What is acceptable DC offset on speaker terminals?

[valterdaw]

Is it possible to name generally acceptable DC offset on amp speaker terminals?
I have couple of perfectly working class A/B amps with 10-30mV offsets in channels. Other channels of these amps (one is PG XS-something and other is JL 500/5) has much closer to zero offset (less then 1mV).
I am sure it's either some semi's are leaking after suffering a heat damage or resistors are out of spec, and it's definitely tough and time consuming to troubleshoot it, so is it anything I should be seriously worried about?

Thank you!!!

[jol50]

I have some that way too and no bias on them. The terminals that actually go to the fets will have more than the 'ground' ones, so if you didn't know a 2ch bridgable will often have one + and the other channel's - be the actual powered output, they will be the ones you use bridged.

[1moreamp]

Its usually the input diff pair thats shot in most cases, and that is the cheapest fix.
I have seen new amps with the issue and found it to be power supply related to the diff pair, and when bridged the amp had 3 MV DC offset. In stereo it was 25 MV DC offset < NEW mind you>

PPI had a DC offset adjustment, and every time I realigned these my clients claimed the amp sound like new and often came back just to tell me how happy they were with my work.

So when ever possible I take any measurement about 5 MV as out of spec and try to resolve it. I recently had a PG M50 come across my bench replacing the diff pair lower the offset by 50% closer to 0 but even then it was still at 23MV offset. I found a leaky driver and output that resolved the rest.

Also check the main rails to see if the read the same with respect to the common of the toroid. mismatched rails also tends to cause issues in this area on some amps.

The PG XS amp most probably has trashed input pairs, especially if they had bad outputs to begin with. At least that is what I have always found. On XS and Tantrum series amps the output feedback to the input in such a way that a blown output causes diff pair damage. Other than that check all the zener diodes for leakage. Most repairs for a channel in these require a complete set of semiconductors to give like new results < i.e. input to output all new devices> There only MPSA-06 56 devices and 2389 and 1038 devices in the input pair. The rest are all N devices MPSA-56's, and 4 zeners 2 6 volt and 2 15 volt in the bias circuitry.

I call it the $2.00 channel as the 56's and 06's only cost 3 cents each in 100 piece lots at most suppliers, and the zeners are about the same. It costs more to buy the diff pair and outputs, but even those only tend to cost like 20 cents each.

So I completely rebuild the channel and keep my mind busy on other things as rather then trying to piece meal my way through the amp.
I get like new results, no headaches, and I move on faster to other things. And my clients seem to be happier, and my rework is always zero, so its a win, win situation in the end run.
Once you go through and complete the rebuild a few times it will start to go like clock work and your time will drop amazing as there will be little to needed rework on final test...Hope this helps, they are great little amps, and very economical to regenerate all new, and reliable....

[valterdaw]

Thank you!!! Looks like I am looking at the channel rebuild, which shouldn't be bad, as I found these amps a pleasure to work on.
I love the idea of labeling components for multi-channel amps, that resistor R123 for the first channel is the same as R223 for second and R323 and R423 for third and fourth Also, traces that are not peeling from board when you turn on your iron within one feet distance