Quad 303 pop on switch on

[mikes08]

I have a 303 that pops in the right channel when switched on. It works fine after that and i have just upgraded with the Dada kit inc new power supply and smoothing capacitors - all 4700mf.
I have the boards out and cannot see anything obviously wrong and comparing readings with a multimeter all seem similar.
Where should i be looking - the power transistors possibly?
I do have some spares tucked away so could try switching these on the right channel?
Any advice appreciated.

[Trevor]

Does it pop when the input is disconnected ? You may have a small dc offset on the input from the pre-amp This would suggest a leaky coupling capacitor

[fetteler]

Good suggestion Trevor - I wonder if the amp was making the pop before the DaDa upgrade? If so that adds weight to your line of thought.

Steve.

[mikes08]

Thanks - i will check that though pretty much ruled out the 33 as i have two other 303's that make no noise on turn on.

Have to admit school boy error in that i didn't run this particular one before the upgrade so not sure but all have been upgraded in the same way.

[Radio Wrangler]

Yes, that's a common oversight. But water under the bridge now.

I've seen reports of thumpy 303s due to problems in the PSU regulator. Also the 303 uses a bootstrap capacitor and problems in this area can give peculiar turn on/off transients.

DAvid

[mikes08]

Thanks David though if the power supply would it not affect both channels as this is shared by both amp boards?
I know the thump you mean but this is more of a transient pop which is what is puzzling..once on the amp it fine..
Could it be one of the power transistors..I have a couple of spare I could swap over...
Mike

[Radio Wrangler]

It could be any of the transistors, so if going down that route, you could start anywhere.

Bootstrapped collector load is a popular way of increasing the voltage stress on the VAS transistor beyond what most people would guess it has to endure. It also serves as a handy route for importing loudspeaker disconnection transients and routing them to the VAS.

VAS transistors have a hard life and ideally you want one safe from 'second breakdown' conditions.

There's probably a good case for fitting an entire set of semiconductors to the suspect channel as a single operation. It dodges the endless circling blowing and replacing the same devices over and over again which is yoour fate if you try power with any dead (even recently killed) devices still in place.

David

[Petedox]

I've had this a couple of times now. A scratchy rustle/pop (as opposed to a thump) on switch on. Both times it was TR102, which I replaced with a BC546B.

[Boulevardier]

Originally Posted by Radio Wrangler

There's probably a good case for fitting an entire set of semiconductors to the suspect channel as a single operation. It dodges the endless circling blowing and replacing the same devices over and over again which is yoour fate if you try power with any dead (even recently killed) devices still in place.
David

You’re right David; it can turn into an endless circle! And rather than removing and testing all those passive components, safer and easier to just replace those too. Oh, and to ensure there’s been no damage caused to the PCB, replace it also with one of those cute, high-quality Chinese copies.
Repairing faulty DC-coupled amplifier boards is a mug’s game, and makes no sense. Just build new ones. Especially as this is such a relatively simple matter with the 303.

Mike

[Radio Wrangler]

At least the 303 gives you the choice of how far to go.

Ideally, fault finding should get the problem down to one or more definitely faulty components. In this case you get the maximum info out of the repair, and learn the maximum eduction from it.

However, a lot of people can pull the transistors and replace them after testing,, but their concept of testing is a quick squirt on one of those little PP3 powered pocket testers that go nowhere near running currents and operaring stresses.

As general advice I sway towards a set of semiconductors as a way of avoiding the effort of looping around repeated failures.... The work in explaining and hand holding isn't insignificant either.

Too many people treat DC coupled solid state amplifiers as demonic on the unrepairability scale. Getting them going and not blowing their confidence is important.

David

[Trevor]

At this point I would look at the small board mounted electrolytics their must be one that li
mits the rate of rise re the output stage

[Radio Wrangler]

I was wondering whether the PSU regulator did that, as a soft start, but then both channels would suffer, and the PSU seems not to have any timing components.

On the power amp board, C100 and C104 look to be involved in controlling the turn-on process. C100 looks like an input DC block, but it's really the DC block for the shunt element of the second part of the feedback divider. C104 does that job in the first part.

