Sony AGC tale

[Studio263]

I picked this Sony TV-110UK off the pile at Mikey 405’s TV Junk Swapmeet; I really like Sony’s monochrome sets of this era. Unlike the colour ones the circuitry is very solid and conservative; there is little in the way of experiment or innovation, just sensible high quality television engineering. Because of this, these sets are amongst the most reliable available and despite huge sales little in the way of actual faults are ever encountered. The TV-110UK was known for poor tube life, but apart from that very little goes wrong with them.

The TV-110UK was the single standard replacement for the TV9-90UB, and as well as being for UHF only it also came with a larger tube, 11” instead of 9”. Much of the circuitry is pretty similar but one area where big changes were made was the AGC system, rather than being mean-level as it is in the TV9-306UB and the TV9-90UB it is a proper line-gated system which samples the line sync pulse amplitude rather than the average picture content. This gives a much more stable amount of picture contrast and less change in black level with picture content, although the video channel is still AC coupled. Since this particular set had an AGC fault these stages would be of interest to complete the repair.

Looking at the circuit diagram the principle of operation is easy to understand. D301 is the vision detector and the output of this is buffered by the emitter follower stage Q401. Intercarrier sound is tapped off by T401 but the rest of the circuits use the complete video waveform which is developed across D401 / R404. The video output stage and sync separator are connected via C501 (where the red circle with a 1 in it is) and Q304 is the AGC gate. This transistor is normally biased “off” (note that it is a PNP type) but when a line pulse arrives at the cathode of D302 (where the black circle with the D in it is) it conducts and builds up a voltage proportional to the line sync pulse amplitude across C330 and C331. This voltage biases the first two video IF transistors Q301 and Q302 and therefore sets the gain of the IF section as a whole. The collector current of Q301 also controls the buffer stage Q305, which provides a delayed AGC fed to the UHF tuner. By dividing up the circuit in this way the gain of the IF stages is reduced before the tuner is attenuated, ensuring the best possible noise performance.

AGC faults are rare but they are normally easy to identify. Faults which cause low gain are normally the result of defective transistors either in the tuner or early in the IF strip. With such a fault the set will give little or no results from a weak signal and perhaps a snowy picture with a strong one. AGC faults normally result in the gain being set at maximum so a signal of any strength will overload the set completely, normally resulting in an inverted picture or no apparent signal at all. Reducing the input level may return an almost normal picture as the set’s now fixed gain more closely suits the reception conditions. A pattern generator with a decent attenuator is a useful tool when investigating faults like these but it is not essential, pulling out the aerial plug a little way and seeing what happens is a good substitute. Despite the apparently simple circuitry AGC faults In portables can fool even the best engineers; look at the mess which LLJ got into with a Hitachi in the December 1977 issue of ‘Television’ in one of his best ever sagas.

This TV-110UK produced exactly these symptoms, using the loop antenna an almost normal picture could be obtained from the leakage form our modulator system, only marred by weak contrast and fluttering lines running across the screen. Plugging the lead in directly overloaded the set completely and resulted in inverted video and sync crushing, classic AGC fault symptoms.

My first check was for the presence of line gating pulses since it many sets it is not unusual for the lead from the LOPT to become disconnected somewhere. The presence of -18V at the collector of Q304 confirmed that pulses were getting through, and the reasonably normal picture suggested that suitable video should be present at the base too. The resistors used in Japanese sets of this period hardly ever change value or fail and the non-electrolytic capacitors don’t give much trouble either so there wasn’t any obvious place to start, other than that all the DC voltages around seemed a few volts out from what they should have been. The voltage across R326 was around 1.5V and didn’t change by much with a strong signal and no signal at all, so the AGC system clearly wasn’t working though. My first move was to replace Q304 with a BC557B but this made no difference, all the voltages stayed the same. Perhaps then D302 was faulty, zapped maybe by a spike from the LOPT? Again no, a 1N4004 tacked in also had no effect.

Why were the voltages wrong? Without much conviction I ran around the resistors in the area with a meter but all were perfect, no surprises there. I then looked at where Q304 gets its bias from, since this would dictate the DC conditions under which it operates. Both the video output stage and the sync separator are AC coupled so the DC source can only be Q401, which is biased via the detector circuit by R316, R317 and R319 amongst other things. The voltage across D401 was nearly 5V instead of the correct 3.7V which suggested too much bias but given the stability of the resistors used it wasn’t obvious where this would come from. An open circuit R319 is one possibility, but these components are inside the soldered-on screens on the IF PCB and a pain to get to, can’t be that then! Meter checks at the base of Q401 showed a plausible DC resistance to ground anyway, so this possibility was dismissed. Q401 was then removed and found to be very slightly leaky, ah ha! This would certainly account for the voltage errors noted so a BC547B was fitted (I’m always short of Japanese transistors…). This returned the voltages to normal and restored the AGC action. Result – good sound and a perfect picture!

What was going on though? Despite the leak the video signal was still going through Q401 and since the following stages are AC coupled they don’t care what the DC component is here as long as it doesn’t get stuck against the supply rail or ground. Q304 (the AGC gate) is DC coupled though, so unless the DC component at its base is correct the line pulses won’t operate the transistor in the intended manner and the line sync pulse amplitudes won’t be sampled properly. Q304’s emitter voltage is set by R331, R322 and R323 and if any of these change value a similar AGC problem will probably occur.

Other than the AGC problem the TV-110UK proved to be in good condition and the tube gives a bright, sharp picture. These sets represent a genuine improvement over the TV9-90UB but I have to say that I prefer the styling and the proportions of the latter. The TV-110UK is better built and much easier to work on though.

[mark pirate]

I do like these little Sony sets, good news that you have a decent CRT. I picked up a mint boxed one from a boot sale some years ago. Having had very little use it has a very good tube and gives a cracking picture. But I still prefer the 990UB as it is a more appealing looking set with the added bonus of 405 lines!
My 990UB is also near mint and boxed, but sadly has lost one of the three knobs on the side.

Sony did produce some very nice kit back then, well built and extremely reliable.

Mark

[Hybrid tellies]

They were very reliable sets, did not realise that they suffered tube problems. I so enjoyed your excellent write up of this abc fault. Thanks Studio263