question about germanium based amps

[ke4mcl]

do bias issues have a particular sound or behavior to them? the amp in question is a 1967 xam integrated made by electro voice. i believe this amp uses germaniums.

at low volume it sounds just like what a speaker with loose voice coil windings sounds like. its almost as if its at the threshold that a transistor is conducting on and off with audio peaks. its full of the epoxy pill shaped transistors which i have heard were problematic.

ideas? unfortunately i dont have a schematic for it. these were made as a house brand to a department store.

[Conrad Hoffman]

No doubt there are people here better versed and more favorably disposed towards germanium. Germanium devices are known for going leaky and freaky. Sometimes you put them on the curve tracer and the whole family points skyward. They go intermittent and can do something different every time you power them up. The parts can look fine with the usual DVM diode check, but unlike a silicon device where a good diode check almost always means the device is good, the germanium will look good and still be bad under power. The circuits are a pita because the low turn-on voltage and leakage make almost all in-circuit measurements suspect. I'd try some freeze mist or heat (use heat very sparingly) to locate a bad transistor. Naturally at the age of the amp it could also have bad caps or other problems.

[ke4mcl]

the amp has been recapped and power supply diodes replaced. the transistors seem to be all original. will try the hot/cold idea when i return to work next week.

[Mooly]

From your description it sounds like crossover distortion.

Many germanium stages were really basic, sometimes just a three transistor affair with complementary outputs and driver.

You need to see how they are biased. Some designs had a preset for current, others a fixed bias. It's a fine line as any excess reduces battery life drastically and can also cause thermal runaway. You are problably looking at around 5ma or so in the output pair.

[toprepairman]

1967 and epoxy pill shaped transistors sound more like early silicon to me, though could be hybrid, (any pictures?).
Agree with Mooly, sounds like crossover distortion, biassing problem?

[ke4mcl]

sure i have pics.

[Mooly]

Thanks for the pics.

They are silicon. For no good reason I assumed it was a battery powered radio...

Power output is obviously limited by just using those TO5 ? outputs.
Are both channels the same as regards the distortion ? A scope check with a generator and correct load impedance would reveal the nature of the distortion.

You still need to determine how the stage is biased and what the standing current is. Next to the output transistors is what looks like a 10 ?? ohm resistor. Is that in series with the emitter or collector of the outputs. If so you can determine current from voltage dropped across it.

[Conrad Hoffman]

Those little ceramic packages with epoxy tops are, IMO, pretty decent. I've seen a lot of FETs in those too. Usually there are part numbers on the perimeters- can you read anything there?

[indianajo]

I find gargly sound can come from old electolytic caps. These look pretty old. My 1970 dynakit ST120 amp had some elna caps that look like that. You can 1. do what A T etc. recommends, pull one leg of each one, measure the capacitance and ESR of each cap. Meters to do that are about $150, or $90 in kit form. Make sure the unit is 60 deg F or below when you do it. or 2. do what I do, and replace them all at about $.07 each. I recapped a 197? FM-AM radio, a PAS2 preamp, a ST70 power amp, and a PV 1.3K power amp last year. The first 3 sound great, the last has issues due to the OT meltdown before I bought it, and a re-assembly mistake I made. (It worked 1 minute last week before motorboating). I recapped an H182 organ in 2009 (70 of them) and have a Wurlitzer 4500 organ and a 10-82 tone cabinet lined up for recap. There might be other issues in the Wurl and TC, but I feel there is no sense wasting time debugging when all these old dried up caps are in there is my policy. In case you don't know, electrolytic caps have a + on one end, or a crimp on the + end. Mark your PCB with + before replacing, and do two at a time and check your work before proceeding, since it sort of works now. If a problem, back up and look over your solder joints etc again. I buy the caps all at once, though, freight is 100X what a single cap price is. You can get 8000 hour rated caps if you check the datasheets before you buy, and never do it again. You might want to get some MPS8099 (npn) and MPS56 (PNP) or other small NPN transistors while you're at it, and maybe some TIP31 (NPN) and TIP32(PNP) TO220 transistors for power. The cost of the transistors is a lot less than the freight, too. Watch the TIP's, they are not EBC leg layout.

[Conrad Hoffman]

With some decent assorted caps and those generic but very good transistors, there isn't much that can't be fixed!

This is far from the last word and I may yet change the emphisis, but here's a page on electrolytic capacitor testing I'm working on-

Testing Aluminum Electrolytic Capacitors