Help please! 1973 Marshall Super Lead - EL34 to 6550s bias question

[Chuck H]

You'll be increasing local NFB in the PI and inside the global NFB loop. I've temporarily put that MV on for some bench tests when I want to decrease PI output. It tends to remove some "chaos" from the tone. It's subtle though. Not like a bad/good switch or anything Easy enough to try it for yourself. Four dots of solder and you test it before drilling a hole or putting the amp back in the cabinet.

[Helmholtz]

Connecting the outputs of a LTPI via a variable resistor does not produce local feedback. It's just an additional PI load causing strongly increased AC currents in the tubes. Actually a brute force method to reduce power tube grid drive. As the lowered PI output voltage decreases PA open loop gain, global NFB gets less. This causes the presence to lose its effect at lower NFB settings.

[Greg_L]

Ok thank you. I mainly just want to make sure it'll work. It seems like it's safe since no components are actually removed. If the pot fails I still have bias! I don't mind the reduced presence/NFB as I'm not trying to turn it into a bedroom amp. I just wanna knock the top off a little so I can sometimes use it live. I have an attenuator also.

I suppose I could put the NFB on a pot too to help balance the NFB when the PPIMV is used?

This amp has four speaker jacks and an unused "polarity switch". So I won't need to drill any holes.

[Chuck H]

Ok... Feedback might not have been technically correct as nothing is fed BACK. Fine. The phenomenon is similar though because the method is frequency cancellation rather than voltage division. In that regard it's the peaks and dips that really get squashed instead of a more relative balance of the same signal just at a lower level. That's what I meant. But good to keep it real.

EDIT: And when you consider the operation of cathode coupling there may actually be addition of local NFB. I'll have to consider it.

[Helmholtz]

the method is frequency cancellation

Well, if you call partially shorting an AC voltage "frequency cancellation"..

[Greg_L]

Well, if you call partially shorting an AC voltage "frequency cancellation"..

Isn't that sort of what the Type 3 does? It mixes the out of phase signals to cancel them out and reduce volume?

[Helmholtz]

Isn't that sort of what the Type 3 does? It mixes the out of phase signals to cancel them out and reduce volume?

Sorry I can't see any signal mixing or cancellation. Where would be the mixed output?

What happens is this:

Let V be the signal voltage at one of the PI outputs. Then -V will be the voltage of the other/inverted PI output. If we connect a resistor between the two, it will see the signal voltage 2*V and an AC current 2*V/R will flow through the resistor. This relatively high (depending on R value) additional signal current flows between the 2 PI plates and causes voltage drops at the internal plate resistances and thus lowers the PI output voltages.

In other words it acts as a dual voltage divider that uses the plate resistances as hidden series parts.

[Greg_L]

Sorry I can't see any signal mixing or cancellation. Where would be the mixed output?

What happens is this:

Let V be the signal voltage at one of the PI outputs. Then -V will be the voltage of the other/inverted PI output. If we connect a resistor between the two, it will see the signal voltage 2*V and an AC current 2*V/R will flow through the resistor. This relatively high additional signal current flows between the 2 PI plates and causes voltage drops at the internal plate resistances and thus lowers the PI output voltages.

That whooshing sound is all of that going right over my head!

I'm looking at it like it's taking a feed from each side of the phase inverter and summing them in the pot. They'll cancel each other out. But I freely admit that I don't know wtf I'm talking about!

[Helmholtz]

I'm looking at it like it's taking a feed from each side of the phase inverter and summing them in the pot. They'll cancel each other out.

The virtual summing/cancellation point would be exactly in the middle of the pot resistance at any setting, where signal voltage wrt ground must be zero (cancellation).
This can hardly explain the variable signal reduction at the plates.

[Greg_L]

The virtual summing/cancellation point would be exactly in the middle of the pot resistance at any setting, where signal voltage wrt ground must be zero (cancellation).
This can hardly explain the variable signal reduction at the plates.

Ok fair enough. I don't pretend to know this stuff.

I thought it was like how much or how little cancellation happens depends on where the wiper is at. Turn it down...lots of cancellation. Turn it up, very little cancellation.

[Helmholtz]

I thought it was like how much or how little cancellation happens depends on where the wiper is at. Turn it down...lots of cancellation. Turn it up, very little cancellation.

If the output impedance of the PI triodes was zero, a floating (variable) resistor between the plates would have no effect at all.
Some effects in electronics need to be explained in terms of current rather than voltage, more precisely using both.

Without being able to explain the details, cancellation is just an empty term.

[Greg_L]

If the output impedance of the PI triodes was zero, a floating (variable) resistor between the plates would have no effect at all.
Some effects in electronics need to be explained in terms of current rather than voltage, more precisely using both.

Without being able to explain the details, cancellation is just an empty term.

Okay sorry. I'll just try it and see what happens.

[Helmholtz]

Okay sorry.

No reason to apologize. I often see the term "phase cancellation" used as explanation for the type 3 MV. To me it's an internet myth.

[Greg_L]

No reason to apologize. I often see the term "phase cancellation" used as explanation for the type 3 MV. To me it's an internet myth.

Understood. Thanks. As long as this type of MV is safe for my application, I'm gonna give it a try.

[Helmholtz]

As long as this type of MV is safe for my application, I'm gonna give it a try.

Sure, why not?

I used this type of MV in my Vox AC30 in the early 70s, not knowing that it would later be called "Type 3".