I now have the voltage control loop running reliably. It’s issuing torch up and torch down commands appropriately and Mach seems to be responding well.

I also found some inefficiency in the LCD display output. I had been only able to read a voltage sample and run the entire control loop once every 5 to 7 milliseconds. After fixing the display code, I now have it running at under 1 millisecond on average.

The biggest issue now is with the voltage divider and op-amp. The op-amp always outputs 1.8 volts, no matter what the input. Additionally, it clips at 3.8 volts. So, the range of the output is too limited. Added to that, my voltage divider appears to be more than 2 volts per analog count which probably isn’t enough resolution.

So, I think I’m going to move from an op-amp filter to a dedicated filter IC. There’s a Linear Technologies chip that can do a 10 hz low pass filter with a single $6 chip and a couple resistors and capacitors. If I couple this with a better voltage divider, I should be able to get better voltage resolution.

So, I need to order the parts for that filter. In the meantime, I can tighten my hysteresis value and see if I get better control.

I should probably take a video of it to post.