Fullscreen HTML5Med Res (??MB)HTML4

Sounding out a video recorder For the most part, TV equipment sound sections tend to be taken for granted; they don't give a great deal of trouble and we don't think much about them. Which is a far cry from the pre-TV days when restoring sound was the main exercise. But modern sound circuits do fail and when they do, they can be quite tricky. machine on that occasion. I had it in the workshop for several days and put it through many recording and replay cycles, without the slightest sign of any problem. The only clue I had was a faulty tape brought in with the machine and this carried a background of the previous sound track at a reduced but still annoying level. Nothing I could do would produce the fault. And since there seemed to be an urgent need for the school to have it back, I simply gave it a routine service, explained the situation and advised them to give it another try and bring it back if or when the fault was more predictable. And that was the last I heard of the device until a couple of weeks ago. I can only assume that it This story concerns a National NV-300 video tape recorder. It belongs to a local private school and had first come to me with the sound fault some two years previously. The complaint then was that it would not always erase the previous sound when a new recording was made, although there seemed to be some doubt in the mind of the staff member as to how serious it was. Some users complained that there was no erase action at all, while others described the effect as a faint background of the previous sound. The NV-300 is one of the older top loading machines but has been a very reliable model. It is also easier to work on than some more modern designs. I didn't have much luck with the • R4038 5~0 TO VIDEO AUDIO OJT l4001 GNO . 470uH BPGOO; ®c,026 Q033 - ____ i I - TP-i001PLAY . 0.5 VI0.5msec . div . : '/ FE H(AO P6005- l GNO P6001-1 __-_ , /,/ . r ' TP<:001 AEC. 0.5VI0.5 msec. d1v. Fig.1: bias/erase oscillator circuit for the National NV300 video recorder. Although a simple arrangement it developed a very tricky fault. Note the missing chassis symbol on pin 8. 34 SILICON CHIP operated normally or, at least, acceptably during that time. Then it was back, brought in this time by another staff member who wasn't even aware that there had been a previous problem. This time the complaint was rather more specific. Granted the failure was still unpredictable but at least when it occurred now, it was complete; there was no attempt to erase the previous signals. I set the machine up at one end of the bench, connected to a monitor and made a half-hour recording over a previous recording. The machine behaved perfectly. So it was going to be one of those faults. I went through this exercise, at approximately half-hour intervals, over the next couple of days, without even a hint of trouble. Then just as I was beginning to wonder whether there was some environmental factor involved, the fault appeared. The previous sound was there at full blast and, just as important, there was only a shadow of the new sound which should have been recorded; just some brief distorted bursts of the louder passages. Well, at least I now knew that the fault was real enough. I had not delved very deeply into the machine at this stage, simply, removing the top cover but avoiding moving anything else for fear of disturbing the fault and causing it to lie doggo. But now I had to go on the attack. I began by removing the bottom cover which provides access to the copper side of the main board. Access to the component side of the board involves removing the front panel and unclipping the sub-boards carrying the clock display and various operating buttons which are on flexible leads. About six screws secure the main board and removing these allows the board to be swung up and propped open. ~f\~'t \ , •• 1-r SU..Or-.l6S TO A 1-OCAl- PR\VAT& SC~OOl-- • .,. The section I was seeking was the audio circuitry and in particular, the bias and erase oscillator, the circuit of which is reproduced here (Fig .1 ). It was fairly obvious that the system was suffering from loss of erase signal but remembering that there had been little of the new signal recorded, it appeared that it was suffering loss of bias signal as well. And since both signals come from the same oscillator, it seemed reasonable to suppose that this was where the trouble lay, rather than, in a lead or plug feeding the erase head. Circuit details As can be seen from the circuit, the oscillator is a simple arrangement; one transistor, two resistors, three capacitors, a choke and a tapped winding forming one side of a transformer, T4001. I decided to put this section under observation and the easiest place to connect the CRO was at pin 1 of plug BP6005, which connects to pin 7 of the secondary of T4001. This tapping supplies the "FE HEAD", or full erase head. This was not the section of interest but it was a convenient monitoring point. The audio erase head is supplied from pin 9, while pin 5 supplies the bias for the audio recording head, this being adjustable via variable resistor R4048. One other point to note is that there appears to be an omission from the circuit, in that pin 8 is shown floating, whereas it is a chassis connection and a vital one. With the CRO hooked up, I set the machine going again in the record mode and observed that there was approximately 45V p-p at the monitoring point, at about 63kHz. And as before, the machine performed perfectly for the next few days. · However, I adopted a routine of recording for periods of about 30 minutes, then shutting the machine down long enough for it to cool, before making another test. My hunch was that the fault was most likely to occur at start-up from cold, rather than while running. And in fact, it proved to be. Suddenly, at one start-up, I had no oscillator signal at pin 7 and a quick check with the CRO at other points gave the same result. But the mere act of making those measurements triggered the circuit back into oscilla- tion. It came up part of the way first, dithered for a second or so and then rose to normal level. From then on it behaved more or less normally but having been alerted, I observed the pattern more closely. Sure enough, it was exhibiting a dithering characteristic whereby the amplitude would vary over a range of about five volts. Then suddenly, it began dropping lower and lower, down to about 10 volts, then failed completely. · Well, that seemed straightforward enough. All I had to do was find out why the oscillator was intermittent and with so few components involved, I didn't imagine it was going to be particularly difficult. My first suspect was the transistor, Q4014. I pulled it out and replaced it but it wasn't long before the machine began dithering again and I knew that I had drawn a blank. That seemed to leave only the two resistors, R4049 and R4050, and the three capacitors, C4010, C4025 and C4024. The resistors were fairly easily checked, in situ and came up spot on. What was more, they showed no indication of any intermittent characteristics in spite of some vigorous pushing and prodding. The capacitors were not so easily checked and I was on the point of pulling them out and replacing them when it occurred to me to try some freezer spray on them. This had no effect on C4010 or C4025 but brought a swift reaction from C4024, the 10,-iF electrolytic; oscillation stopped immediately. So, I was getting somewhere at last. I pulled the capacitor out and replaced it but this produced a strange result. This capacitor no longer responded to the freezer but the original fault was still there! That put me almost back to square one. I pulled out the other two capacitors, tested them, found them to be well within tolerance but replaced them anyway. This achieved exactly nothing; the fault still persisted. Rescued by a drunk By now, there wasn't much else left to suspect; just the transformer in fact and then only by default. The snag was that I had no such device in stock, this being the first time one JULY 1988 .35 issued some two or three years previously and how I missed out on it is rather a mystery. The gist of it was that, in the event of unreliable operation of this oscillator, the 4.70 resistor in the emitter circuit of Q4014 (R4049) should be removed. In fact, the suggestion was that this resistor should be removed anyway, as a routine mod whenever this model was encountered. SERVICEMAN'S LOG A new theory had ever been suspect. But I did have another NV-300 machine on hand an unfortunate victim of a wild Saturday night party during which someone had upset a tankard of beer over it. It's been in the workshop for months and I've been working on it during odd slack periods; the only basis on which any kind of economical repair is possible. In fact, I am close to saving its life and that should make another story. In the meantime I was happy enough to rob the transformer from the drunken machine and try it in the sober one. And that was it. The problem vanished and has not been seen since, in spite of prolonged bench testing and several follow up calls after it went back into service. But that's not the end of the story. For one thing it left several questions unanswered. What was wrong with the coil anyway? Careful testing failed to reveal any obvious faults; certainly nothing in the way of abnormal or varying winding resistance, such as one might expect. It is a small device, apparently of pot core construction, and one thought was that it might have suffered from cracked or chipped ferrite material, a condition which can seriously upset the behaviour of the associated windings. Well, that was one to go on with. 36 SILICON CHIP And what