Debugging 120Hz hum - power supply design flaw?

[pvsage]

I just finished building a custom amp based on the Valve Junior, but with switched components throughout the signal chain and at the cathode bypass caps. It works, but I have a low 120Hz hum that is independent of the signal volume. I think I've traced the problem to the power supply - the bleeder resistor may be allowing some unfiltered rectified current through to B+?

Would moving the bleeder to the cap for the screen grid current fix this?

Peter

[Enzo]

Bleeder resistor has zero to do with hum.

SIngle ended amps do tend to hum more.

You are drawing the plate off the first B+ node, which is commonly done. Add a filter stage before that to clean up the B+ some.

[pvsage]

In other words, make it just like the Valve Junior power supply. I suppose that's the best solution, since it will let me bring the plate voltage in line. (346V right now! ...though I know other amps run EL84's even hotter.) Thanks, Enzo!

What about grounding? I went with the Watts Tube Audio/TurretBoards scheme for their VJr replacement, which ties the 12AX7 filter stage to signal ground. Is this kosher, or would it make more sense to tie it to chassis ground with the rest of the PS?

[stokes]

120hz hum is always related to the B+,the 60hz,when rectified becomes 120hz.When grounding filter caps,the power tube plate and screen caps should be grounded near the PT,along with the PT CT,the power tube cathodes.The rest of the filters should be grounded near or with the preamp grounds.The power tube grounds are "high current" grounds and should be kept apart from the preamp grounds.

[pvsage]

Thanks for the confirmation, Stokes - it seemed fairly intuitive to me that the hum was post-rectifier, but it's good to know for certain.

[Bruce / Mission Amps]

120hz hum is always related to the B+,the 60hz,when rectified becomes 120hz.When grounding filter caps,the power tube plate and screen caps should be grounded near the PT,along with the PT CT,the power tube cathodes....

Adding to this, the 120Hz hum is from a full wave rectifier not a half wave.
You don't see many B+ supplies using a half wave rectifier but a good example of one is the classic Fender 6G15 stand alone reverb unit... amazing to me that it isn't a serious hum box.
IMHO, grounding the center tap of a PT using FW rectifier, at the exact same place as the negative lead of the first and main filter cap, is an excellent way of reducing quite a bit of 120Hz hum right from the beginning.
To reduce hum in a single ended, Class A PA, use the max allowable filter cap your rectifier is rated for followed by a 47 ohm 2 watt resistor and then a much larger filter cap with capacitance of at least 100uF or more.
Or you could use a smaller 22 ohm 2w resistor connected right to the rectifier output followed by a 10uF to 22uF cap, then another 22 ohm 2w resistor with a final, 100uF@450-500v cap.
You could use smaller value capacitors (such as 10uF@450-500v) with 10-22 ohm resistors as far out as you want for extra filtering before the big and final high uF cap.
That's brute force filtering without the choke but the second resistor could be a high current choke for superior filtering.

The voltage drop across the 10-22ohm resistors at normal current levels will not lower the B+ very much at all and the series filter caps will help regulate the B+ and reduce ripple by a large amount.
I know it looks weird but its a kind of like extended "pi filters" with extra filtering nodes that is quite effective.

[pvsage]

I just realized that some people may have trouble viewing the schematic in my initial post because it was in bitmap format - my bad! I'm uploading a jpg, adjusted to show actual grounding configuration. I went with a non-CT power transformer and a FWBR. Is this as good in terms of grounding as FW with a grounded center tap?

If I understand what Bruce is saying, the first solution he proposed is very similar to the original Valve Junior PS, but with a smaller resistor after the first cap...as per the second attached illustration. (Junior is dead silent until you turn up the gain, so Epiphone must be doing *something* right!)

A lot of us Valve Junior modders have been increasing the resistor after the first cap to about 1K/5W to bring the EL84's B+ down to about 300V. I assume this is kosher? Also, a lot of us have been replacing the first capacitor with 100uF and leaving the second as 22uF (stock). Thoughts? Since so many have done this without incident, I assume that isn't too much capacitance for a FWBR built from 1N4007/UF4007's...

Peter