Regulated screen supply

[mashaffer]

Just want to check my understanding here. Pentodes are said to perform best when the screen supply is regulated. I presume this is because the spacing of the characteristic curves changes with changes in screen voltage.

It has also been pointed out that the plate supply should/must be regulated as well. Intuitively this indicates to me that the spacing of the curves must be primarily dependent on the DC potential between the plate and screen rather than that between the cathode and the screen.

Is that a correct interpretation?

If so it would seem to follow that any method which nails down the plate to screen potential would substantially solve the problem. By regulating both supplies one regulates the difference between the two in a fixed bias amplifier. To the extent that cathode voltage can vary in a cathode biased amp this ideal would I suppose be compromised.

So in the case of the cathode biased stage would it not be advantageous to set the screen voltage with respect to the plate and thus ease the requirements for tight regulation in the plate supply itself? I am thinking of either some sort of shunt regulation scheme or possibly a bootstrap of some kind.

Am I off in left field here?

[Eli Duttman]

Mike,

AFAIK, it's OK for g2 B+ to be regulated, while anode B+ is not. Remember, that I_B in pentodes is controlled (for the most part) by the potentials on g1 and g2, not the plate.

What I am certain about is that if anode B+ is regulated, a bias supply must be regulated too. This true for both triodes and pentodes.

Pentode linearity is excellent when g2 B+ is a fraction of anode B+. Bad things can happen, distortion wise, when g2 potential is greater than the instantaneous plate potential. How much of a differential between g2 and the plate is "best" draws more than 1 opinion.

[mashaffer]

Thanks Eli, as I think about it regulating the screen to plate potential just equals triode mode. So as long as the screen is well regulated with respect to the cathode a well filtered (maybe CLCRC) plate supply is all that is required right?

[ray_moth]

Eli says:

What I am certain about is that if anode B+ is regulated, a bias supply must be regulated too. This true for both triodes and pentodes.

True, but in the case of a pentode it is g2, not the plate, that is acting as the anode.

With pentodes there is no point in regulating the plate supply and whoever "pointed out that the plate supply should/must be regulated as well' was wrong.

Regulating the g2 supply is important for output pentodes and, as Eli says, if you're using fixed bias then regulating the bias supply is important too.

In the case of small signal pentodes, g2 'regulation' is commonly achieved by bypassing g2 to cathode with a capacitor, usually in the range of from 0.01 to 0.1 uF.

[Wavebourn]

I have SG4S with 120V Zener in series, so have both +150 and +270 for input and output tube's grids (+270 through IRF730 follower, +150 directly from SG4S through 1K and 10 uF).

[mashaffer]

Ahh, I bet I was remembering the instructions incorrectly. So in fixed bias regulate bias and screen. In cathode bias regulate only screen.