Fullscreen HTML5Med Res (??MB)HTML4

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. Send your email to silicon<at>siliconchip.com.au Epoxy for High Power Ultrasonic Cleaner I want to build your High Power Ultrasonic Cleaner (September & October 2020; siliconchip.com.au/Series/350), but I have found multiple J-B Weld epoxy products and am not sure which one to use. (M. T., Auckland, NZ) • We recommend using the original J-B Weld epoxy. It is sold by Jaycar (Cat NA1518). Charging USB host and OTG devices Is it possible to power a host USB device and USB accessory at the same time? I have found a lot of confusing and contradictory information concerning this question. There seems to be a standard, but I am unsure how the system is actually implemented. (T. F., via email) • Perhaps not all devices implement the standard correctly, but it does support charging both the host and accessory using a “USB Accessory Charger Adapter”. There is no technical reason why it should not work. See Wikipedia for more details: https://w.wiki/t76 https://w.wiki/t77 Updating Colour Maximite 2 firmware I have been enjoying MMBasic and the Colour Maximite 2 (July-August 2020; siliconchip.com.au/Series/348) for some time now. We are up to CMM2 V5.06.00, and things are looking good. The problem I have is with loading new firmware into the Colour Maximite 2 using the USB-A to USB-A cable in the keyboard port. I used this method twice successfully with the STM32Cube Programmer software in Windows 7, but now it will not detect the Maximite any more. I had to take the case apart and slide the BOOT switch to SYSTEM and try again, and now everything is back to normal – the software detected the siliconchip.com.au Maximite, and I could upload the latest firmware. Is there some way to fix this? I prefer to avoid pulling the case apart and switching the boot config switch whenever I need to load new firmware, even though I don’t have to do that every day. (R. S., Epping, Vic) • Geoff Graham responds: There is no known problem related to loading firmware upgrades on the Colour Maximite 2. It sounds as if there is some confusion with the UPDATE FIRMWARE command. This is equivalent to setting the BOOT CONFIG switch to “SYSTEM”, but this will only work when using the alternative method of loading the firmware via serial transfer over USB. It will not work when using the USB-A to USB-A cable in the keyboard port because you cannot use the keyboard to enter the command in the first place. When using this method, you must open the case and set the BOOT CONFIG switch to “SYSTEM”. Flashing LEDs on Ultrasonic Anti-Fouling We purchased, built and installed your Ultrasonic Anti-Fouling units (May & June 2017; siliconchip.com. au/Series/312) two years ago. Recently, the power LED suddenly stopped flashing. After two years of fabulous service, and a lovely clean bottom, we were taken aback. We found the 3A fuse had blown and replaced it. The green power LED came on, but did not stay on. No fuse blew, no fault light showed. We then replaced REG1 (LP2950ACZ) and carefully monitored the slow start-up. The green power light came on for approximately five seconds, went off, and the fault LED started a slow flash. We replaced both 2200µF capacitors, but no joy. Again, the green power LED lit up, but this time, the power light went off quickly, and a few seconds later, the fault LED started flashing. Australia’s electronics magazine Many thanks for this brilliant kit, and any help you can offer to get it back working. (W. B., Vancouver, Canada) • You could have a faulty driver Mosfet (Q1-Q4) or the transformer windings might have failed (unlikely but possible). Check the Mosfets for short circuits between the drain and source pins (if there is one, you will get a low ohms reading regardless of the lead polarity). If they check out OK, the problem may be that the soft starting feature has found fault with the low-ESR bypass capacitors, where there is current leakage. You could add in a resistor across the drain and source of Q5 to counteract the leakage so the circuit will start up. This can be done on the underside of the PCB. A 330W 1W resistor should counteract any leakage. Increasing boost supply output power I want to build a switching DC/DC converter to drive an audio power output valve from a low-voltage DC supply. I was thinking about modifying the supply from your Valve Stereo Preamplifier For HiFi Systems (January 2016; siliconchip.com.au/ Series/295) to do the job. I need about 1-2W continuously at about 250V, and maybe 3-5W peak. What changes do I need to make to your circuit to achieve this? I checked the data sheet for the MC34063 switching regulator IC, but it doesn’t give any details about using it to drive an external Mosfet as you have done in your circuit. The data sheet gives formulas involving the Vsat value of an external bipolar transistor. Do I use the Vgs(th) value