What Exactly Happens When You Pull Two Output Tubes On the Same Side In A Plexi ?

[Cathode Ray]

Asking based on two things I've experienced.. first was a guy who went by VHoholic on the old MetroAmp forum who used to run a Marshall 1959 circuit with only one 6CA7 output tube installed, and more recently Jim Gaustad has said he runs his custom-built EVH spec amp with V4 &V5 pulled.

I've always pulled either the two inner or out tubes to cut power down to 50W, but always been told never to pull tubes from the same side.

These guys claim this works with no harm to the amp - running them this way for decades.

[maxkracht]

What happens: less power, more distortion (not the kind you want), more hum, greater risk of damaging your output transformer, potentially risk damaging other tubes with higher voltages. This is assuming the tubes on the same side are in parallel, you don't need to wire it in the normal fashion where one side is push and the other pull, you could have the tubes staggered.

[Cathode Ray]

What happens: less power, more distortion (not the kind you want), more hum, greater risk of damaging your output transformer, potentially risk damaging other tubes with higher voltages. This is assuming the tubes on the same side are in parallel, you don't need to wire it in the normal fashion where one side is push and the other pull, you could have the tubes staggered.

Yeah, so the scenario here would be a standard circuit - off-the-shelf 68 Super Lead, so to speak.

Jim Gaustad was saying doing this turns the amp into "pull" rather than a "push/pull".

[maxkracht]

If you want a SE amp, probably wise to use a SE transformer. They are made differently. You might get away with running a PP transformer sorta like a SE, but I doubt it sounds very good and the the transformer probably isn't happy.

[Cathode Ray]

.. I doubt it sounds very good ..

You guys tell me.

[martin manning]

I tried pulling two tubes from the same side of a 100W 4x 6L6 power amp sim. Plate dissipation went from 21.8 (72%) to 36.8W (123%) per tube, with a whole lot of second harmonic distortion. I guess if you halve the life on half the number of tubes you would break even.

[Helmholtz]

Plate dissipation went from 21.8 (72%) to 36.8W (123%) per tube..

Interesting.
I assume this is not at idle?

[Roe]

I tried pulling two tubes from the same side of a 100W 4x 6L6 power amp sim. Plate dissipation went from 21.8 (72%) to 36.8W (123%) per tube, with a whole lot of second harmonic distortion. I guess if you halve the life on half the number of tubes you would break even.

do you get any of this effect if the phase inverter is greatly unbalanced, driving one side way harder than another?

[martin manning]

Plate dissipation went from 21.8 (72%) to 36.8W (123%) per tube..

Interesting.
I assume this is not at idle?

I have input sine amplitude set to produce 100W output into a resistive load, then just remove the two tubes on one side and rerun.

[martin manning]

do you get any of this effect if the phase inverter is greatly unbalanced, driving one side way harder than another?

Not sure how this would be relevant given that one side is open circuit. It would just be more or less signal being applied to the working side, no?

[Helmholtz]

I have input sine amplitude set to produce 100W output into a resistive load, then just remove the two tubes on one side and rerun.

Thanks.
I'm surprized by the huge increase in PD.
Grid signal, bias and plate load should essentially be the same.
Leaves less plate and screen voltage sag, but I didn't epect that much difference.

[Roe]

do you get any of this effect if the phase inverter is greatly unbalanced, driving one side way harder than another?

Not sure how this would be relevant given that one side is open circuit. It would just be more or less signal being applied to the working side, no?

No, I was thinking of a case where all output tubes are operational but the driver is greatly imbalanced. I've seen quite few cases of redplating on v4-v5 on 100w marshalls (without any redplating on v6-v7)

[Purgasound]

I can't see where this would have any good outcome. A 100W EL34 based transformer has a winding ratio to suit around a 1.7K primary impedance. Dropping two tubes creates a mismatch since two EL34's want to see a winding ratio suited for a 3.4K primary. That alone turns your 8 ohm tap into your 16 ohm tap, your 4 into 8, etc. When driving a load that's 2x higher than intended the current flow across the voice coil will be halved. That's going to come with other side effects as well.

When one side push/pull amplifier is at its peak current flow the other side will shut off so I can't imagine an incomplete waveform would be pleasing to the ears. I suppose with high levels of crossover and squarewave distortion it may shorten the gaps.

Whatever the desired goal is for tone, I'm sure there is a better way to get there. Of course nothing is wrong with experimentation but this isn't an avenue I would be interested in pursuing personally.

Cheers.

[Roe]

I can't see where this would have any good outcome. A 100W EL34 based transformer has a winding ratio to suit around a 1.7K primary impedance. Dropping two tubes creates a mismatch since two EL34's want to see a winding ratio suited for a 3.4K primary. That alone turns your 8 ohm tap into your 16 ohm tap, your 4 into 8, etc. When driving a load that's 2x higher than intended the current flow across the voice coil will be halved. That's going to come with other side effects as well.

When one side push/pull amplifier is at its peak current flow the other side will shut off so I can't imagine an incomplete waveform would be pleasing to the ears. I suppose with high levels of crossover and squarewave distortion it may shorten the gaps.

Whatever the desired goal is for tone, I'm sure there is a better way to get there. Of course nothing is wrong with experimentation but this isn't an avenue I would be interested in pursuing personally.

Cheers.

how about 4 el34s with a very unmatched driver that redplates v4-v5, but not v6-v7 (a common scenario)

[Purgasound]

how about 4 el34s with a very unmatched driver that redplates v4-v5, but not v6-v7 (a common scenario)

I'm not sure what you're getting at.

Yes, in a Superlead circuit V4 and V5 draw double the current as V6 and V7 when the power amp is driven into full distortion. The red plating is a sign something needs adjusting. If the amp is going to be driven that hard it's a good idea to increase the screen grid resistors to 2.2K and increase negative voltage for the bias. That should take care of any redplating. Cement wirewound for those resistors too. Those old shiny1K green ones and the black metal oxide ones can start to crack open when they get super hot.