly spoil your day. So it was last Tuesday – the birds were singing, the tem­ perature and humidity were just right and I was full of bonhomie when I booked in Mrs Townsend’s Teac CT-M144 34cm TV set. All that was wrong was a broken RF socket on her tuner and she still had the broken parts. It promised to be a simple fix for a simple lad on a sunny day – if only life could always be this sweet. On removing the covers, I felt I might be able to resolder the coax socket onto the tuner in situ. Unfortunately, I soon discovered that I would have to remove the tuner and its covers to resolder the centre pin to the PC board. I pulled the chassis out and placed it upside down, happily whistling a little ditty while I prepared the solder­wick to desolder. Mrs Serviceman wasn’t quite so happy – I don’t know whether it was because she didn’t like my ditty, because it was out of tune and rather repetitive. Or November 1999  29 perhaps it was because I was happy and she wasn’t. Anyway, all this was about to change because, unbeknown to me, the set had been switched on in the last 12 hours or so and when I placed the solderwick braid across the PC board pattern, there was a bright flash, a spark and a crack. Whoops! Well, there was nothing I could do until I had replaced the tuner. It didn’t take long to do this but a degree of anxiety was creeping into me and my whistling ditty had stopped. Where had the spark come from and more importantly, what had it struck? I was praying it was just a direct short across an electrolytic capacitor but unfortunately this wasn’t the case. When I switched on there was no sound or picture. There was EHT and voltage on the CRT filament heaters but not much other activity. Fortunately, I had a schematic diagram and I soon established that all the obvious voltage rails were intact (115V, 24V, 15V, 12V, 5V, etc). And the source of the spark was eventually traced to a residual voltage across C260, a 2.2µF capacitor associated with the video output transistors (Q601, Q602 & Q603). But it was one 30  Silicon Chip thing to know where the dis­charge originated and quite another to know what it had struck. By turning up the screen control, I established that the raster was scanning correctly and touching the pins of the audio IC (IC205) produced noise in the speaker. At this point, I wished I had insisted on having the remote control. Because the fault had seemed so simple, I hadn’t seen the need for it when the set was brought in. Now I hesitated to ask the lady for it, in case she suspected the worst. The front controls were having no effect and there was no effect when external signals were applied to the SCART socket on the rear; neither was there any on-screen display. By now, I was beginning to suspect that the microprocessor IC201 (TMP47­ C434N-R214) and/or the EEPROM had been damaged. I checked that Vcc of IC204 (vertical output) and Vdd of the microprocessor were both getting 5V, and that crystal XT201 was oscillating correctly at 4.19MHz. I changed IC202 (TC89101P) first as it was cheaper and simpler but even replacing IC201 as well made no difference. The CRO confirmed that the video was getting to the TA8717 jungle IC (IC206) from the SCART socket but no further. By switching the IC to the TV mode, I could also put the set into the preset tuning mode and tune stations, using the CRO to monitor the video input to the jungle IC on pin 16. But, as before, the signals were going no further. I checked the voltages to IC206 and then used the CRO to check crystals XT202 and XT204. Unfortunately, this provided no clues and it was now obvious that I had overlooked something, but what and where? I went back to microprocessor IC201 and decided to check each pin. I discovered that even though nothing could be seen or heard, most functions were working and responding to the front controls, and these could be measured on the appropriate pins. I finally checked pin 26, marked HD, and found nothing on it. I did not know what HD - or indeed VD next to it - stood for but they suggested horizontal and vertical pulses. I followed the horizontal circuit back to the collector of Q218 and then checked the base circuit. This was fed from horizontal output transformer T201 (pin 10) and so I expected to see horizontal pulses - but there weren’t any! Following the circuit further, I found a branch feeding diode D233 (MTZ­208) and resistor R330 (10kΩ) to pin 17 of the jungle IC (IC206) which wasn’t getting any pulses either. It took some time to follow the PC track to find where D233 was situated on the mother board but I finally found it nestled right next to connector CN203. And guess what was on pin 1 of this connector? Yes, the 200V rail to the video output transistors (Q601-Q603). This rail is derived from pin 3 of T201 via diode D229 and my old friend capacitor C260. Obviously, the desoldering braid connection had shorted this rail directly to D233, the residual voltage in C260 instantly destroying it and turning it into a short circuit. The next step was to identify D233 and I worked out that it was a 20V zener diode. Fitting a new one restored all the set’s functions. Mrs Townsend was spared my anguish and so remains blissfully unaware of the trials and tribulations involved in fixing her wretched anSC tenna socket. Looking for something different this Christmas? Try our multi-coloured, multi-pattern LED Christmas Tree. It will look great at the top of your Christmas tree or in the front window. By Les Grant* November 1999  31 I N NOVEMBER 1998, we published the Christmas Star as a novelty project and it proved extremely popular. This year, our “just for fun” festive season project is in the shape of a Christmas Tree but the display is a lot more diverse and interesting because it uses bi-coloured LEDs. Not only can each LED produce 16 different colours, the LED Christmas Tree has a fascinating range of ever-changing patterns. In fact, considering that the LEDs are red/green types, you will wonder how they can produce such a range of colours; some of them are quite odd. As with last year’s Christmas Star project, this circuit uses just one IC (OK, one-and-a-bit!) and yet the patterns it produces are seemingly endless. How does it do it? Yes, you guessed it. The Tree is controlled by a microcontroller but this one is different. While it can be programmed by most “high-end” (expensive) chip programmers, it can also be programmed (and re-programmed) by a PC parallel port with minimal hardware. This makes it ideal for hobbyists. If you have been avoiding microcon-trollers because of the cost of the programming hardware, now there is no excuse! And most of the development soft- ware can be downloaded free from the Internet – that avoids another excuse! Circuit description Fig.1 shows the circuit. The key to understanding any circuit is “divide and conquer” – break it down into functional blocks. There are three main blocks in the Tree circuit. The first, the power supply, is straightforward. 9V DC is applied from a plugpack to socket SK1. Reverse polarity protection is provided by diode D1. The 3-terminal 7805 regulator (REG1) then provides a 5V rail for the LEDs and the logic. Bypass capacitors C4 and C5 ensure that the 7805 remains stable. Next is the microcontroller IC1. In the Christmas Star and Heart of LEDs (May 1999) projects, we used the Atmel AT89C2051. However, its I/O port structure is not quite suitable for this application so we have used the similar Atmel AT90S2313. See the section entitled “What’s in the AT-90S2313” for a description of the microcontroller. IC2 is a 24C16 serial EEPROM where the pattern data is stored. While IC1 has some EEPROM on chip (128 bytes), this was not enough for the number of patterns we wanted to provide. The final circuit is the LED matrix. At first glance, the PC board looks like it contains 32 LEDs. In reality, there are 64 LEDs as each is a bi-colour LED capable of glowing red or green. 2-pin bi-colour LEDs were chosen to reduce the number of PC board tracks and microcontroller output pins required. 3-pin LEDs would have been easier to drive but would have required more output pins from IC1. To enable IC1 to control so many LEDs with relatively few output pins the LEDs are multiplexed. Multi-plexing is a switching technique whereby each column of LEDs is activated for a short time during which the appropriate rows are driven. This means that individual LEDs are only turned on for a short time. Provided the rate at which the LEDs are turned on is fast enough, our eyes don’t see any flicker. So for multiplexing in this circuit, we connect the LEDs in a matrix of four columns and eight rows to give a total of 32 LED packages. That enables us to drive the whole matrix with just 12 output pins from IC1. Note that while there are only four columns of LEDs, we have to drive each column twice in each multiplex cycle so that we can activate the red and green LEDs. Consequently, each LED’s timeslot is just 12.5% of the total. This is a practical minimum duty-cycle for adequate brightness from the LEDs. The 100Ω resistors R6-R13 set the What’s in the AT90S2313? The AT90S2313 is a member of the Atmel AVR family of microcontrollers which range from tiny 8-pin packages to a 64-pin feature-packed “monster”. Here is a short summary of the features of the ’2313: • • • • • • • • • • • • • • • • • 118 instructions, most single-clock cycle execution 32 8-bit general purpose working registers Up to 10 MIPS throughput at 10MHz 2k bytes (1k words) of In-System-Programmable Flash for program storage (endurance 1,000 erase/write cycles) 128 bytes of SRAM 128 bytes EEPROM (endurance 100,000 erase/write cycles) May be locked for program and EEPROM data security 1 8-bit timer/counter with separate prescaler 1 16-bit timer/counter with separate prescaler, compare and capture modes and 8, 9 or 10-bit PWM On-chip analog comparator (rail-to-rail inputs) Programmable watchdog timer with separate on-chip oscillator SPI serial interface (for in-system programming only) Full duplex UART Low power idle and power down modes External and internal interrupt sources 15 programmable I/O lines in a 20-pin package 2.7 - 6.0V (4MHz parts) or 4.0 - 6.0V (10MHz parts) 32  Silicon Chip Fig.1: the micro drives the 32 bi-colour LEDs in a 4 x 8 matrix with 4 columns and 8 rows. Each row and column is driven by complementary emitter-follower pairs which can sink or source current. This is necessary because the bi-colour LEDs need to be driven in both directions. November 1999  33 peak LED current to about 14mA. Because there can only be eight LEDs on at any time, the maximum current drawn by the Tree is about 150mA. Any 9V DC plugpack rated at 250mA or more should be suitable. Do not use a 12V plugpack otherwise you will cook the 5V regulator. Unfortunately, the microcontroller can’t drive the LEDs directly because its maximum current ratings would be exceeded. So each output pin is buffered by a transistor connected as an emitter-follower. Because each LED package has two LEDs connected in inverse parallel, the emitter-followers have to be “bi-polar” so they can both source and sink current. So two transistors are used for each output and they are connected as complementary emitter-followers so that they can source or sink current. Software The software for the Tree was written in C and compiled by the Dunfield Micro/C compiler which is available from Grantronics. As each byte of pattern data is read in, it is processed by a simple interpreter. Each byte is an instruction such as “set colour to red” or “set LED 22 to the current colour” or “pause for 500ms”. All the complex light patterns are built up from these and similar simple instructions. If you want to know more about the instruction codes, you can download the software from www.grantronics. com.au Down on the assembly line With all the technical stuff out of the way, let’s get the soldering iron going and start building. Your solder- Fig.2: the component overlay for the Christmas Tree. Make sure that you insert each LED to match the overlay otherwise the colour patterns will not be correct. In every case, the flat on the LED faces the closest outside edge of the PC board. ing iron should be temperature-controlled (about 600°F or 320°C) with a fine tip. First, check the PC board for shorts Parts List 1 PC board with Christmas Tree shape 1 4MHz crystal (X1) 1 20-pin IC socket 1 8-pin IC socket 1 9V 250mA DC plugpack 1 2.1mm PC mounting DC socket (or to suit plugpack) 8 100Ω 0.25W resistors 1 1µF 16VW electrolytic capacitor 3 0.1µF monolithic capacitors 2 27pF ceramic capacitors 34  Silicon Chip Semiconductors 1 AT90S2313 programmed microcontroller (IC1) 1 24C16 programmed EEPROM (IC2) 1 7805 regulator (REG1) 12 BC547 NPN transistors (Q1,3,5,7, 9,11,13,15,17,19,21,23) 12 BC557 PNP transistors (Q2,4,6,8, 10,12,14,16,18,20,22,24) 32 red/green (bicolour) LEDs 1 1N4002 silicon diode (D1) 1 1N914, 1N4148 signal diode (D2) between tracks and broken tracks. As usual, start with the small items such as wire links and resistors. Next, fit the IC sockets, crystal, small capacitors, regulator and the diodes. The regulator should be bolted to the PC board. The transistors should be fitted next. All the BC547s face one way and all the BC557s face the other way. Now you can fit the LEDs. Be careful to insert them the right way and don’t apply too much heat as the leads are very short when the LED is pushed down against the board. By the way, make sure each LED is installed the right way around. While no damage will result if you do put a LED in the wrong way around, the resulting colour pattern won’t be right. You will notice that each LED position on the PC board has a circular * Les Grant is the Engineering Director at Grantronics Pty Ltd. They can supply the programmed microcontrollers and EEPROMs for $15 plus $5 for packing and postage. Send cheque or postal order to Grantronics Pty Ltd, PO Box 275, Wentworthville, NSW 2145. Phone (02) 9896 7150. Complete kits for the Christmas Tree will also be available from all Jaycar Electronics stores. workmanship, connect a 9V DC plugpack. No LEDs should light. Measure between pins 10 & 20 (+) of IC1. You should have +4.8V to +5.2V. If all is well, remove power and plug in IC1 and IC2. Make sure they are correctly oriented and be careful not to bend any of their pins as you plug them into the sockets. Turn your Tree on and the display sequence should start within a few seconds. If it doesn’t work... Use this same-size photograph in conjunction with the PC board overlay at left when assembling the Christmas Tree and you should have no problems. Be careful that the two types of transistors aren’t mixed up! outline with a flat on one side – put each LED in so that it matches the outline. Finally, C3 and the DC power connector should be fitted. Testing Carefully check your soldering – use a magnifying glass and a good light. Mistakes found now are less embarrassing than damaged components later! Don’t plug in the two DIL ICs yet. Do a quick continuity check using your multimeter’s diode check range between pin 10 and every other pin of IC1. There should be no shorts or diode junctions. Reverse the probes and you should see diodes (base-collector junctions) on the 12 pins that connect to the LED matrix. A similar test should be performed with pin 20 as the common pin. This may seem like a lot of work but a solder blob shorting an I/O pin to 0V or +5V may damage IC1 and spoil your Christmas! When you are satisfied with your Modern electronic components are very reliable and faulty new components are very rare. All microcon-trollers and EEPROMs programmed by Grantronics are individually tested so problems with these parts are unlikely. The reality is that the most common causes of problems are soldering, a wrong component or wrong component orientation. So the first step in sorting out any problems is to thoroughly check your workmanship. After that, we need to get more logical. If a few LEDs don’t work, are they all in a single column or row? Maybe they only glow red and not green? The column drivers go high and the rows go low for red and vice versa for green. To help with fault finding, the first few patterns are simple “all one colour” displays. The patterns get more SC interesting after that. AVR Resources on the Internet Manufacturer’s data sheets, application notes, free development software and sample source code are available at: http://www.atmel.com Sample startup code written by Dave Van Horn for the Atmel STK200 Started Kit: http://www.dontronics.com/8515.html More sample code and an FAQ http://www.avr-forum.com/ Email list with an active group of AVR enthusiasts Send an email to atmel-request<at>pic.co.za with the word JOIN in the body of the email. November 1999  35 SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.altronics.com.au SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.altronics.com.au SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.altronics.com.au Neon signs use a straightforward high intensity discharge approach to creating light. But how on earth do they make those signs? Electric By JULIAN EDGAR Lighting Pt.15: Making A Neon Sign The different coloured neon tubes that are commonly available. The colours can be created in one of three ways – different fill gases, different fluorescent coatings, or different coloured glass. The most common technique for generating the different colours is to use tubes with different coloured fluorescent coatings with a fill gas of argon and mercury vapour.  November 1999  39  When the different colours are created by the colour of the fluorescent coating, the unexcited tubes all look white – irrespective of the colour they later glow! This can lead to problems when some of the tubes within a pack are mislabelled. Neon tube is available in 9, 15 and 18mm diameters, and made from soda or lead glass. The soda glass lengths are five feet long, while the lead glass tubes are four feet long.  The layout of the sign is provided as a sketch on paper, which is then transferred to a fibreglass mat. The mat provides a template against which to compare the glass tube bends. The next step is to select a piece of tube and soften it over a heater that uses a flame fed with natural gas and compressed air. The width of the flame can be adjusted to suit the length of tube to be heated and the temperature of the flame can also be changed. Take note of the single, four-foot length of tube that you see here – it ends up a very different shape!  The hot tube is bent to the shape shown by the marked template. The need for a fibreglass mat (as opposed to a paper plan) can be seen here – the tube is hot enough to cause a paper plan to burst into flames! Every bend that you see in a neon sign is formed individually, being compared with the template at each step. It is a time-consuming task that requires patience and skill. ING NEON SIGNS MAKING NEON  Because the letters of the sign are made from the continuous length of tube, often the tube needs to wrap back on itself. Using the least amount of tube and the smallest number of bends means that the sign maker needs to have a complete mental picture of where the tube is to go. Sometimes, this requires that the bending starts from one end; at other times the bending begins in the middle of the length of tube. 40  Silicon Chip  After each couple of bends are formed, the bent tube is heated with a ‘cool’ flame. This removes any built-up stresses within the tube. If this is not done, the tube can crack as it cools. The tube is quite fragile – it can be broken by as little force as being placed firmly on a bench. When mounted in the signs, the tube is supported on sprung fittings so that undue stress isn’t placed on it.  Here’s what that straight section of tube that we saw a few pictures ago now looks like! Remember, each bend that you can see here was formed individually. The cork in the end of the tube closest to the camera is so that the sign maker can blow air into the tube to expand the glass at the bends. Note how closely the tube follows the template. As the tube is bent, it tends to close up. So that the original tube diameter is re-formed, the bender (having corked the other end of the tube!) blows into it. The hot, softened part of the tube then returns to its original diameter. Too much air pressure would balloon the tube out at the bend, so this, too, is an operation requiring a deft touch.  Where a bend is needed near to the end of the tube, another section of tube is joined to provide a convenient handhold. This tube will later be removed, so cheaper clear glass can be used. The ends of both pieces of tube are heated with a natural gas and oxygen flame until just molten and then the tubes are pushed together. It’s quite amazing watching how well they join – much easier than welding steel tube!   Here the new clear section of tube is being used as a handhold while the tube is again heated in preparation for making another bend. Since the beginning of this sequence, about 30 minutes has passed – it’s a slow and careful process. The fluorescent coating on the inside of the tube stays attached, despite the heat that is applied. N SIGNS MAKING NEON SIGNS  This particular assembly (the last three letters of a ‘Pokies’ sign) is formed from two four foot long lengths. The tubes therefore need to be joined, with the join placed so that it falls behind part of another letter. It will be later blacked out by paint, so the change in illumination caused by the disruption to the fluorescent coating won’t be visible.  