Silicon ChipI'm on holidays, but not from servicing! - October 2018 SILICON CHIP
  1. Outer Front Cover
  2. Contents
  3. Publisher's Letter: Seemingly helpful technology may not be that helpful
  4. Feature: Reusable Rockets by Dr David Maddison
  5. Project: GPS-synched Frequency Reference Pt.1 by Tim Blythman
  6. Project: Arduino-based programmer for DCC Decoders by Tim Blythman
  7. Project: Low-voltage, high-current DC Motor Speed Controller by Nicholas Vinen
  8. Serviceman's Log: I'm on holidays, but not from servicing! by Dave Thompson
  9. Feature: Developing CleverScope's high-performance CS448 by Bart Schroeder
  10. Project: Opto-Isolated Mains Relay by Tim Blythman
  11. Feature: Intro to programming: Cypress' System on a Chip (SoC) by Dennis Smith
  12. PartShop
  13. Product Showcase
  14. Vintage Radio: Emerson 838 hybrid valve/transistor radio by Ian Batty
  15. Subscriptions
  16. Market Centre
  17. Advertising Index
  18. Notes & Errata: Steam Train Whistle/Diesel Horn / Arduino Data Logger
  19. Outer Back Cover

This is only a preview of the October 2018 issue of Silicon Chip.

You can view 40 of the 112 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 "GPS-synched Frequency Reference Pt.1":
  • GPS-synched Frequency Reference PCB [04107181] (AUD $7.50)
  • PIC32MX170F256B-50I/SP programmed for the GPS-Synched Frequency Reference [0410718A.hex] (Programmed Microcontroller, AUD $15.00)
  • VK2828U7G5LF TTL GPS/GLONASS/GALILEO module with antenna and cable (Component, AUD $25.00)
  • Micromite LCD BackPack V2 complete kit (Component, AUD $70.00)
  • SMD parts for the GPS-Synched Frequency Reference (Component, AUD $80.00)
  • Firmware (HEX) files and BASIC/C source code for the GPS-Synched Frequency Reference [0410718A.hex] (Software, Free)
  • GPS-Synched Frequency Reference PCB pattern (PDF download) [04107181] (Free)
  • GPS-synched Frequency Reference drilling and cutting diagrams (PDF download) (Panel Artwork, Free)
Articles in this series:
  • GPS-synched Frequency Reference Pt.1 (October 2018)
  • GPS-synched Frequency Reference Pt.1 (October 2018)
  • GPS-synched, lab-quality frequency reference (Part 2) (November 2018)
  • GPS-synched, lab-quality frequency reference (Part 2) (November 2018)
Items relevant to "Arduino-based programmer for DCC Decoders":
  • DCC Decoder Programmer PCB [09107181] (AUD $5.00)
  • DCC Decoder Programmer PCB plus stackable headers [09107181] (AUD $7.50)
  • Small 2A boost step-up regulator module (MT3608) (Component, AUD $3.00)
  • DCC Decoder Programmer sketches (Software, Free)
  • DCC Decoder Programmer PCB pattern (PDF download) [09107181] (Free)
Items relevant to "Low-voltage, high-current DC Motor Speed Controller":
  • Four-channel High-current DC Fan and Pump Controller PCB [05108181] (AUD $5.00)
  • PIC16F1459-I/SO programmed for the Four-channel High-current DC Fan & Pump Controller (0510818A.HEX) (Programmed Microcontroller, AUD $10.00)
  • Firmware for the Four-channel High-current DC Fan & Pump Controller (0510818A.HEX) (Software, Free)
  • Four-channel High-current DC Fan and Pump Controller PCB pattern (PDF download) [05108181] (Free)
Articles in this series:
  • Low-voltage, high-current DC Motor Speed Controller (October 2018)
  • Low-voltage, high-current DC Motor Speed Controller (October 2018)
  • Low voltage DC Motor and Pump Controller (Part 2) (December 2018)
  • Low voltage DC Motor and Pump Controller (Part 2) (December 2018)
Items relevant to "Opto-Isolated Mains Relay":
  • Opto-Isolated Relay PCB plus two extension boards [10107181] (AUD $7.50)
  • Opto-Isolated Relay PCB pattern (PDF download) [10107181] (Free)
Items relevant to "Intro to programming: Cypress' System on a Chip (SoC)":
  • Cypress PSoC4 CY8CKIT demonstration project files (Thermistor/LCD) (Software, Free)
Articles in this series:
  • Intro to programming: Cypress' System on a Chip (SoC) (October 2018)
  • Intro to programming: Cypress' System on a Chip (SoC) (October 2018)
  • Cypress “system on a chip” part 2 (September 2019)
  • Cypress “system on a chip” part 2 (September 2019)

Purchase a printed copy of this issue for $10.00.

