Pentode guys: important to decouple g2 supply between stages?

[leadbelly]

Background: one of the amp ideas I keep mulling around in my head is a variant of Pete Millett's E-Linear amp with pentode finals.

So, since it is commonly (universally?) recommended to decouple the plate supply for the input/driver stage from the plate supply for the output stage, would the same also hold true for a screen grid supply?

Or if you want a specific example to make it more meaningful, let's say we're talking about about an SE amp with a small signal pentode driving a power pentode. Would it be OK for the g2 supply, let's say composed of a resistor/0D3+cap, to supply all 4 g2's for the amp?

I know this may touch on how some feel that only a true regulator should be used for g2, so feel free to expound on that topic if you want TIA

[DigitalJunkie]

I suppose that you could use the same supply for all of the screens,That is,assuming that they are all going to be at the same voltage.
I would think performance may suffer a little bit,but I'm not sure.

I think it's probably better to use separate RC-sections for each screen,tweak the R for the voltage needed,and put a cap to ground for bypassing.

[leadbelly]

Originally posted by DigitalJunkie
I think it's probably better to use separate RC-sections for each screen,tweak the R for the voltage needed,and put a cap to ground for bypassing.

Thanks, but I'm not sure that's a good idea, as the VR tubes are prone to oscillation.

[SY]

I would absolutely regulate, and regulate tightly, the output tube screens. The input tube ones are much easier to deal with since they might be swinging a fraction of a milliamp.

[Miles Prower]

Originally posted by leadbelly
Or if you want a specific example to make it more meaningful, let's say we're talking about about an SE amp with a small signal pentode driving a power pentode. Would it be OK for the g2 supply, let's say composed of a resistor/0D3+cap, to supply all 4 g2's for the amp?

It would be highly unusual to find a situation where the finals and the small signal pentode operated at the same Vsgsg. It probably wouldn't be desireable anyway. The pentode finals will be operating with their cathodes at AC ground. Since that's the case, the Vsg-to-ground voltage will be the same as the Vsgk. No problem there. Small signal pents often do not operate that way. If the cathode is used as a gNFB summing node, then the cathode will not be at AC ground, and the Vsgk will move with the cathode voltage. This is undesireable. Small signal pents are best supplied from a voltage divider (regulation not so good as a regulator, but better than a series dropping resistor) with the screen bypassed to the cathode. This will make Vsgk more nearly constant. Under most operating condidtions, the Isg of a small signal stage doesn't vary nearly as much as the screen current to the finals, so screen regulation isn't such a big issue.

What I'd do is use an active regulator to supply the screens of the finals. It can be hollow state or solid state as you please.

"resistor/0D3+cap,"

I would also watch out for this kind of configuration. Unlike a Zener diode, these gas discharge VTs have a negative resistance characteristic, and adding an RC network will make a nice pulse oscillator. You can't treat 'em like they were Zeners.