The ends of the assembly where the electrodes will be located need to be cut to size. The tube is heated with a direct flame and then stretched to narrow the wall thickness of the tube. After that, a nick from a normal metal-cutting file creates a weak spot, with the tube breaking cleanly at that spot when struck. It was obvious that extreme care was now being taken – it’s apparently easy to destroy the 2 hours of concentrated work!  The electrode comes as a preformed assembly. The tube diameter closest to the camera matches the diameter of the neon tube being used, while the smaller diameter tube is used to evacuate the tube and fill it with gas. Two conductors are connected to the electrode and these are wired to the test transformer in parallel. November 1999  41  The electrode is attached to the tube in the same way as the clear glass extension piece was previously attached. The flexible plastic tube (visible at the top left) connects the upper glass extension of the electrode assembly to the sign maker’s mouth, allowing him or her to suck on the neon tube. With a cork located in the other end, the maker can sense when the electrode join is airtight.  The neon tube is then ready to be filled with gas. Those tubes using just neon require no addition of mercury but those using fluorescent coatings (such as the sign we have watched being made) use a fill gas of argon, with mercury then added. A syringe is used to place the mercury in the glass bulb assembly in the foreground.  At left is the electrode of the neon tube, with the mercury bulb tee’d off from the evacuation/gas fill tube, which leads off at the right to the machine that is used. During both the evacuation and gas filling procedures, the mercury remains in the bulb. However, after these operations are completed, the glass is sealed to the right of the bulb, allowing the mercury to be then added to the neon tube. GNS MAKING NEON SIGNS MAK  When the tube is being evacuated, sheets of mica are placed between adjoining parts of the tube. This is done in case the tube should get so hot that it distorts, allowing adjoining parts to touch and so cause cracking. The charred paper test strip can be seen on the right. 42  Silicon Chip  Once the tube has been evacuated, filled with argon and then disconnected from the machine, the mercury remaining in the bulb is added to the tube. This is done by simply tilting the sign so that it flows out of the bulb and into the main body of the tube.  The tube is then connected to a high voltage source and energised on the bench. The sections of the tube where mercury has mixed with the argon are glowing brightly; the sections it is yet to reach are dull. As the tube heats up, the mercury vaporises and fills all sections of the tube evenly. The neon tube is evacuated by the machine in the background. During this process, 20kV is applied at currents of up to an amp, causing the electrodes to glow red hot. The tube also gets hot; a strip of paper is placed across the tube and when it starts to char, the tube is hot enough! Even though it is not filled with a gas, the tube still glows brightly during this process.  As indicated earlier, either neon or argon can be added, depending on the sign’s application. Argon is much more popular, being used with the fluorescent-coated tubes. Neon is used most often in clear glass tubes, lighting up red when switched on but being clear (and so not very visible) when switched off. This characteristic makes neon signs suitable for use in ‘open’ and ‘no’ (as in ‘no vacancy’) signs.   The vacuum pump can be seen in the foreground and the 20kV, 1-amp transformer can be seen behind it. The machine has been built expressly for the purpose of making neon signs. During evacuation, a pressure as low as 0.001mm Hg can be developed. The pressure within the sign after gas filling is in the range 3-20mm Hg. Thanks to Australian Trade Neon 08 8351 7811 KING NEON SIGNS MAKING NEO  As can be seen here, the whole length of tube is now glowing brightly with the mercury mixed evenly. When the sign is switched off, the well-distributed mercury vapour will condense onto the adjacent walls. This means that all sections of the tube will glow with the same brightness when it is again switched on.  The effect of the fluorescent coating can be clearly seen here. The blue discharge that occurs in the mercury/argon mixture is visible between the electrode and the beginning of the fluorescentcoated tube, which can be seen to be glowing very brightly. Note the truncated evacuation tube on the left of the electrode. All signs are ‘run in’ on the bench. This neon-filled tube shows the characteristic red neon colour and would of course be clear when switched off. The black-painted sections of tube are simply the connecting links between the SC letters. November 1999  43 MAILBAG 250V DC would be dangerous Leo Simpson’s idea of powering DC-compatible household appliances from 250VDC sparks a memory of an article I saw some years back. I am not sure if it was in SILICON CHIP or in Electronics Australia. The subject was the change-over from DC to AC in a particular Sydney suburb. The writer told a story of the days of DC. It seems that he turned on a light and it blew. Normally that would be the end of things but because the supply was DC, the arc struck by the breaking filament did not extinguish. Instead, it travelled up the light globe and was half-way up the flex to the ceiling before the author had enough presence of mind to turn off the switch. The point of the story was that DC can maintain a spark under extreme conditions. This is obviously quite dangerous. Another problem is that an electric shock from DC is far more dangerous than a shock from the same voltage using AC. This is because AC tends to throw the person away, while DC tends to paralyse the muscles, making it difficult for the victim to escape the shock. The higher the voltage, the greater is the paralysing effect and hence the greater the danger. In short, I think the idea needs a rethink. Jonathon Waller (via email). DC has drawbacks for TVs and monitors I read the Publisher’s Letter about DC mains in the October 1999 issue with great interest. However, may I bring up a few points about existing appliances and DC mains? The first is that you can’t use most switches on 240V DC as the distance between the contacts is insufficient to extinguish the arc formed when the switch is opened. Appliances and house wiring for DC mains generally use heavy duty tumbler switches. Second, colour monitors and TVs are unsuitable for DC as the degaussing coils would actually cause magnetisation of the CRT’s metalwork on 44  Silicon Chip power up; quite the opposite of what you want! You would have to feed the degaussing coils off the switchmode transformer & take into account the higher frequency and asymmetrical waveform that would have a DC component. Third, switchmode power supplies and compact fluorescent lamps have a DC supply of typically 320V. When you specify a 250V DC source supplying these things, it would be equivalent to a 180VAC supply. This would be at the limit of regulation for a lot of power supplies. Compact fluorescent lamps don’t get enough drive to their switching transistors which overheat and die with this lower supply (I’ve experimented with this). This is why you shouldn’t use a 240V peak-peak square wave inverter for the electronic kind of CFL. Normal fluorescent tubes can be used with a resistor or incandescent lamp instead of a choke but this reduces the efficiency. It is interesting to note that the widespread DC mains in the UK were subsequently killed off by the National Grid in the 1960s. Likewise in Australia, DC mains were used in some country towns, the overnight power being supplied by batteries charged during the day. Sydney’s CBD lost its DC mains in the mid 1980s but by this time it was only being used for lift motors. I hope these points are of some interest. John Hunter (via email). AC switches not suitable for DC The idea of supplying 240V DC to certain domestic loads was floated in your latest editorial. It might look attractive at first but carries some risk. There is no problem with the logic concerning the equivalence of AC & DC for heaters, incandescent lights and switchmode power supplies used in consumer goods. However, the ratings on appliance ON/OFF switches and domestic light and power point switches are all qualified by the words AC ONLY for very good reason. Some arcing always occurs between switch contacts when any load current is interrupted. Inductive loads create a severe “back emf” and are particularly hard on switch contacts but arcing still occurs even with purely resistive loads. With DC, the only passive way to extinguish such an arc is to provide generous contact clearance – maybe several centimetres. (We are not likely to install automatic compressed air blast arc suppression on every household and appliance switch!) The switch contacts must also be robust enough to tolerate repeated arc attack. Since AC voltage passes through zero 100 times/second, an AC supply offers an inherently reliable way of extinguishing contact arcing with small contact clearance – typically a minimum of around 1mm. This allows switch designs to be compact, safe and cheap. I recall an anecdote related several years ago by the late Neville Williams in Electronics Australia. He vividly described what happened when a pendant light globe, supplied with DC power at the time, failed spectacularly in a factory where he worked in the ’30s. The globe went “pop” and an arc occurred between the two filament supply leads. This arc burnt into the lamp base, up through the socket and then just kept going up the twisted pair rubber and fabric cable into the ceiling rose. Fortunately, the power was cut in time to prevent a major building fire. Noel Erbs, Trafalgar, Vic. DC makes sense in remote areas Yes, DC power in the home DOES make sense. Your editorial in the October 1999 issue, in my opinion, is very sensible. I live about 400km west of Rockhampton. For many years we generated our own 240VAC, then solar panels became available, not quite as cheap as now but affordable. We went into the cost structure very thoroughly. Batteries were the biggest cost but we were lucky in obtaining a couple of sets of ex-Telecom 500A.h 2V cells. These are harder to find now but a 350A.h unit is on the market. So our power unit was trickle continued on page 93 SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.jaycar.com.au SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.jaycar.com.au SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.jaycar.com.au SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.jaycar.com.au SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.jaycar.com.au SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.jaycar.com.au SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.jaycar.com.au SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.jaycar.com.au PRODUCT SHOWCASE Digital Sound Level Meter from Jaycar Jaycar Electronics new QM-1590 Digital Sound Level Meter has a lot of features for its $169 price tag. Applications range from industrial and workplace sound measurement to determine safety levels, through to the measurement and performance-checking of hifi speakers. Until now, most low-cost sound level meters have been analog devices but this instrument has both digital and analog output. In fact, it features dual digital/quasi analog display and has auto and manual ranging, sound level “hold” function and both dB “A” and dB “C” weighting. It also offers fast and slow time weighting and maximum/ minimum level recording. Accuracy is ±2dB under reference conditions and the measurement frequency range is 300Hz to 18kHz. A detailed instruction manual is included, as is a 9V alkaline battery which will give 30 hours operation. The instrument can be examined at any Jaycar Electronics store throughout Australia or can be ordered direct from Jaycar at 8-10 Leeds St, Rhodes NSW 2138. Phone (02) 9743 5222, fax (02) 9743 2066, email techstore<at> jaycar.com.au You can also view details on Jaycar’s website, www.jaycar.com.au D-day is coming – new TV standards announced One of the main standards required for digital television broadcasting in Australia has just been published by Standards Australia. With D-TV scheduled to commence on January 1 2001, the new standard, AS4599-1999, Digital Television – Terrestrial broadcasting – Characteristics of digital terrestrial television transmissions, defines the specifications for the Australian digital TV system. It will complement the standard currently being developed for digital TV receivers and forms the foundations of the “next generation” of television in this country. The standards are based on the European DVB-T standards with necessary modifications for Australia. Meanwhile, Philips aren’t happy... Philips Australia has called on the Federal Government to carefully consider which definition format it adopts for the introduction of digital TV. Philips believes that the proposal by Australian freeto-air TV stations to adopt high definition TV is premature and will be too expensive for average consumers. Harry Van Dyk, General Manager, Philips Sound and Vision, claims that the current proposal means Australians would be forced to pay up to $15,000 for a high definition TV receiver, display and sound system. Philips is proposing that broadcasters should broadcast standard definition television (SDTV) to give all Australians access to digital TV. Another Safety Recall for Fluke T2 Testers Fluke Australia has found a potential product malfunction in the Fluke T2 Electrical Tester. This notice includes all units manufactured before September 1999, with serial numbers lower than 74165430. The safety notice includes instruments involved in a previous safety notice dated November 1998. The T2’s positive battery contact can corrode over time due to exposure to vibration. In certain situations, the instrument can operate intermittently; sometimes it will turn on and sometimes not. When the malfunction occurs, the tester may not indicate that voltage is present, placing the user in a potentially hazardous situation. T2 owners are directed to stop using their T2 Electrical Testers as soon as possible, even if they have not experienced this problem. Owners of units that have serial numbers between 70521601 and 74165430 have two options: (1) they can request a free Field Repair Kit, containing a new battery contact coil spring, two new AA-size zinc-oxide batteries and installation directions; or (2) they can return their T2 Tester to Fluke Australia for repair. Customers are urged to email david.mayhew<at>fluke.com.au or fax their request for a Field Repair Kit to (02) 8850 3300 as soon as possible. If required, telephone (02) 8850 3333. To return a T2 Electrical Tester, send it with a Safety Notice Return Form (available by faxing Fluke Australia) to Fluke Australia, 26/7 Anella Avenue, Castle Hill, NSW 2154. If the T2 owner’s unit was affected by the November 1998 safety notice, has a serial number lower than 70521601 and it does not have an “R” stamped after the serial number, the T2 was not repaired in the November 1998 recall. This unit now MUST be returned to Fluke for repair. November 1999  53 CTRONICSHOWCASELECT 3990 FULL RANGE $ ELECTROSTATIC Now you can afford the legendary clarity, transparency, depth and precision of an electrostatic speaker. The new Vass ELS-5 is a full range electrostatic speaker, able to faithfully reproduce frequencies from 40Hz-20kHz. • 5 Year Warranty • Wide range of custom finishes. • Individually hand built & tested. MicroZed Computers GENUINE STAMP PRODUCTS FROM Scott Edwards Electronics microEngineering Labs & others Easy to learn, easy to use, sophisticated CPU based controllers & peripherals. PO Box 634, ARMIDALE 2350 (296 Cook’s Rd) Ph (02) 6772 2777 – may time out to Mobile 0409 036 775 Fax (02) 6772 8987 1/42-44 Garden Bvde, Dingley 3172 Pyramid subwoofer Ph 03 9558 0970 Fax 03 9558 0082 separately available email: vass<at>hotkey.net.au http://www.microzed.com.au Most Credit Cards OK PRODUCT SHOWCASE – continued SoundMan X1: quality stereo sound for your PC With high-quality stereo being increasingly available on the Internet, Logitech has released a new generation of speakers specifically intended for this growing market. The SoundMan X1 speakers are housed in attractive, contemporary cases yet require minimal space on the desktop. The system comprises a high- performance subwoofer and a pair of two satellite speakers, rated at 25W RMS. With a recommended retail price of $149, the system is said to bring theatre-quality sound to MP3, DVD, CD music, gaming and other popular multimedia applications. Bundled with the speakers is the MP3 Music Centre CD with more than 100 MP3 songs and internet music software. The SoundMan X1 system will be increasingly available from computer specialist and department stores. Hobby/radio control show for Melbourne A show of special interest to hobbyist readers is being held this month in Melbourne. Featuring virtually every facet of technical hobbies, including electron- s entitlet r e h c n ou This vo 50% discou e t h t u o y try to for en y show. b Hob 54  Silicon Chip ics and amateur radio, radio control cars, boats, planes, model railways, astronomy and many more, the Hobby, Engineering and Radio Control Modelling show is being staged by St Kilda Community Projects at 8 Chapel Street, St Kilda from 10am to 5pm on Saturday 6th and Sunday 7th November. Along with plenty of demonstrations, workshops and seminars there are great bargains promised. And while entry is priced at a very reasonable $2 per person, SILICON CHIP readers can get 50% off by presenting the coupon at left (or even a photocopy) on admission. For more information, contact 0411 837 187 or (03) 9523 1037; email isentz <at>labyrinth.net.au NDS Moves Up Namlea Data Systems (NDS) has relocated to a modern new distribution centre in Lane Cove. Located at Unit 4, 11 Orion Road, the new premises offer vastly improved office and warehouse space for the rapidly-expanding company, now one of Australia’s leading data and networking equipment suppliers. Postal address is now Locked Bag 7, Lane Cove NSW 1595, while the phone and fax numbers have changed to (02) 9429 0800 and 9429 0899 respectively. The NDS National Call Number for general customer sales and support remains unchanged on 1300 30 30 69. Intelligent microphone mixing in a half-rack Shure has release a four-channel automatic microphone mixer which puts the performance of its eight-channel (SCM810) system into a half-rack space. The SCM410 mixer is intended for both installed and touring applications and increases the clarity and intelligibility of speech while dramatically reducing feedback, reverberation and the need for comb filtering. The mixer incorporates Shure’s patented “Intellimix” technology which automatically gates microphones on and off to optimise sound quality. It can be used on its own or linked to other mixers. Shure products are distributed by Jands Electronics who can be reached on (02) 9582 0909. TRONICSHOWCASELECTRO • • • • • • • SWITCHMODE POWER SUPPLIES R.T.N Basic Stamps, SX chips and tools. OZ-made boards and development tools Best pricing on temp, a/d, rtc kits New Xilinx PLCC44 development system New OZ made serial LCD module 2*16 Stepper and R/C servo motor chips New super catalog on CD Rom with 40 meg of Stamp related data. Now available via SAE and our cost $4.50, or free with orders over $125 Phone/Fax 03-9338-3306 http://people.enternet.com.au/~nollet Email: nollet<at>mail.enternet.com.au NEW FROM QUESTRONIX DVS5 Video & Audio Distribution Amplifier 25W500W Extensive Range 6 Sarich Court, Technology Park, Bentley WA 6102 Ph: 08 9470 1177 Fax 08 9470 2844 web: www.computronics.com DVS5 Video & Audio Distribution Amplifier VGS2 Graphics Splitter Five identical Video and Stereo outputs plus h/phone & monitor out. S-Video & Composite versions available. Professional quality. VGS2 Graphics Splitter High resolution 1in/2out VGA splitter. Comes with 1.5m HQ cable and 12V supply. Custom-length HQ VGA cables also available. Check our NEW website for latest prices and MONTHLY SPECIALS www.questronix.com.au Email - questav<at>questronix.com.au Video Processors, Colour Correctors, Stabilisers, TBC's, Converters, etc. QUESTRONIX All mail: PO Box 548, Wahroonga NSW 2076 Ph (02) 9477 3596 Fax (02) 9477 3681 Visitors by appointment only UNIVERSAL WIRELESS DEVELOPMENT SYSTEM EMC Technologies' internationally recognised Electromagnetic Compatibility (EMC) test facilities are fully accredited for emissions, immunity and safety standards. EMC Technologies Melbourne: (03) 9335 3333 Sydney: (02) 9899 4599 Linx RF modules from Clarke & Severn Electronics offer a simple, efficient and cost-effective method of making a product wireless. Want to know more? Contact CLARKE & SEVERN ELECTRONICS PO Box 1, Hornsby NSW 1630 Ph (02) 9482 1944 Fx 9482 1309 email: sales<at>clarke.com.au www.clarke.com.au USB Devices