Fullscreen HTML5Med Res (??MB)HTML4

Another song about Sam Following last month's introduction to Samsung TV sets, I have another story involving this brand; a really rare fault and for several reasons, an elusive one to boot. And from the Apple Isle - well, all is not apples! As with last month's Samsung stories, this one involved a CB515F from a local motel. And this meant that I had less background to the problem than if the set had come from a private home. This is a serious problem with motel sets, as noted last month; it is seldom that one person watches the set for more than a few hours one night usually - and if they complain at all, this is often forgotten by motel staff or written off as incompetence on the part of the guest. Anyway, this story started when the manager complained that one of their sets was "hard to tune - or hard to keep in tune". Given such meagre information, I could only suggest that he bring the set in when business was a bit slack. In fact, it was several weeks before anything happened but the set turned up eventually. I set it up, switched to each channel in turn and checked the tuning. They all responded more or less normally, except that the tuning adjustment seemed to be marginally more critical than on other versions of the same model. But it was only an· impression and I couldn't be certain. The truth is, sets which use trimpots to control a voltage to a varicap tuner system are fairly critical to adjust, particularly on the UHF bands. If it were not for good AFC systems, which take over where the manual adjustments leave off, setting up would be almost impossible. Anyway, having set up the channels, I closed the door in front of the adjustments - which switches in the AFC circuit - and switched through all the channels again as a final check. All seemed well so I pushed the set to one side and left it playing. It played perfectly for the rest of the day and for most of the following day. Then, late that day, I decided to go through the channels again. And this was my first sighting of the problem; one channel was off tune, having obviously :,,, -- --- --- -- ---- --------- --------- - _, - - - - - - - - - - - - - - - - - - - - - - - - - - - """ - __ 1 : ,, ,, ., ,: PWB-MAIN-BOARD ..... M, ,.. JCllt Fig.1: channel selector and tuner circuit of the Samsung CB515F. The selector is at left while the tuner, TU001, is immediately to the right on the main board. Resistor SR105 and zener diode SD101 are near the top of the main board section. 44 SILICON CHIP l dritted outside the range of the AFC system, and had lost colour. I retuned it and it seemed to hold, at least for the moment. This behaviour was then repeated several times during the next few days and not just on the one channel. In fact there seemed to be no pattern to the problem; it came and went at random. And even those channels which held under AFC control appeared to have drifted when this was switched off. Likely suspects Well, at least I'd seen the fault and knew what I was seeking. There were two likely suspects: either the tuner (TU00l) or the selector unit with the channel selector switches, trimpots, etc which delivers the signal control voltages to the tuner. Both have given trouble in the past in a number of brands and of the two, I tended to favour the selector unit. Just why these give trouble is something of a mystery. They consists of two parts: a PC board carrying the channel selector switches plus the various other user controls (volume, contrast, brightness and colour), and a block containing the pots, etc. It is this block which usually develops a fault and since it is a sealed, non-repairable unit, it is normally discarded without the exact fault being pinpointed. My guess is that they suffer from either internal leakage or faulty (dry?) joints. Anyway, the first step was to determine which of these two sections was the likely culprit. This was done by monitoring the tuning control voltage at pin VT on the tuner. If this remained steady while the tuning drifted, then the tuner would be at fault. But if it drifted the selector would be suspect. Although this is fine in theory, such monitoring can be a time consuming procedure in the ordinary way. Fortunately, I have a Fluke 87 multimeter and this features a monitoring facility. The existing voltage is used as a reference and the meter then detects a rise or fall from this value, and gives an audible beep. It's worth its weight in gold for problems like this. Anyway, it quickly clarified the situation - the voltage to the tuner ~,," SE,S WH\Ct-\ use: 'TR \~'PO,S -f"O CON""rROL A VARl CPtP TVNE.