SERVICEMAN'S LOG I'm on holiday, but not from servicing! Dave Thompson* It’s been 11 years since we’ve visited my wife’s hometown in Croatia and a lot has happened in the meantime. At home, we’ve suffered through a global financial meltdown and something of a physical meltdown in the form of 14,000-odd earthquakes, both of which were devastating to almost everyone in Christchurch. On the other hand, Croatia (and the Dalmatian coast in particular) has seen a huge boom in the number of people visiting and is enjoying the economic benefits this has produced. However, all these tourist dollars come at some expense. My wife’s hometown has an off-season population of around 25,000. At the height of the season, around half a million live here with another million or so passing through during the summer months. In order to house all these extra people, what started as a (literally) cottageindustry of folks renting out an empty room or two to the passing tourist is now a global business, with hundreds of new, multi-story apartment buildings dotting the coast, courtesy of wealthy European investors looking to cash in. Many locals who once enjoyed pristine, panoramic Adriatic views now look out upon some cinder-block wall literally a few metres away. I suppose you could consider that progress but the locals might disagree! One thing that hasn’t progressed at the same pace is the Internet infrastructure. In Christchurch, we have enjoyed three significant broadband speed upgrades over those 11 years. The last time we visited Croatia, my wife’s family were enjoying then-revolutionary 2Mb/s (megabit per second) copper-wire based broadband pipe they’d just had installed. It certainly beat the pants off the dial-up they’d been using previous to that, and that faster internet revolutionised almost everything here. But nothing has really changed since then. Economic growth spurts Countries emerging from conflict often benefit from rapid progress and development and Croatia was no exception. While they weren't exactly behind the iron curtain, it certainly had an effect on them; the Socialist ethic of the day shunned outside influences. Once that veil lifted, things changed rapidly. German über-telco T-Mobil stepped in and offered to rebuild the telecommunications infrastructure. This saw Croatia at the bleeding edge of telecommunications technology in Europe. Their Internet services were equal to or better than those we enjoyed at home. We had a similar telecommunications boom in New Zealand. Before that, we were regarded as being 20 years behind everyone else in just about everything. But then several overseas companies came along with massive investments in hardware and infrastructure, which kicked our creaky old analog systems into the 21st century. Without them, we’d still be in the dark ages, so to speak. Sadly, in Croatia, the lustre of those new investments has now well and truly faded; my relatives use exactly Items Covered This Month • • • • • The Kiwi takes flight Solar-related failures Swarfed up stepper motor Honda SUV failure to spark Palsonic TFTV3920MV TV repair *Dave Thompson runs PC Anytime in Christchurch, NZ. Website: www.pcanytime.co.nz Email: dave<at>pcanytime.co.nz siliconchip.com.au Australia’s electronics magazine October 2018  61 the same modem they’d been supplied with back then and the speeds they get from the now way-oversold system are often deplorable. Given the state of the ancient copper lines and the sheer difficulty of installing a fibre network into an ancient city, where the 2,000-year-old Roman ruins down the road look like a new subdivision, it seems their Internet isn’t going to get faster any time soon. The serviceman's curse returns Homes here are built with very thick walls, using concrete, bricks and steel reinforcing, all of which creates a rudimentary Faraday cage. WiFi signals reach from the router to about as far as the next wall. The last time we were here, my brother-in-law and I spent a few gruelling hours under the intense Dalmatian sun installing a temporary, 25 metre long local area network (LAN) cable, running it up and over the roof through eaves, always-open windows and shutters and through wherever it would fit to link us to that router. Sadly, with the installation of airconditioning and other renovations, that cable had long gone and the route was no longer viable. And so the first thing asked of me when I arrived (after the usual family stuff) was if there was any way to speed things up. In an effort to see what I was dealing with, I tried to wirelessly connect my laptop to their router, which is situated about eight metres away as the crow flies, in an office on the other side of the house. Despite the short distance, the connection was very poor; virtually useless. Since we had a lot of catching up to do with various people, it was a while 