about the freezer upsetting one capacitor but not the one that replaced it? I was still turning these questions over in my mind when I encountered one of my colleagues at a social gathering and during the inevitable shop talk, I related this story. He was suitably impressed but commented that he seemed to recall some kind of modification sheet dealing with cranky bias oscillators in that model. He promised to try to find it and let me know. He rang me the next day to say that he had found it. It had been TETIA TV TIPS HMV 12613 (Braddon) Symptom: No distinct red and somewhat dirty greens. Colour bars look vaguely normal but careful inspection shows red is really a muddy brown . The monochrome picture is perfect. Cure: R541 (4700 0.5W) open circuit. This resistor supplies Vee to IC503, one of two reference oscillator chips in this unusual circuit. In this fault, IC502 (the 8-Y oscillator) probably supplies some drive to the red and green circuits but it's phase is wrong, hence the weak and bad colour. Well, that threw a whole new light on the situation. After mulling it all over for a few hours, I finally came up with what I feel is the most likely explanation. My theory is that the design of the original oscillator circuit was a bit dicey; that is, while it would maintain oscillation under ideal conditions it was only just making it, and was on the verge of dropping out at any time. It was the reference to the emitter resistor which gave me the clue. The use of an unbypassed resistor in the emitter circuit of a transistor oscillator - or the cathode circuit of a valve oscillator in olden days was a favourite trick to improve the oscillator waveform. Because it is unbypassed, it is a simple way of providing a degree of negative feedback, thus holding down the strength of oscillation and reducing the tendency for the transistor (or valve) to be driven to cut-off and saturation, two conditions which seriously distort the waveform. And it can be very effective, the only snag being that it is a compromise arrangement; the negative feedback provided by the emitter resistor is fighting the positive feedback which is fundamental to the oscillator circuit. So, if the idea is carried too far, the whole arrangement can turn cranky. So, was the original oscillator coil faulty or not? It's hard to say. All I know is that it would not work in a circuit which was otherwise fault free. It's possible, of course, that the "fault" in the suspect coil may have been nothing more than a normal spread within otherwise acceptable tolerances. Along with other component spreads it could easily have been the last straw which sank the camel's hump (to mangle a wellknown phrase). UHF REMOTE KEY SWITCH (EA, JAN.87) This proven and reliable unit is *guaranteed* to work. It has countless applications, 40-50 metre range, multiple outputs, and is available at a fraction of the cost of the other equivalent unit. Beware of "short form" kits that include little more than the PCB! <at>;;;~1~ ... ••• IN Tlt£ tl\E.AN"r\W\E., I WAS HAPP'{ "'rO "ROB "n\e:. Ttlt.1-\1\lSFORIV\e:R FROM 11\£ ~UNKeN MACHl>J£ ••• As for the original machine, I'm keeping my fingers crossed. But if I can contrive to get my hands on it again, for any reason, I'll whip that resistor out, just to make sure. In the meantime, I suggest you make a note of that mod which, I understand, also applies to the model NV370. Let's hear it from J.L. To change the scene, here is a story from my regular contributor, J.L. of Tasmania. It is one which is not only intriguing technically but also emphasises that when all else fails, it is important to be able to analyse how a circuit is supposed to work. This is how J.L. tells it. The set was a 63cm National, model TC2652, which was completely dead. However, the owner had observed that it made a brief noise when switched on. When I tried it I found that there was a five second burst of normal sound before the set shut down. This kind of overload is often caused by a faulty tripler and disconnecting it will let the set run normally. Unfortunately, this chassis uses a diode split output transformer rather than a tripler, which rules out this trick. A CRO check of the horizontal output transistor, Q502, showed that it started up OK but died as the waveforms approached full value. And there was not even a flicker of EHT, even while the transistor was running, briefly, at about two thirds its normal voltages. I wired a 60W lamp in series with CHOPPER T801 0801 +111V Our kit includes all the parts needed to make a working link. Even the transmitter battery is included! ON SPECIAL UNTIL END OF THIS YEAR AT ONLY $41.95 Plus $2.50 pack & post For a complete Remote Controlled Car Alarm, combine the UHF Remote Key with our Ultrasonic Car Burglar Alarm (SILICON CHIP, May 1988) or our Shock Detecting Car Alarm Module (ETI, March 1988); or perhaps with both of them. Interconnecting diagrams available from us. For more information, see our advert on page 96. * If upon completing any one of our kits you cannot get it to operate correctly, we will repair it for a maximum fee of $1 0. 