Trimming paraphrase inverter ... hum question

[Diverted]

Hi all,

A friend of mine dropped a Benson Monarch Reverb at my house a few weeks ago, and I ended up tracing out the schematic and built my own. It's a very nice amp and everything sounds good.
One issue: As built, the original Benson had a 10KL pot to ground on the PI (V3B) grid, taking place of the 10K voltage divider resistor to ground (R2 in this schematic). The pot circuit is shown below.
When I built mine I used a resistor, not the pot.
Consequently I was getting some hum ... Pulling V1 reverb and V2 preamp had no effect. Only removing V3 PI eliminated it.
So I revisited that 10K pot and paralleled another 10K resistor across R2 to give me about 5K. This reduced the hum considerably.
There is still a tiny bit of hum left and I'm sure I could eliminate it more. My question is, what is the tradeoff? Am I in danger of shutting off one half of the phase inverter if I go much lower? Seems like that would be a possibility with that 10K pot, if someone just turned it all the way counter-clockwise, it'd be basically be grounding out that grid.

[Merlin]

It's not clear whether your hum is mostly arriving at the grid of V3A and then being amplifed by it, or whether it is mostly leaking down from the B+ and arriving at the anodes of V3, or both. These would have different solutions.
What happens when you ground the grid of V3A?

[Diverted]

It's not clear whether your hum is mostly arriving at the grid of V3A and then being amplifed by it, or whether it is mostly leaking down from the B+ and arriving at the anodes of V3, or both. These would have different solutions.
What happens when you ground the grid of V3A?

Grounding out V3A grid doesn’t quiet the amp. It changes the tone of that hum to a lower, fatter version if that makes sense.
Grounding out V3B grid does about the same.
Grounding V3 shared cathode eliminates it.
I just replaced the shared cathode RC for one on each cathode (10uf/820 ohms). No change.

[pullshocks]

When I had my Monarch on the breadboard, I had the PI balance pot on the front panel.  I tried various settings. 


I measured the SPL of the amplifier hum at various settings of the PI pot.

[Diverted]

Interesting. Doesn’t seem like a giant impact but still, it’s there. Neat.

[Willabe]

I think it's your grounding. Probably V3. How did you ground the amps circuit? Random grounding?

When you added the || 10K you cut the PI's output, so the humm was less, but so was the amps volume at the same setting.

A member here just got rid of the humm in his Magnatone vibrato/reverb preamp build using Merlin's grounding scheme.   https://el34world.com/Forum/index.php?topic=32605.msg361692#msg361692

Read Merlin's page on grounding, the bussed multi star grounding works great;

https://www.valvewizard.co.uk/Grounding.html

[Diverted]

I think it's your grounding. Probably V3. How did you ground the amps circuit? Random grounding?

When you added the || 10K you cut the PI's output, so the humm was less, but so was the amps volume at the same setting.

A member here just got rid of the humm in his Magnatone vibrato/reverb preamp build using Merlin's grounding scheme.   https://el34world.com/Forum/index.php?topic=32605.msg361692#msg361692

Read Merlin's page on grounding, the bussed multi star grounding works great;

https://www.valvewizard.co.uk/Grounding.html

I followed the grounding scheme and layout of the original. This amp has five grounding points. Maybe I’ll shuffle them around and see how it reacts.

[HotBluePlates]

Grounding out V3A grid doesn’t quiet the amp. It changes the tone of that hum to a lower, fatter version if that makes sense.
Grounding out V3B grid does about the same.
Grounding V3 shared cathode eliminates it.
I just replaced the shared cathode RC for one on each cathode (10uf/820 ohms). No change.

Got any many-µF caps?  like 100-250µF?  I would tack the big-cap across each cathode resistor for the inverter.  This will tell you whether the tube is leaking AC from heater-to-cathode (and I think that is the case based on the hum being killed when you directly-grounded the cathodes).

I'm not sure anything useful is done by separating the cathodes of a paraphase inverter.

As built, the original Benson had a 10KL pot to ground on the PI (V3B) grid, taking place of the 10K voltage divider resistor to ground (R2 in this schematic).

FACT:  The pot would allow getting either equal-output from the sections of the inverter, or counteracting some imbalance of the output tubes.  "Perfect Balance" gives the loudest clean signal the amp can manage.

OPINION:  The useful aspect of the paraphase is that it is not balanced, and can contribute some even-harmonic distortion to the output stage because of that.  I would choose a paraphase when I want intentional-imbalance (which is partly the "fatter sound" you heard earlier).

[pullshocks]

FACT:  The pot would allow getting either equal-output from the sections of the inverter, or counteracting some imbalance of the output tubes.  "Perfect Balance" gives the loudest clean signal the amp can manage.

OPINION:  The useful aspect of the paraphase is that it is not balanced, and can contribute some even-harmonic distortion to the output stage because of that.  I would choose a paraphase when I want intentional-imbalance (which is partly the "fatter sound" you heard earlier).

In the whole 33 page Benson Monarch thread on Amp Garage, I don't recall that the Benson factory setting of this pot was ever discovered or revealed.  I would love to know what your friend's amp is set at.


The AmpBooks paraphase calculator shows about 4K is balanced for this amp Paraphase Resistor Calculator


One thing I recall, playing guitar with the PI the pot  turned way down, volume was a lot lower, as you would expect with only 1 6V6 running.  Almost like turning it into a single ended amp.

[HotBluePlates]

One thing I recall, playing guitar with the PI the pot  turned way down, volume was a lot lower, as you would expect with only 1 6V6 running.  Almost like turning it into a single ended amp.

It was "almost like turning it into a single-ended amp" because, in fact, turning that all the way down will turn it into a single-ended amp.

The output transformer will have a tube on either side to pull "balanced idle current," but only one of those tubes will be driven with an AC signal.  The net effect is "a single-ended amp with a non-ideal load" (which causes it to deliver somewhat less power than if the load were "ideal").

[Merlin]

Grounding V3 shared cathode eliminates it.

That suggests you could use a much bigger cathode bypass cap to kill the hum.

[tubeswell]

Grounding V3 shared cathode eliminates it.

That suggests you could use a much bigger cathode bypass cap to kill the hum.

Or that the tube is bad

[HotBluePlates]

Grounding V3 shared cathode eliminates it.

That suggests you could use a much bigger cathode bypass cap to kill the hum.

Or that the tube is bad

A lot of tubes have heater-to-cathode leakage, and will hum in the absence of a heavy bypass from cathode to ground.  I'm convinced that Fender used a 250µF bypass cap on the tweed Bassman to kill hum, not because they needed a lower cut-off frequency.

But most stages of most preamps have a bypass cap.  So it's only when we are unlucky with a tube-sample, and also unlucky with a socket/circuit that lacks a bypass, that we notice the tube humming.  So the "bad tube" might give acceptable performance in a different socket (not unlike throwing a microphonic tube in a blackface Fender amp's trem oscillator socket).