- All the resistors check out fine and I don't have a way to accurately measure capacitance unfortunately, especially a few suspect SMD ones. The electrolytics seem to check out fine though. (I do believe there might be something wrong with a snubber circuit though)
- Their is essentially two power supply circuits in here, both using the UC3845B SMPS controller. The smaller one powers the VBIAS and 5VSB circuits and is where I believe the problem is. Topology I believe is a flyback (SMPS's are far from my area of expertise so could be wrong on that).
- Waving a scope probe doesn't help too much in this case unfortunately, as basically everything in this supply switches at 100kHz (PFC, both SMPS).
- The CMRR is a problem when using a DIY solution unfortunately, looking at the quality of a 15V rail or some MOSFET gate switching at ~-100V DC bias doesn't really work.
I believe to have isolated the problem to this area of the circuit. Or the signals which drive this circuit.
Seeing the below measurements, the yellow trace represents the transformer output of T2 for the 5VSB rail while the purple trace represents the output of the 5VSB rail. The switching noise (or some of it at least) is correlated directly to the switching output of the transformer, driven by the bias circuit. The ringing on the output of the transformer makes it seem like a snubber circuit isn't doing a very good job. Particularly C92 - R145 snubber circuit maybe?
And everything is pretty much the same but worse on the VBIAS rail. What I'm unsure about is, is it really the snubber circuit(s) (C92-R145 and C103-R149) made of a MLCC and resistor that's the issue? Or could it be something else? No replacement alternatives available locally, so any ideas before I place a Digikey order?
To note: The resistors measure fine and inspecting the MLCC's from under the microscope they look spotless. The electrolytics highlighted and with a comment next to them have all been replaced. The drive MOSFET and current sense resistors have all been replaced, so has the actual switching IC from the previous repair.