Bush TR 82C repair advice wanted

[Mark Hennessy]

The comment "I didn't measure and DC voltage" was when the set was powered up I measured 0.016 Vdc downstream of C21 with 1.1Vdc upstream, I thought i might have measured more if the cap was leaking voltage, just out my head.

In general, measuring the DC voltage at either end of a capacitor doesn't really tell you anything on its own about the health of the capacitor; such measurements can only be meaningful when considered in the context of the circuit.

In this case, the positive end is connected to the chassis, so the reported 16mV seems quite plausible, depending on where you placed the other probe. The voltage on the negative end goes straight into the base of VT2 (via the secondary of IFT1), and will vary according the signal strength - with no signal, R14 is "pulling up" the voltage, which makes VT2 conduct more and have more gain; as the signal comes in, a positive DC offset will be formed by the detector diode, and this opposes the negative bias from R14, reducing the voltage on the base of VT2 and reducing its gain. That's AGC action in a nutshell...

So a reading of around 1V on the negative end of C21 is entirely plausible.

What would be more instructive here would be to see if there was an AC signal at the negative end of C21. When in good health, the capacitor should present a short-circuit to ground for the audio signal that is present at the detector diode - so the signal at the negative end of C21 should be nothing but DC. But as it ages, it might allow AC to form, and this gets fed back to the VT2 and causes the distortion that I mentioned earlier. In an ideal world, you'd use a 'scope, but the DMM set to AC should give you an idea...

I woud also imagine the my DMM (Fluke 189) would typically read high on a 12V electrolytic given it is only charging the cap to 3V, again just a hunch. Any cap I have changed in the past (not on radios, industrial DC drive controllers and instrumentation equipment) it was visible or I had them changed anyway.

That's a nice DMM - I treated myself to one of those about 10 years back. Only complaint is their voracious appetite for batteries! Being the top of the Fluke line back then, it's definitely over the top for basic maintenance work like this - bear in mind that folk repaired radios with little more than an AVO 8; you don't need a basic accuracy of 0.025% for this sort of work! Still, extra precision doesn't hurt - what matters is the interpretation of the results.

From what you've written, there's obviously some confusion here. When talking about "reading high", we're talking about the capacitance value. Nothing to do with the voltage.

When you read about how they manufacture electrolytic capacitors, you have to marvel at the whole game. It's amazing that anything working drops out at the end of the production line - let alone anything with any sort of specification and life-span. The tolerance on electrolytic capacitors is very wide indeed - back then it might have been -10%, +50% for a decent one. So an 8uF reading 13uF is not a cause for concern - it's just beyond the top of that tolerance range I suggested, but that won't hurt the operation of this circuit.

But this is microfarads. Determined by the area of the overlapping plates, their spacing, and the properties of whatever is separating them. Nothing to do with the rated voltage.

The voltage rating is determined by thing that separates the plates. Exceed the "dielectric strength", and it might break down, allowing a spark to jump across. But that's not likely in a set like this.

So the Fluke applying 3V across it is honestly neither here nor there. I've yet to see a capacitor rated at less than 3V - no-doubt someone will have! - so 3V seems like a sensible value to pick to avoid the risk of damage. But again, that's unrelated to the point I was trying to make.

Ask yourself: just how would you measure an unknown capacitor? There are several approaches, but one might be to charge the unknown capacitance with a known current, and then time how long the capacitor takes to charge. That's an easy thing to do in a DMM because you've already got current sources for the resistance functions. And measuring time is very easy for the computer inside the DMM to do. So, while I don't know for sure, I wouldn't be surprised if that's how the 189 works...

Next, consider what leakage looks like to the DMM. Imagine a "perfect" capacitor, but with a resistance in parallel. This resistance will be able to pass a DC current, and this is what causes the leakage. Of course, you won't have an actual resistor inside the capacitor, but that's a valid way to "model" the leakage that might be taking place in the imperfect capacitor.

So when the DMM applies its known current to the capacitor, most of that current will charge the capacitor in the normal way, but some of that current will flow via the leakage. What this means is there is less current to charge the cap, so the capacitor will take longer to charge than you might expect. So, the computer in the DMM measuring charge time will conclude that the capacitor is a higher value that it actually is. Which might be the scenario here.

That's what I meant when I said "If it is leaking, any indicated deviation from the actual value depends very much on the capacitance meter - they are all different.". If a meter measured capacitance differently - perhaps using an AC technique, or a bridge method, then the leakage might affect the value in a different direction and by a different amount. Or not at all... Does that make sense?

So I believe all your advice is spot on, don't think too much, just do it. I haven't got the test equipment for ESR anyway.

To be honest, an ESR meter is not essential for radio repairs.

As a capacitor ages, the electrolyte within gradually becomes less "good", and as a result, the capacitor degrades. In practice, you observe a rise in ESR first, and then the capacitance value might fall, or the leakage current might rise. In certain applications - like modern switched-mode power supplies - the ESR is a critical part of the circuit, and when it rises, it causes real problems. However, in vintage radios, the ESR is rarely that problematic, and a below-par capacitor can soldier on until leakage or a fall in capacitance becomes a problem. So, for now, the Fluke would probably get you by...

Oh, a new problem now the on/off switch is now sticking off following me squirting some contact cleaner in to clean the crackly pot track, WD40 type. I used to use Electrolube in the olden days, I'm sure it slightly lubricated as well as cleaned, the WD40 stuff is pure volatile solvent. New switch/pot?

As has already been said, use the right stuff on these. Pots and switches are one of the biggest problems with old radios, and modern equivalents are hard to come by and generally inferior. A while back I wrote about potentiometers - the article was aimed at Hacker radios (which you really should investigate!), but the advice might be of interest to you: http://www.markhennessy.co.uk/articles/h...ntrols.htm

It might take a few re-reads before you can absorb all of what I've written, but I hope it helps.

All the best,

Mark