of the Mosfet instead? (J. H., Glasgow, Scotland) • First, let’s cover your question about the Mosfet. When you’re substituting a Mosfet for a bipolar transistor, replace Vce(sat) with Rds(on) multiplied by the expected drain current (Id). Presuming that Vce(sat) was calculated using the peak current value, use that same current figure as Id. You February 2021  107 will need to know the Mosfet’s operating Vgs to determine the correct value for Rds(on). As for boosting the output power of our circuit, we calculate that the total current drawn by the 12AX7 in our design is 4mA, so at 250V, that’s 1W. And as described in that article, the converter is running flat out to achieve that. Increasing that to 2-3W is not going to be a simple job, but it’s probably possible. Start by reducing the value of the resistor between pins 6 and 8 of REG2 and loading up the output to see what it can deliver. We suggest the first test should be with a 0.1W resistor. If you can’t increase the output power to your desired level, the next step is to substitute a larger inductor for L1. Try a 100µH 3-5A toroidal inductor. You might get better results with a different inductance value. You can probably lower the resistor value further with a larger inductor. You will find it easier to achieve the desired output power with a higher input voltage, up to the maximum that the MC34063 and input capacitors can handle. That limit is 24V with the circuit as presented, or up to 40V if you increase the capacitor ratings. NiMH cell recharges suspiciously fast Recently, I decided to test several AA-size NiMH used cells that had been out of use for several years. It was no surprise that many were totally flat, but several showed an open circuit voltage of 0.6-0.8V. I have a mains charger with a 25V AC input that can charge two or four AA cells. The charger has a red & green LED indicator that flashes slowly when the cells are placed in the charger, then a steady red for about 20 seconds, then a steady green indication which suggests that the cells are fully charged. I find it hard to believe that halfflat AA cells can be fully charged in 20 seconds, although I measured their open-circuit voltage at 1.33V! Given that I know very little about the charging characteristics of NiMH cells, could you refer me to an article that would answer my query? What is a safe rate of charge for AA sized cells? Have you published an article or a construction kit with instructions to build a reliable charger with enough capacity to charge up to 108 Silicon Chip D-size NiMH cells? (R. W., Loxton, SA) • NiMH cells are typically charged at a ten-hour rate. So a 1000mAh cell would be charged over 10 hours at 100mA. Fast chargers require an endof-charge detection method, typically either using the drop in voltage of a cell (dV/dt) once charged, or via a rise in temperature (dT/dt) at the end. The cell(s) that charged in 20 seconds would have little charge capacity and are probably high-impedance and therefore no good. The voltage would drop as soon as a load is placed on it. A good cell would not rise to 1.3V so quickly. We have published many suitable chargers over the years, including the following: ■ SuperCharger for NiCd & NiMH batteries, November & December 2002 (siliconchip.com.au/Series/111) ■ A Fast Charger For NiMH & Nicad Batteries, September 2007 (siliconchip.com.au/Article/2337) ■ Float charger for NiMH cells, June 2010 (siliconchip.com.au/Article/180) ■ Burp Charger For NiMH & Nicad Batteries, March 2014 (siliconchip. com.au/Article/6730) ■ Intelligent Charger for Nicad & NiMH Batteries, July 2015 (siliconchip. com.au/Article/8677) Repairing speakers with substitute tweeters I have a pair of Sansui SP 1000 speakers, rated at 50W/8W. Each has two 20W 16W tweeters connected in parallel, and all four are open-circuit. Trying to source replacement/substitute tweeters is difficult, so would it be OK to use two 20W 8W tweeters in series? These speakers are 49 years old and have some sentimental value. (R. S., Humpty Doo, NT) • It is difficult to answer that question without knowing more about the original tweeters and the replacements, specifically, their respective sensitivity ratings in dB/W at 1m. You need those figures, plus the impedance numbers, to figure out how to correctly match the new tweeters to the existing speakers. For example, say that you connect the two new 8W tweeters in series, then connect a high-power 16W resistor in parallel with the pair. That will give you the same 8W impedance as presented by the original pair, but with 6dB less signal going to the tweeters Australia’s electronics magazine (the rest of the power will be dissipated by the resistor). If the new tweeters are 6dB more sensitive than the old ones (which is possible), that would be a good arrangement, giving you a similar balance of high and low frequencies as before. Otherwise, the result could