00. This offer applies only to properly constructed individual kits which haven't been overloaded in incorrect supply voltages. OATLEY ELECTRONICS ERROR AMPLIAER 0803 8809 Fig.2: simplified circuit of the National TC2652 power supply, as drawn by J.L. T801 and Q801 form a ringing choke oscillator, with regulation provided via Q803 and Q802. 5 Lansdowne Pde, Oatley West, NSW 2223. Phone: (02) 579 4985. Bankcard, Mastercard and Visacard accepted with phone orders. Send mail orders to PO Box 89, Oatley, NSW 2223. JULY 1988 37 e 1'HE.. (\\\f\\111 se., w~s C.OM?L.E."i"EL"'( t>EA'O•••. the collector of Q502 and ran the set for about five minutes. By that time the output transformer was quite warm and giving off a typical sour "brown" smell. I ordered a new transformer. This duly arrived and was fitted but resulted in a set which was even deader than before; there was not even the five second burst of sound. As before, the set would run with the lamp in series with the output transistor but the 111 V HT rail was up to 120V. Without the lamp the voltage shot up to 120V before the power supply shut down. By feeding the set from a Variac I found . that 120V HT resulted from only 100V Philips Tuners ECL 2060 $17.00 Exchange or repair. Three months warranty . Send cheque or money order plus $2.00 pack and post anywhere in Australia. J.V. Tuners, 216 Canterbury Road, Revesby, Sydney, NSW 2212. Telephone (02) 774 1154. 38 SILICON CHIP AC input. Clearly the power supply was not regulating. This power supply is a self oscillating chopper circuit, which controls its ouput voltage by varying the chopper "on" time via an error amplifier. The trouble was that the three transistors and the reference zener diode involved all checked OK. Also, the HT ( + B Adjust) control worked in a seemingly normal way, varying the voltage by about 20V a not unreasonable range. At this point I put on my thinking cap and reviewed what I knew about the problem. (1). The set worked OK when fed with low AC input. This restricted the fault to the power supply circuits. (2). The chopper transistor Q801, regulator transistor Q802, error amplifier Q803 and reference diode D811 were all OK on static test and gave the proper responses under load. (3). The HT ( + B) adjusting trimpot worked over a logical range but at a higher than normal level. Correct HT output should occur with the chopper operating at about 50% duty cycle but in this case it was on for close to 90% of the time. This suggested that the error amplifier was delivering the wrong information. There are not a lot of parts in this circuit but I found its operation difficult to follow. In fact, I could make no sense of it at all until I redrew the circuit in simplified form, omitting every component that was not vital to its operation. Fig.2 shows my simplified diagram. As far as I can see, it works as follows: Q801, T801, D806 and their associated components form a ringing choke oscillator, with voltage regulation via Q802 and Q803. Imagine that Q802 is turned on. This will tie the base of Q801 to its own emitter and thus turn it off. In normal circumstances, Q802 is held in dynamic conduction by error amplifier Q803 using zener D811 as a reference. When the supply rail rises above a preset level, Q803 and Q802 turn on and Q801 turns off. So much for my theory. I already knew that all the transistors and diodes were OK. I also knew that C809 and a number of other capacitors were, at least, not shorted. They might have been open but that would not produce these symptoms. At this point I should have been able to find the fault with just a little logic. With hindsight, I can see it sticking out a mile. In fact, I decided to try an experiment. I removed Q802 and switched on. There was absolutely no difference in behaviour whether the transistor was in or out of circuit. Since it wasn't Q802 it had to be the only other component in this part of the circuit, R808. Sure enough it was open circuit and a new resistor restored normal operation. One final point. Why did I get some output adjustment when the regulator was patently not working? This was the result of Q803 varying its resistance, as part of a divider chain also involving R803/820, R809 and R810. It was only a superficial action but was enough to mislead me in the early stages. Well, that's J.L.'s story and I think we should thank him for the trouble he has taken to set down his analysis of the system and provide a simplified circuit. While this is not the kind of approach we can afford to adopt with every tricky set that lands on the bench, it is one we should be prepared to adopt whenall else fails. ~