To lighten the stress on the zener string, one could increase the value of R66.
I also trimmed R68 to get the U2 output voltage to within the value designated on schematic. This stopped the overheating of V8.
I did notice some instability after a while, with U2 output toggling at subsonic frequency, so I added a 25uF cap across the output to ground and that seems to have stabilized it.
I also trimmed R68 to get the U2 output voltage to within the value designated on schematic. This stopped the overheating of V8.
I did notice some instability after a while, with U2 output toggling at subsonic frequency, so I added a 25uF cap across the output to ground and that seems to have stabilized it.
I increased R66 to 220K to lessen the stress on the zener string.
I was chasing down a subsonic (0.2Hz) oscillation and found that replacing D13/15 helped cut that amplitude down by over 90%.
The stabilizing of subsonic oscillation was also achieved with the addition of a R/C network with a 6.2V zener at the grid of V8. Prior to adding this, there was a large amplitude swing at subsonic frequencies at grid. Completely stabilized by the zener/cap. Prior to addition of parts, this voltage would swing between 1.3V and 21.6V at a subsonic rate, causing phase modulation of the audio output, V8 runs MUCH cooler now, as well.
I'm capped at 4 hours labor by customer budget, so this will have to be the final solution. Posting in case it helps others with this oscillation problem.
I was chasing down a subsonic (0.2Hz) oscillation and found that replacing D13/15 helped cut that amplitude down by over 90%.
The stabilizing of subsonic oscillation was also achieved with the addition of a R/C network with a 6.2V zener at the grid of V8. Prior to adding this, there was a large amplitude swing at subsonic frequencies at grid. Completely stabilized by the zener/cap. Prior to addition of parts, this voltage would swing between 1.3V and 21.6V at a subsonic rate, causing phase modulation of the audio output, V8 runs MUCH cooler now, as well.
I'm capped at 4 hours labor by customer budget, so this will have to be the final solution. Posting in case it helps others with this oscillation problem.