This is a shunt-feedback amplifier and so the output Z of the preamp is involved, though the 2-section divider limits the extent. It's good for reducing the effect of non-linearity in the input stage, but less good from the viewpoint of sensitivity to pteamp Z and also carries a small noise disadvantage.

David

[mikes08]

Thanks all - that is useful as relatively new to this. I do have a couple of spare Chinese boards as it happens as i re-built one of my 303's via this route about 10 years ago and must have been sent a couple of spares. I am ok at building and replacing components - it is just the fault finding and theory that has me scratching my head unless something has obviously burnt itself out! I will try TR102 first as a relatively easy fix and if not solve all the TR's on that board and also look at the small capacitors. As i have three 303's I am comparing one against another for anything obvious.

[mikes08]

Originally Posted by Radio Wrangler

I was wondering whether the PSU regulator did that, as a soft start, but then both channels would suffer, and the PSU seems not to have any timing components.

On the power amp board, C100 and C104 look to be involved in controlling the turn-on process. C100 looks like an input DC block, but it's really the DC block for the shunt element of the second part of the feedback divider. C104 does that job in the first part.

This is a shunt-feedback amplifier and so the output Z of the preamp is involved, though the 2-section divider limits the extent. It's good for reducing the effect of non-linearity in the input stage, but less good from the viewpoint of sensitivity to pteamp Z and also carries a small noise disadvantage.

David

C100 and C104 have just been replaced on both boards as part of the Dada upgrade kit. I guess i could swap them across boards and see if the fault follows. I have ordered some transistors so will swap those first and see where that leaves me.
Thanks
Mike

[Hartley118]

When my Quad 405 had a hissy fit, it was clear that conventional fault diagnosis was leading me nowhere. So I resorted to the old trick of using the ‘diode test’ function on the multimeter to check whether each transistor still had two functioning junctions present, and not a short or open circuit. With the amp switched off, a silicon transistor should exhibit two junctions each with a forward-biased voltage drop of between 0.5V and 0.8V. No need to remove the transistors from the board.

This test soon showed a couple of transistors which were no longer capable of transistoring and needed replacement. I can’t guarantee that this diagnosis method works in every case, but it’s solved a few problems for me, both on the Quad and on items of test gear that have no service data. It’s quicker and cheaper than blanket transistor replacement.

Martin

[radio69]

I do exactly the same as Martin above. That is always my first step. My second step, with the equipment powered-up, is to check that any small signal stages are 'transistoring' (love that word!) - exhibiting a voltage drop of between 0.5V and 0.8V between base and emitter.
My third step is to take and record voltage readings.

When faulty transistor(s) and/or components have been replaced, temporarily replacing an on-board fuse with a 100R resistor for power up can be beneficial.

I have to be honest; replacing all transistors in one go is a last option for me, at least.

[Sideband]

Not in the same class but I sometimes use an Akai amplifier and that went through a session of pop and crackles on switch on and off (in fact when switched of, a rustling could could be heard for about 15 seconds gradually getting fainter. It was noisy transistors...pre-drivers actually and I was able to replace them with BC368. Perfect after that and no more problems.

[nutteronthebus]

I have restored many 303 and they are of an age where epoxy rot sets in on the front end transistors ( they lose gain and go noisy ) one from HFE of 300 down to 35 see this link for replacements http://quadrevisionspot.blogspot.com...ansistors.html

[mikes08]

Thanks all - yes i have been using the diode test on transistors and nothing obviously wrong on readings (i am comparing one board against the other? though I guess when powered up one may break down? I have some spare power transistors both RCA vintage and some new made in Mexico! Going to start with on board ones - at least only chasing fault in channel. When the amp is working no noises at all and sounds fine - just this annoying pop/crack when i turn it on. Just worried for my aging BC1's
I did find one resistor 4.7k which was reading a little low and replaced this but no effect.

[RichardGM]

If the amp works fine apart from the initial thump at switch-on, replacing loads of components or even starting again with a new pcb seems to me like serious overkill. Why not try a scope to look at a few voltages during switch-on? With luck that might reveal what's not quite right.