BUSINESS FOR SALE: • • • • • • • • • Escape to the sun in beautiful Coffs Harbour! Stable electronic retail business Easily run by husband and wife team. Agent for GSM carrier Access to large electronics suppliers (niche market). Very strong customer base inc Government depts and schools etc. Five year rental option on current highway premises. Full figures available. Current owners (12 years) are moving to a new business. Price only $55,000 + SAV. Enquiries: phone (02) 6652 5684 or fax (02) 6651 3731 Do you want YOUR product or service showcased to Australasia's most important electronics marketplace? CALL ME: RICK WINKLER on (02) 9979 5644 and let me explain how cost effective the SILICON CHIP ELECTRONICS SHOWCASE can be for YOU! Just PLUG them in! Looking for ways to protect your peripheral investment, while moving to USB-based systems? You can even network two computers via their USB ports & save on the cost of network cards. You're ready to take full advan­tage of USB, but your desktop system does not have USB support? Not a problem, our USB card can add 2 USB ports to any PC with a spare PCI slot. If your motherboard already has USB port pin connectors then you can add 2 USB ports to the backplane. USB type A & B cables also available. Vamtest Pty Ltd trading as Microgram Computers A.C.N. 003 062 100 Web site: Email: www.mgram.com.au info<at>mgram.com.au Hubs, Adapters & Converters, Cables, Scanners, Bar Code Readers, Video Capture & more Check out our web site or Unit 1, 14 Bon Mace Close give us a Berkeley Vale NSW 2261 November 1999  55 Phone: (02) 4389 8444 call Fax: (02) 4389 8388 Setting Up An Email Server If you want to streamline the email service in your office, setting up an on-site email server is the answer. It lets users send local email with ease, without initiating separate dial-up sessions. A popular solution for the Windows platform is the MDaemon mail server package. By PETER SMITH The standard dial-up account offered by most ISPs (Internet Service Providers) is often the starting point for many small businesses when email access becomes a requirement. A modem, telephone line, email application software and dial-up account are all that’s required to get online. Standard dial-up accounts generally include one email address with a username of choice. The domain name part of the address (everything after the ‘<at>’ symbol) is common to all dial-in users of the particular provider for this type of account. For example, a Big Pond dial-in customer would have an address of username<at> bigpond.com Of course, most businesses will want the company name or some derivative of it as part of the domain name. In addition, they also usually want multiple email accounts, including a general company account plus individual user accounts. That’s where the ISP comes into the picture. An ISP can host a suitable domain name on a company’s behalf Fig.1: a minimum system requires the MDaemon Server and the MDaemon Documentation and Help components. Fig.2: during setup, you have to enter the IP address of your DNS server. This address is provided by your ISP. 56  Silicon Chip and can also offer multiple email accounts, each account existing as a separate mailbox on the ISPs’ server. Individual users can then access their email via separate modems but this quickly becomes unwieldy if there are more than three accounts involved. For this reason, most businesses use some kind of sharing technique, so that users can access their respective mailboxes through a common Internet connection point on the local area network. Often, this connection is made via a single 56K modem, which is shared using proxy server software such as WinGate, WinProxy or SyGate, etc. Hardware-based proxy servers are also available. We looked at WinGate in detail last month and we’ll be looking at a couple of hardware solutions in a future article. Adding an email server Although this type of email setup works well in many small businesses, there are several disadvantages. Fig.3: select the option shown here so that MDaemon automatically starts and runs as a system service. Fig.4: first-time users should run the configuration wizard, as this greatly simplifies the setup. All settings can be changed later if necessary. Fig.5: enter your domain name in this dialog box and ignore the default company.mail entry recommended by the wizard. Fig.6: the ISP’s mail server address (POP host) and the logon name and password are entered here (for option A in Table 1 only). Fig.7: leave this dialog box blank if you entered your real domain name in the dialog box shown in Fig.5. Fig.8: select the option shown here if you want MDaemon to automatically connect to your ISP. Fig.9: this is where you assign a name to your dial-up profile and enter your logon name and password. First, adding, removing or otherwise modifying user accounts often requires a call to the ISP and there can be delays while the changes are implemented at their end (a few of the more progressive ISPs do provide web-based account management, however). Second, depending on the ISP, there may be costs involved each time account changes are made. And third, because the email server actually resides at the ISPs end, email between local users must go via the Internet connection. This not only adds to the call costs but can also cause significant delays and inconvenience for companies that depend on fast message delivery. The solution is to install an email server on the company network, effectively relocating the user mailboxes from the ISP’s server to the local LAN. No changes need be made to the type of Internet connection and all existing email client software is retained. The ISP still provides the domain name hosting but rather than manage multiple email accounts, all email for the company is either placed in a single mailbox or forwarded directly to the email server whenever it is connected. One big advantage of this scheme is that account administration is now performed locally, so users can be added or deleted at will. More importantly, local email never leaves the company – instead, it is transferred directly to the recipient’s mailbox on the email server and can be collected immediately. Other important administrative tasks like redirecting important email when employees are out of the office also become much easier. manent Internet connection, which can be a costly proposition for small businesses. By contrast, one of MDaemon’s key features is its ability to operate effectively over a standard dial-up modem connection. By using a feature called DomainPOP mail collection, MDaemon is able to collect all email from a single POP3 mailbox at the ISP and distribute it to local user accounts, which is just the shot for a small business. With this scheme, individual users retrieve all their email from the local server which means that they don’t have to make separate phone calls. Similarly, when individual users want to send email, it too goes to the local server. MDaemon then subsequently sends and retrieves email to and from the ISP at designated intervals (eg, two or three times a day, or even once every hour). MDaemon Most ISPs and some large corporations run their email servers on UNIX operating systems or derivatives but a small businesses will want something that runs on a more familiar platform – either Windows 95/98 or Windows NT. One popular solution is an email server program called MDaemon. This runs on the Windows platform and provides many advanced capabilities, including web-based email access, auto-forwarding, mailing lists, remote administration, auto responders, multiple domains and many other features. Many email servers require a per- System requirements MDaemon is available for both Windows 95/98 and Windows NT 4. It requires a PC with a 486 processor or higher, 8MB of memory minimum (we recommend 16MB) and approximately 30MB of free hard disk space November 1999  57 Fig.10: this dialog box is normal. Just click OK so that you can proceed with the setup. this type of service. The two other commonly available services can be roughly categorised by their connection type, as shown in Table 1. MDaemon supports all of these options. By the way, the services offered and the prices vary significantly between ISPs, so its worthwhile finding out what your ISP has on offer. Installing MDaemon Fig.11: you only have to enter your ISP’s mail server address here. The installation wizard automatically completes the other entries for you. Fig.12: MDaemon automatically formats new user accounts that you create later on using the information entered here. (plus space for any mail that will be stored). The TCP/IP protocol must be installed on the network and, of course, an Internet connection is required. In addition, Dial-up Networking (DUN) must be installed for Windows 95/98 and Remote Access Services (RAS) for Windows NT 4. All popular email client software will work with MDaemon, the only requirement being that they support both the POP3 and SMTP protocols. Compatible email applications include Outlook, Outlook Express, Internet Mail, Netscape Mail and Eudora, etc. MDaemon will work with most types of Internet connection methods (modem dial-up, ISDN, cable modem, etc) but consideration needs to be given to the type of email service provided by the ISP. As mentioned above, MDaemon can be configured to collect all email from a single POP3 account. This is the most cost-effective method, as all the ISP need do is redirect (or alias) all email for the particular domain into one mailbox. Internet connection time is reduced to a minimum too, because the company’s email server need only be online for long enough to retrieve all email from the POP3 account and send any email queued for delivery. However, not all ISPs provide Table 1: Common Email Service Options Option Service & Connection Type A B C Description Single POP3 email account The server connects to the Internet at assigned and standard dial -up intervals, collects all emai yrom a singl e POP3 modem connection. account and sends any waiting email. The server connects to the Internet at assigned interval s to send and recei ve email. At the ISP's end, SMTP email servi ce with emai l recei ved for the company whil e the connection non-permanent modem i s off-line i s held in a queue unti l the next connection connection. i s made. This method i s si mil ar to option A but there are more overheads at the ISP's end. The server i s permanentl y connected to the Internet. SMTP email servi ce with The advantage of thi s system i s that there i s no del ay permanent modem in sending and recei ving email but it i s more costl y connection. than options A and B. 58  Silicon Chip Because of its wide range of features and options, installing and configuring this product could be a rather daunting task for some. Certainly, we don’t recommend that you attempt it unless you have a basic understanding of Internet and networking concepts. On the positive side, MDaemon includes a configuration wizard to help get it up and running with a minimum of knowledge. Comprehensive documentation is also available and can be down­loaded separately in a variety of formats (Microsoft Word, Adobe Acrobat, etc) from http://www. mdaemon.janteknology.com.au Because product improvements occur frequently, you should verify that you have the latest version of MDaem­on (2.8.5.0 at time of writing). The software is available for download from the above listed website and can be evaluated free of charge for 30 days. Before installing MDaemon, it is a good idea to configure and test the Internet connection. Make sure that the system is capable of dialling up and logging in automatically (ie, without the user having to manually enter a username and password each time). When you start the installation, you will be prompted to enter some basic information to get things started. Note that unless you have a permanent connection to the Internet, you will probably not require all of the components listed in the “Select Components To Install” dialog box (see Fig.1). A minimum system requires the MDaemon Server and the MDaemon Documentation and Help components. After all files have been installed, you will be asked for the IP address of your DNS server (see Fig.2). Enter the DNS address provided by your ISP here. Note that MDaemon has the ability to run as a system service. This is the preferred option because MDaemon starts and runs automatically, regard- Truscott’s !RESELLER FOR MAJOR KIT RETAILERS !PROTOTYPING EQUIPMENT !COMPLETE CB RADIO SUPPLY HOUSE !TV ANTENNA ON SPECIAL (DIGITAL READY) !LARGE RANGE OF ELECTRONIC COMPONENTS Professional Mail Order Service Fig.13: double-clicking the envelope icon in the System Tray brings up the Message Router window. This lets you add and delete user accounts, gives a running update of all email activity on the network (in the righthand pane) and lets you change any of the settings. less as to whether anyone is logged in to the server or not (see Fig.3). At the completion of the installation, you have the option of running the configuration wizard (see Fig.4). This is a good idea for first-time MDaemon users, since it greatly simplifies the setup. Wizardry The following examples show the settings we used in the wizard’s various configuration windows. Note that we selected option A in Table 1 as our service type of choice. As shown in Fig.5, the wizard recommends using the default domain company.mail. However, we recommend that you enter your real domain name here. If necessary, the domain name can later be changed in the Setup, Primary domain menu accessed from the main MDaemon window (ie, the Message Router window). If MDaemon will be using Domain­ POP mail collection (option A in Table 1), enter the ISP’s mail server address (POP host) and the logon name and password for the POP3 account (see Fig.6). If not, you can leave all these fields blank. Next, if you used the default company.mail domain name, enter your real domain name in the dialog box shown in Fig.7. Conversely, if you previously entered in your real domain name (as recommended), leave this field blank. The next two windows deal with dial-up connection settings – see Truscott’s Come In And See O New Storeur ELECTRONIC WORLD Pty Ltd ACN 069 935 397 Ph (03) 9723 3860 Fax (03) 9725 9443 27 The Mall, South Croydon, Vic 3136 (Melway Map 50 G7) email: truscott<at>acepia.net.au www.electronicworld.aus.as SMART FASTCHARGERS® 2 NEW MODELS WITH OPTIONS TO SUIT YOUR NEEDS & BUDGET Now with 240V AC + 12V DC operation PLUS fully automatic voltage detection Use these REFLEX® chargers for all your Nicads and NIMH batteries: Power tools  Torches  Radio equip.  Mobile phones  Video cameras  Field test instruments  RC models incl. indoor flight  Laptops  Photographic equip.  Toys  Others  How MDaemon Manages Multiple Email Accounts Using A Single Mailbox At The ISP Let’s say, for example, that an organisation has a domain name stooges. com and an accompanying mailbox three<at>stooges.com hosted at their ISP. The organisation arranges for their ISP to alias, or forward, all email messages sent to the stooges.com domain (regardless of the username) to the three<at>stooges.com POP3 account. This means that all email messages with the stooges.com domain name received by the ISP – eg, larry<at>stooges.com, curly<at>stooges.com and moe<at>stooges.com - are deposited in the three<at>stooges.com POP3 account. MDaemon can then log into this single POP3 email box, retrieve all email messages, sort them out by user name (ie, larry, curly and moe), and send them to the matching POP3 mailboxes that you’ve defined in MDaemon. Larry, Curly, and Moe can then log into their individual MDaemon POP3 accounts on the local server and send/receive their email as usual – from the MDaemon web site. Rugged, compact and very portable. Designed for maximum battery capacity and longest battery life. AVOIDS THE WELL KNOWN MEMORY EFFECT. SAVES MONEY & TIME: Restore most Nicads with memory effect to capacity. Recover batteries with very low remaining voltage. CHARGES VERY FAST plus ELIMINATES THE NEED TO DISCHARGE: charge standard batteries in minimum 3 min., max. 1 to 4 hrs, depending on mA/h rating. Partially empty batteries are just topped up. Batteries always remain cool; this increases the total battery life and also the battery’s reliability. DESIGNED AND MADE IN AUSTRALIA For a FREE, detailed technical description please Ph (03) 6492 1368; Fax (03) 6492 1329; or email smartfastchargers<at>bigpond.com 2567 Wilmot Rd., Devonport, TAS 7310 November 1999  59 Fig.14: you just enter the user’s real name in the Account Editor and the other fields are automatically completed for you. You can then manually change the entries if you wish. Fig.15: the Account Alias Editor lets you nominate (or alias) an account as the postmaster. The postmaster receives “undeliverable” email and is notified of any problems. Fig.8 & Fig.9. Most users will want MDaem­on to automatically connect to their ISP. Note that the connection times can be chosen by the administrator and we look at this in the final step of the installation. At this point, the MDaemon engine automatically starts if you are running NT, while Windows 95/98 users have to reboot to achieve the same result. As soon as the engine starts a small envelope icon appears in the system tray. Our installation reported “missing or damaged” settings when it was first launched (see Fig.10). However, based on the author’s previous experience, this appears to be normal for a new installation, so just click OK. MDaem­on now presents three additional screens to allow inspection of the default settings. In the Domain Configuration dialog box, under the Message Delivery section, enter your ISP’s mail server address (see Fig.11). The installation 60  Silicon Chip wizard automatically completes the other settings, so you don’t have to worry about these. The next dialog box is the “New Account Defaults”, as shown in Fig.12. This lists a number of templates and MDaemon formats new user accounts according to the information entered here. Careful setup of the templates greatly speeds up the creation of new accounts later on, as MDaemon will do most of the work for you (as we shall see later)! We only needed to change two settings for our installation. The default Mailbox setting generates mailbox names from the first initial and last name of each user; ie, the Mailbox entry is $USERFIRSTINITIAL$$USERLASTNAME$. This means that if we later create a new user called Peter Smith, this would produce PSmith as the mailbox name and the email address would be PSmith<at>siliconchip.com.au. However, we wanted our addresses to use full names, with a dot as a separating character. To do this, we simply changed the Mailbox entry to: $USERFIRSTNAME$.$USERLASTNAME$ (note the full stop between the two dollar signs in the middle). Therefore, when we later create our email account, we get Peter.Smith<at>siliconchip.com.au Note that you should never use spaces in email addresses; use an underscore (_) or dot (.) character instead. The default POP password setting is defined so that it generates passwords from the first initial and last name of each user. For testing purposes, we changed this setting to use the string “pass” for all passwords. Note that POP passwords are case sensitive; user names are not. This dialog box is also accessible from the main MDaemon window (ie, the Message Router) via Accounts, New account defaults. The next dialog box shows the Miscellaneous Options settings. No changes are required here for a basic installation, so just click OK to close this window. Putting it work We now need to add at least one user, define who the “postmaster” will be and tell MDaemon when to connect to the Internet to send and receive email. You do this by first double-clicking the envelope icon in the System tray to open the main MDaemon window. This is called the Message Router window and is shown in Fig.13. Next, select Accounts, New Account from the menu bar to display the Account Editor (see Fig.14). When you enter a user’s real name, you will notice that MDaemon automatically completes the other fields according to the account templates we defined earlier. Generally, you will not need to alter items on other tabs in the Account Editor, so click on the OK button to create the account. Internet email systems also require a “postmaster”. This user receives notification of “exceptions” that occur within the system, such as undeliv­ erable email. Note that MDaemon does not require a separate account for the postmaster. Instead, you can “alias” this user to any other defined user. Select Accounts, Account Aliases from the menu bar to display the Account Alias Editor (see Fig.15). In the Alias field, enter “postmaster”, then from the Mailbox drop-down list, select the appropriate user and click on the Add button. Finally, you need to tell MDaemon when to connect to the Internet. Open the Send/Receive Scheduler window by selecting Setup, Send/ receive scheduler from the menu bar (see Fig.16). If you are using either options A or B in Table 1, select which days and times you wish email to be transferred. In our example, we have chosen to connect at 8am, 12pm and Fig.16: this is where you set up a schedule so that MDaemon automatically connects to your ISP to send and receive email (options A or B in Table 1 only). Glossary Of Common Terms Looking for an old valve? DNS: the Domain Name Service is an Internet service that translates domain names into IP addresses. Names are obviously easier to remember than long strings of digits and names can be cleverly chosen to represent the owner’s interests. or a new valve? Domain Name: a name that represents one or more IP addresses. Domain names form part of all resources, including email addresses, on the Internet. For example in the address http://www.mdaemon.com, the domain name is mdaemon.com. In email addresses, everything after the “<at>” symbol is the domain name. IP Address: a string of digits that identifies a computer or device on a TCP/IP network. The format of an IP address is a 32-bit numeric string written as four numbers separated by periods (full stops). Each number is in the range 0-255. For example, 203.2.191.122 could be an IP address. Regulatory bodies assign all IP addresses used on the Internet, primarily to avoid duplicates. ISP: Internet Service Provider. POP3: Post Office Protocol (version 3) is a protocol used by most recent email applications (also called email clients) to retrieve email from a server. Proxy Server: a server that acts as an intermediary between a client application, such as a Web browser, and the Internet. Proxy servers often provide security, administrative control and local caching of web pages. SMTP: the Simple Mail Transfer Protocol is a protocol for sending email messages between servers. It is also usually used to send messages between email clients and servers. Most email systems connected to the Internet use the SMTP protocol. TCP/IP: an Abbreviation for Transmission Control Protocol/Internet Protocol, TCP/IP is the defacto standard for communication on the Internet. 