62 Silicon Chip before we got around to visiting the over-worked local computer guy at his shop. His workshop is about the size of my bedroom and is stacked from floor to ceiling with old laptops, desktops and printers. It appeared that every tourist visiting the city was in there complaining (in 10 different languages) that something was wrong with their mobile device. From what I could gather, most of them were simply unaware of the requirements for WiFi passwords. The poor tech was trying to explain in his best pidgin Euro that all they needed to do was enter a WiFi password and they’d be able to connect in their hotel, apartment, camper-van, Ražnjići stand etc. I felt his pain, but to be honest I was somewhat relieved that I wasn't him! Once I got a chance to talk to him, I asked for his best WiFi access point and paid about half what I would have shelled out in Christchurch for an Asus Router/Wireless Access Point. This gave me a few options and just to be safe, I also purchased another 25 metre Cat5 network cable. Repeaters don't work that well I’d used routers in wireless repeater mode before and while that would have been an easy solution, I’ve had little real success using them in this way. The idea is that the router picks up whatever WiFi signal is available and then relays it, to provide better coverage. But the problem is that the weaker a WiFi signal gets, the more packets get dropped and the slower the connection goes. So even if we had a good connection to this new, stronger and faster WiFi network, the router still has to relay the packets back to the original network, which is as slow as a wet week. Also, when a router is used in repeater mode, its bandwidth is halved because it has to deal with double the amount of network packet requests and receipts. So while this configuration means no pesky cabling, it is clearly not ideal. The best option is to use the router as a network access point but then the access point must be hard-wired into the original router and placed close to where we will connect to it. That means running a cable at least part way; a challenging prospect but one I’d have to overcome. Australia’s electronics magazine I started by walking the proposed cable route with an eye to boring holes in either the timber door frames or the masonry itself. I’d talk over the options with the homeowner later; right now, I had to see how viable it would be to run a cable this way. I had three doors to circumvent and while it seemed I’d be able to run a cable through the gap under two of the doors, things came unstuck on the last door which separated the two halves of the house. On this older part of the house, while the upper door frames were timber, they were actually very thin, merely decorative strips, so drilling anything through them was going to be impossible. There was also a bottom strip, like a small step about 25mm high that the door closes against, and this is made of a marble-like polished stone that is bonded to the floor and fits perfectly into the wooden door frame; so running a cable under the door was not an option. Drilling a large enough hole to take a Cat5 cable and plug through a 500mmthick concrete wall wasn’t something I was prepared to do and besides, the owners didn't like the idea. If I pushed the point I might have swayed them but, as non-technical types, they regarded the work as non-essential. So I’d have to go another way. Just before that inaccessible access door is a spare bedroom and it has the newer type of door with no marble step, so I came up with the idea to run the cable into that room instead and sit the access point high on top of the dresser, on the opposite side of the wall to where we’d be using the laptop. I rolled the cable out and temporarily positioned and set up the AP to check the signal, and it was good; I’d put it there. Back to work after a nap After 5pm, I wandered back down to the computer shop. Like many European stores during the summer, he opens at 8am, closes at 1pm and re-opens at 5pm, trading until 9 or 10pm. This is simple practicality; it is so hot during the early afternoon that nobody ventures out anyway. Most locals have their main meal at around 2pm, then have a kip before going back to the office. It is all very civilised, though a little alien to us Antipodeans. The Spanish siliconchip.com.au “Setting the standard for Quality & Value” Established 1930 ’ CHOICE! THE INDUSTRY S CNC Machinery Metal Working Sheet Metal Fabrication Wood Working Workshop & Automotive Lifting Handling Cutting Tools Measuring Equipment Machine Tool Accessories HSS Industrial Centre Drill Set Metric Precision HSS Jobber Drill Set Imperial Precision HSS Jobber Drill Set Metric Precision HSS Jobber Drill Set • • • • • • • • • • • • • • • • • • 5 piece set No. 1, 2, 3, 4, 5 HSS M2 grade Industrial quality 40 $ 47.30 (D508) $ HSS Sheet Metal Step Drill Set • • • • • • 3 piece set For drilling thin material HSS M2 grade 4-12mm x 1mm steps 6-20mm x 2mm steps 6-30mm x 2mm steps 75 $ $ 82.50 (D1071) 25 piece set Precision ground flutes HSS M2 bright finish Range: 1~13mm 0.5mm increments 89 $ 95 $ 99 (D1272) $ HSS Countersink Set • • • • • • • 4 piece set HSS M2 grade 45° angle Ø2 ~ Ø5mm Ø5 ~ Ø10mm Ø10 ~ Ø15mm Ø15 ~ Ø20mm 65 $ $ 29 piece set Precision ground flutes HSS M2 bright finish Range: 1/16 ~ 1/2" 1/64" increments 115 $ 104.50 (D1282) • • • • • • • • • 6 piece set HSS M2 grade 45° angle Ø1.5 ~ Ø6.3mm Ø6.3 ~ Ø8.3mm Ø8.3 ~ Ø10.4mm Ø10.4 ~ Ø12.4mm Ø12.4 ~ Ø16.5mm Ø16.5 ~ Ø20.5mm 75 126.50 (D1285) $ Metric HSS Combination Tap & Drill Set HSS Countersink Set $ 71.50 (D1051) $ 51 piece set Precision ground flutes 1-6mm in 0.1mm increments HSS M2 grade • 7 piece set • Designed for up to 3.2mm sheet metal • HSS M2 grade • M3, M4, M5, M6, M8, M10 • Includes hex snap-on drive 60 $ 82.50 (D1061) $ $ 66.55 (T0191) Digital Caliper Digital Caliper EF-5S - Engineers File Set Metric HSS Hole Saw Set • • • • • • • • • 200mm hardened and tempered files • Second cut: Flat, 1/2 Round, Round, Square, Triangular • Includes carry case • 11 piece set • M42 Bi-Metal high speed steel • 19, 22, 25, 32, 35, 38, 44, 51, 57, 64, 76mm • Includes 3/8" & 1/2" arbor shank & pilot drill 150mm / 6" Metric, inch & fraction 4-way measuring Includes battery 33 $ $ 38.50 (M738) 200mm / 8" Metric, inch & fraction 4-way measuring Includes battery 49 $ $ 35 $ 59.40 (M739) 99 $ 42.35 (F100) $ $ 110 (D102) RSP-500 Pneumatic Roller Seat GSP-795 Pneumatic Stool TCS-3 Mobile Tool Cabinet Seat WCH-6D Workshop Series Tool Chest • • • • • • • • • • • • • • 600 x 260 x 340mm • 6 drawers with protective liner • Ball bearing slides • Key lockable drawers & lid 380-500 seat height Ø300mm padded seat 360º swivel wheels Moulded tool tray 39 $ $ 44 (A360) ALL & THIS E MOR & 675-795mm seat height Ø360mm padded seat 360º seat rotation 135kg capacity 99 $ $ 99 $ 110 (A359) Staff Member $ 110 (A001) 89 $ $ 99 (T690) ATBG280/6 Industrial Bench Grinder BD-325 Bench Drill PD-325 Pedestal Drill • 150mm wheels • Fine/coarse grit • 0.37hp, 240V motor • • • • • • • • • • TORE IN S INE ONL - CAM 3 x drawers with ball bearing slides 420 x 235mm padded seat 2 x magnetic side trays 406mm seat height 360º swivel wheels 129 $ $ 16mm drill capacity 2MT spindle 12 spindle speeds Swivel & tilt table 1hp, 240V motor 319 $ 143 (G150) $ 352 (D590) Be a Mate & 16mm drill capacity 2MT spindle 12 spindle speeds Swivel & tilt table 1hp, 240V motor 359 $ 396 (D592) $ How to Enter 1 SPEND $100 2 LOG INTO YOUR ACCOUNT www.machineryhouse.com.au/SignUp UNIQUE PROMO CODE 3 ENTER ONLINE SC1018 T&Cs apply. Visit www.machineryhouse.com.au/Win-a-Harley Permit No. LTPS/18/23950 SYDNEY BRISBANE Australia’s electronics magazine (02) 9890 9111 1/2 Windsor Rd, Northmead (07) 3715 2200 625 Boundary Rd, Coopers Plains PERTH MELBOURNEOctober 2018  63 (03) 9212 4422 4 Abbotts Rd, Dandenong (08) 9373 9999 11 Valentine St, Kewdale Specifications & Prices are subject to change without notification. All prices include GST and valid until 31-10-18 10_SC_270918 ONLINE OR INSTORE! siliconchip.com.au www.machineryhouse.com.au/Win-A-Harley call this “siesta” and it is colloquially known as that here too. While the big malls and supermarkets stay open from 7am until 11pm during the tourist season, many tourists from outside Europe grumble about smaller, local shops being closed at these seemingly odd hours. But I’m OK with it. I bought some small plastic cable clamps with a nail at one end. These were masonry nails, designed for the type of walls used in almost all houses here. I also purchased some cable ties because the last part of the cable route follows some copper central-heating pipes and I could tie the cable to the existing clamps. Back at the house, I started at the router and laid out about five metres of cable. I used cable clamps to tack the wire to the walls every 300mm or so. The odd nail would hit a stone in the concrete and either ping out and fly across the room or simply folded up, ruining the cable clamp. But the majority hammered in easily and held fast. I routed the cable up and down door jambs, underneath and up again and around corners until I had the cable wired into the room where the access point would live. I coiled the remaining cable up and sat it beside the access point; the cable I got was a bit too long but that's better than it being a bit too short! The acid test was whether our computers and phones could now connect and actually reach the internet. While we could connect periodically before, surfing the web was a lesson in frustration. Now, once I connected to the access point I got five bars and web pages loaded in double-quick time. First world problems I need remote access