sound too bright or too dull. Your suggested configuration will probably change the way the crossover works as the tweeters will have twice the original source impedance. Without knowing the details of the crossover design, it’s hard to say what effect that will have. Regardless of what you end up doing, you might need to make further changes (eg, adding padding resistors) to match the tweeter volume to the other drivers in the system. Higher power valve amplifier wanted I am currently building my second Currawong valve amplifier (November 2014-January 2015; siliconchip. com.au/Series/277). In past issues, you mentioned that you might develop a higher-powered version. Have you found ways to increase its power beyond 10W/channel? (C. J., Samson, WA) • The limiting factor in the output power of the Currawong is the pair of 15W output transformers. The 6L6s certainly should be capable of considerably more than they are delivering in this design, although they would likely need a higher anode voltage. When we looked at this in the past, we found that higher wattage output transformers were prohibitively expensive. The Currawong is costly enough to build in its current form already, so we didn’t think it was worthwhile to do the engineering work to design a higher-power version as it would probably cost over $1000 to build. A keen constructor might be able to figure out how to fit upgraded output transformers, change the power supply to deliver higher voltages to the 6L6s (but not the 12AX7s!) and obtain perhaps 20-30W per channel. But we just don’t think it’s worthwhile when solid-state amplifiers with much higher outputs power ratings, lower distortion, lower noise and with flatter frequency responses can be built for significantly less money. siliconchip.com.au Building a mains-based PortaPAL-D with effects I would like to build your PortaPAL-D portable PA system (December 2013-February 2014; siliconchip. com.au/Series/177), but I want it to be powered from the mains, not a battery. I also want to add the Digital Effects Processor (October 2014; siliconchip. com.au/Article/8033). My question is concerning the power supply. If I install the transformerbased dual rail power supply for the power amplifier, how can I get the +12V single rail supply for the Microphone input PCB and the Digital Effects Processor? I don’t think I can just step down one side of the dual rails, as the transformer ground will be shorted to the signal. What do you suggest? (V. S., via email) • That should be possible. The PortaPAL-D is based on the CLASSiC-D amplifier which originally used a mains power supply (that we published in December 2012) which produced ±57V and optionally ±15V rails from an extra set of transformer secondary windings, or a second lower-voltage transformer. Those extra components were not shown in the article, as the CLASSiCD did not need them, but the PCB has provision for them. The 7815 on that board could be changed to a 7812 to produce a +12V rail. A separate +12V supply from something like a plugpack or open-frame switchmode supply could certainly be used, as long as the two supply grounds are joined. Model train controllers damaged by short circuit I have purchased a few PWM train controllers off “fleabay”, but for some reason that no-one can tell me, I have lost the ability to control the speed. They give full power all the time. Do these controllers “blow up” if a train derails and causes a short circuit? Have you designed, or can you point me in the right direction for a topnotch PWM DC controller? I would like to use PWM on my two model train layouts, taking full advantage of the slow running speed that the usual transformer won’t allow. Scale speed is what I am after. I would also like a forward/off/reverse switch and speed control potentiometer. siliconchip.com.au Others who have tried to help me suggested adding a self-resetting thermal cut-out (whatever that is). I wouldn’t have a clue where to put one, and I cannot seem to find a 1A version anyway. (R. L. B., Pine Mountain, Qld) • It is possible that a short circuit could have damaged the PWM controllers you bought. You could try connecting a PTC between the controller and one of the tracks. It would need to be rated for a trip current slightly higher than the controller’s rated current. But we aren’t sure that this would prevent those controllers from being damaged. We published the Li’l Pulser PWM train controller in July 2013 (siliconchip.com.au/Series/178) and a revision in January 2014. It has short circuit protection and forward and reverse as well as speed control. We think it will do exactly what you want, and will not be damaged easily. Ignition system failure in an older car I built your High-Energy Ignition System for Cars (November & December 2012; siliconchip.com.au/ Series/18) from a Jaycar kit (KC5513) and installed it in a classic club car. It appeared to be