I'm having a hard time even finding the spec's for the PM3302, but it seems people suggest using the PM3305 schematic, I put it below. I see it has some digital stuff inside. But I'm guessing the vertical trace position is still some analog circuit.
So does the peak to peak voltage look right on all divisions, but moving the trace up/down barely works ? Have you tried on all V/div ?
I guess some ladder of transistor's, like in a push-pull output, some part of that must have a problem. Or 1 of the signals for it's biasing. Maybe the input voltage from the PSU has a problem too, so always worth checking (but yeah watch out for the crazy HV area's).
If you have another scope, or even a good DMM and a low frequency signal source, even from a 555-timer, you could try tracing the signal over to the output section. With the power off, check all the resistors, diodes, cap, around there. I'd spend the time and make my own schematics some sections. And I'd probably remove a bunch of caps/ and even transistor's, to really make sure everything read's as expected. And I'd make a model in a simulator, just to see some reasonable biasing of the output drive.
But yeah I'm mainly thinking of some stereo push-pull amplifier, so no high voltages to worry about there.
The PM3305 schematics is over here, it's too big to upload
http://bama.edebris.com/manuals/philips/pm3305