Separate bias voltage

[Enzo]

You proposed one pair of tubes set right at cutoff so they immediately start conducting, and as you cross zero they hand off to the other side right at zero. That is class B. The extra tube on each side is biased warmer so it is on more than half the time - class AB. So what you have is a pair of seriously mismatched tubes in essence. The tubes themselves may be matched, but the circuit makes tham act as if they were not.

In yours, on each side, two tubes conduct the entire positive half cycle, then as we cross zero to the negative, one of the two shuts off immediately, the remaining one continues to conduct some portion of the cycle, but not all. (If all, it would be class A)

Have I stated it incorrectly?

[Chuck H]

I'm pretty sure the OP is talking about a separate bias for each of the four tubes. And Mike is talking about the possibility of biasing one tube on either side of the push/pull one way and the other tube on each side another way. Similar to my suggestion above.

[mozz]

Maybe dual bias controls and 2 humpots.

[Mike Sulzer]

You proposed one pair of tubes set right at cutoff so they immediately start conducting, and as you cross zero they hand off to the other side right at zero. That is class B. The extra tube on each side is biased warmer so it is on more than half the time - class AB. So what you have is a pair of seriously mismatched tubes in essence. The tubes themselves may be matched, but the circuit makes tham act as if they were not.

In yours, on each side, two tubes conduct the entire positive half cycle, then as we cross zero to the negative, one of the two shuts off immediately, the remaining one continues to conduct some portion of the cycle, but not all. (If all, it would be class A)

Have I stated it incorrectly?

The tube that is off at idle turns on only as the tube on the other side (that is part of the AB) turns off. Thus two tubes are on all the time.

[Enzo]

Oh, so maybe you are saying we have an AB tube on one side, and what I have been calling a B tube opposed, which is biased way beyond cutoff so it only comes on late in the opposing sycle. So if AB is on 70% of the cycle, the opposing tube is on 30%. Is that your premise? And then the remaining two tubes are the same relation, but on the other side of center?


SO you have invented something that hands off hard as class B does, but way assymetrically, like my 70/30% example? So you are asking the 30% tube on each side to only do anything on peaks? And in the meat and potatoes middle amplitude stuff, we are only using two of the tubes?

[Chuck H]

I think he means this:

[Mike Sulzer]

I think he means this:

That is it, modulo exact settings. First I would try lower than normal bias current in the AB tubes since we have the B tubes to turn on and contribute. The B tubes would be biased more off than normal B in my first attempt, but the possibilities are various. Somebody must have done this; anybody have a reference?

[g1]

Boogie 'simul-class' ?

[Mike Sulzer]

Boogie 'simul-class' ?

I guess not. Wikipedia: "Mesa/Boogie's "Simul-Class" system, where two of the power tubes (always 6L6s) run in class ABpentode while the other two tubes (either 6L6s or EL34s) run in class A triode."

[Chuck H]

I guess the notion of running one tube on each side of the push/pull in a different class is the similarity. Your idea would use the notion to potentially improve performance as well.

[überfuzz]

Thanks for all the input, so much I'm having trouble keeping up...

I have tried to simulate different set ups where I have separate bias to the power tubes in a 6L6 power section. It runs smoothly when I have separate bias to each side in a push pull setting, but when I try to have separate bias to each tube i get instability. As of now I don't know whether this is caused by something wonky in the simulation or plain instability in the circuit.

The attached picture is the current simulation. I'm still tinkering with the values on resistors and caps, so go easy on those. I use this schematics to visualize how things is going to sit in the amp, hence the topology.

[Malcolm Irving]

... The B tubes would be biased more off than normal B in my first attempt, but the possibilities are various. ...

Isn't that class C ? Just asking.

[Gregg]

...but when I try to have separate bias to each tube i get instability.

I don't know what kind of instability you're talking about but it works in practice.

[überfuzz]

I don't know what kind of instability you're talking about but it works in practice.

You've tried it?

[Mike Sulzer]

Isn't that class C ? Just asking.

Maybe, but n true Class C a tube conducts for significantly less than half a cycle with the waveform smoothed out with, for example, a resonant filter.

But calling it class ABC... does have its appeal.