rickdiculous
New member
I’ve now built two Lowballer (Bassballs) boards from PedalPCB and both have failed in exactly the same way. Instead of passing audio, I just get hum, and when I check VREF it has about one to two volts of AC ripple riding on it. This happens whether I power it from a wall wart or a battery, so it’s not coming from the supply. If I clamp VREF to ground with a 1 kΩ resistor, the ripple disappears completely and things quiet down, but of course that drags the DC reference down toward zero and nothing biases correctly.
To troubleshoot, I’ve already tried dropping the divider from the stock 47k/47k to 10k/10k, I’ve swapped the VREF cap up from 10 µF to 100 µF and even larger, I’ve paralleled in ceramics, and I’ve added local decoupling at the op-amp area. I’ve pulled the op-amps out entirely to rule out oscillation, and the ripple is still there. Grounds all show continuity to the jack sleeves, the diodes and resistors all check out, and nothing looks mis-stuffed. I’ve also compared this against Aion PCBs I’ve built with the same parts, and those work flawlessly, so the issue doesn’t seem to be my build technique or component choice.
At this point it feels like the PCB’s VREF implementation is just too high-impedance or has a poor return path, because the behavior matches a floating reference. The fact that the 1 kΩ clamp kills the ripple instantly seems like the smoking gun. Has anyone else run into this on PedalPCB boards, and if so, is there a known fix? Should I be looking at bodging in an external VREF (like a little daughter divider with heavy filtering or a TLE2426 rail splitter), or is there a simpler jumper/trace tweak that others have used successfully?
To troubleshoot, I’ve already tried dropping the divider from the stock 47k/47k to 10k/10k, I’ve swapped the VREF cap up from 10 µF to 100 µF and even larger, I’ve paralleled in ceramics, and I’ve added local decoupling at the op-amp area. I’ve pulled the op-amps out entirely to rule out oscillation, and the ripple is still there. Grounds all show continuity to the jack sleeves, the diodes and resistors all check out, and nothing looks mis-stuffed. I’ve also compared this against Aion PCBs I’ve built with the same parts, and those work flawlessly, so the issue doesn’t seem to be my build technique or component choice.
At this point it feels like the PCB’s VREF implementation is just too high-impedance or has a poor return path, because the behavior matches a floating reference. The fact that the 1 kΩ clamp kills the ripple instantly seems like the smoking gun. Has anyone else run into this on PedalPCB boards, and if so, is there a known fix? Should I be looking at bodging in an external VREF (like a little daughter divider with heavy filtering or a TLE2426 rail splitter), or is there a simpler jumper/trace tweak that others have used successfully?
