Help please: Revox PR99 MkIII Input Amp problem

[shawthing]

Hello all, can anyone please help me solve a problem on channel 1 of a Revox PR99 MkIII Input Amp? I think IC1 is the problem but if someone can advise how to be sure I'd be very grateful. I can offer the following readings for IC1, feeding a 1K signal in:
Pin 1 shows no signal (where IC2 pin 1 shows signal).
Pins 2, 3, 5 & 6 show low level signal, the same as IC2
Pin 7 shows good signal, same as IC2.
Pin 8 shows 20.5v, same as IC2.
Looking at resistors around IC1 - R112, 113, 114, 108, 109 - they show no signal, neither does the RA101 pot. R110 and C106 do show signal.
I think this means that IC1 has failed: but, because I don't fully understand the circuit, it occurs to me that it may be some other component(s) causing the problem...
I've checked all resistors on the board and they all appear to be at or reasonably near correct; no caps show any signs of deterioration and all give reasonable resistance readings, but all these items have only been quickly checked, still on the board, so if there are any that you think I should re-check, or lift a leg and then check, please let me know.
Chas

[julie_m]

I haven't seen this recorder, but it sounds as though IC1 has 8 pins and pins 1 and 7 are outputs. That is certainly consistent with a dual op-amp. Either it's dead, or something external is shunting the signal to ground (which might or might not have finished the IC off). If socketed, lift pin 1 and see what voltage it has on it (it will swing hard to the limits with the negative feedback disconnected, but you should see be able to see the frequency is the same as the input). Search for any parallel components which may have failed short-circuit.

For a quick substitution test, any old dual op-amp should more or less work; although you probably want an ultra-low-noise part in a quality tape recorder.

[Restoration73]

If its a TBA931 there is an official fix for this obsolete chip

ftp://ftp.studer.ch/public/products/...ement_Rev2.pdf

[shawthing]

Hi, thank you both for your replies: to Julie, IC1 (which btw I should have said is an RC4559P dual op-amp as you suggest, apologies for not specifying) is not socketed so I can't easily lift a leg - unless there's a trick for that I don't know? All resistors on the board show correct values (or within specified tolerances) barring a handful which differ slightly from the schematic but which exactly match the equivalent ones in the working Ch 2. Likewise all all lytic and ceramic caps appear to test OK (i.e. they show rising ohms and then decreasing volts). There are 4 polypropylene caps which I can't get to show either ohms or volts, but as two are in each channel & Ch 2 is fine, I'm assuming they are fine but would need testing in another way.
If there are any components that you think would be particularly likely to be the cause I can double-check my measurements in case I've missed something subtle, but otherwise, swap out the IC?
Thanks again, Chas

[Ted Kendall]

Pretty sure the TBA 931 was long gone by this time...fortunately.

[shawthing]

Hi Ted, yes, none in the PR99 AFAIK. Plenty in my Revox A700 though... C

[julie_m]

The pin is in the corner, so you should be able to snip it right on the bend and then separate the cut ends gently.

Before you do that, though, measure the resistance from pin 7 to ground and from pin 1 to ground. It might not prove much if there is a short inside the IC. (But if you get a low resistance from pin 1 to ground, high resistance from pin 7 to ground and, after cutting pin 1, high resistance from the stump that is not connected to the IC to ground, that probably is the problem.)

If, when you power up and apply the oscilloscope you get a signal on the stump of pin 1 connected to the IC, that means the problem lies somewhere in the external components. Otherwise, the IC has had it; in which case, you may as well just cut the remaining pins and desolder them individually. Fit a high quality turned-pin socket for the new IC. If you can't find an exact replacement, try the more-readily-available 5532.

[shawthing]

Julie, thank you very much for this! I shan't be able to try it until tomorrow or Tuesday evening but that sounds like it should help a lot. I'm very much hoping it's the IC, because I haven't been able to detect any faulty components elsewhere, whereas I can source a replacement IC, so it'd be job done! Embarrassed to say I hadn't thought snipping & separating the pin - that was the 'trick I didn't know'!
I'll report back asap, thanks again, Chas

[Radio Wrangler]

Snipping the pin is a good old trick, and a blob of solder can restore things afterwards.