6pm from Monday to Friday. In the same window, click on the RAS Setup button. If you are using option A in Table 1, you do not need to change any settings (see Fig.17). If you are using option B , leave the “Keep Sessions Alive For At Least” option selected and enter the number of minutes that your server needs to be connected each time it dials in. Your ISP will be able to tell you what the minimum time is but from our experience it is usually around 10 minutes. Finally, if you are using option C, select the “Once Established, MDaem­ on Will Not Close the RAS Session” option. By the way, you can force MDaemon to connect immediately and send/ receive email. You might want to do this for urgent email or for testing purposes, for example. To do this, either hit the F9 key or select Queues, Process local and remote queue and MDaemon will connect immediately. Well, that’s about it for a basic setup. If you’re running WinGate or BUYING - SELLING - TRADING Australasia's biggest selection Also valve audio & guitar amp. books SSAE DL size for CATALOGUE ELECTRONIC VALVE & TUBE COMPANY PO Box 381 Chadstone Centre VIC 3148 Tel: (03) 9571 1160 Fax: (03) 9505 6209 Mob: 0417 143167 email: evatco<at>mira.net Silicon Chip Binders REAL VALUE AT $12.95 PLUS P&P Fig.17: the entries in this dialog box depend on which option you are using from Table 1. They determine how a dial-up session terminates. other software that shares the Internet connection, additional changes can be made to optimise connection usage. Check out the MDaemon user’s manual or visit the web site at http:// www.mdaemon.janteknology.com. au for tips. SC  Heavy board covers with 2-tone green vinyl covering  Each binder holds up to 14 issues  SILICON CHIP logo printed on spine & cover Price: $A12.95 plus $A5 p&p each (Australia only) Just fill in & mail the handy order form in this issue; or fax (02) 9979 6503; or ring (02) 9979 5644 & quote your credit card number. November 1999  61 Build the addacom and add extra stations to any existing 2-way intercom Do you have a 2-way intercom in your home? Would you like to add extra stations at very little cost? The solution is to build the Addacom. It contains a modest amount of switching and circuitry to add the extra stations. The Addacom is a little black box with four pushbuttons which sits next to the master station of a 2-way intercom. If an extension calls in, a light emitting diode (LED) lights up on the Addacom and then you press the corresponding button to talk to that extension. In reality, the Addacom consists of the bank of four pushbuttons and a little circuitry to indicate which extension is calling the master station. Before we describe how the Addacom works, we need to have a look at the circuit function of a typical 2-station intercom. There is really very little to them. Most push-to-talk types consist of a master and one or more slave stations. The active master station unit con- Design by Paul Hoad* 62  Silicon Chip tains an amplifier, a small loudspeaker, a battery and a switch. The slave station is much simpler, consisting of another small loudspeaker, a switch, capacitor, a diode and a light emitting diode (LED) The small loudspeakers do double duty, acting as microphone or loudspeaker, depending on whether the stations are talking or listening. The basic intercom circuit is shown in Fig.1. We have one pair of wires and one amplifier (IC1) so only one end can speak to the other at any given time. Switch S1 at the master station swaps the connections to the two speakers to allow for two-way communication. The telecommunication term for this is “half-duplex” – one station can listen while the other talks and then the first station can talk while the other listens. Telephones are actually “full duplex” devices since both parties can simultaneously talk and listen. The other type of communication is “simplex” and is one-way only, eg, broadcast TV, radio etc. You cannot talk back (swearing at the TV when the ad break comes on is not counted!) Signaling Let’s look at the manner in which a remote station signals the master station. Much of the terminology is “borrowed” from telecommunica- Fig.1: this is the circuit of a typical 2-way intercom. The push-to-talk switch (S1) swaps the remote and local speakers so that they can talk or listen. tions so we’ll start by looking at what happens in a telephone circuit. When a telephone is in the idle state it has a high DC resistance (although modern phones draw a tiny current to maintain their memory for stored numbers). The exchange or PABX supplies 48V DC. When you lift a phone handset off its cradle, the hook switches complete a circuit and current flows back to the exchange. The exchange senses this current and sends you a dial tone. This is called loop signaling. In our typical 2-way intercom, 9V is sent from the master to the remote Addacom opened out, viewed from the component side of the PC board. It is powered from a plugpack supply with battery backup. November 1999  63 Fig.2: the Addacom is really just a 4-way switch-bank and the rest of the circuit just shows which extension has called in so the right button on the switch-bank can be pressed. station. When the button is pressed at the remote station, switch S2 creates the loop condition by shorting out the capacitor in series with the speaker. The master station responds to the loop by producing a tone which is heard at the master end. On hearing the tone, the user at the master end would turn on the intercom via double-pole switch S3 and press the momentary contact switch S1 (press to talk). The amplifier is now powered on via D1, and S1 reverses the line polarity and both LEDs are turned on. What happens now is that the loudspeaker in the master station is connected by switch S1 to the amplifier’s input to act as a microphone while the speaker at the remote end becomes the loudspeaker load for the amplifier. The user at the master end is now able to talk to the remote station: “Yeah, mate?” On releasing the press-to-talk switch, the loudspeaker in the remote 64  Silicon Chip station becomes the microphone while the speaker in the master station becomes the loudspeaker load. The remote user is then able to speak: “About time, mate! Is it beer o’clock yet?” By pressing S1 the user at the master end can speak again. Note that when the user at the extension speaks, he does not have to press the button; he can talk “hands-free”. We won’t carry on the scintillating conversation but you get the picture. For the purpose of this exercise, our typical remote station only has those five components shown in Fig.1: loudspeaker, LED, diode, a 100µF bipolar capacitor and a pushbutton switch. If we want to add extensions, each one will need those five components. You probably won’t need to go out and buy them because they will already be in your parts collection. Switching & transmission A telephone exchange will analyse your sent digits, then switch you to the appropriate destination. These days this is achieved using digital techniques. The switching used in this project is via a switch bank. You decide which extension needs to be switched by seeing which LED is lit. There is also an audible indication. The voice signals are restricted to a frequency range between 300Hz and 3.4kHz and sent down a single pair of wires to your phone. In some of the very large cables you may have another 3000 pairs in the same cable travelling side by side for several kilometres. Each pair of wires is colour coded so that they can be identified. Because all the cable pairs are so close together, won’t the lines be noisy? Well, no. By twisting the pairs, interference tends to be cancelled. Twisting the pairs results in the noise signals having the same polarity and magnitude on each leg. Without a potential signal difference between the two legs, no noise current flows. With our intercom we should also try to twist the cable pairs joining each station. Our main enemy here is hum from your home’s electrical wiring which is amplified along with the weak signal from the microphone (speaker). Circuit description Fig.2 shows the circuit of the Addacom. Each extension is wired to the screw terminal block and then to switches S1, S2, S3 or S4 which are interlocked as a switch-bank. The interlocked switch-bank connects only one extension to the master unit at a time. Let’s assume that extension 1 is selected and the others are not. In this condition, extension 1 is switched straight through to the master; terminals 1 & 2 on the terminal strip are switched via S1a & S1b through to terminals 9 & 10, the connections to the master station. For the other three extensions which are not selected by the switch bank, switches S2a, S3a and S4a, connect the positive supply line to one side of their respective remote switches. Now, if extension 2 calls by having its remote button pressed, the resulting line (loop) current via S2b will turn on transistor Q2 via diode D3 to light LED2 and sound the buzzer via diode D4. The pulsating current in the buzzer also causes a tone to be heard in the speaker for extension 2. OK. So now the person at the master station sees that LED2 is alight and presses switch S2 to select that extension. So extension 2 is connected to the master station and extension 1 is automatically disconnected (desel-ected); that is the nature of a switch-bank. The master station and Extension 2 can now talk to each other, as described above, ie, the master station has to be turned on as before and its press-to-talk switch pushed to allow the master to speak and so on. The same process would occur if extension 3 called in. In this case, the loop current via S3b would turn on Q3 via diode D5, lighting LED3 and powering the buzzer via D6. The user at the master station sees LED3 lit, presses S3 and then exten- Addacom viewed from the copper side of the PC board. The terminal block in the foreground is used to connect to the master station and up to four add-on stations. 6-pair telephone cable connects the PC board to the terminal block. sion 3 can talk. the unit will be powered at all times, regardless of blackouts. Each of the four transistors has a 47µF capacitor following its respecA couple of points need to be extive diode and before its 100kΩ bias plained about the extension stations resistor. The capacitor allows the re- and for clarity we have repeated the spective transistor circuit in Fig.3. to remain on for First, the 100µF about 15 seconds coupling capaciafter the extension tor for the speakhas called in. er needs to be a This gives the non-polarised (NP user at the masor BP) type. ter station enough This is because time to identify the switching of the extension and the master station Fig.3: this is the circuit of each select it. reverses the supply extension. The 100µF capacitor Power for the to the extension. must be a non-polarised (NP or Addacom comes For the same reaBP) type because the Addacom from a 9V DC plugson, the LED in the switching causes the supply pack via diode extension station polarity to be reversed. D10 and resistor must be connected R5 or from a 9V in series with a battery via diode diode because in D9. Zener diode ZD1 in conjunction the idle condition, the LED will be with R5 provides crude regulation of reverse-biased by the positive supply the supply to 10V. line from the Addacom. This supply is then available at Construction terminals 11 & 12 to power the master station if this is desired. If you have The Addacom was built in a standboth battery and plugpack supply, ard plastic utility box measuring 130 x November 1999  65 68 x 45mm. We’ll assume that you’re building the Addacom from a kit so all holes will be punched in the plastic lid for the switches and LEDs. The interlocked switch-bank is mounted on the PC board which makes the wiring relatively simple but it does take up a lot of the board space. For this reason, some of the components must be mounted on the copper side, as we shall see. Do not mount the switch-bank first; it is left till last otherwise it is just too difficult to mount and solder the components. You can start with the four transistors; these are mounted on the copper side of the board, as shown in Fig.5. Take each transistor and push its leads through the holes on the copper side. Then bend the transistor back so that the curved part of the transistor body touches the board. Just solder the emitters at this stage to hold them in. Mount and solder all other components, as shown on Fig.5 & Fig.6 except for the LEDs, diode D9 and the switch-bank. Make sure that you clip off the excess leads from the copper-side-mounted components. It is important that the capacitors be perpendicular (straight up) from the board otherwise the switch actuators will hit them. Before you install the switch-bank, terminate and solder the four extension pairs, the master pair and the positive battery wire to the tagstrip side of the switch. Is it an intercom station? It is now but it started life as a flip-top computer disk box (from Jaycar). The photo at right shows how the components were mounted inside. Even better, a cheap computer speaker (far right) could be modified to suit. 66  Silicon Chip Fig.4: these are the wiring details for the cable pairs from the switch-bank to the 12-way terminal strip. To avoid shorts it is best to hook the wire across the top rather than wrap it around the tags. Twist each pair as you solder them and follow the colour code. Attach the M3 screws and 12mm standoffs to the switch and solder its terminals to the board. Bend the leads of the four LEDs around the shaft of a small screwdriver to achieve the 90° bend as shown in one of the photos. Put the LEDs in but don’t solder them in at this stage. Attach the switch-bank to the front panel and align the LEDs with their holes, then solder them to the PC board. The buzzer is held in place with two 3mm self-tapping screws. Next, solder the battery snap connector, buzzer and negative supply wire to the PC board. The positive terminal of the DC power connector is wired via diode D10 to the PC board. Parts List Fig5: these components are mounted on the copper side of the PC board. Fig.6: these components are mounted on the top of the board and must be inserted and soldered before the switch-bank is soldered in. Put a kink in one of D10’s leads to allow for some flexing. Next, attach the 12-way terminal block to the plastic case. All the wires pass through a 1/4-inch hole to the terminal block. Terminate all the wires according to the diagram of Fig.4. You can attach the battery to the inside of the case with a piece of double-sided adhesive tape. Making the extensions There are several possible approaches to building the extension stations – we show just two. One shows the speaker and the circuit components of Fig.3 wired into a standard 3.5-inch disk box – cheap and cheerful. A more elegant approach is to purchase a pair of cheap computer speakers (we got ours from Woolworths at just $6.99 pair!) and wire in the same components. Alternatively, you can use just about any small plastic case that comes to hand and install the components into it. Perhaps the biggest job is running all the cable pairs from the extensions back to the Addacom which will probably be mounted quite close to the master station intercom. We’ll leave you to figure out the 1 PC board, 100 x 35mm 1 plastic utility case, 130 x 68 x 45mm, with punched lid 1 4-way interlocking switch-bank (S1-S4) 1 9V or 12V DC plugpack 1 DC connector socket to suit plugpack 1 9V battery and snap connector 1 electronic buzzer 1 12-way insulated terminal block 2 12mm tapped spacers 4 M3 6mm screws 2 6mm self-tapping screws 4 47µF 16VW PC electrolytic capacitors Semiconductors 4 BC549 NPN transistors (Q1,Q2,Q3,Q4) 8 1N914, 1N4148 small signal diodes (D1-D8) 2 1N4004 diodes (D9,D10) 1 10V 400mW zener diode (ZD1) 4 5mm red LEDs (LED1-LED4) most suitable installation. Testing Place a short across the cable pair for each extension on the terminal strip and make sure the buzzer and the relevant LED lights up. If the buzzer sounds when you connect an extension, reverse the cable pair. If the buzzer sounds without any extension calling, then you have a short somewhere in your cabling. To find it, remove one wire of each cable pair and then replace it; repeat until you find the shorted pair. Next, plug the Addacom into the extension of your existing intercom. Turn the intercom off. Select each extension with the switch-bank. If the intercom buzzes then you may need to reverse the wires. If you intend using one of the extensions as a baby monitor then you won’t need any switching for that station – just wire in a speaker and its 100µF coupling capacitor. If you only need a couple of extensions then you can use the spare switch positions for other devices. For example, you might connect a microswitch on a roller door, to tell when it is open. This is handy because it stops the SC door being left open at night. Resistors (0.25W, 1% or 5%) 4 100kΩ 1 15kΩ 1 220Ω 2 47Ω Extensions 4 speaker cabinets (see text) 4 8Ω miniature speakers to suit cabinets 4 100µF bipolar electrolytic capacitors 4 red LEDs 4 1N4001 diodes 4 momentary contact pushbutton switches Miscellaneous Telephone cable for extension wiring, double sided tape, solder. Where to buy the kit * The copyright for this project is owned by Hoad Electronics. They can supply a complete kit which includes a punched and screened front panel for the plastic box. The price is $34 including postage and packing anywhere within Australia. Contact Hoad Electronics at 19/314A Pennant Hills Rd, Carlingford, NSW 2118. Phone/fax (02) 9871 3686. email: hoadelectronics<at>one.net.au http://web.one.net.au/ ~hoadelectronics November 1999  67 VINTAGE RADIO By RODNEY CHAMPNESS, VK3UG The case of the disappearing TV sets Vintage black and white TV receivers from 1956 onwards are now almost impossible to find. If we don’t start saving the remaining examples now, these sets may well go the way of the Tasmanian Tiger and the Moa. Valved black and white TV sets are now quite rare animals. Is black and white valved television a part of our electronics heritage? You bet it is and the time to start collecting is now, otherwise the sets will be gone for good. Back in the March issue of “Silicon Chip”, in the Publish­er’s Letter, Leo Simpson suggested that it was time to save those old black and white TV sets. Leo and I have discussed this sub­ject on a number of occasions and this article will be followed by others later on. I would like to dedicate this 68  Silicon Chip particular article to the late Rex Wales of the Historical Radio Society of Australia (HRSA) who was encouraging members to get into vintage television restoration. Sadly, he died before much work could be done to find and restore these old sets. Why write about vintage television! And what has television got to do with vintage radio anyway? Directly, probably very little. However some members of the Historical Radio Society of Australia (HRSA) and the New Zealand Vintage Radio Society (NZVRS) have started to show an interest in preserving another aspect of our electronic entertainment medium. Therefore, an article on vintage tele­ vision sets is very appropriate at this time. Unfortunately, most black and white valve TV sets have been consigned to the rubbish heap. In the process, quite a few were scavenged for parts, power transformers, valves and other odd bits and pieces. If you don’t have a 6BX6 or a 6BL8 in your radio junk box, you have never wrecked