Fullscreen HTML5Med Res (??MB)HTML4

SERVICEMAN'S LOG Some jobs aren’t worth the trouble Some jobs really aren’t worth the trouble, particularly if the equipment is old or if the fault is intermittent. The trouble is, my persistent nature often stops me from giving up on jobs that have become uneconomic. I thought that fixing Mr Hilda’s JVC HR-D750EA VCR would be relatively simple. His story was that he went away on holidays and switched it off when he left but when he came back, it wouldn’t switch on straight away. Instead, he could only get it to work intermittently. Finally, it wouldn’t start at all, giving only a “squeak” and then nothing. I quizzed him further and established that he was trying to get it to start by switching it off and on using the power point switch on the wall. He wasn’t using the remote control or the power switch on the front of the VCR for this job. To me, this all suggested that one or more electrolytic capacitors had gone leaky in this now 10-year-old VCR, especially the start-up capacitor (which ever one it was) in the switchmode power supply. In this machine, the main section of the power supply con­ sists of a separate module in a metal cage, in the far righthand corner of the chassis. The rest of the power supply is on the main circuit board and consists mostly of IC regulators. My initial attention was drawn to a regulator in the cage, which was fairly easy to remove due to good access to the printed side of the board. The first thing I did was scrape off the liberal quantities of corrosive brown goo that had been applied to both sides of the board. The component side proved to be a bit of a nightmare as the parts were packed tightly between various metal screens and heatsinks. That done, I examined the service manual which is marked “switing regurator”. Anyway, there were two electrolytic capaci­tors on the primary side of the switchmode transformer (T1) that looked like the suspects I was seeking. These were C14 1µF 50V and C13 180µF 16V. I could see that all the electrolytics had been replaced some years previously but these two were now well and truly dried out due to their proximity to the heatsinks. I fitted two new EXR 105°C capacitors, substituting a 220µF unit for the 180µF capaci­ tor and also increasing the working voltage of both values. The new capacitors were smaller than the originals so this was straightforward. By the way, the EXR range of electrolytic capacitors is specially designed for high-frequency switchmode power supplies. They have very low impedance and low leakage, typically around 4µA, which should give extra long life. The equivalent (effective) series resistance, or ESR, of an electrolytic capacitor defines its performance and life; the lower it is, the better the capacitor. An Australian engineer, Sets Covered This Month Fig.1: part of the switchmode power supply circuitry in the JVC HR-D750EA VCR. There were several problems that prevented the supply from starting. •  JVC HR-D750EA VCR •  National TC1407 portable TV set •  National NV-H70A VCR •  Mitsubishi CT2584AS stereo TV set March 2000  65 Serviceman’s Log – continued Bob Parker, has designed an ESR meter kit (available from Dick Smith Electronics, Cat.K-7204) to measure this. Having just bought and built one of these, I was itching to give it a go. The two capacitors I had taken out both measured high im­pedance, the 180µF unit reading 44Ω and the 1µF unit not reading at all. I had fitted an additional buzzer circuit, designed 66  Silicon Chip by Mark Stevenson, that gives an audible sound if the capacitor is OK (ie, below 1Ω) and a slight noise if it is less than 10Ω. While I was at it, I measured all the remaining electrolyt­ics in-circuit with the ESR meter and the buzzer indicated that they were OK, with two or three giving the lesser noise, (ie, possibly borderline). I then resoldered any suspicious joints before putting it all back together and switching it on, confid­ent that it would work. To my surprise, nothing happened. However, I was still convinced it was really an electrolytic capacitor that was at fault, so I replaced the three units that had given a doubtful reading on the ESR meter. That didn’t fix it either. Finally, I took the advice of a colleague who always main­tained that it was quicker and more reliable to just change the lot at one go. I did this, replacing all 11 of them, and bombed out again! Next, I checked the high value resistors R2, R3, R4, R7, R8 but still no joy. I then checked the secondary output rails for shorts and was finally rewarded by finding that D25, a 39V zener, was short circuit. But this still did not fix the fault. To make matters worse, I then introduced a red herring by unplugging the power supply from the VCR (in case the load was too much). This time the power supply finally fired up and I had vol­tages on all the rails. However, when I plugged it back into the VCR, it died again and so I checked for shorts on the main VCR rails. Again, nothing was found – this sorry tale was just too frustrating but worse was to come. I reconnected the two plugs while the power supply was on – fairly risky, I know, but I was desperate – and the whole video burst into life! I checked all the functions and apart from lines across the screen in playback mode, everything worked fine, including the power on/off on the