-R 'S ¼'STE:M A.'R~ 'F~\'RLY C'R\1\CAL- TO AD:!"u'S, ""., was varying. This virtually ruled out the tuner and cast suspicion on the selector block. Since the set was still under warranty, I simply ordered a new unit from Samsung. This arrived in a couple of days and I lost no time in fitting it. The only snag was it made no difference. So was there a faulty batch and had I been saddled with a second faulty unit? Stranger things have happened but I needed help. I repacked the new unit and returned it to the service department with an explanation of the problem and a request that it be checked. The response was a telephone call a couple of days later from one of the technicians, who insisted that there was nothing wrong with the unit. And while I was trying to digest the full implications of this, he went on to suggest that the fault was probably in the IF section. I wouldn't buy that. Although not impossible, there was the varying voltage to the tuner which could not be ignored and the fact that the drift occurred on individual channels. When I raised these points, the technician tended to brush them aside. As far as he was concerned, I should look at the IF section. There seemed to be little point in arguing; that was his idea and he was stuck with it. More to the point, I was on my own. But I had to concede that, if there was nothing wrong with the replacement unit, then the chances were that the original unit was also in the clear and I had to look further afield. My problem now was that the replacement selector was back at Samsung and all I had was the original. But luck was with me. Almost immediately, another CB515F came in for one of the routine faults described last month. FEBRUARY1990 45 SERVICEMAN'S LOG -CTD That was soon fixed and I had a golden opportunity to try another selector, which I knew was OK. It was a simple job to swap the units and that settled it; the fault was just as much in evidence as before. This was gratifying technically but a bit of blow to my pride; I'd been so sure that this was another selector fault that I had hardly considered any alternatives. Now I was forced to. The most likely alternative putting aside suggestions that it was the IF strip - was the supply rail to the selector. According to the circuit, this is derived from the main HT rail (125V) via a 12k0 2W resistor (SR105) and zener diode SDl0l (UPC574) which delivers a nominal 33V to the selector. The exact voltage isn't critical - just so long as it is stable. So how stable was it? A convenient check point was at plug and socket combination SN0l which feeds the selector, at the pin marked 33V. But the reading was nothing like 33V; it was 70V plus and quite apart from the main fault, was clearly the reason I had found the pot settings unduly critical. pick this up. The set would work, all channels could be tuned, and only ari astute tester would pick the marginally more critical adjustment. So out it went into the field. Fortunately, I had a suitable zener diode on hand and this gave a rock steady 33V rail. Naturally, I had to retune all the channels but that done, the set was literally better than new. So the technician at Samsung was both wrong and right. He was wrong about the IF system but right about the selector. When I contact him again, I'll put him in the picture. Who knows, there might be a few more sets out there with the same problem. But enough of my problems here's the latest effort from J.L. in the Apple Isle. How to make work A moment's carelessness · can cause hours of unnecessary toil. This happened with a Thorn 9503 but it could equally have been any AW A or Thorn "G" chassis. The initial complaint was bad horizonatal sync and I tackled the job in the customer's home. The picture was weaving from side to side What's missing? I wasn't really familiar with this part of the board layout but I spotted the 12k0 resistor easily enough. I checked it as a matter of course and it was OK. I couldn't find the zener diode, though. I scanned the board a couple of times without success, then happened to glance at the board in the other set. And there was the zener, not far from the 12k0 resistor, in a TO92 package. Closer inspection of the faulty board revealed a couple of holes and the marking "SD101" in small print - but no zener. What was more, an examination of the copper side of the board confirmed what I already suspected; there never had been anything connected to that part of the board. In simple terms, somebody omitted to fit the zener diode during board assembly and it