to our machines back in Christchurch and with the old WiFi connection, I couldn’t even reach the logon screens. Now, clicking the remote desktop icon for any of our three remotely-accessible machines resulted in our familiar desktop being displayed almost instantly. While it takes a while to get used to the slower pace of the Internet here (2Mb/s compared to our 900Mb/s at home), at least we are back online. It isn't all roses, though. When I logged into my email server, I saw I had over 120 emails waiting in my personal inbox and I haven’t even looked at 64 Silicon Chip my work email yet. All this after only six days offline! While being away from it all is appealing, we have to be locatable by the likes of the airlines, friends and family, so it just isn’t feasible to unplug and walk away. At least I don’t have any more work to do... yet! Editor's note: running a network cable gives the best performance but if you don't want to (or can't) then power line networking is often the easiest way to add extra WiFi access points. A series of solar-related failures Several years ago, N. D. retired and decided to move from suburban Perth to the countryside of Western Australia. He and his partner decided to install a solar-powered hot water system and solar photo-voltaic (PV) system with grid-tied inverter at their new property, employing a local company to install them. Recently, he ran into problems with these systems... The solar hot water system circulates water between a storage tank under the house and a roof collector panel. It also has a natural-gas powered booster which initially masked some of the failures. Early on, we had to call the installers to fix the circulation pump as it was running constantly, undoing all the good work done by the sun of heating water in the day by re-radiating it at night. After a lot of fiddling about, it was pronounced fixed by the installers and they went on their way. Six months later, I discovered the pump was not running at all. An inspection of the controller showed fault lights that indicated the roof sensor was open-circuit. Meanwhile, the installers had gone out of business! An internet search revealed that new sensors were $75 each plus postage; a bit steep I thought, but if that was the problem, it would save me the call-out fee for a plumber. So I ordered one and when the sensor turned up, I fitted it and the system went back to working correctly. Two years passed until one day I was under the house and could hear the sound of a relay clicking on and off, followed by a pause of a minute or so, then more relay noises. But it was not the hot water service this time, it was the PV inverter. The inverter was displaying an error code that an internet search showed Australia’s electronics magazine was a common problem for this particular brand and model. The inverter was still under warranty and luckily, despite the fact that the installers who sold it to us were no long a going concern, the manufacturer (a large German company) arranged to exchange it for a new unit at no cost to us. So that solved that problem. But another six months down the track, on the first hot day of the summer, we were startled by loud growling and bubbling noises coming from the roof space. It turns out that the water in the solar hot water system was boiling! A trip under the house showed the circulation pump was not running and the controller was dead. Fortunately, the pump plugged into the controller used a standard mains plug and connecting that directly to the mains got the pump working and water circulating again. More internet searching determined that a new controller was $240. That also seemed pretty steep, especially after I took a look inside and saw that the major component was a printed circuit board with an IC, a couple of transistors, a relay and a handful of components. Notably, there was a small SMD device marked “F2” in series with the primary of a PCBmounted transformer; presumably this was a fuse. It was showing signs of distress and measured open circuit. Although I’ve never been employed in electronics industry, I’ve maintained a keen hobby interest over the years building many projects so I could not let this go by without attempting to fix it. Despite doing more internet search- siliconchip.com.au es, I could not find a circuit for this controller. It was obvious that it controlled the relay that turned the pump on and off. The IC was most likely a comparator with the two inputs being from the tank temperature sensor and the one on the roof. The PCB-mounting transformer was marked to show it had a nine volt output. The secondary was connected to a bridge rectifier, feeding an electrolytic capacitor rated at 25V. The relay showed it had a 12V coil so that probably meant the supply voltage to the comparator and a nearby switching transistor was