set. But when I switched it off at the power point and then back on, it was dead again! By now I was feeling considerably older – I think my own electrolytics were past their use-by date! The only thing left I hadn’t tried was IC1 (STR1006), which I really doubted could be the problem – but naturally it was. It only has three transis­tors, one zener and three resistors but it was the culprit. My theory is that the two original electrolytics failed, causing the output voltages to rise. And this in turn destroyed D25 and, presumably, the zener diode in IC1. But I still wasn’t out of the woods – I still had the lines on playback, although they slowly weakened, the longer the set was on. From experience these are also usually caused by faulty electrolytic capacitors, this time on the 5V and 12V rails feed­ ing the head amplifier. As a result, I checked C802 to C807 with the ESR meter and they all checked OK. The ESR meter makes it so much quicker, as they can be checked in-circuit and the meter was proving to be very accurate. It was then that I remembered a previous repair I had done on a JVC HR-D400EA with the same fault. After a long saga, the fault had been traced to the underside of a small soldered metal shield on the main video board. This shield conceals a patch of brown goo which holds a 1kΩ resistor between pins 1 and 32 of IC102 (pin 3 being +5V Vcc and pin 1 9V Vss (PB12V) for the MS6967RS 1H delay). As before, removing the goo and corrosion also removed the final problem. However, my persistence with this job had been a little uneconomic and I really should have drawn the line sooner. I was, however, very impressed with the ESR meter and can recom­mend it to you – good one, Bob! National TV set My next story involves a complete change of scene. The local motel brought in a National portable TV set, complaining of horizontal white lines across the picture but only on Channel 7. I wasn’t all that keen about repairing such an old (1984) chassis – a TC1407 (M12H) – but was intrigued by the symptoms. Initially, I suspected the CATV (cable antenna television system) antenna used in the motel but when I connected my own antenna to the set, the same fault occurred. When I tuned to Channel 7 there were bright white horizontal retrace lines about two thirds the way down the screen from the top. Because it was only on Channel 7, I reasoned that this was because this is the only channel transmitting Teletext during the vertical flyback interval. The motel owner, while appreciating that the set was over 15 years old, was prepared to pay up to $100 to get it fixed. Apart from its current problem, the set produced a good picture and had been very reliable. Well, how difficult could it be to find a vertical blanking fault? Surely this would be an easy $100, I foolish- ly told my­self. In fact, I still had an original service manual for this set, although the circuit diagram is a bit of a mess and diffi­cult to follow. I discounted the AGC and IF stages and concentrated my efforts on the blanking circuit between IC401 (the vertical output IC, AN5521), IC601 (the chroma decoder IC, AN5625) and IC301 (the video output IC, AN5615). By using an oscilloscope, I thought that it wouldn’t take long too find out where the blank­ing pulses disappeared. My assumption was that it was probably a leaky electrolytic capacitor somewhere that was causing the problem. I started at pin 6 of the vertical output IC (IC401) and measured waveform 20, which is a 28V peakto-peak vertical pulse (the DC value being 0.3V). This turned out to be spot on. This pulse then goes through C416 and R424 where it joins the horizon­ tal blanking pulse from pin 2 of the horizontal output transform­ er via R553, C650, R678, R642 and D601. The horizontal waveform (waveform 37) started out at 30V peakto-peak on pin 2 and is reduced to about 9V on the anode of D601 but no information on the waveform at this point is supplied in the service manual. However, the 5ms vertical pulse, embedded between the horizontal pulses, was clearly visible. Because I was expecting something dramatic, I was rather perplexed to find that this waveform reached pin 18 of IC601 and pin 11 of IC301 quite correctly. This waveform (24) shows the horizontal pulse as 6.2V p-p, which was spot on at a DC voltage of 1.2V. The vertical pulse was still there too, so where was the obvious fault? In my notes of previous repairs, I had recorded that a TC1408 (M12C) had displayed the same symptoms due to a faulty AN5615 (IC301). I now felt sure this must also be the culprit here so I changed IC301, then IC601 and then IC401 – all without result. Clutching at straws, I then replaced C414, C416, C420 and C650. This made no difference either, so I measured the resistors and diodes in circuit and all read OK. By now I was really frustrated – by rights, I should have fixed this supposedly simple fault and moved onto another job. Unfortunately, I still didn’t have a clue but I was determined that it wasn’t going to beat me. My next theory was that perhaps it was the width of the vertical pulse that was the significant factor. An hour later I abandoned this idea in a bad temper, even though I could vary the number of retrace lines by carefully adjusting the vertical hold. March 2000  67 Serviceman’s Log – continued Fig.2: this circuit section for the National