Question on DC Offset

[davodavo]

Hi. This isn't really a DIY question, but I'm hoping your collective expertise will instantly know the answer. I searched the forums, didn't see it...

I've got a fairly old quasi complementary symmetry, direct coupled amp that has 140 mV DC offset on one channel. The amp works, but that offset is about 100 mV higher than you'd expect. Naturally, there is no adjustment for either bias nor DC offset.

The transistors all seem OK, no burned components, no obvious cap problems.

For this kind of amp, which transistors most directly effect DC offset? where would I look to tweak a resistor value to adjust?

Circuit diagram is up at www.taberconsultling.com/circuit.jpg

[mjf]

hello.
slip of pen.
http://www.taberconsulting.com/circuit.jpg

when parts are old they are often out of tolerance,especially caps and res.
you can measure the resistors and compare the values of the two channels,the better one with the bad one.perhaps you find some different values.
and you can recap the little elcos with new ones.
greetings..............

[nigelwright7557]

Personally I wouldnt be getting too upset at 140mV DC offset.

The transistors blowing is another matter.
It could be the bias is far too high.
You need to measure the volts across output resistors to get an idea if the bias is faulty or not.

[lineup]

One simple try you can do is to let input pair, Q3 and Q4,
change places.

Another better try is to find and buy 10 pieces 2N4036.
Then measure volt B-E, at a current of 1mA, to find a good matching pair.
Or just try different combinations of those exemplars until you get low offset at output.

One resistor that has got some effect on DC-offset is R12, 27k.
But is not optimal to begin to change the value of this resistor
because probably the amplifier should have 1+1 mA in input pair.
Best is if they share exactly 50-50%
And so my guess is that the Supply is around 2x55 Volt DC.
( 54V/27k = 2 mA .. and each input transistor takes 1 mA )

----

I agree with previous poster. 140mV is not too much to worry about.
Will not harm the quality of the output or hurt your speakers.
So, there is not much to gain from fixing this moderate Dc-offset.

[taj]

Just curious, trying to learn. What is Q1 and Q2 used for before the LTP? It looks like an extra gain stage to my under-educated eye.

..Todd

[gni]

Q1 and Q2 look like gain but could be just buffer. . . those caps should
stop DC before LTP. . . Q3 and Q4 are probably not matched. . . .keep an
eye on the 140mV. . . .not really bad. . .but it could be much lower. . . .

I like the idea of swapping them. . .that would be simple and then you would
know. . .

Would love to hear how it works. . .

[Tony]

Originally posted by gni
Q1 and Q2 look like gain but could be just buffer. . . those caps should
stop DC before LTP. . . Q3 and Q4 are probably not matched. . . .keep an
eye on the 140mV. . . .not really bad. . .but it could be much lower. . . .

I like the idea of swapping them. . .that would be simple and then you would
know. . .

Would love to hear how it works. . .

hi,
Q1/Q2 forms a non-inverting gain buffer with a gain of around 3 if i am right....

with today's sources, maybe not required...the main amp istelf has a voltage gain of about 12 sp overall gain is around 36 which just about good for a PA amp....

[mjf]

hello.
at the base of q3 there is a res without designation........may i guess..............that is the one for trimming the offset?
greetings.................

[Tony]

that resitor should be 10k like R17.....

[mjf]

yes,that is a point........in a nearly perfect working amp these res values should be equal............;
amazing that r-c network across r17...............seems to give some extra punch in the bass tones...........