is unlikely that there was any test that would 46 SILICON CHIP TETIA TV TIP Kriesler 37-104 (Sharp 9C140 chassis) Symptom: Set hiccupping, just like earlier Kriesler and Philips sets with a shorted line output transistor. This one shows some leakage to ground but it is one way only and is due to an internal diode - the transistor is quite good. Cure: R644 (2.20 ½W) open circuit. This resistor feeds the 115V rail to the line stage and when it opens it takes the line load off the power supply. This supply hiccups with over-voltage, not over current as do most other Krieslers. TETIA TV Tip is supplied by the Tasmanian branch of The Electronic Technicians' Institute of Australia. Contact Jim Lawler, 16 Adina St, Geilston Bay, Tasmania 7015. like a hula girl in full swing. Sometimes it broke up completely; at other times it was almost stable with only the slightest flicker. My first thought was that it was another manifestation of the "edge connector gunk" used on these sets. I have found an extraordinary number of different faults that can be cured just by cleaning the edge connectors of the various plug-in boards. I pulled each of the boards and cleaned their contacts and the sockets. This seemed to be enough at first, because the set came good and played perfectly for the rest of the day. But that night the customer called to say that the set was as bad as ever. This time the problem showed itself to be mechanical. It would shift between "weaving" and "breakup" each time I tapped the board, which seemed to indicate a broken joint of some sort. In this model, the sync separator is on the video/chroma output board so I pulled that board and gave the contacts an extra good scrubbing, just for luck. I didn't have any (luck) though; the fault was as bad as ever. I pulled the board again and began a detailed search. The first thing I looked at was the sync separator transistor, Q401. A multimeter test showed the transistor to be OK but while I was checking the board from the copper side, my fingers on the component side felt something that wasn't at all right. In fact, I knew what I had found without even looking. It was a old style half watt resistor that had once been very hot. It had swollen in the middle and split along its length. When I traced its tracks, I found that it was the feed resistor to the emitter of the sync separator, R405 . Its value is nominally 1000 but was in fact varying between a few thousand ohms and infinity. It took very little time to change the resistor and replace the board for a test run. Only I wasn't careful enough and my carelessness cost me dearly. A loud "psst" The set was standing against the wall in a fairly bright room. However, the back of the set was in shadow and I couldn't see all that well as I pressed the board into its socket. I have fitted these boards many times before and wasn't particularly worried. But I should have been more careful because when I switched on, there was a "psst" from inside the set and then nothing. I realised straight away that I hadn't installed the board correctly. When I turned the set around into better light, I saw that it was sitting high in its socket and more importantly, it was about half the width of a contact out of line. At least some of the contacts were touching adjacent pads as well as their own. I refitted the board, properly this time, but no joy. There was a faint raster and a quiet hiss from the speaker but nothing else. Whatever I had done, it was a workshop job from here on. Back at the shop, my first test was to see if there was any video information coming into the video/ chroma board from the IF board, PCB-HF. I didn't expect anything because of the lack of sound so I wasn't really disappointed when I found no signal. Next, I pulled plug FA, which is the input to the IF board from the tuner. I tried injecting a 1MHz square wave from a small function generator into the IF. In many sets, the harmonics from the square wave will pass through a functioning IF system and give an idea of how well it is working. In this case there was only a click as the contact was made and then nothing. Either the strip was totally dead or the harmonics weren't strong enough. It was time for some better test gear. Among my homemade test gear is a portable TV set with a modified link between the tuner and IF board. It enables me to extract IF direct from the portable' s tuner or to feed the portable with IF from a suspect tuner. It's an I :;~ It ~ ©• g~ ~ <. r,.- -- ;, .1 ---- : ~ o l~~~-;i~,~ s~ ~ g8, r=-:-...:.