NV-G30 proved to an effective substitute for a National NV-H70A. In particular, it allowed me to identify transistors Q6005 and Q6006, both of which had been “cooked”. I even changed IC501, the jungle IC (AN5435) but the retrace lines were still visible on Channel 7. In the past, I have had similar symptoms due to poor smoothing of the 198V rail to the RGB outputs so I replaced C556 (10µF 250V). When that didn’t work, I checked the screen volts but I was getting nowhere fast. I had obviously overlooked something but I couldn’t think what it was. All I could do now was go back over what I had done and recheck my work. A previous fault I had encountered with another set with no colour had turned out to be a leaky diode (D602) which had dis­torted the horizontal pulses to the burst gate. I remembered that, at the time, I couldn’t measure this in-circuit to deter­mine its leakage. Therefore, I felt it would be a good idea to measure D402, D601 and D602 out of circuit with the ohmmeter on the 100kΩ range. And it was when I measured D601 that I found the answer – there was significant reverse leakage. Replacing it with a 1N4148 fixed the fault and the vertical pulse on its cathode was double what it was previously. I richly deserved the $100 I charged 68  Silicon Chip for this job but when will I ever learn? National NV-H70A VCR Mr Peterson’s ageing National NVH70A VCR came into the workshop with the complaint that a tape was stuck inside. He neglected to mention that the set was otherwise dead but on removing the covers the cause was fairly obvious. F1102, a 2A fuse, was open circuit on the UNREG 18V rail that also supplies several other rails: 12V, 7.2V, 6V and 5V. Replacing the fuse restored all functions and I could do whatever I wanted with the tape. Next, the phone rang and I had to deal with an enquiry that took some time. Suddenly, towards the end of the conversation, I began to smell burning. I wound up the call as fast as I could and started sniffing around for the source. And when there is a lot of electrical gear spread out and all switched on, it can be hard to trace the source of a smell. Fortunately it didn’t take long to trace this one; it was coming from Mr Peterson’s VCR! Although it had been doing all its tricks minutes earlier, the tape was now firmly stuck inside and wouldn’t budge. Sniffing carefully, I traced the smell to two transistors on the right­ hand side of the motherboard. These had become so hot that they had un­ soldered themselves and fallen out of the PC board so that they were now resting on the bottom of the cabinet (the PC board is horizontal, with the wiring pattern on the top and the compon­ents underneath). Pretty neat trick, I thought – at least I wouldn’t have to unsolder them. But that was the easy part. After scooping them up, I was faced with the problem of identifying them; they had been carbonised and I didn’t have a circuit diagram. My approach was to find a similar National Panasonic cir­ c uit – one with the same microprocessor chip set used on the main CBA (Circuit Board Assembly VEP03309). That way, I would have a good chance of identifying the transistors used. The best substitute circuit I could find was for the NV-G30 model. The clue was IC6001, an MN15342VEB, which is used in both units. And the transistors in this part of the NV-G30 circuit, Q6005 and Q6006, turned out to be 2SB790s, which are general-purpose PNP transistors. Before replacing these, I checked the circuit board for burn marks. The two transistors are controlled by the micropro­cessor, with Q6006 switching a regulated +12V rail via the Record Safety Switch and Q6005 driven by pin 43 D-REC (Delayed Record). Similarly, Q6008 is driven by pin 42 DA-REC (Delayed Audio Record). Apart from the two transistors there was no other apparent damage so why did they get so hot and fail? This wasn’t easy to deduce but with the common denominator being the regulated 12.3V supply, the source of the problem had to be in the power supply itself. Access to the power supply is not easy, with a lot of short leads and metalwork in the way. Once it was out, I decided to adopt a blanket approach and replace all 10 electrolytics with EXR 105° types. It was a fiddly job but most of the capacitors looked pretty sorry anyway, especially C1101, C1107 and C1104. Finally, I left it on soak test and I’m pleased to report that there were no more pyrotechnics. Mr Peter­son is once again a happy man though I did advise him to get a new one if it played up again. He would be surprised how cheap they are now compared with what he paid some 13 years ago. Mitsubishi stereo TV Mr Crane requested a service call for his 1992 59cm Mitsubishi CT2584AS (ASV59S/AS2 stereo chassis), which he said had an intermittent crackle in the sound. Initially, I was emphatic that an intermittent fault would have to be fixed in the workshop but he was equally emphatic that it really wasn’t intermittent and that because he was 76 years old, he couldn’t possibly get it down the stairs of his duplex and into his car to deliver it to me. I saw his point and relented. When I arrived the next afternoon, I switched it on and sure enough, there was a faint crackle on all channels. He assured me that it was normally a lot louder. I knew I was going to regret it later but I decided to take a quick look anyway, in the hope