stereo amp cleaning

Q411 collector CANNOT be at +33 volts unless there is a major DC offset at the output of the amp. Same for the collector of Q409, and Q413.

Also, the negative rail from the main negative filter cap is shown as -42.6 volts. Therefore, the voltage you gave (+33.8) at the emitter of Q413 must be wrong, or else resistor R445 would have SMOKED with better than 70 volts across the 100 ohm resistor. Also, if it was just a typo, and it should read -33.8, then I'm concerned that the main filter cap C9 might be open, and not filtering.You might mesure the collector voltages at both channels' output transistors. Since there are dual supplies, one open cap would show a significantly lower voltage at the collector of the one channels' PNP outputs versus those on the other channel, if the one cap were open-circuit, that is.

I did see this, but wonder if you might have measured the -DC from the other channel...

I also see the the voltages at the emitters of Q401 and 403, low though they are, seem unbalanced - this supports my theory that you have a major DC offset in the amp channel. Please check this and repaort back to me. It's not uncommon the have a bad differential pair at the amp input (Q401/403). This could have been your main problem all along.

Mark Z.

Um, if you read back in this post you'll see that replacing time-worn components is what started this mess in the first place... I replaced all of the electrolytic caps in the amp and pre-amp and have and have had nothing but trouble ever since. I also cleaned up all of the pots and switches...

as -42.6

volts

channels'

a

open-circuit,

other

they

Yes yes yes.

Yes, R445 _IS_ smoked. Well, it looks fine but it's open... +33V on one side and -43V on the other.

Yes, there is +33VDC at the emitter of the output transistors on this channel, zero on the other channel.

There is 0VDC at the output of the preamp and at the base of Q401.

I checked all my voltages again and yes, all the +33VDC ones are correct. Confirmed the +0.2 and +0.4, steady as a rock, at the emitters of Q401 and Q403, respectively.

How do I tell where the DC offset is being introduced? Do I pull Q401/Q403 and test? If I replace them do I need to worry about matching just the pair, or should I look at matching the gain with the other channel pair?

Thanks.

Dave

I think it's worth just replacing the bad resistor. Sometimes that's all there is. Do check the various drivers and outputs for any shorted or open junctions first, though.

R465, R485 Q415, Q417, Q419, R459, R457 all should be checked, as well as the output transistors and the .22 ohm emitter resistors R471 and R473.

Don't forget the small value resistors like R469 and R461.

Mark Z.

Mark Z.

All the resistors off the power rails are listed as "1/4W, Fuse". Does this mean I've got to find 1/4W fusible resistors, or will a regular old 1% 1/4W

100-ohm metal film resistor do the trick?

Dave

Metal film would be fine. I'd usually avoid carbon types for this application.

Mark Z.

I checked ALL of the resisitors in circuit and when I found one that didn't jive with the expected value, I tested the same component in the other channel. The resisitors used must be all 1% or better, they are BANG ON their reported values... everything under 100 ohms was within 0.3 ohms and this is a 20-plus-year-old amp.

R419, R421, R439, R445 and, strangely, R453 and R461 all tested open whereas their duplicate in the other channel tested OK. R462 in the opposite channel also tests open... It looks like about half the transistors should be testable in-circuit, if one tests bad I'll pull and test it(oscilloscope/curve tracer would save some time here).

Might take awhile.

Dave

Dude... (Hurley voice here...)

You got a blown amp channel.

Mark Z.

You know, for the amount of time I've got into this, I could have realistically spent about $30 on parts and replaced every transistor and resistor in the channel... hopefully the outputs are ok, they're pretty robust compared to the resistors that blew. Theyr'e also fairly expensive.

What causes an amp channel to blow other than shorting the outputs?

"Dave" wrote in news:8gC5i.59374$V75.23360@edtnps89:

That is what is commonly done in service shops, even when the real problem is located. Since the remaining parts have been badly stressed. It is also a good idea to repalce the good channel components so the unit does not have the other channel die a month later...

failed output transistors.

