Since there's no decoupling between output and driver stages in the B+ supply, I would be amazed if it didn't oscillate. A Zener is very low impedance and provides no decoupling. Don't use one there; use a resistor.
All good fortune,
Chris
All good fortune,
Chris
Congratulations. Welcame to the real world. Simulation are pretty fine but real world is far from the software simulations.
The upper resistor (the one between boot cap and Ebb) is the load for the boot cap. Thus their combination must regard capacitive coupling rules: time constant almos 10 time larger than the lower frequency period.
The upper resistor (the one between boot cap and Ebb) is the load for the boot cap. Thus their combination must regard capacitive coupling rules: time constant almos 10 time larger than the lower frequency period.
Interestingly I didn't find a decoupling cap to help with my issue. Zener is doing just great for me on its own. Might add it back for future troubleshooting, even if the LEDs have stopped blinking this journey seems far from over 🫠
Alright, but I don't have that extra resistor, there is a zener there to drop the 450V to 250V and before that I guess the 3.3K to drop the 500V to 450V?
Either way I'm getting output but it's pretty weird. Going to redo the meassument tomorrow but this is what I got on the scope today(yellow is input into both channels 1kHz 150mVrms signal and blue and purple are the left and right outputs):
Listening too it, there is a clear sinus tone in both ears but with one side being much quieter.
Looks like one of the tubes in the drive stage of one channel is taking a day off and the other chanel the 8136s seem like they aren't doing much of anything... Still can't find a single wrong connection or bad solder joint anywhere when poking around with the multimeter. Even reflowed a few that looked suspicious but to no effect. I'm suspecting the bias of the 12AX7s as I added back the(what I at least believed to be fixed) CCS in the tail instead of the resistor.
At this point I've changed a few things so I should really post a new schematic of the amp and power supply so everyone is on the same page 🙂
But you could see that we were not lying. The proper use of bootstrapping is essential. I learned it of my own an a personal project not documented here yet. And learned from the patents I suggested.
My next step would be removing leds and using classic RC bias networks.
My next step would be removing leds and using classic RC bias networks.
Decoupling requires a significantly large series impedance and a significantly small shunt impedance. Yours is a push-pull stage driving a push-pull stage, so both stages have both polarity effects on and from the power supply possible. One polarity's effects are negative feedback, the other positive feedback. This is inherently unstable, and the Zener is definitely not doing just great on its own. Your power supply needs decoupling between stages and no amount of denial will save you.
All good fortune,
Chris