AMB Laboratories DIY Audio

[amb]

What DMM do you have anyway?
Why remove the heatsinks? Do you feel them getting warm at all?
Now compare your voltages against the operating points PDF again. Anything look significantly out of spec?

[ntropic]

A Mastercraft 52-0052-2, as detailed here: http://www.tnt-net.ca/ve6kq/mastercraftdmm.htm

The heatsinks aren't warm in any way whatsoever. I don't think removing them would do a thing, but at this point, I'm stumped short of replacing everything, which I would bet would do nothing. I'll have another shot at the operating points.

[amb]

Weird DMM -- it does not have a DC or AC mV range. Nevertheless, if the meter is any good, you should see some 0.0xxV readings when there is some millivolts of voltage.

If the heatsinks don't warm up at all,then they're not getting sufficient bias. Let's see what your voltage measurements show.

[ntropic]

Aaaaaaaand we're back. Several months later.

Question: If I had to play the replace the parts game, at this point, I'd be attempting to replace the BC550s and BC560s? At this point, the only thing left I can think of to try is to replace those transistors.

[amb]

Do you have the correct value R23/R24 and R47/R48 output resistors? They should be 0.47 ohms. Measure them.

[ntropic]

Yes, all 4 resistors are the correct values.

[amb]

So the issue is still that you couldn't achieve a proper quiescent current by adjusting the trimpots? What is the voltage across C7 when you have max'ed out the trimpot rotation anyway (I asked this some posts ago). Before you start blindly replacing transistors, please perform the BJT sanity check on all transistors except the input JFETs.

[ntropic]

Must have missed the bit about C7/C16. They both measure 0.959-0.961V.

BC550 NPN:
Q5/18 and Q8/21 check out.

Q4/17 fail +C/-E test, pass +B/-E test.

Q8/21 fail -C/+E test.

Q9/22 both fail -B/+C/+E tests. Abnormal voltage in +B/-C/-E test (0.3/0.5V).

Q10/23 both fail -B/+C, but not -B/+E. Normal voltage in +B/-C/-E test.

Q9/22 and C10/23 fail the +C/-E and -C/+E tests.

BC560 PNP:
Q6/19 check out.

Q1/14 fail -C/+E test, pass -B/+E test.

Q11/24 fail +C/-E and -C/+E tests.

Q11/24 fail +B/-E test, but not +B/-C. Normal voltage in -B/+C/+E test.

Q7/21 fail +C/-E test. Normal voltage in -B/+C/+E test.

I'm not entirely sure what this means, but if diodes are only supposed to pass current in one direction, does that mean that Q1/14, Q11/24, Q7/21, Q4/17, Q8/21, Q9/22, Q10/23 are all dead?

[amb]

Must have missed the bit about C7/C16. They both measure 0.959-0.961V.

Seems a bit low, which could explain why you're not able to achieve the desired quiescent current.

BC550 NPN:
Q5/18 and Q8/21 check out.

OK.

Q4/17 fail +C/-E test, pass +B/-E test.

This is fine, because these junctions are explicitly shorted (see the schematic, and the BJT sanity check doc also mentions this).

Q8/21 fail -C/+E test.

Hmm, you said above that Q8/21 passed? So which is it.

Q9/22 both fail -B/+C/+E tests. Abnormal voltage in +B/-C/-E test (0.3/0.5V).

This may be ok because of interaction from resistors R12/13 R36/37.

Q10/23 both fail -B/+C, but not -B/+E. Normal voltage in +B/-C/-E test.
Q9/22 and C10/23 fail the +C/-E and -C/+E tests.

You already mentioned Q9/22, but Q10/23 looks suspicious.

BC560 PNP:
Q6/19 check out.

OK.

Q1/14 fail -C/+E test, pass -B/+E test.

Same comment as Q4/17 above.

Q11/24 fail +C/-E and -C/+E tests.
Q11/24 fail +B/-E test, but not +B/-C. Normal voltage in -B/+C/+E test.

This doesn't seem right either.

Q7/21 fail +C/-E test. Normal voltage in -B/+C/+E test.

The right channel version of Q7 is Q20, not Q21. Q21 is NPN. I assume this is just a typo and you really meant Q20. This doesn't seem right either.

What about Q12/25 and Q13/26?

Double check again that you don't have BC550C and BC560Cs mixed up on the board, and the output transistors too (are they swapped, or installed backwards). If an amp doesn't work af initial setuip time, and both channels exhibit the same problems, it is likely that you assembled the two channels at the same time and made a repetitious mistake somewhere on the board. This may have caused some of the transistors to be damaged, but check the board very carefully first.

[ntropic]

Q8/21 fails -C/+E test. A typo there as I thought it passed at first, but it turned out I wasn't contacting the part correctly because flux.

Q12/25 and 13/26 will fail some tests, but once I pull them out and do the tests, they'll pass fine. Found that out when I replaced the parts.

All transistors are in the places they should be.

I made you a chart to simplify things. Q9/22 continue to give odd readings. The rest, while not necessarily passing the tests, are at least consistent.

http://dl.dropbox.com/u/5703631/BJT%20S ... 20Test.xls

Sidenote: What causes the rising voltage phenomenon when I do the tests? The capacitors charging in the circuit?