a B & W TV set. Have you considered how rare early black and white tele­vision sets really are? Could you lay your hands on one of the original 17-inch Astor, Philips or AWA TV sets, for example? Very Below: these four sets are all AWA P1s and have yet to be restored. Often, it’s necessary to strip parts from one set to get the others going. few of us could. Fortunately, some sets been stored in garages (or even under the house), so there are still a few sets around. Most vintage radio collectors have probably shunned col­ lecting black and white TV sets for several reasons: (1) the sets are usually bulky (there aren’t any mantle set size B & W TVs!); (2) they haven’t considered B & W TVs as being “vintage” sets. We didn’t think of old radios as vintage until about 15 years ago and we are now waking up (almost too late) that B & W TV sets are vintage as well. (3) B & W TV sets are perceived as being complex – which they are compared to an AM radio receivers. However, this and the follow­ing articles may help to dispel some of the mystery. (4) Replacement parts such as picture tubes, line output trans­formers and deflection yokes can be difficult to obtain. However, even if specialised parts are hard to get, it doesn’t mean that sets shouldn’t be collected – after all, static displays of our early TV heritage are much better than no dis­plays at all. For this reason, I hope that this article and later ones will help readers to get into this exciting “new” aspect of vintage radio/television collecting and restoration. This photo shows an AWA P1 with the cabinet removed. This set is compact, has an 11-inch picture tube and uses 13 valves. A concise history Television in its various forms has been around for quite a long time. However, like wireless (radio), it has taken quite a few decades to evolve into the sophisticated communications medium that it is today. When the question is asked as to who invented television, the usual answer is John Logie Baird in the 1920s. But although he was at the forefront in developing the mechanical television system, there were many others who had also experimented with mechanical systems, including here in Australia. Baird pushed for his system to be accepted by the relevant authorities but it was never going to be suitable for domestic use for many reasons – the prime one being that it was a mechani­cal nightmare. But although the mechanical system was unsuitable for use by the general public, it did show that moving pictures could be sent by radio waves. This opened the way for a fully electronic method The AWA P1 set is quite easy to service, as the PC board swings out to give access to all the parts. The picture tube can be replaced in around 15 minutes. to take over a decade later. As a matter of interest, the mechanical scanning type of TV system only had 30-60 lines to convey the picture. By contrast, a modern PAL TV system has 625 lines and between 150 and 600 times the definition of the mechanical systems. The early pictures were sent on the AM broadcast band and on nearby frequencies and only required 10-40kHz of bandwidth, depending on the number of lines used in the particular system. Television became practical only when a fully electronic system was developed in the mid 1930s. In fact, Britain had a working electronic tele­ vision system by 1936, using a 405 November 1999  69 The AWA 242 was an up-market 21-inch set that was made during the mid 1960s. It had 21 valves and was fitted with a rotary VHF tuner. line system. Of course, a number of quite complex problems had to be overcome before this became a reality, including the development of the first practical cathode ray tube (picture tube) by V. K. Zworykin in 1929. Note that, in those early days, the pictures were rather small, being only a few inches across. Australian & NZ television Television for the masses came to Australia in 1956, in time for the Melbourne Olympic Games. So, we’ve had TV in Austra­lia for over 40 years! Who’d have thought that it was as long ago as that? Before that, there were some early experiments following the 1920s with mechanical low-definition systems which were transmitted mostly on the broadcast band. New Zealand had experimental TV transmissions in Auckland from 1957 onwards and television for the masses by 1960. Much earlier experiments probably took place there as well but I have no information on that subject. It’s interesting to note that “Radio 70  Silicon Chip & Hobbies” ran a television course in instalments during the early 1950s. This would have helped many servicemen later in the decade, when it came to servicing the new technology. Before the introduction of TV, servicemen were used to the 4/5-valve mantle AM receiver and the occasional 8 or 9-valve multiband receiver which had a few extra bells and whistles on it. What a shock TV was! Sets typically had up to 25 valves (eg, the STC 730-SU1), with several of these valves having at least two functioning sections (eg, the 6BL8). Some later sets used as few as nine valves but many of these were multifunction “compac­trons”. In short, the circuitry and its functions were a whole new ballgame for many servicemen of the era. Some older servicemen hastily decided to retire but many others went back to school and learnt all about the new marvel. Early TV sets In most cases, the black and white valve TV sets were con­soles, as were the more elaborate radio sets of the late 1920s through to the late 1940s. The average wage in the early 1960s was around 20 pounds a week and I remember large console sets costing up to 300 pounds at that time, or about 12 weeks pay. Today a colour TV set can be purchased for one week’s average pay. In this first article, I have no intention of describing the restoration of any particular set. Instead, the accompanying photos are intended to show readers what some of the sets looked like from the outside and to give a few glimpses of the internal circuitry. Note that the sets shown here have yet to be restored. The AWA 242 console in the photographs was one of the more up-market sets during the mid 1960s. It had 21 valves and the picture tube was nominally a 21-inch unit. It was fitted with a VHF tuner as only VHF stations were available in Australia until the advent of colour TV. The AWA P1 is my favourite valve black and white portable TV set. It is compact, has an 11-inch picture tube and has a total of 13 valves. The set is quite easily serviced, as the PC board swings out as shown in the photographs. In fact, it is so easy to service that a picture tube can be replaced in around 15 minutes. I have a total of five P1 sets, which means that I will be able to get at least one operational using one or more of the others for spare parts. I suspect that this is the approach most vintage TV restorers will have to adopt when it comes to restor­ing sets to working condition. Another very interesting set shown in the photos is the Healing. This is a solid state receiver made towards the end of the black and white era. Technical details TV receivers are very different from AM radio receivers, although there are a few similarities between them. For example, TV sets are superhets as are most vintage radios from around 1935 onwards. TV sets use much higher operating frequencies, however. During the black and white era in both Australia and New Zealand, the frequency range tuned by the sets was in the very high frequency (VHF) band, from around 44MHz to 225MHz. This is considerably higher than the highest frequency tuned on most dual-wave AM receivers, which usually don’t go past 18MHz. As in AM broadcast receivers, the local oscillator operates higher than the frequency to which the set is tuned by a fixed amount. This is known as the intermediate frequency, or IF. The IF for the picture carrier was 36MHz, and the sound carrier IF is 30.5MHz. These frequencies are slightly different today. In some early TV sets, two separate IF amplifier stages were used for these two separate parts of the TV signal. However, apart from some rare exceptions, all the sets made in Australia (and, I imagine, New Zealand) used only one IF channel for both sound and vision. In a normal AM broadcast receiver, the IF bandwidth re­ quired for high-quality music reproduction is around 20kHz. However, a vision signal requires a very much wider IF bandwidth for quality pictures to be reproduced – around 7MHz in fact, for both the sound and picture. We’ll explain why this much bandwidth is required in a later article. The signals are usually detected by a germanium diode, after which the audio and picture signals are separately pro­cessed. The audio is fed to an IF amplifier stage on 5.5MHz and thence to an FM detector and audio amplifier. At the same time, the video (picture) information is amplified by a video amplifier which usually has a response from DC to about 5.5MHz. A number of other important components also appear at the detector. These include the vertical and horizontal synchronising (or sync) pulses. These pulses are processed and ensure that the picture is “locked” vertically and horizontally into position on the screen. The horizontal sync pulse peak level is also used to provide automatic gain control. As an aside, most people will have seen pictures that roll vertically or tear horizontally. This is usually caused by a fault in the vertical or horizontal sync circuitry. The other two important components are the vertical and horizontal blanking pulses. These pulses are necessary to blank the screen at set intervals, so that retrace lines aren’t visible when the electron beam jumps to the start of a new line or to the top of the screen. This view shows the chassis layout of the AWA 242 TV receiver. The set is fairly easy to work on, with good access to most of the major parts. This portable b&w TV set carries the Healing brand name. It was made towards the end of the B&W era and uses solid state (transistorised) circuitry. As can be appreciated from this, quite a bit of circuitry is required to process a “composite” video signal to achieve the quality of picture and sound that we have become accustomed to. In later articles, simplified descriptions of how the sets work will be presented, to assist restorers in the task of re­storing their black and white TV receivers. As Leo pointed out in his Publisher’s Letter, these sets are worth restoring and are part of our history, so start collecting even if a complete restoration is beyond you at this stage. Finally, a few words of caution – be very careful how you treat the picture tube. A leather apron and protective glasses should always be worn when working with a picture tube, to give protection if the tube implodes. SC November 1999  71 Foldback speakers are essential to any live performance in a large venue when high power amplification is employed. They enable each musician to hear his or her own playing, over and above the general sound level. Build these and save a bundle of dollars compared with commercial units. By JOHN CLARKE FOLDBACK LOUDSPEAKER BUILD A T HIS VERSATILE FOLDBACK loudspeaker is suitable for stage musicians, vocalists, entertainers and performers. It uses readi­ ly available loudspeaker drivers and can be built using quite basic hand and power tools. Foldback loudspeakers, commonly called “wedges” because of their shape, are used by musicians so that they can hear them­selves over and above the general noise level in a performance venue. Often there are several foldback loudspeakers on stage, for example, one for the keyboard player, one for the vocalists and one or two for guitarists (bass and lead). Without foldback loudspeakers, the performers would have to rely on the sound projected to the audience via 72  Silicon Chip conventional “front of house” loudspeakers. However, they may find it difficult to hear from their location behind the speakers and there can also be a considerable delay before the sound reaches them. This delay will cause the performance to become slow and very deliberate as the performers attempt to sing or play and then wait to listen. With foldback speakers, the performance can be kept tight and lively. Foldback loudspeakers are designed with quite different criteria compared to conventional types and as a consequence they look and sound different. While conventional loudspeakers pro­ject the sound away from the performers and toward the audience, a foldback loudspeaker does the opposite and projects the sound toward the performers themselves; this provides the perfect recipe for acoustic feedback. To avoid this problem, foldback loud­ speakers are designed with a sloping front baffle to project the sound directly toward the performers’ ears. This means that they beam the sound to the rear of the microphones which are usually designed to have a minimum pickup from the rear, so as to mini­mise feedback. The design described here enables the loudspeaker baffle to be set to 35° or 55° to the horizontal, depending on how the box is placed on the floor. This will suit either close-up use (35°) or more distant listening (55°). Sloping the baffle is done for two Two 200mm “Redback” woofers (one shown) and a single Motorola KSN-1141A piezoelectric horn are used in the Foldback Speaker. The woofers are protected by steel mesh grilles. other reasons. First, the sound is beamed at the listener so that they receive the brighter “on-axis” sound. Second, it reduces floor reflections back up to the microphones where their rejection of sound is far less than directly from the rear. Angling the baffle does not solve all acoustic feedback problems though. Feedback can still occur when the microphones are spaced out on the stage, producing the potential for a micro­ phone to receive off-axis sound from an adjacent foldback loud­ speaker. However, this new foldback design has lobes and nulls in its off-axis response and by placing adjacent microphones in the “nulls” of the foldback speaker, acoustic feedback can be further reduced. The SILICON CHIP foldback loudspeaker is 546mm wide, 335mm high and 408mm deep. The box is made of 18mm MDF (medium density fibreboard) and covered in loudspeaker carpet. The loudspeakers are protected with steel mesh grilles and moulded plastic corners provide protection for the enclosure against rough handling. Two 200mm woofers and a single piezo ceramic speaker with a 50mm x 150mm wide dispersion horn are used. The woofers are placed either side of the vertically mounted horn. This arrange­ment produces a symmetrical on-axis frequency response ideal for foldback. If the loudspeaker box is mounted on end, the woofers will be arranged vertically and the horn horizontal. This allows the speaker to be used conventionally for projecting sound toward an audience. The specified woofers (Altronics Cat. C-3060) have a power rating of 60W and a nominal impedance of 8Ω. Their Thiele-Small parameters are Vas Specifications Frequency range: 45Hz to 20kHz <at>-6dB Nominal Impedance: 16Ω Power rating: 200W into 16Ω (equivalent to 400W into 8Ω) Mass: 18kg Dimensions: 546mm (W) x 408mm (H) x 335mm (D) Baffle slope: 35° or 55° Fig.1: only two components are used in the crossover network for the Foldback Speaker. The 1.8mH coil (L1) prevents high frequencies from being fed to the woofers, while the 0.33µF capacitor (C1) reduces the output of the tweeter by about 6.4dB. November 1999  73 Fig.2: these graphs of Fig.2 show the CALSOD response predictions for the woofer and tweeter combination. The solid line on the upper graph is the on-axis frequency response while the dotted line is the impedance plot. The solid line on the lower graph is the phase response. Fig.3: this graph shows the predicted horizontal off-axis response at 30°. The notch in the re­sponse at around 1.2kHz occurs at different frequencies depending on the off-axis angle. 54l, Qts 0.323, Qes 0.398 and Qms of 1.72. Their resonant frequency (Fs) is 34.7Hz and sensitivity is around 89dB at 1W and 1m. The two woofers are connected in series to provide a nominal 16Ω impedance. This means that we can parallel up several foldback loudspeaker units together without overload­ing the driving power ampli74  Silicon Chip fier. Two foldback units will produce an 8Ω load, three units a 5.3Ω load and four units in parallel a 4Ω load. The tweeter is a Motorola KSN­ 1141A piezoelectric horn which incorporates protection circuitry to allow its use with amplifi­ers rated at up to 400W. The protection comprises an incandescent lamp and a positive temperature coefficient (PTC) resistor in series. These components increase their resistance at high power levels to protect the tweeter element. A common design approach when using a piezo tweeter is to simply connect the speakers without a cross­ over. The natural rolloff of the woofer at higher frequencies and the tweeter at lower frequencies are supposed to compensate each other and produce a smooth response. This rarely works well and inevitably the response is markedly louder in the 1kHz to 5kHz region as the sound level is summed from both woofer and tweeter. Usually the tweeter is also quite a bit more sensitive than the woofer and so such systems are often excruciatingly bright at the higher fre­quencies. For our design, we used a series 1.8mH inductor to roll off high frequencies to the woofers. While this will produce a theo­retical rolloff of 6dB/octave above 1.4kHz, in practice the rolloff will be somewhat less than this due to the woofers’ own voice coil inductance. In addition, to compensate for the sensitivity mismatch between the tweeter and woofers, the tweeter is fed via a series 0.33µF capacitor. Since the piezo tweeter itself has a ca­pacitance of 0.3µF, there is a capacitive voltage divider effect which reduces the signal level by a factor of 0.48 or 6.4dB. The speaker circuit diagram is shown in Fig.1. Note that the tweeter is connected out-of-phase with the woofers, to ensure a flat response. An in-phase connection pro­ duces a null in the response at the crossover frequency. The phase was predicted using a computer simulation and took into account the distance between drivers, the offset behind the baffle and the phase response of the drivers and crossover. Some readers may be puzzled by the two jack sockets shown on the circuit. One allows the cable from the amplifier to be connected while the second allows another foldback loudspeaker to be connected in parallel. In practice, you could “daisy chain” four of these foldback speakers together, to provide a nominal 4Ω load to the driving amplifier. Design software The crossover and low frequency ported response was mod­elled using Fig.4: the overall dimensions for the Foldback Loudspeaker. Use a jig saw to cut out the holes and rebate the woofer holes using a router. an Australian developed loudspeaker design program called CALSOD 1.40 (Computer Aided Loudspeaker Sys- tem Optimisa­tion and Design). This allowed the response to be adjusted for optimum smoothness and match- ing between drivers. The program allows off-axis predictions to be made using multiple drivers. November 1999  75 Fig.5: cut the two sheets of MDF as shown in this diagram, to make the sides, base, back, top and baffle. The material can be cut using either a jig saw or a circular saw. You can model and optimise cross­ overs, produce phase and response curves and also position the drivers on the baffle. Copies of this DOS based program can be obtained from Audio­ soft, 13 Beatty St, Ivanhoe, Victoria 3079. Fax (03) 9497 4441. Email audiosoft<at>netwide.com.au. The price for the budget version CALSOD 1.40 with on-disk users manual is $119 including postage and handling. Professional versions which allow importing measured data (CALSOD 3.10) are available from $369. The graphs of Fig.2 show the CALSOD response predictions for the woofer and tweeter combination. The solid line on the upper graph is the on-axis frequency response while the 76  Silicon Chip dotted line is the impedance plot. This shows the expected double hump at low frequencies for the ported design while the value rises rapidly above 5kHz due to the parallel tuned circuit formed by the woofers’ inductance and the capacitance of the piezo tweeter. Note that the impedance does fall back to lower values for fre­quencies above 16kHz, as can be seen at the top righthand corner of the graph. The solid line on the lower graph is the phase response. The abrupt changes from -180° to 180° at 40Hz and 7kHz does not mean that there is a sudden phase change; it is simply drawn that way so the phase plot fits on the graph. Fig.3 shows the predicted horizon- tal off-axis response at 30°. The rolloff above about 3kHz is an estimate for the attenuation in sound level at this angle. The notch in the re­sponse at around 1.2kHz occurs at different frequencies depending on the off-axis angle. Construction The overall dimensions of the fold­ back loudspeaker are shown in Fig.4. It is made from one 600 x 900mm sheet and one 450 x 1200mm sheet of MDF. This is cut as shown in Fig.5, to make the two sides, the base, back, top and front pieces. We used a jig saw and a straight edge guide to make all the cuts although you could also use a circular saw. The baffle is made from the material remaining after the base and front pieces have been cut from the 450 x 900mm sheet. Adjust the jig saw or circular saw so