·1 ~L. ___J ::, ~ <t g~ ~;g > - - - - ---+-'V\1\----+-'jf----t i~ C ! ! ~ :~· ~ ~.. r- -~~~-·~ -~--J - ---'J .';.~~~~~~~~~"!.~~~~ . Fig.2: video IF and sync separator circuitry for the AWA "G" chassis. The tuner input is at top left, the video IF IC below it, and the IF and 1st video transistors to the right. The video/chroma hoard and sync separator are at extreme right. FEBRUARY1990 47 This signal then passes to Q101 (a 2SC383 HF transistor) and then to the video detector, in can T108. The first video amplifier, Q102 ideal tool for this type of inAlong with audio, H and V sync, (2SC710), is also on this board and vestigation. and an adjustable negative bias from it the signal passes through In this case, a signal from the supply, it comes close to delivering pin 3 on plug FD to the video/ portable's tuner would not activate every signal needed to test any part chroma output board. the big set, nor could the big set's of a TV set. I had to probe my way through tuner drive the portable. It looked this signal chain in an attempt to Dead IF strip as though both the IF strip and the locate the trouble. In the event, the In this case, I needed the variable tuner had suffered a major blowup. multiple outputs of the TV analyser Way back in the closing days of IF output. . I fired up the old proved quite valuable although, analyser, connected its output to monochrome TV, I had bought a looking back, I realise I could have piece of equipment called a ''B & K the IF input and switched on. found the fault with nothing more Television Analyser". It was Nothing. I adjusted the input level than a simple multimeter. from 500µV to lOmV. Still nothing. I designed for testing valve type Another thing that was working monochrome sets but is still useful swept the IF from 30MHz to against me was that bloke Murphy. 40MHz. Still no response. The IF for working on modern colour sets. He made sure that I started at the The analyser provides a monostrip was totally dead. wrong end of the chain and wasted chrome video pattern on a variable This IF strip consists of one IC an hour in fruitless poking and IF, or on any selected VHF channel. (IC101) and one transistor (Q101). prodding. It also supplies variable amplitude IC101 is an M5183P which accepts My first test was of the 20V rail positive or negative video at an adinput on pin 1 and puts out a proto the IF board and the 12V rail justable horizontal sweep rate. cessed signal between pins 7 and 8. derived from it. Both were correct and the specified voltages were present on pin 11 of the chip and the collectors of both transistors. Next, I injected IF into pin 1 of the IC. This produced no response, so I moved the probe to test point 12, the output (pin 8) of the chip. This did produce a result but not quite what I expected; an extraordinarily loud burst of sound from the speaker but no picture. (I must learn not to leave the volume control flat out when working on this kind of fault!) This was a bit confusing at first because the sound is usually taken off with the video at the video detector. This set is different in that the sound IF is generated in a separate detector attached to the collector of the video IF amplifier transistor. At first I considered that the video detector diode might be shot as this would account for the loss of picture information. I found the . diode to be undamaged, then realised that this had been a silly deduction. If this diode was the cause, then the sound would have been normal because of the two detectors. Next, I injected video at test point ~WA.'1 'jsACK lN 1'\-\~ CLOSING "DA'-(S 15, at the input to the 1st video amplifier. There was still no sign of OF IV\ONOC-H'R0IV\E:. TV, 'I HA"C> ts0UGl-\".,a picture. Finally, I fed the video to p.._ '°Pl!ZCE:" 0~ E"Q~W'('(\E..l\.lT CP-.LLED pin 3 on plug FD, the input to the A ":B&\<. TELEV\'StON AN~LYSE:."R'~ •• video/chroma board. (This is where SERVICEMAN'S LOG -CTD 48 SILICON CHIP I would have started if Murphy had not poked his nose into my affairs). Immediately, the screen lit up not with any recognisable picture but with some kind of scrambled information that looked as though it could be resolved into a picture. This was better than anything I had seen so far. Transistor checks I pulled out the IF board and tackled the video amplifier transistor. A static test with a multimeter showed no problems it had no leakage and normal junction resistances. The resistors on base, emitter and collector all read close to the circuit values, and I had already checked the collector voltage as correct. So what was going on? I refitted the board and freed some of the