Hi everyone! I have a strange hum/buzz issue that has just popped up on my Crack with Speedball. The hum has shown up during the course of some upgrades - and I can't see how the upgrades might have changed anything, so I need a sanity check as to potential causes and solutions....
Basically, the Crack has a hum/buzz, but only in one channel, only at very low positions on the volume knob. The noise progressively disappears as the volume knob is increased. It happens whether or not there is music playing, whether or not the source is turned on, and whether or not the source is even connected. I first noticed it when switching over to a brand new input tube immediately after my upgrade spree, but have since been able to hear it in several other input tubes, albeit some are better/worse than others. So, probably not the input tube. If I switch out the power tube from my beloved 5998 to a Sovtek 6AS7G, the buzz is reduced, but still audible. So, probably not the output tube.
What I don't understand, is why the buzz only happens at very low volumes. Perhaps something noisy in the ground, that the input tube can see when the volume pot is shorted to ground - but wouldn't a noisy ground show up in both channels? Perhaps something wrong with the pot - but why would the noise still be there when the source is disconnected?
Swinging back to the upgrades I performed immediately prior to noticing this noise. I switched out the stock rectifier diodes for Cree Schottky's (using the PCB from Parts Connexion), I replaced Axon 2.2uF film bypass caps in the PSU with Mundorf 2.2uF, and I replaced Axon 91uF output caps with Mundorf 100uF caps. So the only direct signal path change is the output caps; I could perhaps understand that one of the output caps is faulty or poorly connected, but I still can't understand why the noise would only be present at low positions on the volume knob. Only other thing of note related to grounding is that the amp does have a ground loop breaker installed, in the form of a 10R 5W resistor paralleled with a 0.1uF cap between safety ground and input ground. Pot is a TKD 2CP2511 50k, padded with 50k series resistors for about -6dB pre-attenuation.
I'll probably reflow all the solder joints in the noisy channel, double check my voltages (they were fine when I first noticed this) and try to see if my multimeter can see the noise frequency or amplitude at various points along the ground and signal chain. But other than that, I'm stumped.
Thoughts?