around 12V. I hooked my current-limited variable power supply across the output of the bridge rectifier and powered it up. Success! The indicator lights came on with about 50mA of current drawn from the supply. The relay clicked in and plugging in the sensors showed them to be working. That left as suspects the bridge rectifier, the transformer or the fuse. The bridge rectifier appeared to be OK as determined by in-circuit testing of its diodes. I decided to replace the fuse and try it again. SMD fuses are not something that I carry so I replaced it with a 1W ¼W resistor, figuring that it would quickly destroy itself if there was a fault in the transformer, and that’s exactly what happened; it went up in a cloud of smoke! Measuring the transformer primary showed a resistance of around 14W which is way too low. With hindsight, I should have tested the primary resistance first. A replacement transformer of the same brand did not appear to be available locally (the controller was made in Israel) and none of the regular suppliers had one with the same pin spacing. However, with a bit of pin bending, I managed to fit one from Altronics to the board and as that had an internal fuse I did not bother to try and source a replacement SMD fuse; I simply shorted it out. The controller is now back in place and working until the next thing goes wrong with the hot water system or solar inverter. Or should I sell the house? It seems to be cursed… Faulty stepper motor G. C., of Nelson Bay, NSW, had a faulty stepper motor in his 3D printer that intermittently jammed. He siliconchip.com.au couldn't find any information on Google about what might be causing this problem so he decided to investigate... I fitted a new extruder to my 3D printer which uses a stepper motor to feed the plastic filament into the heater. Unfortunately, it quickly started to jam up and I discovered that it was the motor that was at fault. When it jammed, it wouldn't rotate until I manually rotated it in reverse. It would then work for a little while before jamming up again. I decided to bite the bullet and see if it was repairable. After some fiddling with the four screws holding its case together, I pulled the stepper motor apart. It's a relatively simple design. The body comes apart in three pieces; there's also a stator, the rotor, two ball bearings and some washers. The rotor has a very strong magnetic field. Having gotten it apart, I still couldn't see a reason for this intermittent stopping but I suspected that the rotor was the likely culprit. It is, after all, the only part that actually moves. Careful checking it, using my iPhone as a magnifying glass, revealed a possible cause. I could see tiny pieces of metal swarf stuck to the very strong stator magnet. I used long-nose pliers, tweezers, compressed air, percussive maintenance (hitting it) and managed to remove many, many pieces of swarf. But every time I reassembled and tested the stepper motor, it kept on jamming. It seemed likely that the tiny metal pieces were hidden away inside some other part (a bearing?) but during operation, they were pulled into the rotor by the intense magnetic field, causing it to jam again. The problem remained as these tiny pieces kept resisting my best efforts to remove them. Then I had an epiphany – if I surrounded the rotor with epoxy, filling in the gaps between it and the stator (without actually causing any friction), there would be nowhere for the swarf to get in between the two and jam it up. So I put the rotor in a battery drill, mixed up a big blob of five-minute epoxy, liberally covered the rotor in epoxy and started the drill at a very low speed. This rotated the rotor, preventing any big blobs from forming at the bottom as it flowed down due to gravity. I waited an hour until the epoxy Australia’s electronics magazine The stator magnet had small pieces of metal swarf stuck to it; the likely suspect for the cause of jamming. The disassembled stepper motor comes in three major pieces: the stator, rotor and two ball bearings. An epoxy coating was applied to the rotor and then machined so that swarf could not get in and jam it. October 2018  65 was fully hardened, put it in my lathe and removed the excess epoxy so that the rotor and stator would have sufficient clearance. After carefully cleaning the rotor, I carefully reassembled everything (yet again) and, success! No more jamming. So it was rubber to the road time. I put the repaired stepper motor into my 3D printer. Everything worked this time, and I've been printing successfully for several days now, with no problems and no jamming. Honda CR-V ignition system lacking a bright spark The very same G. C. also had some family drama when his 17-year old granddaughter was getting in some last minute practice an hour before her driving test and the car died. Luckily, the family had