that I could fix it on the spot. I pulled the set away from the wall, removed the back and started tapping around, looking for a dry joint or a bad connection. After a few minutes, I emerged from behind the set having achieved nothing except that when I looked at the picture it was line tearing. “Oh yes”, said Mr Crane, “it does that too sometimes”. I secretly sighed with relief – I hated to think I might have been held responsible for this “new” problem. “Well”, I said authoritatively, “that settles that, I will have to take it back to the workshop”. I lent him a portable set and with immense difficulty finally got his set into the car and back to the workshop. The set must have enjoyed the ride because when I switched it on, all the faults had cleared and the picture was excellent. Such is life, I thought and left it to soak test. The crackling in the sound refused to come back but the line tearing did occasionally. I took the chassis out and reworked the soldering for any potential dry joints – there were none that were significant. I also noticed that the heatshrink tubing on some of the electrolytics had peeled back. Initially, I replaced three capacitors in the power supply secondary – C917 and C453 on the 28V rail to the vertical output IC (IC451) and C920 on the 15V input to the two 12V regulators (IC902 & IC103). I also replaced C552 (1µF 160V) which connects to the line driver Q551/T552. The two significant capacitors were C917 and C920 as they smelt “fishy” when I unsoldered them. They had spilt their electrolyte and this had attacked the copper tracks on the board. Anyway, this finally fixed the line tearing but I didn’t know what to do about the sound. I left it on soak test for another two weeks before Mr Crane finally demanded I return it. Reluctantly, I agreed but didn’t have much faith in the long-term reliability of the set and told him so. The set bounces As I expected, Mr Crane was back on the phone just two months later to let me know that didn’t last long. And despite my previous explanations, he seemed to think that it was all my fault (which I also expected). I called the next afternoon and listened attentively to the elusive crackle. This time I established that it wasn’t due to interference from an DON’T MISS THE ’BUS www.siliconchip.com.au SILICON CHIP’S Do you feel left behind by the latest advances in com­ puter technology? Don’t miss the bus: get the ’bus! Includes articles on troubleshooting your PC, installing and setting up computer networks, hard disk drive upgrades, clean installing Windows 98, CPU upgrades, a basic introduction to Linux plus much more. ORDER NOW: Use the handy order form in this (02) 9979 5644, 8.30-5.30 Mon-Fri with your 132 Pages 9 $ 95 * ISBN 0 95852291 X 09 9780958522910 09 9 780958 522910 COMPUTER OMNIBUS INC LUD ES FEA TUR E LIN UX A collection of computer features from the pages of SILICON CHIP magazine o Hints o Tips o Upgrades o Fixes Covers DOS, Windows 3.1, 95, 98, NT NO AVA W DIRE ILABLE C SILIC T FROM ON issue or call just $ CHIP 125O INC credit card details. RT P&P external source. Instead, it sounded as though it was due to arcing somewhere inside the set. With the speakers switched off I couldn’t hear it at all but I thought it might have been arcing internally inside the flyback transformer or even the deflection yoke and that the interference was finding its way into the audio chain. I disconnected the yoke momentarily and it wasn’t that. I also unplugged the CRT socket in case it was arcing inside the tube guns but it wasn’t from that either. Finally, I decided that it was the flyback transformer that was the cause of the problem and told Mr Crane that it would be expensive to replace. It was time now for him to put up or shut up. He decided to invest in a new transformer. It took the set back to the workshop to wait for a new flyback transformer. And although I had sounded confident, it was really only an intelligent guess. I knew that if this didn’t fix it, I would have to keep at it until I had cracked it. Finally the new part arrived. I unsoldered the old one using a solder sucker but I ran into problems when I tried to remove the EHT final anode/ ultor cap to the tube. The type of rubber that Mitsubishi uses is quite hard and resilient compared to others and getting a screwdriver underneath it was difficult. Then I had to push in the side clips that hold it on to the tube before one side came away and I finally got it off. The reason I had problems removing it was that it was extremely rusty underneath, with heaps of fine brown-red rust powder everywhere. This mystified me, as there was absolutely no sign of rust or water damage anywhere else but it got me thinking – was this the cause of the invisible arcing and crackling in the sound? I cleaned up the rust with a wire brush and CRC 2-26 and fitted the new flyback transformer – I didn’t have the time or patience to refit the old one and because of the intermittent nature of the fault, I felt that the result would have been inconclusive. As additional insurance, I replaced all the electrolytics feeding the audio output stages and reworked the sound module. Once again I soak tested the set before returning it to an anxious Mr Crane. It is now over three months since it went back so I feel confident that it has really been fixed. SC March 2000  69