ok I got some time so I put the new outputs in (MJE15032/33's) and funky light bulbs.
I used some jumper wires and a ton of to-220 heat sinks so I could run this outside of the chassis as it is very difficult to assemble/solder in. there is no way to access the PCB with the factory wires soldered on.
So, before when it worked, the headphones played but ran out of headroom quickly and sounded a bit tinny. likely due to the lack of power, high impedance etc
so I put it together and listen on headphones, sounds full and good. when turning way up, I imagine hearing distortion.
I connect my work speakers up, turn up - volume! then I go more and it starts distorting.
I poke prod, trace - all seems good. The thing is a ground loop nightmare.
but I cannot find an apparent reason, so using the SCM, I hook up power and grounds only
I have all my voltages, off perhaps .25v but the power supply is over the spec by half a volt and its non adjustable. I had slightly low on the base of the bias unit but I wonder resistor differences. MEasured every single one and all were within spec. The PO apparently replaced them all with more modern units.
So its time to hook up the sig-gen and the scope, I injected 1Khz (1013 according to my meter, its analog controlled and its not like this is an IF or anything), I set my ancient BK scope to read the sig gen, plugged that into aux in, then I tested the pre-amp P400. tested fine. I even jumped the pre-amp over to my sx890 and not one unruly noise.
so while listening to the output (not a very catchy tune) I started probing legs of various stages. At low volumes see the screen of the scope (sorry, I have 2 laptops, the unit, all my junk on the desk, the scope went on the floor tonight)
That wave is the same shape (with soft points) as the generator itself.
And see what happens when you turn it up a bit over low volume, look at how the upper part of the wave is shaded, it has the shape, but I guess this is representation of noise or breakdown.
what the holy heckfire is going on. it should be noted, that the hfe between the MJE15032 and the 33 is almost twice. and a set of MJE15030/31s were even worse. either my meter does not like to switch pnp/npn quickly or wth?
Im half temped to glue in a set of KSA940/KSC2073s to see what happens (150v, 1.5A 25W, 3Mhz Hfe)
any ideas? ever see a trace like that? I listened to it on headphones for over an hour and it sounded great, but add speakers with lower impedance and stuff and it falls flat, and yes, I cleaned and tested the speaker switch, checked the resistors on the selector network etc.