that it is set for a 35° cut. This will enable the baffle to mate flush with the inside of the front piece. Cut the baffle edge at 35° using a straight edge as a guide. Measure 300mm from the inside of this bevelled edge and draw a line across the 510mm length. Now cut this edge at 55°. Assemble the base, back, sides and front pieces together using 8g x 30mm countersunk wood screws. Do not glue the pieces at this stage. Fit the front baffle and top piece in place and check for fit. The bevelled edges may require some adjustment using a plane to produce a good fit. Now secure them with screws. Check that all right angle edges are square and that the straight edges provide a close fit. Mark out the pieces, indicating their orientation and positioning to adjacent pieces. This will make it easier to reassemble later on. Now disassemble the pieces. Cutting the speaker holes Mark out and cut the baffle as shown in Fig.4. Use a router to rebate the woofer hole and a jig saw to cut out the holes. The cutout for the tweeter horn was made deliberately small to allow a greater amount of wood between the woofer and tweeter hole. The 63mm diameter cutout is to allow the piezo element of the horn to fit through the baffle. Check that the speakers and port tubes fit into their respective holes. Now assemble the box using PVA glue on all mating surfaces. Assemble one side, the base, back, top and front first, followed by the baffle, using the screws to secure the pieces in place. Now glue the second side in place. This assembly method will ensure that the baffle can be glued on its side. Wipe any excess glue off with a damp rag. We fitted braces made from 31 x 13mm timber to the inside of the base and back. A 500mm length was used along the base spanning from side to side and a 290mm length along the back spanning from base to top. These were located offset from centre. For extra strength in the enclosure, we also recommend using 12 x 12mm cleats on all right angle joints and along the baffle to side joints. These Fig.6: this diagram shows how the main pieces are fitted together to form the box. The inside of the box is also fitted with braces and cleats, for added strength – see text. should be secured with PVA glue and screws. Smooth the box edges with a rasp or plane to produce a small 2-3mm chamfer and round off the corners neatly. This will allow the plastic corner protectors to fit correctly. Test these for fit before finishing this process. It is probably unnecessary to sand the box since the carpet covering will mask any imperfections in the surface. However, remove any large protrusions from the box surface such as glue runs, screw heads, etc. You will need to drill a 25mm hole in each side of the box for the 6.35mm socket adaptors. The same sized holes are also required if Neutrik panel sockets are used. Standard 6.35mm sockets, while relatively inexpensive, are really not rated for driving high powered speakers. Also they are not airtight and may introduce extraneous noises as the air passes through them. While you can seal them using a cover and silicone sealant, Neutrik locking chassis jack sockets are preferable because they are sealed and attach more solidly to the case. Standard 6.35mm sockets also have a tendency to fall inside the speaker box if the securing nut becomes loose and falls off. Alternatively, you could use Neutrik “speakon line” sock­ ets. These leaded sockets feature a 30A rating, locking plug, rugged line plugs and solid wire clamping on the plug. Covering it with carpet We covered the whole speaker box in carpet and attached it with contact adhesive. The carpet covers the box in four sec­tions. It is cut out using a cutting mat, metal straight edge and a sharp utility knife (eg, Stanley knife). Start by cutting a 510 x 300mm piece and coat the baffle with contact adhesive. Place the carpet over the baffle and then immediately lift it off again. This will reveal the baffle cutout areas on the carpet which do not require coating with adhesive. Coat the required areas on the carpet and recoat the baffle where the contact adhesive has lifted off. Wait for the adhesive to dry, then place the carpet in position over the baffle, making sure that it is oriented correct­ly. Now press the carpet down firmly until it is fully attached. The carpet will then need to be trimmed to reveal the cutouts on the baffle. For the woofer cutouts, trim the carpet to the outer diamet­er of the rebate using a sharp knife. The port cutouts will need to be recut to the November 1999  77 Parts List 2 200mm woofers (Altronics C-3060) 1 Motorola KSN1141A piezo ceramic speaker and horn 2 50mm adjustable speaker ports 2 200mm speaker grilles 1 strap handle 7 box corners 2 6.35mm jack sockets (Altronics P-0071) plus mounting cup (Jaycar HS-8025) or 2 x Neutrik locking chassis mount jack sock­ets (Jaycar PS-0196) or 2 x Neutrik “Speakon” sockets (Altronics P-0790, Jaycar PS-1094) 1 1.8mH air-cored inductor 1 0.33µF 200V polyester capacitor 1 5-way 30A mains terminal strip 1 piece of 1m x 1.8m x 3mm speaker carpet 1 1m x 500mm piece of speaker wadding 1 1200 x 450mm sheet of 18mm Medium Density Fibreboard (MDF) 1 600 x 900mm sheet of 18mm MDF 1 790mm length of 31 x 13mm dressed timber (pine or meranti) 1 250ml tin of contact adhesive 1 100ml container of PVA adhesive 1 2m length of speaker sealant or adhesive backed draught exclud­er 46 8g x 25mm countersunk bronzed wood screws (for corner protec­tors speakers and ports) 2 10g x 25mm cheese head bronzed wood screws (for handle) 50 8g x 30mm countersunk wood screws (for securing box panels) 3 6g x 20mm cheese head wood screws (for terminal strip and inductor) 4 6g x 20mm countersunk bronzed wood screws (for 6.35mm sockets) 1 1.5m length of red 15A hookup wire 1 1.5m length of black 15A hookup wire 78  Silicon Chip Fig.6: this wiring diagram shows how the 5-way terminal block is used to terminate the leads from the jack sockets, the woofers, the tweeter and the crossover components. outer diameter of the port mounting flange. This can be done by placing the port in position and then cutting the carpet to the flange diameter. Similarly, the tweeter horn can be fitted and the carpet cut around its perimeter. The second and third pieces of carpet required are for the sides and are initially cut to 335 x 408mm. Glue these to the sides and then trim the carpet along the sloping edge so that the carpet will fold over to meet the baffle. Cut holes in the carpet at each side for the 6.35mm jack socket or adaptor. Insert the socket (or adaptor if this is used) and trim the carpet around its perimeter. The final piece of carpet needs to be 550 wide by 1100mm and fits over the top, back, base and front of the box, starting at the top of the baffle and going around to the base of the baffle. Trim the carpet at one end so that it meets the carpet already applied to the baffle and sides. Apply contact adhesive along those edges where the carpet joins will be and on the carpet itself. Fit the carpet in place when the glue has dried. Now coat the bulk areas of the box on the top, back, base and front and the carpet and fit it in place. The corner protectors can now be fitted with countersunk screws. These mount on each right angle corner. We cut another corner protector to provide two separate flat pieces and these were fitted along the sloping edge on each side of the box to allow it to be stood on side for normal loudspeaker use or for stacking during storage. Attach the handle onto the front of the box close to the baffle and central to the width of the box. This should provide the best balance point for carrying the box. Wiring Use 15A hookup wire and solder a 300mm length of red wire to the positive terminal of each speaker and a 300mm length of black wire to the negative terminals. Then connect a 300mm length of red wire to each tip connection of the jack sockets and a black wire 300mm in length to the ring terminals. The crossover components are mounted using a 5-way terminal block Above: both the terminal block and the inductor are secured to the bottom of the box using wood screws. Note that the inside of the box is lined with wadding which is stapled in place but this must be clear of the port holes. which is secured to the inside of the box with two wood screws. Mount the inductor using a wood screw also. Wire up the components by passing the wires through the holes allocated for each component as shown in the diagram of Fig.6. Apply a layer of speaker wadding around the edges inside the box and secure it in place with staples. Make sure the wad­ding is clear of the port holes. Fit the jack sockets to the side of the case with screws and secure the speakers with sealant between the baffle and speaker mating surfaces. Cut the two 50mm ports to a length of 125mm and secure them in place. Fit the protective mesh grilles over the woofers using the supplied clamps and wood screws. Then vacuum the outside of the box to remove wood shavings and sawdust. Any contact adhesive on the outside of the carpet can be removed with mineral turps and an old toothbrush, before it sets hard. Then you are ready to have a SC listening test. Enjoy. A 2-metre length of speaker sealant (or adhesive-backed draught excluder) is used to seal the loudspeakers and the port tubes, to prevent air leaks. November 1999  79 A remote controlled throttle for model railways PART 2: By JOHN CLARKE & LEO SIMPSON BUILD THE RAILPOWER Last month, we presented the circuit details of our new Railpower model railway speed control. This month, we describe the circuit of the IR remote control and give the construction details. L AST MONTH, we completed the circuit description of the Railpower except for the infrared remote transmitter and this is shown in Fig.5. It comprises a single IC, two transistors, an infrared LED and a few passive components. The 80  Silicon Chip IC’s internal oscillator is set to 455kHz by ceramic resonator, X1, con­nected between pins 12 & 13. The 455kHz oscillator frequency is divided down by 12 to give a 37.9kHz carrier frequen­cy for the infrared LED (IRLED1). Current drive for the LED is provided by the Darlington-connected transistors, Q1 & Q2. When any pushbutton is pressed, it pulls the corresponding input of IC1 low and this causes the output at pin 15 to deliver a uniquely coded stream of pulses (at 37.9kHz). The pulse codes can be changed using different combinations of links LK1 and LK2 so that you have the option of using up to four separate Railpow­er controllers which operate independently on the same layout. This can be a boon to realistic operation on large layouts with cab (block) switching. Naturally the receiver coding on the main PC board must match the respective remote control transmit­ter in order to operate. However, if you only intend to use one Railpower controller on your layout, you can omit the two links on both the transmitter and the main circuit board. The transmitter circuit is powered by two AAA cells con­nected in series to provide a 3V supply. The IC draws only about 1µA on standby, when the switches are not pressed, so the batter­ies should last for virtually their shelf life. Construction The Railpower may have a relatively complicated circuit but it is very straightforward to build. All the circuitry in the case is installed on a PC board measuring 216 x 170mm and coded 09310991. Before you begin assembling components on to the PC board, check that it fits properly into the base of the instrument case. Enlarge the corner mounting holes in the PC board to 3mm or 1/8" if these have not been drilled to size. Check that the holes line up with the integral pillars in the case. Then check the PC board for shorts between tracks or for any breaks. Make any repairs now, if required. Check the holes for the 0.1Ω resistor and the power diodes (D15-D18) as these may need to be enlarged to accommodate their thicker pigtails. The component overlay for the PC board is shown in Fig.6. Start by installing all the links on the PC board using 0.8mm tinned copper wire. Most of the links are 12.5mm long so you could bend these over a suitable former about 12mm wide, to create a uniform appearance. There is a longer link near IC5 and more links adjacent to the power transistors Q16, Q17, Q20 & Q21. Install the PC stakes in position and then insert and solder the resistors. Use the accompanying colour code table as a guide to selecting the resistor values. Better still, use a digital multimeter to check each value before it is inserted. The 0.1Ω 5W wirewound resistor should be raised above the PC board by about 2mm before soldering the leads. Next, install the diodes. Several types are used although most are the small switching diodes (glass encapsulated). 1A types (black resin body with silver stripe) used for D7, D8 & D14 while the 1N5404 power diodes Fig.5: the transmitter encoder IC has an internal oscil­lator set to 455kHz by the ceramic resonator, X1, connected between pins 12 & 13. The 455kHz is divided down by a factor of 12 to give a 37.9kHz carrier frequency for the infrared LED (IRLED1). When any pushbutton is pressed, it pulls the corre­sponding input of IC1 low and this causes the output at pin 15 to deliver a uniquely coded stream of pulses at 37.9kHz. used for D15-D18 are larger again. Note that you only need to install diodes D15 & D16 if the trans­former is a centre-tapped 24V unit. The capacitors can be installed next, taking care to orient the electrolytic types with the polarity shown on Fig.6. Note that you should only Improving The Speed “Hold” Time Following last month’s article we have had a chance to do some serious testing of the new Railpower on a large HO layout and it came through with flying colours, except for one aspect: the speed “hold” time. When you set the speed with the remote control, you expect it to stay set indefinitely. In practice, that is not possible with the “hold” circuit involving IC4b but the circuit did need improving so that the speed setting did not drop noticeably after a few minutes. Therefore, we are recommending a change to the value of C1. Instead of using a 2.2µF tantalum or low leakage electrolytic capacitor, C1 should now be a 22µF tantalum type. At the same time, the 10MΩ resistor associated with IC5a should now be 1MΩ while the 4.7MΩ associated with Q3 should now be 470kΩ. In other words, C1 is now ten times larger and the associated charging discharging resistors are one-tenth of their original values. These changes have been incorporated into the component overlay diagram of Fig.6. With the new values, a given speed setting can be expected to drop by 36% after 15 minutes or there­abouts. This should be more than adequate, even for the lar­gest layouts where protracted running at a given speed is re­quired. November 1999  81 Fig.6: the component overlay for the PC board. Install C3 or C4 (not both) depending on whether you want the circuit to power up in the forward or reverse mode. For forward mode, install C3; for reverse mode, install C4. install C3 or C4, not both. Install C3 if you want the circuit to power up with the controller in the forward mode. 82  Silicon Chip To power up in reverse mode, omit C3 and install C4. Now insert and solder the ICs, tak- ing care to orient them correctly. Be sure that each is in its correct place before soldering the pins. Secure the mains wiring using cable ties and be sure to sleeve all exposed terminals with heatshrink tubing as detailed in the text. Install the transistors and regulators next, taking care to insert the correct one in each position. REG2 and transistors Q16, Q17, Q20 & Q21 are mounted at full height with about 1mm of their leads extending below the copper side of the PC board. Six trimpots need to be installed, Capacitor Codes  Value IEC Code EIA Code  0.1µF   100n   104  .01µF   10n  103  .001    1n (1000p or 102) Resistor Colour Codes  No.    1    1    1    1    1    8    1    1    3  35    1    2    3    7    6 Value 10MΩ 4.7MΩ 560kΩ 220kΩ 120kΩ 100kΩ 47kΩ 39kΩ 22kΩ 10kΩ 4.7kΩ 3.3kΩ 2.2kΩ 1.2kΩ 1kΩ 4-Band Code (1%,5%) brown black blue brown yellow violet green brown green blue yellow brown red red yellow brown brown red yellow brown brown black yellow brown yellow violet orange brown orange white orange brown red red orange brown brown black orange brown yellow violet red brown orange orange red brown red red red brown brown red red brown brown black red brown 5-Band Code (1%) brown black black green brown yellow violet black yellow brown green blue black orange brown red red black orange brown brown red black orange brown brown black black orange brown yellow violet black red brown orange white black red brown red red black red brown brown black black red brown yellow violet black brown brown orange orange black brown brown red red black brown brown brown red black brown brown brown black black brown brown November 1999  83 Fig.7: wiring details inside the case. The 240VAC mains wiring should be run in 250VAC-rated wire, with green/yellow striped wire for the Earth. The buzzer is fixed to the board with double-sided tape. and again, make sure that the correct value is installed in each position. Finally, the nine LEDs can be installed. These are all oriented with the Anode (longer lead) to the right and are mount­ed so that the cathode lead is about 1mm below the copper side of the PC board. This will allow the leads to be soldered and have the maximum height above the PC board. The LEDs will eventually need to be bent over at right angles so that they 84  Silicon Chip can be inserted into the front panel holes. Front and rear panels Mark out and drill the front panel to provide an access hole for the infrared detector, the LEDs, the power switch S1 and the analog meter. The meter is supplied with a cardboard template to assist in marking out its cutout hole. The front panel artwork should be used as a guide for the hole positions. The rear panel is aluminium and it needs to be drilled for the four transistors, REG2, the access hole for auxiliary cir­cuits and the IEC power socket. Before the rear panel can be drilled, you will need to mark the positions of the four power transistors and REG2. To do this, you need to sit the assembled PC board in posi­tion on the base of the case. You will need to shorten some of the integral pillars with a large drill bit so that the PC board can sit correctly on the corner mounting pillars. Now slide the aluminium The H-bridge transistors (Q16, Q17, Q20 & Q21) and the 3-terminal regulator are mounted on the rear panel, which provides the necessary heatsinking – see also Fig.8. rear panel into place and mark the mounting hole posi­tions for the power transistors and regulator. At the same time, mark out the hole positions for the terminals and the auxiliary output lead hole which will need to be fitted with a grommet. The IEC mains socket should be mounted as far to the side as possible to allow clearance for the leads around the trans­former. Mark out the holes required for this and also for the adjacent earth screw terminal. Place the transformer on the PC board and determine its optimum mounting position. The mounting hole positions for this should not encroach onto the track area on the board. Drill and file out the holes on the rear panel and do the same for the transformer mounting holes on the PC board. We mounted the transformer with rubber grommets inserted into the holes in its mounting feet. The holes may need to be reamed to a larger size for this. The transformer is secured to the PC board with a screw, a flat washer and a nut for each foot. The nuts are tightened down sufficiently to anchor the transformer but not so tight as to prevent free movement. Secure the PC board in the case with self-tapping screws. Attach the IEC socket to the rear panel using 3mm screws and nuts and Fig.8: mounting details for the power transistors and 3-terminal regulator. Use mica or silicone washers and insulating bushes to isolate the metal tabs from the metal panel. This close-up view shows how the PIC12043 IR receiver (IC1) is aligned with a hole in the front panel, so that it can pick up the IR pulses from the remote control unit. November 1999  85 The remote control transmitter board has just a handful of parts and should only take a few minutes to assemble. Note that this board is supplied by Oatley Electronics and is substituted for the existing board inside the transmitter case. fit the binding post terminals. You can also attach the meter and switch S1 to the front panel. Fig.7 shows all the wiring details inside the case. The 240VAC mains wiring should be run in 250VAC-rated wire, with green/yellow striped wire for the Earth. Use a plastic insulating boot for the IEC socket terminals and a length of insulating tubing for the terminals of switch S1. Some of the wires will need to be passed through this insulation boot before soldering them in place. The earth connection to the transformer is made by scraping away the coating on the transformer mounting foot and soldering the wire in place. Be sure it is soldered properly, with a hot iron, and check that the solder has flowed onto the exposed steel. The earth terminations to the rear metal panel are made using crimp lug eyelets or solder lugs, each secured in place with a 3mm screw, star washer and nut. Fig.8 shows the details of how the power transistors and regulator are mounted to the rear panel. Use mica or silicone washers and insulating bushes to isolate the metal tabs from the metal panel. Use heatsink compound between mating surfaces if mica washers are used but this is not necessary for the silicone types. Check that the transistor tabs and regulator are indeed isolated from the case by testing on the ohms range with your multimeter. The reading should show open circuit. The binding post terminals are wired with short lengths of heavy duty hookup wire as shown. Wire up the front panel meter with hookup 86  Silicon Chip wire and connect the buzzer wires in place. We se­cured the buzzer to the PC board with a piece of double-sided tape. Alternatively, it can be glued in place with silicone sealant, contact adhesive or similar. The transformer secondary is wired as shown in Fig.7, using the three connections on the PC board for the 24V centre-tapped transformer type or without the centre tap terminal on the PC board (CT) for the transformer with two 12V windings. Transmitter construction The transmitter is assembled into a small case which con­tains an existing PC board. This unit is supplied by Oatley Electronics. You first need to prise open the case and remove the board. You then have to remove the battery clips and 455kHz ceram­ ic resonator from this board and install them on a new PC board that comes with the transmitter. Fig.9 shows the component layout for this new board. Insert the supplied components as shown, taking care to install the correct transistor in each position. Do not forget the wire links and the insulated wire link from the positive supply up to the 4.7Ω resistor. Be sure to orient the IRLED correctly, with the longer lead being the Anode (A). It is placed so that its body sits comfortably over the integral moulding in the case. When completed, you can attach the front panel label and cut the holes out with a sharp utility knife. Insert the PC board into the case and clip the case together. You will need two AAA cells to power the transmitter. Testing Fig.9: this is the parts layout for the remote control transmitter PC board. Check all your work thoroughly before applying power. When power is applied, the neon in the power switch (S1) should light and some of the LEDs should light. In particular, the forward or reverse LED should be on as well as the Stop LED. Check the supply rails on the circuit. Connect the negative lead of your multimeter to the 0V binding post terminal and test the voltage on the positive terminal. It should be +12V. Now check at pin 14 of IC2 for a reading of +5V. If either voltage is low, you should suspect a short on the PC board somewhere. Find it and fix it before going any further. There should be 12V between pins Fig.10: use this actual size artwork to check for etching defects on your controller PC board. 