working fine. However, after about ten hours of driving, the system failed. I found that REG1 (LM2940CT-5) had overheated and its ground return track had fused and burnt the PCB. Also, the label affixed to IC1 had melted away in its centre. Do I need to make changes to the circuit for better reliability? (B. C., Dungog, NSW) • The High Energy Electronic Ignition Module is generally very reliable. We think the return current for the coil ran through the PCB tracks rather than the connection to the case, due to a poor ground connection. Having fixed that, it also wouldn’t hurt to incorporate the extra protection components that we used in our Improved Jacob’s Ladder project from February 2013 (siliconchip.com.au/ Article/2369). It includes extra protection for the regulator, especially where the coil connection lead is adjacent to the power supply leads. Triggering a PIC from a high voltage source I am having a most aggravating time Australia’s electronics magazine with your High-Energy Ignition System (November & December 2012; siliconchip.com.au/Series/18). I have blown up three PIC microcontroller chips, and cannot understand why. I am hoping you can help. I have a dated two-stroke engine driving a vital piece of agricultural equipment, and the CDI ignition system has failed with no spark. I am unable to source a replacement part. I have been trying to manufacture a replacement ignition system using your High Energy Electronic Ignition System project. There is a signal on the HT lead, around +90V going rapidly to about -90V when unloaded, as the magnets on the flywheel pass the poles of the CDI unit. I hoped to use this as a timed trigger signal. I have attempted to condition this signal with increasing severity, but on each attempt, the PIC microcontroller input still fails. I currently have a 5V zener to clamp the voltage at the PIC pin, with a series resistor, plus capacitors before and after the resistors and another 27V zener with a series resistor closer to the signal source. Can you offer any advice or suggestions as to why my attempts to clamp this signal to 5V have failed? I still have one unused PIC chip. (D. L., St Andrews, Vic) • The unloaded CDI coil might be producing brief transient high peak voltage that destroys the PIC input despite your zener clamps. Zeners don’t always have a sharp ‘knee’, and the voltage across them can be significantly higher than expected if enough current is applied. Perhaps a better way to protect the PIC would be to use a transformer to step down the voltage, such as a strobe trigger transformer (eg, Jaycar MM2520) with the secondary connected to the CDI coil, and the primary to the PIC input (via the protection zeners and shunt capacitors). That might give you sufficient signal to drive the PIC input. Note that the CDI coil should be loaded with some resistance to reduce voltage transients. Alternatively, use a 6N138 optocoupler (a 4N28 might be fast enough) to provide voltage isolation. You would still need to have sufficient voltage protection for the LED in the optocoupler using zeners, a limiting resistor and a shunt capacitor like in your circuit. February 2021  109 CDI wanted for a two-cylinder engine I’m looking for a type of capacitordischarge ignition system (CDI) to install on a small two-cylinder engine that could still retain the original points. I’m just looking to take the high power draw off the points. Would your replacement CDI Module for Small Petrol Motors from May 2008 (siliconchip.com.au/ Article/1820) work for this? Do you have preassembled units or parts kits? I could not find any of this when I was directed away from your legacy website. (M. W., via email) • The May 2008 CDI Module is designed for motors with a trigger coil and a high-voltage generator coil. If you have points, then that CDI unit is not suitable. We don’t sell fullybuilt versions, but we can supply the PCB – see siliconchip.com.au/ Shop/?article=1820 For points-based ignition systems, our High-energy Ignition System (November & December 2012; siliconchip. com.au/Series/18) is suitable, provided there is a 12V supply available. Soft Starter for halogen lamp I have been having trouble with a bedside lamp fitted with a 28W halogen candle globe. We have a pretty constant 250V here, and this lamp has frequently been blowing globes at switchon. I remember reading quite a while ago about a soft starter for lamps, so I searched and found the Soft Starter listed in April 2012 (siliconchip.com. au/Article/705). The article I was thinking of was much earlier, but I went ahead and built this project anyway, and it seems to be working OK. However, there is no visible difference when first turned on and when the relay cuts in. It is just as bright as without the soft starter. I looked back at the article and realised that it was designed for highcurrent applications, and using it for a lamp may not be the best