Failing audio PA transistor biasing measurements...

[Al (astral highway)]

Originally Posted by Herald1360

I can't see any antiparallel diodes in that circuit, only series connected ones.

Chris, ah, I was worried that adding the circuit diagram would cause confusion. It has nothing to do with the Yamaha set I'm working on - please see my note to that effect I added the diagram to try to conceptualise the relationship between the configuration of the power devices and the single 45V HT+ rail - I looked around and this seemed to fit the topology described by Mark and David en passant.

By antiparallel diodes, I was referring to the thermal regulation diodes discussed earlier in the thread - the little black oblongular things on the heatsink in one of the photos.

I know I couldn't take them out of circuit before testing them, but they tested as antiparallel diodes would. This, of course, could be an erroneous conclusion...

[mhennessy]

Originally Posted by astral highway

I'm trying to visualise the relationship between the power supply and the output transistors. Since I can't find (nor can anyone else!) the circuit diagram, I've hunted for a conceptually similar one that may help us to visualise this.[/B]

Yes, this isn't a bad approximation. The Yamaha has an extra small-signal transistor at the input end, but apart from that, there are probably close enough for our purposes...

I've attached a circuit used by Hacker which is probably even closer. The only difference being the output stage bias. This one uses a Vbe multiplier (T2) instead of your diodes. But I reckon that everything else will be a closer match for the Yamaha.

T1 is the input stage. The incoming signal is applied to the base, and negative feedback goes to the emitter. T3 is the VAS (voltage amplifier stage) that should have the full output signal on its collector. T4/T6 deal with the positive-going parts of the signal and are a Darlington pair, and T5/T7 deal with the negative-going parts. These are a CFP, and "look like" a PNP power transistor.

C6 is providing bootstrapping for the VAS collector load (very common). C7 is obviously the output cap. C5 in the negative feedback path reduces the gain to unity at DC. There is some complication in the way feedback is arranged with this circuit (R2 bootstraps the input, raising the input impedance, and R15 takes some of the feedback from after the output cap, reducing distortion), but otherwise this is pretty "textbook".

Originally Posted by astral highway

I'm trying to account for the 22.5V across one DC blocking cap and only 16 odd across the other, you see. We're agreed on a common supply rail of it seems 45VDC.

There are many possibilities at this stage - we're not there yet. But soon

Originally Posted by astral highway

It really looks virtually impossible to remove the anti-parallel diodes to test them out of circuit. I tried desoldering them, very patiently but the tracks are thin and the leads were pressed down sharply in manufacture. I could easily damage the silk screen. Alternatively, I could cut them out from the top, but then they'll need some kind of fix before I can re-attach them, assuming they test ok. How difficult would it be to obtain replacements?

Don't worry about these. A standard silicon diode and a blob of heat sink compound would no-doubt do the job just fine...

I'd be surprised if they were anti-parallel, but I suppose such a thing might exist. It would make the PCB layout easier, perhaps! Chances are you're seeing something else in circuit when you measure - perhaps one of the driver transistors? Either way, let's not worry about them yet. First step is to bridge out the biasing arrangements, and see where that gets us.

[Al (astral highway)]

Thank you, Mark. Very helpful to have your Hacker design as a conceptual aid.
I'm recapping some electrolytics and there'll be a delay before I can do the bias circuit bypass test...probably at the weekend!

Thanks again in the meantime

[Al (astral highway)]

Work-in-progress...

I'm adding new photos at a much higher resolution. These show the power supply stage, now completely stripped of electrolytics and ready to re-cap shortly. You can now clearly see the pass transistor with its clip on heatsink.

There's also another general view of the amplifier, in case anyone is interested in following the topology described at length by Mark and David.

I'll get soldering now and at some point later in the weekend, I'll be back!

[Al (astral highway)]

Power supply re-capped and two electrolytics on the main board, too.

All voltages test exactly as before...

only now, there's no audio output whatever from any input! Note to self: argggghhhhh!

[mhennessy]

Oh dear! Is the PSU OK? That's the top section of the first photograph...

[Al (astral highway)]

Hi Mark!! It seems to be. I measured the voltage across the output caps. I'll measure on the PSU test points tomorrow.

[Radio Wrangler]

OK, take it easy.

What you have in front of you is a reasonably simple transistor amplifier made entirely out of discrete components. All of the bits are available, or else close enough equivalents are. There is nothing irreplaceable or terribly expensive in it. You're spared that nasty feeling that some people get with other types of amplifier: "Dear god it might be the PX25!" That beats your "Aaaarrrrrggggh" any day of the week for fear-factor.

1) Is it repairable?

Almost certainly yes.

2) Is it worth repairing?

Well, that's your decision. There are no expensive bits involved. There probably will be a fair amount of time involved. If you were billing someone, they might not be very happy. But it's your time, and what you get out of it is a a lump of solid experience outside of your current comfort zone. That's more valuable than a working amplifier. I doubt that a working music centre is really worth the effort. You could earn more money than it's worth spending the time on the day-job. However, I think the business of getting you comfortable, confident and capable on transistor amplifiers is definitely worth it. The poor little things are terribly misunderstood and are overshadowed by all the over-hyped valve jobs.

There are two ways of approaching this repair job. The long way round is to get or trace a circuit diagram and to work out how it works, what voltages should be where and why, then to compare measured values against this, then work out which bits are certainly live, which bits are certainly dead and which bits are unknown. Maybe you can pull and test the unknowns and put them into known categories. Then you rebuild the thing with all the dead bits and unknown bits replaced. It should then work.