a spare car and she passed her test. But their 2000 Honda CR-V was sitting dead at the end of the street... Armed with some ether starting fluid and a digital multimeter, I walked to the car and tried to start it. Naturally, Murphy was having fun with me and it started first time, so I drove it home straight away. I was deciding exactly where to park it when it died once again. At least it was parked out the front of our house at the time. I tried the normal car diagnostic techniques, starting with removing the air filter and squirting in some starting fluid but the CR-V showed not the slightest sign of starting. That suggested that it probably wasn’t a problem with the fuse system. I then checked the ignition system by connecting an old spark plug to one EHT lead but there was no spark. So it seemed that the ignition system was the culprit. There was no alternative but to remove the distributor cap, rotor and plastic dust cover, to expose the electronics inside. This revealed three main components: a crankshaft position sensor (presumably a Hall-effect device), a sealed electronic ignition system and one coil. I could now determine that there was no output from the EHT end of the coil, so it was time to (literally) drag out my old Tektronix CRO and a 20 metre extension lead. This showed that there was 12V to the coil positive terminal but no signal at all on the coil negative terminal, which should have shown +12V to 0V transitions as the crankshaft turned. So either the position sensor was faulty, or the ignition system was dead. I removed both (fighting some unnecessarily tight 4mm Posidrive screws) and took them inside to my workshop. I tried connecting and disconnecting 12V across the coil manually, which gave a noticeable spark upon disconnection, so the coil was OK. Then I made up a simple test jig (just three wires and some connectors) and applied 12V from a regulated and current-limited power supply but there was not the slightest spark when I grounded the ignition system trigger input, which has an internal pull-up resistor. Figuring I had nothing to lose, I removed the top from the ignition system module using a screwdriver and 25mm diamond saw in my trusty Dremel. This device was made by NEC and marked "MC-8541". Inside I found a small PCB, a transistor/Mosfet and a sticky, jelly-like substance. Presumably this was intended for protection against vibration but it had never hardened – I don't know if this was on purpose, or just some dodgy silicone that never set properly. Based on my previous experience, I suspect the latter. There were seven wires connecting the PCB to the terminals and main transistor with four spot-welded to the PCB and external terminals and the other three running between the PCB and switching transistor. The five slightly thicker wires were OK but two of the wires were extremely thin, which combined with the failure of the potting compound, had resulted in both wires breaking off their spot welds. Using the biggest tip I had on my Hakko soldering iron and some aggressively fluxed aluminium solder, I managed to replace one broken wire with some more robust wire (stripped from an old Cat5 network cable). Luckily the other wire was the tachometer output and is not used in this model, so I removed it to avoid any future problems. Left: the inside of the Honda CR-V. Above: The ignition system module made by NEC. 66 Silicon Chip Australia’s electronics magazine siliconchip.com.au I re-connected everything to my testjig, grounded the ignition system input and a very healthy spark appeared, so my gamble paid off. After that, I carefully cleaned as much of the jelly-like gunk off as I could, then potted everything with some 5-minute Epoxy. The conclusion was a bit of an anti-climax; I reinstalled everything, turned on the ignition and the Honda started right up and has been running perfectly ever since. As a post-script, this adventure finally pushed me to order a phonesized $30 DSO150 oscilloscope so that I wouldn’t have the hassle of dragging out my CRO next time something like this happens. Palsonic TFTV3920MV 39-inch LCD TV repair They say one man’s trash is another’s treasure and that certainly seems to be true for S. G., of Bracknell, Tasmania. He recently came across a nice looking TV that had been discarded and decided that he could probably fix it... When analog television broadcasting stopped, many working CRT TVs appeared at the tip. These days you even see flat-panel televisions in seemingly good condition that have been thrown away. Having moved to a small country town in Tasmania, one day I took a load of garden waste to the tip and spotted a 39-inch Palsonic television (TFTV3920MW) that someone had thrown out. The set only took my eye because it was white. It did not look too bad, just a bit dusty, and as far as I could see, the LCD screen