choice. The lamp has a steady current of 125mA and a cold resistance of 250W. The 10W thermistor with a current capacity of 15A may not have enough resistance to limit the current at switch-on. I thought maybe I should replace the thermistor with one of about 80W or 110 Silicon Chip so. This may provide more of an initial voltage drop to protect the lamp. (B. D., via email) • You certainly could do that. Try the MF72-400D9. It is cheap and rated for mains use. Its maximum steady-state current is 200mA so should be sufficient, and its cold resistance of 400W will reduce the initial current by about two-thirds, giving a much more gradual filament warm-up. Soft Starter modifications In reference to the Soft Starter from April 2012 (siliconchip.com.au/ Article/705), could I use a more readily available 12V DC coil relay instead of a 24V type? Obviously, the X2 capacitor and zener would need to be adjusted, and some resistors to keep the delay constant. In terms of decreasing the temperature and/or longevity of the thermistor, to use it with SMPS, LED lighting and computer equipment with a maximum continuous current of around 1A, could several thermistors be put in series (or parallel)? (B. A., Dee Why, NSW) • We did not use a 12V relay because it doubles the power drawn from the mains and requires a considerably larger capacitor. You would have to increase the X2 capacitor to at least 330nF. If it does not work reliably, try 470nF. As you said, you would also need to change ZD1 to 12V. You would also, as stated, need to roughly halve the value of the resistor which charges the delay capacitor for a similar delay to the original design. You can put several thermistors in series or parallel, but note that the softstart effect will be stronger if they are in series and weaker if they are in parallel. If the equipment is only drawing around 1A, then a series connection is the best option. A series/parallel combination of four thermistors could also be used and would give the same soft starting capability but with much less heat per thermistor. That’s assuming you could fit them all in the box. Using the Soft Starter with a bore pump I am looking at the Mains Soft Starter for Power Tools project from July 2012 (siliconchip.com.au/Article/601), and Australia’s electronics magazine I am wondering whether it is suitable for powering a single-phase (capacitor start) bore water pump of about 1.5kW nameplate rating. There is a noticeable torque/kick that physically moves the pump every time the low-pressure switch needs to activate – which is many times per day. I think it would be beneficial for the general longevity of the motor and pump components if a soft start controller could be implemented. Whether this circuit is exclusively for universal serial wound motors only hasn’t been conveyed in the original article, and would guide my attempt to implement this as a workable solution. (C. T., Sunnybank, Qld) • It’s possible that the Soft Starter would help in your case, but we do not think so. You have three things working against you trying to use a simple soft-starting circuit with an induction motor: 1) You’re reducing the voltage/current but not the supply frequency, so the motor torque will be very low during the soft-start phase; it probably won’t be enough to get it spinning, which means that it will still hard-start once the relay switches on. 2) The pump is presumably always primed, so it’s starting under load and therefore will need to draw a significant current to spin. The Soft Starter was intended more for use with motors which start up off-load or have a very brief initial current draw like most power tools, or devices with switchmode power supplies. 3) During the initial phase, if the motor doesn’t spin, it’s going to draw a lot of current and get rather hot (although the limited soft-start time means that it’s unlikely to be damaged). The Soft Starter is not exclusively for universal motors, but it is far from ideal for induction motors. It might work with some small induction motors, such as the shaded-pole motors often used to drive fans. Our 2012 1.5kW Induction Motor Speed Controller (siliconchip.com.au/ Series/25) would do what you want as it has a soft start feature, but it is much more complicated and expensive and only just rated for your application. Substitute low-noise PNP input transistors I am gathering parts to build a stereo continued on page 112 siliconchip.com.au Notes & Errata Busy Loo Indicator, January 2021: at the bottom of the left-hand column on p79, where the text says the inputs of IC1b are normally high, it should read IC1d instead. Mini Digital AC Panel Meters, January 2021: if the current transformer secondary is not terminated with a low impedance, it will generate a very high (and potentially dangerous) voltage if any significant AC current is flowing in the primary. So make sure to connect the secondary