Little Class AB Amplifier Hiss

You’ve been getting good advice from other members, so I don’t want to intrude. I only suggest taking advantage of partial disassembly to thoroughly understand all possible ground paths. Eg. with all connectors open, confirm there’s no ground conduction (no unrecognized shorts to heatsink). No shorts of transformer to chassis? Both PCB holes offer paths to chassis? Does the 4.7R have a clean path to a quiet ground point? If its path can be independently controlled, that’s an opportunity for experimentation.

Good luck.
I checked all the ground paths and they do check evenly between both boards.

I did also run a 1000hz sine wave through the amplifier and hit it with the o scope. Results below. I cannot zoom in any further with thr cheap o scope. We have a good one in our lab but it's over at the office, not in the shop, so I can go check it with that on Monday.

The second picture is right after the sine wave cut out.
 

Attachments

  • PXL_20250504_023739203.jpg
    PXL_20250504_023739203.jpg
    402.7 KB · Views: 16
  • PXL_20250504_023751564.jpg
    PXL_20250504_023751564.jpg
    404.1 KB · Views: 16
A burned up zobel resistor suggests high frequency oscillation. Is the heatsink grounded to power supply ground? Grounding the heatsink sometimes corrects some oscillation issues.
At the time of of it burning up, no, the heat sink was not grounded the chassis. I was attempting to isolate various areas as I troubleshot the unit.

I think I have it boiled down to the 60 Hz hum. I have a filter on the way. Hopefully that can eliminate that. If it doesn't, I will try a filtering EMI power strip. If that doesn't work.....I suppose I'll start checking components with the LCR
 
Played around with the DMM this morning. The resistor that burned up does not ring out to ground.

Sometimes the resistors surrounding that one do ring out, and then sometimes they don't. Sometimes legs of the larger mosfets/transistors ring out and sometimes they don't. Seems like it's switching or something. I get this on both boards. I am sure someone in our embedded department at work could explain this phenomenon but I do not understand this. I used two different DMMs. One is a $4300 fluke with the ability to generate signals and test thermocouples. It's very accurate as we have tested it with the lab calibrated o scope before.

I assembled the whole thing and it plays music just fine. It plays LOUD too. It just has that hum