5150 style build - diagnosing DC heater imbalance

[mxrshiver]

finishing up this 5150 style custom build, and all the voltages are just where i want them to be. the only issue is, there's a loud hum/buzz from the DC preamp heaters, that ONLY comes in only when i crank the primary voltage up past about 100VAC.

the heater circuit is attached. power tubes (2x KT90) are fed from 6.3VAC, and preamp tubes (5x 12AX7) from the 6.3VDC at the filter cap terminals.

with <100VAC primary voltage at the input of the power transformer, everything's hunky dory - both the AC and DC heater supplies are at roughly +2.5V/-2.5V and perfectly balanced, noise floor is nice and low, no perceptible hum or buzz even when the controls are cranked, just a normal amount of hiss/white noise.

however, when i turn up the primary voltage more, the AC heaters stay perfectly balanced, and the total DC voltage between the terminals stays about the same as the AC... but the DC develops a bad imbalance which creates a loud hum/buzz throughout the preamp. the negative voltage stays at around -2.5VDC, and only the positive voltage continues to increase as the primary voltage is increased. with one multimeter probe to ground, probing the positive terminal with the other slightly improves the hum/buzz, or at lower voltages where the hum/buzz isn't too loud yet, can completely erase it... and probing the negative terminal makes it worse, which is the only part of this that makes sense to me lol.



any ideas? i'm a bit at a loss. triple checked my wiring and solder joints, chopsticked it, re-tested the tubes, replaced the fuse... i definitely don't have more than one heater ground reference... since the AC side stays perfectly balanced all the way up to 125VAC on the primary, i would think it has to be an issue with the rectifier diodes, or filter cap, but they're all brand new. could this all just be the circuit saying 'please put my ground reference on the DC side'?

[pdf64]

Maybe one of the noval valves has a h-k short, what happens to the DC side if they're all removed? eg do the + & -ve voltages become balanced?

[mxrshiver]

good question, i replaced them all one by one but no difference, haven't tried it with none of them installed yet, i'll try that today!

[mxrshiver]

tested with no preamp tubes, balance remains perfect up to full primary voltage. but putting even a single tube back in, and the imbalance returns, albeit to a lesser extent.

i think the filter cap has to be getting some sort of false ground reference... i realized it's a push-fit cap, but i soldered 20AWG stranded directly to the terminals, assuming that would be alright as they're pretty large. i pre-tinned the wires, didn't use a long dwell time or anything, but maybe that could have damaged it?

[pdf64]

...
i think the filter cap has to be getting some sort of false ground reference... i realized it's a push-fit cap, but i soldered 20AWG stranded directly to the terminals, assuming that would be alright as they're pretty large. i pre-tinned the wires, didn't use a long dwell time or anything, but maybe that could have damaged it?

Being as the cap's can is at a different potential to the chassis, it seems much more likely that the can (internally referenced to its negative terminal) is finding a resistive path to the chassis, via the clamp and an imperfection in the film sleeve.

[mxrshiver]

ugh... i think you got it. turns out the clamp made a little dent in each side by the mounting bolts. i was careful not to over-tighten the screws that close the clamp around the cap, but i didn't account for how the fine placement of the mounting bolts, could still cause those corners to bite into the cap body, when i tightened the mounting bolts down to the chassy. time to get another $20 freakin can cap *facepalm*... gonna be extra careful this time, and add a couple layers of thick heat shrink around the body before it goes inside the clamp. thanks for the help!!