Hacker RP33 Autocrat

[pwdrive]

No consolation I know but transistors don't like it up 'em, give me valves anyday.

Lawrence.

[Joe]

My thoughts exactly, Lawrence. I can't see how I could blow three valves, apart from by excess heater voltage, at least without an indicative bang and/or flash.

- Joe

[Yorkie]

Sorry to hear of your misfortune with the Hacker Autocrat Joe - they don't like it up 'em!

I can tell you how you could 'blow' not three but four valves even with the set not plugged in, and many people have done, then done the same again with the replacements!

By pulling them out of those diabolical valve sockets on DAC90As which have spring rings which are meant to keep the valve in place, but which simply destroy the glass pip on the valve. Those rings have absolutely no useful purpose - even if the set is turned upside down the B9A valves won't drop out. Dire.

It's an expensive lesson to learn.

It doesn't matter how many times the cautionary advice is given to people to remove and bin the springs before attempting to remove the valves, you still get people saying 'I've just taken the vales out of this set and they've gone white inside - why is that - does it mean they're no good now?'

The DAC90A is my least favourite set, which to me, is a dismal reminder of the post-war austerity years of ration books, power cuts and queues. The valveholder springs are just one more reason for me to dislike them! It only abounds in such numbers as it was a rental set. If I ruled the world I'd order them to be confiscated and consigned to landfill after having removed the spring rings and had them thrown into the furnaces of hell

[Mark Hennessy]

Hi Joe,

We've all been there! It's funny how even the most logical brains sometimes refuse to work to form. Clearly, the amplifier was fine. And the tuner. Just not both at the same time! It's a shame that we'll probably never learn what was causing the two to play up when working in consort, but never mind...

First lesson here: transistor gear needs a different approach. Much calmer and less cavalier than seems to be the norm for valve gear. To that end, what you need is a power supply with a variable current limit, so that the output will fall at the first hint of trouble. Actually, that's a nice little project, and something I would willingly help you with separately...

When working on these things, I normally set the current limit to around 50mA. That's less that the maximum collector current of any transistor in the amp, so nothing bad can happen (in theory )

I know it feels cursed, but stick with it. I doubt you'd be able to sell it on with an IC amplifier anyway, and if you don't mind waiting, I can probably send you a known-good amplifier PCB at some point in the future. I plan to make a good Autocrat from the 3 I have here, and once done, I'll have some spares...

Wishing you better luck with the Sovereign

Mark

[pwdrive]

Joe, I said earlier that I don't like transistor stuff, and that is true, the main reason being is that i worked on so many sets with repeated monotany when I was in the trade, that said I would not give up on it either if I had one, maybe take a fresh look at it in a week or two, post the diagram up if and when you do as I feel sure it can be cracked, Marks suggestion for a current limiter is an excellent one and would be worth pursuing.

Good look whichever way.

Lawrence.

[Joe]

I've thought about the current-limiting thing myself, and indeed have one I use for charging batteries, but have never known what to set it to plus I've never understood how current can be limited without cutting the voltage.

I'm trying to be philosophical about it and think that at least it's a chance to play about with alternative transistors. The current arrangement works but obviously wouldn't for long so I need to find some that can handle more power.

- Joe

Edit: Thanks, Lawrence I've added the circuit to my earlier post from this afternoon. I'm not for giving up by any means - I've said before that you learn far more from your failures than your successes. I'm loath to shell out for new 'Correct' transistors as, as Mark says, the actual problem remains un-diagnosed.

[pwdrive]

Most manuals will give the quescent current, as a rough guide (it's been a long time) set the current limit to the quoted quescent current plus a working current of say 20-30 ma for starters, but as Mark says..you can set it to a reasonable safety margin (say 30%) below the maximum current the output transistors are rated at , that figure can also be derived from the max power rating of the device at its normall operating voltage.

Lawrence.

[Mark Hennessy]

I've thought about the current-limiting thing myself, and indeed have one I use for charging batteries, but have never known what to set it to plus I've never understood how current can be limited without cutting the voltage.

Ah - when the current limit is reached, the output voltage does indeed fall. Sometimes like a stone! But that's the point - better that than let the smoke out

I'm trying to be philosophical about it and think that at least it's a chance to play about with alternative transistors. The current arrangement works but obviously wouldn't for long so I need to find some that can handle more power.

That's the spirit

Actually, once you've sorted the quiescent current, the BC327/337 would probably be OK. If you leave things as they are, but simply put bigger transistors in there, then all you'll do is eat batteries...

One thing to check; with at least 2 examples that I've looked at, the biasing arrangements around T3 were different to the manual. Rather than the NPN AC127, there was a PNP something or other... As a result, R13 was connected to a different part of the circuit. Anyway, it's worth checking - what was the T3 that you removed - an AC127 or something else?

Edit: Thanks, Lawrence I've added the circuit to my earlier post from this afternoon. I'm not for giving up by any means - I've said before that you learn far more from your failures than your successes. I'm loath to shell out for new 'Correct' transistors as, as Mark says, the actual problem remains un-diagnosed.

These transistors are getting harder to find, so it's a wise idea to experiment with others...

Out of interest, how do you know the originals failed? I must admit that I'm not sure how T3 could...

Also, don't forget that AC127s et al aren't totally immune to the "tin whisker" disease...

BTW, I know I sound like a clever-clogs all the time, but if it's any consolation, I'm struggling with what should be a simple fault that I've fixed many times in the past. I get more than my fair share of difficult things here! And the trouble is, it's easy to forget the countless easy jobs that sailed through when these things are getting you down!

Cheers,

Mark

[Joe]

I'll double-check when I get up but I'm sure that T3 was a AC127.

As for testing them, I've done it with an LED and a couple of resistors. Two of them light the LED with their bases disconnected, ie are short circuit collector to emitter, and the other doesn't light it at all.

I don't know how they can have blown either; I'd have expected the BC108 to go first. Maybe one was faulty before I started?

I thought that BC227/BC237s should be OK as they can stand 800mA and are getting less than that. Why the sods are getting warm is something that needs looking into.

As for the current limited PSU how about this:-

   

I've got all the bits, apart from the FETs but am not bothered about being able to go right down to 0V. One point: what should RSC be?

To be honest I didn't realise that what I was doing to that audio board was dangerous.

Thanks for the support

Joe

[Mark Hennessy]

The transistors are getting hot because the quiescent current is too high - power = volts times amps. Yes, those transistors are rated at 800mA, but because they're only in a TO92 package, they're not intended to dissipate much power. In other words, they can pass 800mA, but only if the voltage drop across them is very small - i.e. they're really good at switching loads.

You'll find that's often the case - a given transistor will have a voltage rating and a current rating, but there's no guarantee it'll do both at the same time. The power rating is the first check - 0.625W in the case of the BC327/337 - but there are second-order effects as well, which are best "parked" for now

Why did the output pair fail, but not the "weedier" BC108? Well, that's because of where the transistors are in the circuit, and their ability to pass damaging currents. The output pair are directly across the power supply, so the only limit is the power supply - probably in excess of 1.5A if you used your LM317T PSU. The BC108 is surrounded by "padding", and even when saturated, the maximum current that can flow is only a few milliamps.

That PSU looks reasonable enough, and of course you can retro-fit it to your existing design. Without thinking too hard about it now, I'd say that Rsc should be ~0.8 ohms (1.25V / 1.5A). A couple of 1.5 ohm resistors in parallel would probably do the job - best to use 1W or higher wire-wound resistors for that. The FETs are just being used as current sources, and normal bipolar transistors can do that job with the addition of a couple of extra components.

Cheers,

Mark