leads of the CT to the panel meter before any current is allowed to flow through the primary. Radiating test antenna for AM Radios, Circuit Notebook, January 2021: the ferrite rod is 200mm long, not 400mm as stated in the text. Vintage Battery Radio Li-ion Power Supply, December 2020: on page 28, the text refers to a 220µF capacitor being charged via a 220W resistor. The capacitor value is actually 10µF. Colour Maximite 2, July & August 2020: the SD card socket specified for this project (Hirose DM1AA-SF-PEJ(21)) is being discontinued by the manufacturer. Instead, use the DM1AA-SF-PEJ(82) which costs the same and fits the existing footprint on the PCB. The March 2021 issue is due on sale in newsagents by Thursday, February 25th. Expect postal delivery of subscription copies in Australia between February 23rd and March 12th. Advertising Index Altronics...............................21-24 Ampec Technologies................. 37 Dave Thompson...................... 111 Digi-Key Electronics.................... 3 Emona Instruments................. IBC Jaycar............................ IFC,53-60 Keith Rippon Kit Assembly...... 111 LD Electronics......................... 111 LEDsales................................. 111 Microchip Technology............ OBC Ocean Controls........................... 5 Silicon Chip Binders............... 111 pair of Ultra-LD Mk.3 amplifier modules (March-May 2012; siliconchip. com.au/Series/27). The transistors specified for Q1 and Q2, 2SA970 lownoise PNP bipolar transistors (BJTs), seem to be no longer available. I am considering using KSA992s as a substitute. Do you think this will compromise the performance of the amplifiers? (I. S., Mitcham, Vic) • The KSA992 looks OK. It’s hard to be sure because the way they specify the noise voltage in the data sheet is not very useful. We don’t think you will notice the difference (if any). We can see some online sellers offering 2SA970s, but we think many of them are counterfeit parts. Some people have said that the clone parts work well, while others say they are not low-noise types; it probably depends on the luck of the draw. Your suggestion of using KSA992s is safer since you can get them from a reputable supplier. Old remote preamp not recommended I was browsing old issues of Silicon Chip and found a project by John Clarke called the Stereo Preamplifier with IR Remote Control (September & November 1993; siliconchip.com.au/ Series/168). Since then, many newer designs have been published like the Ultra-LD Stereo Preamplifier & Input Selector 112 Silicon Chip that I am happily using together with the Ultra-LD Mk4 amp. What caught my attention was the absence of a volume pot and the LED display. I was wondering if that design is still valid and if I could build it? I’d need to find a few replacement components that’d be obsolete by now. I checked that I could source the microcontroller online; however, I have no way to program it. Do you have any advice on this? (O. A., Singapore) • That Preamplifier would be extremely difficult to build at this late date, as many critical parts would be very difficult to obtain. The microcontroller would be difficult to program as we don’t have the facilities for that processor anymore. We do not recommend that you start building this project. Note that we are working on a new digital preamp design with remote volume, bass and treble controls; however, it is not yet finished, and we don’t know when it might be published at this stage. Problem with 3-channel Rolling Code Remote I built this project, described in your August & September 2009 issues (siliconchip.com.au/Series/39) from a Jaycar kit, Cat KC5483. All functions are operational, but the range is only about 3m. I have checked the antennas on both Australia’s electronics magazine SC Micromite BackPack............ 96 Silicon Chip PDFs on USB....... 87 Silicon Chip Shop.................... 97 The Loudspeaker Kit.com........... 7 Tronixlabs................................ 111 Vintage Radio Repairs............ 111 Wagner Electronics................... 51 the transmitter and receiver but cannot find a fault in the construction. Do you have any suggestions? (G. P., via email) • The most likely cause is the soldering to the coiled wire antenna. The wire is enamel-coated, and unless this is scraped off well before soldering, it may not form a good connection, reducing the effectiveness of the antenna. You probably have already checked these connections. However, a multimeter measurement of resistance from the antenna input on the receiver module (or output for the transmitter module) to the free end of the antenna will verify if this is a low-ohms connection, as expected, or high-resistance/ open-circuit. Another thing to check is that there is the full 5V DC supply to the transmitter and receiver modules and that the transmitter supply stays at 5V when transmitting. SC siliconchip.com.au