4 & 11 of IC3 & IC8; bet­ween pins 4 & 8 of IC4; between pins 8 & 16 of IC5 and between pins 7 & 14 of IC6, IC7 & IC9. Now check the operation of the remote control. Pressing the control buttons should operate the front panel LEDs. The Aux1 & Aux2 LEDs should light when the relevant remote buttons are pressed, with Aux1 staying on or off, after each button press. The Inertia and Stop functions November 1999  87 RAILPOWER SPEED SETTING SC D OVERLO A FORWA RD E REVERS UT LOCKO STOP INERTIA OFF AUX2 AUX1 POWER Fig.11: this full-size front panel artwork can be used as a drilling template for the front panel of the main control unit. RAILPOWER TRACK SILICON CHIP 40 60 AUX1 INERTIA AUX2 STOP REV FOR 80 10 0 SPEED 88  Silicon Chip 20 0 SPEED Fig.12: here are the actual size artworks for the remote con­trol handpiece and the meter scale. also should go on and off with alternate pressings of the pushbuttons. The speed (+) and speed (-) buttons should alter the meter reading but with this yet to be calibrated you may not obtain good results. The lockout LED should also switch on as the speed setting is increased. The track LED should gradually light up as the speed is increased and show a different colour depending on whether the forward or reverse mode is selected. It may come up very slowly in brightness because of the inertia setting. You can switch the inertia out for a faster response to the Speed buttons. Note also that the forward or reverse mode can only be selected when the lockout LED is off. Now connect your multimeter to the track terminals. Adjust the speed up fully by holding the speed (+) button down for about 10 seconds, then adjust VR1, the maximum speed trimpot, for a reading of 12V. If you are running an N-scale layout in which the model locomotives normally run at a maximum of 9V, then use this as the maximum speed setting instead. Adjust trimpot VR2 fully anticlockwise. Use the speed (-) button to reduce the track voltage to its minimum setting and then connect the controller to the track on your layout. Place a locomotive on the track, select inertia out (LED lit) and rotate VR2 until the loco just begins to move. Then rotate VR2 slightly anticlockwise from this setting. Now use the speed buttons to start the loco and bring it to a halt again. If the loco still tends to creep at the minimum speed setting, adjust VR2 even further anticlockwise and then check the settings again. Next, adjust trimpot VR6 so that the meter shows full scale at the maximum speed setting. You may want to remove the loco when doing this. Trimpots VR4 and VR5 are adjusted for the required amount of inertia for starting and stopping. The Inertia function will need to be selected (Inertia LED out) to adjust VR4. Trimpot VR3 is adjusted so that the forward/reverse lockout LED lights at the train speed below which you consider it safe to suddenly reverse the track voltage. Next month, we will give the details of how to wire the Railpower without remote control. This will give you the ability to plug a handheld controller into any place on your SC layout to control trains. ORDER FORM BACK ISSUES MONTH YEAR MONTH YEAR PR ICE EACH (includes p&p) Australi a $A7.00; NZ $A8.00 (airmail ); Elsewhere $A10 (airmail ). Buy 10 or more and get a 10% discount. Note: Nov 87-Aug 88; Oct 88-Mar 89; June 89; Aug 89; Dec 89; May 90; Aug 91; Feb 92; July 92; Sept 92; NovDec 92; & March 98 are sol d out. All other issues are currently i n stock. TOTAL $A B INDERS Pl ease send me _______ SILICON CHIP bi nder(s) at $A12.95 + $5.00 p&p each (Australi a only). N ot avail abl e elsewhere. Buy five and get them postage free. e & Get Subscrib count is D A 10% on ther Silic e O ll A n O is d n a h rc Chip Me $A SUBSCRIPTIONS  New subscription – month to start­­____________________________  Renewal – Sub. No.________________    Gift subscription  RATES (please tick one) 2 years (24 issues) 1 year (12 issues) Australia (incl. GST)  $A135  $A69.50 Australia with binder(s) (incl. GST)**  $A159  $A83 New Zealand (airmail)  $A145  $A77 Overseas surface mail  $A160  $A85  $A250 Overseas airmail  $A125 **1 binder with 1-year subscription; 2 binders with 2-year subscription YOUR DETAILS Your Name_________________________________________________ GIFT SUBSCRIPTION DETAILS Month to start__________________ Message_____________________ _____________________________ _____________________________ Gift for: Name_________________________ (PLEASE PRINT) Address______________________ _____________________________ (PLEASE PRINT) Address___________________________________________________ State__________Postcode_______ ______________________________________Postcode_____________ Daytime Phone No.____________________Total Price $A __________ Signature  Cheque/Money Order  Bankcard  Visa Card  Master Card ______________________________ Card No. Card expiry date________/________ Phone (02) 9979 5644 9am-5pm Mon-Fri. Please have your credit card details ready OR Fax (02) 9979 6503 Fax the coupon with your credit card details 24 hours 7 days a week Mail order form to: OR Reply Paid 25 Silicon Chip Publications PO Box 139, Collaroy 2097 No postage stamp required in Australia November 1999  89 ASK SILICON CHIP Got a technical problem? Can’t understand a piece of jargon or some technical principle? Drop us a line and we’ll answer your question. Write to: Ask Silicon Chip, PO Box 139, Collaroy Beach, NSW 2097. Loudspeaker design software I am looking for articles on subwoofer and speaker en­closures. There was an article in the June 1996 issue on BassBox loudspeaker software. I would like to know if you subsequently published a design using the compound speaker concept featured in this article, as I would like to build a subwoofer enclosure using two 10inch subwoofers for my car. If not, are there any plans to feature something along those lines in the future? (C. E., via email). • We have featured only two sub­ woofers designed using the BassBox software. They were the Bass Barrel in the August 1997 issue and the Bass Cube in the April 1999 issue. Flickering flame doesn’t flicker I recently built the Flickering Flame as described in the October 1997 issue of SILICON CHIP and have a problem trying to run it off mains power using a 240VAC AC transformer rectified to Setting up the capacitance meter The Digital Capacitance Meter described in the February 1999 issue looked very useful to me, especially with its pF range. Unfortunately, having built the unit, I cannot null the output of IC2a and this ruins the low range readings. For testing, I have numerous 2% capacitors ranging from 47pF to 0.9µF, a borrowed capacitance meter, a DVM, oscilloscope and frequency counter. I have noted the Errata in May and June 1999 issues and have checked the unit carefully. The voltages seem correct as do the scope waveforms. It can be made to work tolerably well on the nF and µF ranges but 90  Silicon Chip 12V DC. The problem is the light does not flicker; it just stays on. But when I connect it to a car battery it works cor­rectly. Could you please help me with this problem? (G. W., via email). • Our guess is that you have insufficient filtering on your rectified DC supply. You would need a capacitor of at least 4700µF at 25VW. Temperature controller can’t heat & cool I gather from the article on the Switching Temperature Controller in the August 1999 issue that can only run one Peltier Effect device at once. Is this so? So in order to keep the inside of an Esky used as a brooder for raising baby birds, two circuits would be needed – one for cooling and one for heating? I assume that a Peltier device mounted on a CPU cooling fan heat­ sink would blow cool air? (T. A., via email). • You are right in that the circuit can only be used for heating or cooling. If you try to do both, both the heating is not acceptable on the pF range. The voltage at pin 11 of IC2a does not dip (to 5-10mV) and rise again but ranges from 66-89mV when varying trimpot VR1. Changing IC1 alters this slightly but does not fix the problem. The frequency at pin 8 of IC1 with the first 74HC132 measures 12,124Hz and 12,400Hz and with the replacement 11,109Hz and 10,900Hz, approximately. The text refers to 16,160Hz and 16,000Hz. Is this significant? (R. B., Flaxton, Qld). • We will answer to your queries in order: (1) The frequency is not critical but if it is too low then the readings on the µF range may be erratic. Once you get the unit working, reduce the value of the .01µF ca- and cooling sides of the circuit will be powered all the time which is highly inefficient. If you wanted to do heating and cooling, to maintain a constant temperature for a wide range in ambient, the circuit would have to be extended with extra comparators to switch in the heating or cooling function. A Peltier device mounted on a heatsink could be used to blow cool air but it would only be slightly cooler than ambient. Problem with remote central locking In order to add remote central locking to my car, I purchased the UHF remote control kit, as published in the January 1993 issue of SILICON CHIP. I also purchased a Jaycar LR-8835 Power Lock Relay and an LR-8833 Slave Door actuator. The problem I have is that when the transmitter is 30cm or more away from the receiver, the channel that is being operated starts bouncing on and off erratically. This in turn, via the Power Lock Relay, pulses the actuator pacitor at pin 2 of IC1a until the frequency is closer to the nominal value of 16kHz. (2) The varying stray capacitance of different PC boards, especially if they have a screened overlays and solder resist can mean that VR1 may not have sufficient range. The cure is to increase the 12pF capacitor on pin 13 of IC1d to 22pF or 33pF until VR1 can be set close to centre. (3) As you have an oscilloscope, simply clip the probe on pin 11 of IC2a and carry out the set zero adjustment. The pulse width at pin 11 will get narrower and narrower until it disappears, then if you keep adjusting VR1 it will reappear and get wider. The correct adjustment is when it just disappears. in the same direction. With the trans­ mitter within 30cm of the receiver, the actuator operates normally. When the actuator is removed from the circuit, the receiver and Power Lock Relay operate fine, even when the transmitter is a good distance away. This happens on the test bench and when fitted to the car. I have one channel doing unlock and the other doing lock. The receiver is set up for momentary operation. I have earth coming from the two NO contacts in the receiver and going to the Power Lock Relay brown and white inputs. Some things I’ve already tried are: filtering the receiver’s 12V input, supplying power for the receiver from a different source, a new transmitter battery and even wrapping the actuator in aluminium tape. The car already has central locking but the driver’s door only has a microswitch that operates the other doors, no actua­tor. The new actuator operates the lock mechanism which in turn operates the microswitch. I have actually gotten the system to work using the NC receiver outputs. Now the Power Lock Relay inputs are normally at earth but when a button is pushed, the Rx operates and removes the appropriate earth from the Power Lock Relay. When the button is released and the receiver switches off, the earth is then reap­plied to the Power Lock Relay which drives the actuator. But I want to get it to work the way it’s supposed to so I hope that you will be able to help me. (S. B., Amberley, Qld). • Your problem appears to be with the remote control receiver resetting as the heavy current drawn by the door actuators reduc­ es the supply voltage. This can be cured by several means. First­, you need to use a separate supply lead for the actuator positive supply. The receiver will then have minimum voltage drop when the actuators are powered. Second, the receiver circuit can be improved to reduce the incidence of resetting. This involves reducing the 6.8V zener diode dropping resistor from 1.8kΩ (R4) to 390Ω. This will maintain the zener voltage even if the 12V supply drops to 9V. You can also replace the 10µF capacitor C4 with a 100µF 16VW electrolytic to improve supply rejection. You may also need to connect a large value capacitor across the door Converting a standard PC power supply There has been some discussion this year on using old computer power supplies as general purpose bench units but there are many variables; one that I have hit is that one power supply would not start without a decent (15-20W) load on the 5V line. One comment that I saw was that one reader rebuilt the 5V section using the 12V components to get a 220W power supply at 13.8V (after adjustment of the output voltage). That is one hell of a good bench supply for anyone involved in car radio, amateur radio, CB or similar activities. It would be a good supply for 12V garden lighting without the risk of the voltage going up as lights blow. I have a 100W transformer driving a string of 8W lights and if more than actuator solenoid to damp down transients. Also a re­verse connected diode (1N5404) across the actuator coil may help. Loud direction indicators wanted I need a simple circuit for people who are hearing im­paired, to amplify a vehicle’s blinker noise, as the “click click click “ of a normal can under the dash is too soft to be heard. (J. C., via email). • Why not just wire a buzzer so that it is powered from both sides of the traffic indicator switch can via a pair of diodes? That should do the trick. DC-DC converter for military gear Have you ever designed a 12V to 24V converter? I wish to use some military equipment in a “civilian” car for demonstration purposes. Trouble is, all our communications gear is designed to run off 24V DC nominal (28V with the vehicle running). Any ideas? (I. B., via email). • One possibility may be to modify the 2A SLA battery charger (published in the July 1996 issue) so that it delivers 24V in­stead of 13.8V. You three blow then the rest will blow within 15 minutes! PC power supplies are available here in Canberra for $5 each with one proviso; you have to take the whole computer! Can you devise a circuit board and parts list that a mid-range experimenter could work from to build such a beast. Tracing out a switchmode power supply is beyond me; I know because I have tried. (B. W., via email). • We have had a look at the idea of a circuit to convert a standard PC supply to something more useful but the problem is that all these supplies are different inside, even though they all tend to use one or another standard switch­mode IC. The only way we could do it is to specify a particular switchmode PC supply and then go from there although that tends to defeat the purpose of letting people use any surplus supply. would do this by changing the 22kΩ and 2.2kΩ feedback resistors to 39kΩ and 1.8kΩ, respectively. You would also need to add a 4700µF 35VW capacitor to the output. We es­timate that it should be good for about 1A at 24V. (Note: we have not tried this). Programmable ignition for a Landrover I was reading the June 1999 edition and I came across the programmable ignition article. The Lumenition control module on my Landrover has just failed and I wondered if it would be possi­ble to replace it with a PIT module and adapt the optical pickup that is already fitted to the vehicle. (N. W., via email). • Yes, it is possible. We published a brief note about using the Lumenition module in the May 1994 issue. DC-DC inverter for car amplifier I wish to build an amplifier to put in my car. It needs supply rails of ±37.5V or ±40V. Do you have a circuit design for a DC-to-DC converter to give the required supply rail from a 12V car battery? (M. N., via email). • We published a 100W DC-DC ConNovember 1999  91 Display problems on signal generator I built the Low Distortion Audio Signal Generator described in the February & March 1999 issues of SILICON CHIP but it has a problem I can’t explain. When I went to set up the generator, the oscillator worked well but the display didn’t. I eventually found that the 0.18µF capacitor associated with S2 and IC1b was earthed and suspected that it wasn’t meant to be, according to the circuit diagram. I cut the track and then the display worked well, except on the 10Hz to 100Hz range where it dropped to 0000 at about 3/4 setting. This range could be made to work by adjusting VR3 fully counter-clockwise. But in doing this, the other ranges had a point midway in their range that dropped back to 0000. I did notice that if I touched verter for cars in the December 1990 issue and a 600W design in the October & November 1996 issues. Command control decoder I am building the February 1999 version of the Command Control Decoder For Model Railways but I am having difficulty finding the required BD433 and BD434 transistors. Any suggestions? (K. R., via email). • You can substitute transistors such as BD675/677/679/681 or BD263 for the BD433 and BD678/680/682 or BD262 for the BD434. These are readily available from kitset suppliers. Waiting for turbo timer defeats purpose I have just built the turbo timer as published in the November 1998 issue and tested it on the bench. It works fine. I then did a temporary hook-up to my car. Guess what? You cannot take the key out of the ignition while the engine is running, so one has to sit while the timer runs out. I am certainly not going to leave the key in the car while I do my shopping. Anyone else having the same problem? Oh, the car is a 1992 92  Silicon Chip the back of 0.18µF capacitor associated with IC1a, the display worked perfectly on all ranges. I have since put a 10MΩ resistor across this capacitor and it works fine. I am very happy with the end result but cannot for the life of me work out what I have done to cause the problems mentioned. I checked all the components at least twice with the multimeter. Maybe you can shed some light on this problem? (A. B., via email). • With regard to the 0.18µF capacitor being earthed on your PC board, apparently one of the PC board manufacturers decided that a shield track (earthed) should have been connected to an adjacent pad on the PC board and so they changed our original pattern. This meant that some kits did have a faulty front panel board. This was corrected on kits as soon as the error was discovered. Nissan Bluebird AWD Turbo. (G. W., via email). • If your turbo timer needs the ignition switch on to run, you have taken the +12V from a point switched by the ignition. You