was intact. So this set followed me home and into my workshop (“Can I keep him? Please!”). Now that I have a man cave with a good bench, power and lighting, I felt confident that I could fix whatever was wrong with it, that had made the original owner discard it with such disdain. It did not take long to remove the back cover. After a good check over, I found that the power supply ran and was producing 12V DC and also 90odd volts for the backlight. Turning the set on produced not a flicker on the screen and after about 60 seconds, it would shut down. So it seems that the backlight had gone out to lunch. That made it very hard to see if the rest of the set was also dead. siliconchip.com.au In the past, in cases like this, I have shone a high-powered torch on the screen at an angle to see if there is any display. One has to be quick in cases like this, what with the set shutting down by itself. Using this technique, I was able to determine that the set-up menu was indeed being shown on the screen. You just couldn’t see it because the backlight wasn’t working. So I proceeded to totally dismantle the set. This took a fair bit of time as I had to remove many screws. A clean workbench helps since you won’t lose any of the parts; you also need somewhere to store the many parts so you won’t lose them before it’s time to reassemble it. I removed the plastic trim from around the screen and flipped the set on its back so that I could remove the LCD panel. This is the hardest part, as the LCD panel can be damaged very easily and that would make the set a complete write-off. I managed to do that successfully and this revealed a couple of light diffusers and a thick plastic piece that acts both as a diffuser and to help keep the whole set rigid. I could also see a white plastic sheet that covered the LED backlight so I removed that too. Finally, I could see the backlighting LEDs. These are arranged on three boards with 12 LEDs per board. They are connected as a series string; if one goes open-circuit, the whole string will fail to light. That turned out to be the cause of the failure in this set. I used Google to determine how to test the LEDs. The suggestion was to connect a battery made from two "D" cells; this would provide enough voltage to light each LED individually while they were still soldered to the board. Luckily, the boards had test points to make this easier. So I connected the battery across a pair of test points with both orientations until one of the LEDs lit up, then I proceeded to test all 36 LEDs across the three boards. I found two that would not light up across two different boards. There are many types of LEDs on the market so I did a web search for the part number used in this set and I came up with a Chinese supplier offering replacement LEDs boards. They also had the original remote control for this set in stock. But I baulked at the price, as did my wife; I wanted to make sure that the set would work 100% before pulling the trigger. That's when I had a brainwave; holding the LCD panel up to the workshop light allowed me to check it to make sure it wasn’t damaged. Thankfully, it looked good. I could not see any cracks or scratches and with the wife's blessing, I soon ordered the parts. They arrived in just six days. Fitting the replacement LED boards and reassembling the set was not easy; I had to remember where all the screws went and I also had to reassemble the parts in the correct order. After re-connecting all of the plugs and taping back down the many looms, it was time for the big test. With the set now back on its own stand, I applied power and connected the antenna lead, turned the set on and waited. Soon the on-screen display appeared, followed then by one of the local television stations. I did a factory reset, followed by a re-tune (scan) and then the only thing left to do was to clean the LCD panel. For this, I used a clean cloth (actually a cloth nappy; I have a small stockpile of these for cleaning glass). A circular motion and a bit of elbow grease (not pressing too hard) and the set came up like a new one. Another plus for this set is that it has a built-in DVD player which still works fine. So was it worth the time and effort? Given the number of hours I spent fixing it, arguably not, but I did enjoy taking an electronic device that had thrown into the rubbish dump and turning into a fully working and practically new-looking TV. SC Servicing Stories Wanted Do you have any good servicing stories that you would like to share in The Serviceman column? If so, why not send those stories in to us? We pay for all contributions published but please note that your material must be original. Send your contribution by email to: editor<at>siliconchip.com.au Please be sure to include your full name and address details. Australia’s electronics magazine October 2018  67