cables so that I could do some dynamic measurements with the board clear of the cabinet. And this time I beat Murphy. The first check was at pin 3 of plug FD, which is effectively the emitter of the video transistor. It was supposed to be 6.3V but was in fact only 0.8V. I then went right to the transistor and found that although the collector voltage was correct at 13V, both base and emitter were very wrong at 0.8V each. This was another one of those transistors that passes a static test but shorts when working voltages are applied. A minute later and I had a new transistor in place and a picture, of sorts, on the screen. I say "of sorts" because there was no sign of sync, either horizontal or vertical. And there was no colour either. I had cured the IF problem but still had the original sync troubles, or something very like them. As mentioned earlier, the sync separator transistor is on the video/chroma board and this is not easily accessible when the set is operating. What is needed is an extension lead to enable the board to be operated outside the set. I had recognised this need some years ago and used parts from a junked set to make up my own extension leads. I was showing my handywork to a colleague who surprised me by saying "I've got a full AUDIO TRANSFORMED set of those leads and I've never used 'em. Do you want 'em?" Murphy again! Why didn't I ask before I spent hours making my own? Anyway, I soon had the board out and powered up. First, I used the CRO to check the video into the noise canceller Q205 and then into the sync separator. Both were perfect, as near as I could judge. But the output of the sync separator showed nothing. Even with the CRO turned up to maximum there was only a faint ripple on the trace and nothing resembling separated sync. Again, a static check of the transistor showed nothing wrong but it was a different story when it was checked dynamically. The failure was exactly the same as in the case of the video amplifier transistor. And because this was in a pulse type circuit, it was harder to prove that the transistor was faulty. The emitter voltage was just on 19V as shown on the circuit and the base was also around 19V because the sync separator stage is normally biased off except when a sync pulse is present. The sync pulse continued on page 91 TV TEST EQUIPMENT (AUSTRALIAN MADE) SHORTED TURNS TESTER Built in meter to check EHT transformers including split diode type, yokes and drive transformers. $78.00 + $3.00 p&p HI-VOLT AGE PROBE Built in meter reads positive or negative 0-50kV. For checking TVs, microwave ovens. $84.00 + $5.00 p&p. 1 ,.i TELE-TEST TUBE TESTER & REJUVENATOR Removes cathode grid shorts & rejuvenates picture tube. Suitable for all colour tubes. 12 months warranty. $420.00 + p&p. FOR PROFESSIONAL, OEM OR AMATEUR Broadcast quality audio transformers ex-stock More th an 70 standard types available Fast prototyping service for non-standard types Comprehensive data available on request Locally manufa ctured Competitively priced HARBUCH ELECTRONICS PTY LTD 90 George St .. HOR NSBY NSW ~077 Phone (0~)476-5854 NEW PRODUCT DEGAUSSING WAND: A must for all workshops. Efficient design, strong magnetic field, low heat, double insulated with momentary on/off. $75.00 + $1 o p&p. ALSO TUNER REPAIRS FROM $17.00. Cheque, Money Order, Bankcard or MasterCard -:r.V.TuNERs) 216 Canterbury Road, Revesby, NSW 2212. Phone (02) 77 4 1154 FEBRUARY1990 49 The Serviceman Know About Capacitors; Telephone Bell Monitor/ Transmitter; 32-Band Graphic Equaliser, Pt.2; Led Message Board, Pt.2; Fluke's New 80-Series Multimeters. May 1989: Electronic Pools/Lotto Selector; Synthesised Tom-Tom; Biofeedback Monitor For Your PC; Simple Stub Filter For Suppressing TV Interference; LED Message Board, Pt.3; Electronics for Everyone - All About Electrolytic Capacitors. June 1989: Touch-Lamp Dimmer (uses Siemens SLB0586); Passive Loop Antenna For AM Radios; Universal Temperature Controller; Understanding CRO Probes; Led Message Board, Pt.3; Coherent CW - a New Low Power Transmission Technique. July 1989: Exhaust Gas Monitor (uses TGS812 Gas Sensor); Extension For the Touch-Lamp Dimmer; Experimental Mains Hum Sniffers; Compact Ultrasonic Car Alarm; NSW 86 Class Electrics; Facts On the PhaxSwitch - Sharing Your Phone Line With A Fax Machine. August 1989: Build A Baby Tower AT Computer; Studio Series 20-Band Stereo Equaliser, Pt.1; Garbage Reminder - A 7-day Programmable Timer; Introduction to Stepper Motors; GaAsFet Preamplifier For the 2-Metre Band; Modern 3-Phase Electric Locomotives. September 1989: 2-Chip Portable