need to pick up +12V from a fuse not switched by the ignition key. Jumbo LED clock problems I am having many problems with the Jumbo LED Clock described in the March 1997 issue. I obtained the displays from Jaycar and I had to drill new holes in the board as they did not line up correctly. The project did not work and I eventually found out that the displays had different connections. I rewired the displays using separate wires between the two boards. I could not find a supplier for the watch crystal, so I obtained six from old watches that I had to hand. Checking with my DFM, one crystal oscillated at 16kHz, one didn’t work and the other four worked at 31.25kHz. The clock worked but it ran slow, losing three seconds each minute. With the oscillator running too low, I would have thought it would have run too fast. Unfortunately though, the published pattern for the front panel (page 64, March 1999 issue) does show a thin line, making the connection from the ground track to the 0.18µF capacitor pad. This is due to a glitch which sometimes occurred when the Protel file for the artwork was imported into our drafting package, Generic CAD. The component overlay diagram on the same page does not exhibit the glitch. The original artwork sent to board manufacturers also did not have the shielded track connected. The lack of output at the low frequency end would mean that the LDR was not receiving sufficient light to maintain control. This was probably due to the IR LEDs not providing full light coverage over the LDR surface. If you are happy with the result with the 10MΩ resistor across the 0.18µF capacitor, then leave the oscillator as is. Pressing the hour button clocked up the hours OK and the AM-PM indicator worked. Also the 1-second decimal points worked. But on first switch on, it indicated -004, the righthand number being any value on switch on. When pressing the minute button, the number clocked up while the button was pressed but on release the numbers all changed to an arbitrary figure. Adjusting the oscillator trimmer from minimum to maximum only varied the frequency from 31.35kHz to 31.25kHz and the oscillator stopped with the trimmer set at exact minimum. I used machined IC sockets for all the ICs so I can change them easily. I have an oscilloscope and a DFM. I used an earth on the PC and a wrist strap as I thought the ICs might be static sensitive. (S. F., Yangebup, WA). • We suspect that the 4526 prescaler ICs, IC2 & IC3, may be causing the problems of strange results from the minute pushbutton and incorrect timekeeping. Check that the DP1-DP4 inputs for IC2 and IC3 are correctly tied to the ground or positive supply. Erratic behaviour may occur if they are floating. Also check for shorts between tracks around these two ICs. You should compare your PC board pattern with the published pattern (page 49, March 1997) to check that you have not soldered two pads together. We are not aware that the pinouts for the large 7-segment displays are any different to the ones we used in the prototype. These pinouts follow an industry standard. The exact frequency of operation for a crystal oscillator cannot reliably be tested with a probe and frequency meter. This is because the probe’s capacitance can load the crystal and slow down its frequency. Normally, if you want to reliably measure the frequency of a crystal oscillator, you must do it via a buffer stage. A suitable crystal can be obtained from Farnell Electronics, phone 1300 361 005. Their catalog number for the crystal is 569-914. Sync output for TV pattern generator I would like to know if the Colour TV Pattern Generator described in June & July 1997 has or can have a separate sync output as well as the composite video. I have an application that requires this. (D. P., -via email). • The composite sync signal is available at pin 16 of IC10. Since its amplitude is 5V peak-to-peak, you may need to attenuate it to the required level for your application. Notes & Errata Daytime Running Lights for Cars, August 1999: a modification to allow the circuit to be used with cars having headlight switching in the negative line is published in Circuit Notebook this month. PC Monitor Checker, August 1999: circuit modifications to give more ideal scan frequencies are published in Circuit Notebook this month. Mailbag: continued from page 44 charged at about 5A by two 42W Solarex panels. These are now nearing 30 years old but the batteries have been replaced. All lights in the home are fluoresents modified to run from 12VDC. My amateur station (VK4KAL) is operated from batteries, direct where possible or by DC/DC converters. A 1500W inverter is on standby, although rarely used. The 486 DX2/66 computer I am using is also running from 12V DC. We are now connected to the grid but only for the deep freezer and washing machine (not an automatic – these are water wasters in our dry area). The welder has to be AC-operated although the frame of our home was welded using battery power – 36VDC (we live in white ant country!). We use gas for cooking and our hot water is solar-heated. Where possible, every gadget we use is 12VDC operated. We don’t have blackouts, although recently the system in our area was out for three days due to 3km of mains being blown down in a freak storm; but our lights were on. For what it is worth, invest in a smallish Solarex panel and a 12V deep cycle battery. I run as a “emergency” an 18W panel charging a deep cycle marine battery to “fire up” an FT 747GX which draws 20A on transmit. This also lights my “shack” if ever needs be. A couple of amateurs living in Brisbane have similar small setups running 12V fluoros because of blackouts. Don’t use inverters unless absolutely necessary. They waste power while idling. I throw DC plugpacks in the rubbish bin. Why shouldn’t I, with 3kW of DC power at my disposal? We have the best of both worlds. A. Loveday, Rubyvale, Qld. DC concept is worthwhile I have just read your “Publisher’s Letter” for October 1999. I must say that you have touched on a point that I have been trying to make for quite some time, much to the amusement of my friends and colleagues. I believe that there is little to be gained by having domestic dwellings on the 240VAC mains. Apart from heating, all appliances could be powered by low voltage DC, be it 12, 24 or 32 volts. Heating, and by this I mean cooking, space heating and hot water systems, can be more efficiently be handled by gas. There are quite a few domestic appliances (eg, TVs, videos, etc) that are dual voltage and cooling is not a problem with 12VDC fans, made for large trucks, on the market. Refrigeration could also be run from 12VDC because of the relatively new 12V compressors available and of course, the large 12V absorption refrigerators have been around for a while. Air conditioning can again be via 12V compressor or evaporative, as can all manner of power tools. Even audio amplifiers can be low voltage powered; those massive “thumper” systems in cars are testimony to that. You did mention some of these in your article and my knee-jerk reaction was “now someone will listen to me!” I could go on and on about this subject but I had better stop now and say thank you for a timely article. The amusement mentioned above? I am a qualified electrician! Maybe a traitor to my trade? J. Smith, Middleton, SA. WARNING! SILICON CHIP magazine regularly describes projects which employ a mains power supply or produce high voltage. All such projects should be considered dangerous or even lethal if not used safely. Readers are warned that high voltage wiring should be carried out according to the instructions in the articles. When working on these projects use extreme care to ensure that you do not accidentally come into contact with mains AC voltages or high voltage DC. If you are not confident about working with projects employing mains voltages or other high voltages, you are advised not to attempt work on them. Silicon Chip Publications Pty Ltd disclaims any liability for damages should anyone be killed or injured while working on a project or circuit described in any issue of SILICON CHIP magazine. Devices or circuits described in SILICON CHIP may be covered by patents. SILICON CHIP disclaims any liability for the infringement of such patents by the manufacturing or selling of any such equipment. SILICON CHIP also disclaims any liability for projects which are used in such a way as to infringe relevant government regulations and by-laws. Advertisers are warned that they are responsible for the content of all advertisements and that they must conform to the Trade Practices Act 1974 or as subsequently amended and to any governmental regulations which are applicable. November 1999  93 MARKET CENTRE Cash in your surplus gear. Advertise it here in Silicon Chip. FRWEEBE YES! Place your classified advertisement in SILICON CHIP Market Centre and your advert will also appear FREE in the Classifieds-on-the-Web page of the SILICON CHIP website, www.siliconchip.com.au And if you include an email address or your website URL in you classified advert, the links will be LIVE in your classified-on-the-web! S! D E I F I S C LAS EXCLUSIVE TO SILICON CHIP! CLASSIFIED ADVERTISING RATES Advertising rates for this page: Classified ads: $11.00 (incl. GST) for up to 12 words plus 55 cents for each additional word. Display ads: $27.50 (incl. GST) per column centimetre (max. 10cm). Closing date: five weeks prior to month of sale. To run your classified ad, print it clearly in the space below or on a separate sheet of paper, fill out the form & send it with your cheque or credit card details to: Silicon Chip Classifieds, PO Box 139, Collaroy, NSW 2097. Or fax the details to (02) 9979 6503. Taxation Invoice ABN 49 003 205 490 _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ Enclosed is my cheque/money order for $­__________ or please debit my ❏ Bankcard   ❏ Visa Card   ❏ Master Card Card No. Signature­­­­­­­­­­­­ ________________________ Card expiry date______/______ Name _____________________________________________________ Street _____________________________________________________ Suburb/town _________________________ Postcode______________ 94  Silicon Chip FOR SALE TELEPHONE EXCHANGE SIMULATOR, SC Feb. 1998. Test equipment without the cost of telephone lines. $190. MAGNETIC CARD READER, SC Jan. 1996. Holds up to 8 cards. Use as a door lock. $65. Melbourne 9806 0110. ELECTRONIC/MECHANICAL DESIGN AND CONSTRUCTION: we offer a complete design service for electronic and mechanical devices. Most work is done in house and you deal directly with the designers. No job is too small and can be to prototype or “turn key” stage, in one offs or for future production. Simply send us an email at vladimir<at>u030.aone.net.au with your questions or requirements and we will get back to you. VIDEO STEREO TRANSMITTERS & RECEIVERS $149 Multi Channel up to 500 metres. www.allthings.com.au WEATHER STATIONS: Windspeed & direction, inside temperature, outside temperature & windchill. Records highs & lows with time and date as they occur. $420.00 complete plus sales tax if appli­cable. Optional rainfall and PC interface. Used by Government Departments, farmers, pilots, and weather enthusiasts. Other models with barometric pressure, humidity, dew point, solar radiation, UV, leaf wetness, etc. Just phone, fax or write for our FREE catalogue and price list. Solar Flair/ Ecowatch ph: (03) 5968 4863 fax: (03) 5968 5810, PO Box 18, Emerald, Vic., 3782. ACN 006 399 480. RAIN BRAIN AND DIGI-TEMP KITS: 8 station sprinkler controllers, 60 channel temp monitor uses DS1820s over 500 metres. Has PC Data logging. Mantis Micro Products, http://www.home.aone.net.au/mantismp PRINTED CIRCUIT BOARDS for all magazine projects, then go to http:// www.cia.com.au/rcsradio RCS Radio – Bexley (+61 2) 9587 3491. PC-CONTROLS: Receiver 144148MHz (PLL), DS2401 ID-Reader, Temperature Recorder (DS1615), AF Generators, Temperature measurement, I/O cards, Data Logging, ActiveX. Ph/Fax (02) 9482 1565. http://www. ar.com.au/~softmark 1/3 PRICE HEATSHRINK TUBING from 20 cents 2 : 1 Shrink CSA UL 125C GP 600 Volt. www.allthings.com.au C COMPILERS: everything you need to develop C and ASM software for 68HC08, 6809, 68HC11, 68HC12, 68HC16, 8051/52, 8080/85, 8086, 8096 or AVR: $155.00 each. Macro Cross Assemblers and Disassemblers for above CPUs + 6800/01/03/05, 6502 and 68HC12 for $78. Debug monitors: Need prototype PC boards? Positions At Jaycar We have the solutions – we print electronics! Four-day turnaround, less if urgent; Artwork from your own positive or file; Through hole plating; Prompt postal service; 29 years technical experience; Inexpensive; Superb quality. We are often looking for enthusiastic staff for positions in our retail stores and head office at Rhodes in Sydney. A genuine interest in electronics is a necessity. Phone 02 9743 5222 for current vacancies. Printed Electronics, 12A Aristoc Rd, Glen Waverley, Vic 3150. Phone: (03) 9545 3722; Fax: (03) 9545 3561 Call Mike Lynch and check us out! We are the best for low cost, small runs. Satellite TV Reception International satellite TV reception in your home is now affordable. Send for your free info pack containing equipment catalog, satellite lists, etc or call for appointment to view. We can display all satellites from 76.5° to 180°. KITS-R-US PO Box 314 Blackwood S.A. Ph/fax 08 8270 3175 FMTX2A Universal Stereo Coder $49 FMTX2B 30mW Xtal Locked 100MHz Transmitter $49 FMTX1 1-3 Watt Free Running Transmitter $49 FMX1 200mW Full Broadcast Transmitter, built & tested $499 FM220 10-18 Watt FM BGY133 Philips Linear $499 FM1525 25 Watt Discrete Linear FM Band $499 FM2100 110 Watt Discrete Linear FM Band $699 FM3000 300 Watt Discrete Linear FM Band $1499 Philips 828E/A VHF Receiver Boards (6 metres) $9 AWA 721 VHF Receiver Boards (2 metres) $9 AWA 721 VHF transmitter boards 1 watt (2 metres) $19 Philips 323 UHF transmitter boards 500mW (70cm) $19 AEM 35 Watt Little Brick Audio Power Amp $15 Digi-125 200W RMS Audio Power Amp $39 CA Clipper Compiler, new in box $49 6dBd Gain Colinear FM Band Antenna $999 Roll Smart-1 FM Station Audio Processor $999 Free catalog on disk of discounted surplus components Same day shipping, credit cards OK, circuits supplied. SPECIAL STEAM BOAT KITS $14 FREE PC DIGITAL VIDEO Recorder Software with 4 CH Capture Card * VIDEO Transmitters from $119 * 90 mW IR LEDS * DIGITAL PC VIDEO RECORDER - TIME LAPSE - MOTION DETECTION from $113 * PIR DETECTOR inbuilt concealed PINHOLE Mono or DSP COLOUR Camera, Microphone & Timer/Controller for VCR - Lights - etc from $139 * BULLET CAMERAS 480 Line 0.05 lux SONY CCD or DSP COLOUR from $132 * QUADS 4 Pix 1 screen from $256 HI-RES better than SUPER-VHS Quality * Modules 32 x 32 from $76 COLOUR DSP Pinhole from $155 with Tiny Hi-Sens On-Board MICROPHONE * MINI CAMERAS 36 x 36 from $85 - SONY CCD $102 - COLOUR DSP $162 * DOME from $88 - SONY CCD $105 - COLOUR DSP $164 * Video BALUNS from $7 * DIY PAKS: 4 Cameras, Switcher & Supply from $499 - with 14" Monitor from $601 with MUX for FULL SCREEN / RESOLUTION RECORDING from $1209 * 4 COLOUR CAMERAS, SWITCHER & POWER SUPPLY from $807 - with COLOUR QUAD 4 Pix 1 Screen from $1211 * With MUX $2033 * COLOUR QUADS from $512 * COLOUR DUPLEX MUX from $1329 * 14" MONITORS from $218 - with Inbuilt 4 Ch SWITCHER from $256 * SEE-in-the-DARK CAMERA & INFRARED ILLUMINATOR Kits from $19 * ANCILLARY EQUIPMENT * DISCOUNTS * Ask for our Catalogue & New Enquiry Offer * www. allthings.com.au * T 08 9349 9413 AV-COMM P/L, 198 Condamine St, Balgowlah, NSW 2093. Tel: 02 9949 7417 or 9948 2667. Fax: 9949 7095; www.avcomm.com.au Silvertone’s RC Receiver Still the best little performer available! $78 for 6 CPUs. All compilers, XASMs and monitors: $480. 8051/52 Simulator (fast, now incl. 80C320): $78. Try the C-FLEA Virtual Machine for small CPUs, build a “C-Stamp”. Demo desk: FREE. All prices + $5 p&p. Atmel Flash CPU Programmer: Handles the 89Cx051, the 89C5x and 89Sxx series, and the new AVRs in both DIP and PLCC44. Also does most 8-pin EEPROMs. Includes socket for serial ISP cable. $199, $37 tax, $10 p&p. SOIC adaptors: 20-pin $90, 14-pin $85, 8-pin $80. Credit cards accepted. GRAN­TRONICS PTY LTD, PO Box 275, Wentworthville 2145. Ph (02) 9896 7150; Fax (02) 9631 1236; or Internet: http://www.grantronics.com.au SOLAR PANELS: 120 watt $995.00, 80 watt $650.00, 60 watt $510.00, 40 watt $395.00 (all with 25 year guarantee). UNBREAKABLE PANELS: 64 watt $550.00, 42 watt $420.00, 32 watt $340.00, 11 watt $190.00, 5 watt $120.00, 1.25 watt $80.00. WIND Still only $129.50 AM or $149.50 FM. May be used with most ppm transmitters. This and many other radio control products available from: Silvertone Electronics, PO Box 580, Riverwood 2210. Phone/Fax (02) 9533 3517. www.silvertone.com.au GENERATORS: 400 watt $950.00. INVERTERS: sinewave inverters, inverter/chargers, mod. Sinewave inverters, call with requirements. AUST­RALIA WIDE DELIVERY (Free on orders over $500.00). TASMAN ENERGY: (03) 6362 3050 Fax (03) 6362 3054. SATELLITE TV DIGITAL RX NTSC to PAL MPEG-2 FTA EPG Encryption CAM from $399. www.allthings.com.au November 1999  95 Silicon Chip Binders Keep your copies safe, secure and always available with SILICON CHIP binders: they’re cheap insurance! Altronics................................. 36-38 REAL VALUE AT Av-Comm Pty Ltd.........................95 PLUS P &P Coffs Harbour Electronics............55 $12.95  Heavy board covers with 2-tone green vinyl covering Advertising Index Clarke & Severne........................55 Computronics Corporation..........55  Each binder holds up to 14 issues so that you can include catalogs Dick Smith Electronics........... 12-15  SILICON CHIP logo printed in gold-coloured lettering on spine & cover Harbuch Electronics....................54 EMC Technologies.......................55 Instant PCBs................................95 Janteknology Distribution..........IFC Price: $12.95 plus $5 p&p each (available Aust. only) Jaycar .............................. 45-52,95 Order by phoning (02) 9979 5644 & quoting your credit card number; or fax the details to (02) 9979 6503; or mail your order with cheque or credit card details to Silicon Chip Publications, PO Box 139, Collaroy, NSW 2097. Kits-R-Us.....................................95 Microgram Computers..............3,55 MicroZed Computers...................55 Premier Batteries.........................11 Printed Electronics...................... 95 HAMMOND ORGAN NOT WORKING: contact Eric Warren 02 4787 6836 email elwarren<at>ozemail.com.au to inspect and make offer. toria, for sale $70,000 WIWO. Suits technician. Good return, low rent. Call BH (03) 9743 7505 PH/FAX. 0411 149897 MOBILE. Questronix...................................55 PCBS MADE, ONE OR MANY. Low prices, hobbyists welcome. Sesame Electronics (02) 9554 9760 sesame<at>internetezy.com.au http:// members.tripod.com/~sesame_elec WANTED RobotOz......................................95 KIT ASSEMBLY ANY KITS assembled/repaired: professional, speedy service. Phone Nev­ille Walker (07) 3857 2752. BUSINESS FOR SALE ELECTRONICS RETAIL/REPAIR BUSINESS in western suburbs, Vic- VINTAGE AND CLASSIC hifi/audio, Leak, Quad, Beam Echo, Rogers, Marantz, Fisher, Dynaco, Heathkit, McIntosh, Radford, Pye, Lowther, Altec, Klipsch, SME, Garrard, Loftin-White amps, etc in any condition. Radio and audio valves. (02) 6255 2333. WE PAY UP TO $60 for contributions to Circuit Notebook. Send your circuit with a brief description to Silicon Chip Publications, PO Box 139, Collaroy, NSW 2097. HELP SAVE THE NIGHT SKY! We are losing our heritage of starry night skies. Poor, inefficient outdoor lighting is causing glare and “light pollution”. This wastes energy and increases greenhouse gas emissions. You can help by joining SYDNEY OUTDOOR LIGHTING IMPROVEMENT SOCIETY (SOLIS). SOLIS aims to educate and inform about quality outdoor lighting and its benefits. We also lobby councils, government and other bodies to promote good lighting practice. SOLIS meetings are held third Monday night of each month at Sydney Observatory. Individual membership is $20 pa. Donations are also welcome. Cheques payable to “SOLIS c/- NSAS”, PO Box 214, West Ryde 2114. Email: tpeters<at>pip.elm.mq.edu.au 96  Silicon Chip Resurrection Radio......................61 Robotic Education Products........55 R.T.N............................................55 SC Binders..................................11 SC Computer Omnibus...........OBC SC EFI Tech Special..................IBC Silicon Chip Bookshop........... 24-25 Silicon Chip Subscriptions...........89 Silvertone Electronics..................95 Smart Fastchargers.....................59 Solar Flair/Ecowatch....................94 Telelink Communications.............55 Truscott’s Electronic World...........59 Vision Beyond 2020.......................9 Zoom EFI Special........................11 _____________________________ PC Boards Printed circuit boards for SILICON CHIP projects are made by: • RCS Radio Pty Ltd, 651 Forest Rd, Bexley, NSW 2207. Phone (02) 9587 3491. • Marday Services, PO Box 19-189, Avondale, Auckland, NZ. Phone (09) 828 5730.   Own an EFI car? Want to get the best from it? You’ll find all you need to know in this publication                                          ­      € ‚  ƒ   „ †       €   ‡   ƒˆ ƒ   „   ‰               November 1999  97