Fixed that boom box I’ve been trying to build. I had the amp module up and running on the bench, and playing music. Finally without the dim bulb, variac, or oscillations after a few false starts. It would only work properly with Cdom about 2X what it outght to have been, but I was through arguing with it. Go to hook up the car CD player to it and it wouldn’t come up at all after 7 years of storage. Eff it. Buy another, they’re cheap. Put it all together on the dim bulb - sounds great. But every time you change tracks (on USB even) it draws current, drags the voltage down and resets everything. Get out the scope. RF/digital hash everywhere. Add some more RF suppression caps to the preamp board. Still does it. Ok, what’s the worst that could happen taking it off the dim bulb so it doesn’t drag down and reset? Change tracks - then huge burst of oscillation from the amp, goes into DC protect after running the BOFUs to maximum excursion for 50 milliseconds, and the whole place smells like “fresh oranges”. Predrivers, darlington outputs, vbe multiplier, emitter resistors, and the whole SOA circuit burnt to a crisp. Nearest I can figure all that damn digital hash getting into everything caused massive cross conduction in the output transistors. In an RF-free environment it was fine.
So what do you do? Install FASTER output transistors (and Everything they took out). Shouldn’t be necessary for fidelity, but eff it this thing has to work in the presence of digital garbage. Can’t buy just a basic mostly analog piece of equipment anymore - it’s gotta have Bluetooth, Wi-Fi, cellphone connectivity, and a computer to run every single function. Swap out the TIP142/147 for Sanken D2560/B1647. Have to add freewheeling diodes and Rbe on the output darlingtons (since these only have the Rbe an the actual output and no diodes) on the back side of the PCB. It’s been working for the last 2 hours without a glitch.