AM Stereo Radio (uses MC13024 and TX7376P) Pt.1; Alarm-Triggered Telephone Dialler; High Or Low Fluid Level Detector (uses LM1830N); Simple DTMF Encoder (uses Texas TMC5089); Studio Series 20-Band Stereo Equaliser, Pt.2; Auto-Zero Module for Audio Amplifiers (uses LMC669); A Guide to Hard Disc Drives. October 1989: Introducing Remote Control; FM Radio Intercom For Motorbikes (uses BA 1404 and TDA7000) Pt.1; GaAsFet Preamplifier for Amateur TV; 1 Mb Printer Buffer; 2-Chip Portable AM Stereo Radio, Pt.2; Installing A Hard Disc in the PC; A Look! at Australian Monorails. November 1989: Radfax Decoder For Your PC (Displays Fax, RTTY and Morse); Super Sensitive FM Bug; Build A Low Cost Analog Multimeter; FM Radio Intercom For Motorbikes, Pt.2; 2-Chip Portable AM Stereo Radio, Pt.3; Floppy Disc Drive Formats & Options; The Pilbara Iron Ore Railways. December 1989: Digital Voice Board (Records Up To Four Separate Messages, Uses Texas TMS3477NL and 256K RAMs); UHF Remote Switch; Balanced Input & Output Stages; National Semiconductor LM831 Low Voltage Amplifier IC Data & Applications; Install a Clock Card In Your PC; Index to Volume 2. January 1990: Service Tips For Your VCR; Speeding Up Your PC; Phone Patch For Radio Amateurs; High Quality Sine/Square Oscillator; Active Antenna Kit; The Latest On High Definition TV; Speed Controller For Ceiling Fans. Note: November 1987, December 1987 & January 1 988 are now sold out. continued from page 49 drives the base more negative, turning the transistor on. In fact, the transistor was leaking internally with normal operating voltages applied, and was turned on permanently. Under these conditions, the collector voltage fell from the correct figure of 4.1 V to something like 0.1 V. I fitted a new transistor and found that the set now worked perfectly; stable sync, full colour and all. Post mortem But why had the set been dead from the tuner on? That 6.3V at the emitter of the 1st video amplifier also supplies a bias to the video IC at pin 6. So the whole IF strip was inactive when the fault was really in the video amplifier. And the tuner was inoperative because it needs an AFC voltage which is developed from a signal provided by the IF chip. Which brings me back to the point made earlier. If I had been a little more careful when replacing the board in the first place, then none of this story would have eventuated. ~ Bose noise cancelling headphones for obvious reasons, maintained a tight seal to the pilot's head. The ear surround seal in the Bose headset uses a combination of silicone gel and soft foam which follows the contours of the head to create a better seal, with only a slight amount of headband pressure. Even without the noise-cancelling electronics, the Bose headsets still boast better passive noise attenuation than many of the ordinary types on the market, especially at low frequencies. Specifications Each earpiece has its own independent noise-cancelling system, allowing for both mono and stereo use. The sound pressure servo system has been built using surface-mount technology, allowing Circuit Notebook - from page 11 it to fit inside the earcup. The impedance of the earpiece is 1500 and the system has a frequency response of 100Hz to 6kHz. The power supply required is 150mA at 11-16V or 22-32V DC. Applications A number of helicopter and aircraft manufacturers are now offering Bose ANG headsets as an optional feature but Bose see the product as having wide application in any situation where people have to work for long periods while subjected to high levels of noise. And who knows, maybe in the future your W alkman will come with noise cancelling headphones. l!t] Acknowledgement: our thanks to Bose Australia Inc, for their assistance in preparing this article. continued from page 26 represent a variable inductance. By varying dual gang pot VR1, the resonant frequency of the inductor is changed and so the circuit can vary the frequency to be boosted or cut. The filter is an LC type based on the .001µF capacitor and the variable inductance. Note that VR1 is wired so that when VR 1 a increases its resistance, VR1b reduces its resistance. As shown, the centre frequency can be tuned from around 32Hz to 17kHz and the available boost and cut is about ± 12dB. The output of the circuit is taken from pin 7 of IC1b via a 1k0 resistor and 1µF capacitor. Darren Yates, French's Forest, NSW. ($30) (Editor's note: while the circuit shown here is run from a single supply, there is no reason why it could not be modified to run with balanced positive and negative supplies). FEBRUARY1990 91