----== Posted via Newsfeeds.Com - Unlimited-Unrestricted-Secure Usenet News==----

The #1 Newsgroup Service in the World! 120,000+ Newsgroups

----= East and West-Coast Server Farms - Total Privacy via Encryption =----

You do need the check the outputs. They can fail for a variety of reasons. A bias diode can go open-circuit, even momentarily. Bad solder at a bias transistor. Shorted speaker wires or defective speaker(s). Abuse by the user. Other failed components or solder connections.

Mark Z.

P.S.

At this point I'd say it's a virtual certainty that the outputs ARE bad on that channel. Probably the only reason it's not just blowing fuses is that one or both of the emitter resistors, R471/473 are bad also.

Mark Z.

Checked the outputs... they test OK on my transistor tester! No leakage, both PNP and NPN had gain, although not anywhere near the same amounts... I assumed they'd be "about the same"... anyway, the outputs ARE NOT shorted and R471 and R473 DO NOT read open circuit although I haven't pulled the resistors yet to test.

Started pulling transistors from the beginning of the signal path, tested Q401 through Q413 so far, 5 more to go and they all test okay, no shorted junctions and all show gain.

If it's a heat-related issue with a failing transistor, I'll never find it. I don't think it is, because the DC bias on the outputs shows up right away when the amp is powered on.

I'm out of ideas if the transistors all check out okay. I'm inclined to blow $8 and replace all but the outputs, replace the half-dozen smoked resistors, put it all back together and see what happens.

Dave

For what it's worth, the speakers and wires are fine, they're being driven by another amp while this one is being repaired. The amp was not abused/overdriven prior to failure. I think I'd better get out the magnifying glass and start looking at solder connections...

OK, well you're lucky then, the outputs are OK. Replace the bad resistors, test the drivers and predrivers carefully, and test the amp while monitoring the bias as measured across the emitter resistors on a millivolts DC range of your DMM. A variac would be most useful. Lacking that, one can pull the main fuse and substitute a 100 watt light bulb which will limit current in the event of a hard overload. If the lamp glows bright, all is not well. A light glow after initial charging of the filter caps would be normal.

If the amp appears to be running OK, with no more than about 25 millivolts across the emitter resistor pair with no load, I would recommend turning your attention to the bias diode D405 which looks like it ought to have about 1.2 volts across it in normal operation. If the diode opens up, the voltage across it will increase, maybe slowly, maybe suddenly. If this happens, that would be proof the diode is bad. You can compare the voltage to the one on the other channel. They should be within a few millivolts of each other.

Mark Z.

That was my next question to you... the varistor. How does one test it? My understanding of varistors is that they are a high-value resistor at low voltages, and a low-value resistor at higher voltages. Kinda' like a zener. It's listed under diodes on the parts list and its' schematic icon looks like a diode, which would indicate polarity, no? Failing to find a datasheet or specs on this particular part, I'll just check it against the one in the other channel for resistance with no power. Maybe hook it up to

15V with a series resistor and measure the voltage drop across it vs. the drop with a 3V source? The "12" in the part number along with the expected voltages across it (

Next question. If the diode/MOV D405 is shorted, am I ever going to see the correct voltage across R471/R473?

Dave

The bias diode is essentially two diodes in series. (Sometimes three, actually). In this case two, looks like. A diode check function of a DMM would be OK for a static check, and if it became necessary you could use the DMM to help match a replacement pair of diodes to replace it with. However, using regular diodes might not work as well in terms of correctly responding to heat changes. The proof is in the operation, and in this case that means the actual voltage drop across the device in-circuit, while the amp is running. Bias diodes opening up is uncommon but it does happen, and when it does, catastrophe is the result.

In your case the resistors may have burned because of a momentary short at the speaker wires, etc although this would ordinarily have destroyed the output transistors as well.

Mark Z.

If the diode shorts, I would expect to see nothing across the emitter resistors. Some other amp designs would be the opposite, slamming full bias if the diode shorts.

While we're on the subject, I recommend replacing the bias transistor Q417 no matter what. Just get the polarity right...

Mark Z.