The quick way is to take short cuts. People like short cuts. A quick-fix is very desirable.You trawl the internet looking for people who had problems with this circuit... Hey! Joe Q McSporran III in Edinburgh, Indiana reported a fault on the same amplifier and he replaced Q6009 Q1234 and D23 and it worked. So you replace the same and it doesn't work. So you must have had a different fault. But as is the way with transistor power amps, some of your new parts are killed by the remaining fault. Next, you find a report by a chap in LLanwrtyd Wells and his fault description sounds right on. He even gave some 'before' voltage readings which match yours. He replaced Q2001, Q1234, D23 and R11. A Ha! you think, I've already replaced two of those! so you next replace Q2001 and R11. No joy, and silently, your recently murdered Q1234 takes out the new Q2001 the moment you applied power.

Sorry for the Dick and Smithy approach, but these things naturally send people round in circles and naturally breed confusion. The only defence is an orderly approach.

One perfectly sound way to repair them, considering the relative costs of time and parts, is to take all the semiconductors out. Test all the resistors and capacitors, then fit a new set of semiconductors. A compromise is to test the semiconductors when they're out. I don't mean a quick squirt with one of those little Peak testers, but something which takes them to full voltage.

The problem with short cuts is that they don't always work, and a few failed quick-fixes can add up to as much time as going the long way round would take and leave you in a low-confidence state.

Immanual Kant wrote an essay "On the innate animosity of inanimate objects" Clearly that guy knew a thing or two about transistor amplifiers... a great achievement given the period he lived in.

Step one is to trace the circuit. The general structure is quite common, so the Hacker earlier is a guide. There's also a 25W similar design in Mullard's "Transistor Radio and Audio Circuits" Blue book from around 1970.

I've put PCBs on photocpopiers and flat-bed scanners before now. A really good photo of each side would help. Computer mirror-flipped and aligned and you can start to see what goes where. Draw transistors as circles. Don't bother with E B C symbols, they either become obvious later or need later research.

Draw several diagrams, each getting neater and closer to the usual layout. Check it carefully.

You can also pass the finished diagram on to help others.

David

[Al (astral highway)]

Thank you, David. Wise words!

I'm in a bit of an impasse. I started mending the turntable as a favour to someone who sells vintage audio in a 'hipster' shop. There was no mention of an audio fault at the time, but this was obvious as soon as I'd fixed the turntable.

I do have more time on my hands than usually as I'm recovering from a big surgery and illness, but I'm not sure if I can go as far as retracing the circuit and subbing all the semiconductors, a major undertaking indeed.

You're absolutely right that the payoff is a hugely increased understanding of transistor amplifiers of this era. It's only the second one I've encountered of this high build quality (not saying it's hi-fi, just reasonably sophisticated compared to the humdrum valve Dansettes which seem curiously in vogue)

I'll have a think about how far I want to get enmeshed in this. I'm certainly keen to share any insights I gain along the way, so keeping this thread going for the time being.

Thanks again for your spot-on analysis, wisdom and support!!

[Al (astral highway)]

Originally Posted by Radio Wrangler

1) Is it repairable?

Almost certainly yes.

2) Is it worth repairing?

Well, that's your decision. ...However, I think the business of getting you comfortable, confident and capable on transistor amplifiers is definitely worth it. The poor little things are terribly misunderstood and are overshadowed by all the over-hyped valve jobs.

Hello again David, Mark in particular, also Andy and everyone else who contributed generously to this unexpectedly popular thread.

I've taken a step back and made the decision not to go ahead with any further intervention. David, I think you're absolutely right that only a step-by-step repair along the lines you advocate, is likely to get us anywhere.

I don't, however, feel like dissecting the semiconductors and caps and tracing the circuit.

Still I wanted to highlight some interesting 'take-home' learning from my experience.

1) A 'hunch', when nothing is obviously at fault, and no circuit diagram is available, may prove to be a red-herring. Notably, my initial impression that I should dig around the power transistors. I ordered a new bunch unnecessarily.

2) Check quiescent current as a priority in failing transistor amps of this era.

3) Temperature compensation arrangements in transistor amps of this era are critical since they control bias arrangements. They may fail for a variety of reasons, including mechanical failure or ageing preset pots.

I have missed anything else, before we close this out?

[mhennessy]

An understandable decision - you can't win them all

I agree that it's been useful to chat about the fault-finding techniques, and it's nice to see what you've taken away from this. What would I add?

4) DC conditions must be correct before you think about the AC signal. Your point 2 (quiescent current) is a big part of that, of course. The next priority after that is the voltage at the mid-point (+ve end of the output capacitor if one is fitted).

5) Despite the seemingly diverse range of designs out there, most follow the same 3-stage topology.

6) Some amps have current-limit transistors that won't have 0.6V across their base-emitter junctions.

7) A schematic makes a big difference

The next one will be easier

All the best,

Mark

[Al (astral highway)]

Great summary, Mark. Thank you.

In my limited experience (I've mended two designs of this era and one now has the better of me), transistor amps integral to home music systems of this era may have surprisingly good power supply regulation and reasonably sophisticated design, completely outclassing by a mile some of the valve circuitry that is now enjoying a comeback that isn't explained by acoustic quality. (There are two-valve - rectifier plus single audio device- designs from Dansette all over the place in local 'hipster shops, for comparison.)

So I'm not giving up on these when they pop up in the future. In fact, my curiosity is even more awakened. All's well that ends well!