Less well known is that with some specialist test gear, it's possible to measure the voltage drop down an IC leg and to measure the current being taken. This sounds quite ludicrous but it's a great method to have in reserve when you meet a complex PCB with something somewhere generating a short. Specialist test gear makes it sound terribly expensive... yup, it once was, but some very good DVMs crop up on the second hand market occasionally. Being aware and having such a thing in that wouldn't-it-be-nice shopping list in the back of your mind allows you to pounce if you see one. I got one for a fiver. No-one else knew what it was or what could be done with it. It just sat on a stall at a rally. A DVM with more digits than anyone thought were necessary. Old, dirty and by a maker no-one had heard of. Standards-lab grade DVMs are expensive if you go looking for one, but you wouldn't want to miss one if it turned up at a bargain price? Not needed often, but when they are...

David

[shawthing]

Hi David, thanks for the suggestion, this wasn't something I'm familiar with (or equipped for yet) but I can see how it might help too! Chas

[shawthing]

Hi Julie, results as follows:

Resistance from pin 7 to ground initially about 220k, falling fairly rapidly to about 6k, then rising more slowly to about 16-18k and hovering around there. Pin 1 started about 6k, rising to about 16-18k and hovering around there. (These results are exactly the same on the healthy IC2 btw). Snipping pin 1, resistance of the stump in the board to ground is the same - starts about 6k, rises to late teens and hovers.

After snipping, there is no proper signal present on the IC stump of pin 1, but there is an extremely low level wave: I'm not good on reading scale on the scope I'm afraid, but if I adjust it so that the signal on pin 1 of the healthy chip fills the screen top to bottom, then the signal appearing on the IC stump of the cut pin 1 is less than one 1cm division. Also, where the healthy chip's pin 1 signal is a very clear sine wave, what's on the snipped pin 1 is hugely fuzzy. I can't help wondering whether this is just breakthrough?

So by my understanding, this shows that the IC is failed, and the similar resistance to ground readings for the two chips' pins 1 & 7 and the board-stump of the snipped pin 1 show that there are no other shorts occurring in surrounding components. Does that sound right?

Thanks again for your help, Chas

[julie_m]

The signal at the cut pin 1, even on a good IC, will be very messy, because there will be no negative feedback. It should be square, spending most of the time at one or the other power rail; whereas on a good amplifier, it will be a nice, small sine wave (assuming the input is a sine wave).

So I'd expect the signal at the cut pin 1 to fill the screen, with a smaller signal at the good pin 1.

It certainly sounds as though it is IC1 that is faulty, and not the peripheral components.

[shawthing]

Hi Julie, very many thanks: I shall order a replacement and fit it (socketed) and see what happens.
Hugely appreciate your help!
Chas

[julie_m]

If you have one to hand, you could try any old 8-pin dual op-amp just for a quick test to see if it works (it will probably be noisy, compared to the proper part).

Make sure to use a high quality, turned-pin socket in order to avoid problems with poor contacts. Cheap IC sockets are more trouble than they are worth.

[shawthing]

Nothing suitable in my stock - never had one go down in a Revox before - but I have some RC4559P on order, and some decent turned-pin sockets too as the only one I had was a very flimsy affair! Depending on delivery times I should be able to move this on either at or just before the weekend - looking forward to it!
Thanks again, will report back, Chas

[julie_m]

Ah, well, it was just a thought. If you'd had something like an LM358 about the place, it would have been fine for a quick go/no-go test. But obviously, not everyone's parts drawer is as deep as mine

[shawthing]

Hi Julie, success, though not at first. The replacement 4559 worked fine, but led to my discovering two cold solder joints elsewhere. I then found that although the signal was fine on the Input board, I was still getting nothing on the Rec amp board... a couple of dodgy looking caps didn't solve it, but the collector of Q2 (which should be about 11v) was at 2v. Replaced that (which was one thing I did happen to have in my spares drawer!) and - finally - full success! Very many thanks for your help, hope to return the favour one day, Chas