voltage divider for screen grids

[ChrisM]

Yes, when I have measured the voltages to ground the plate and screen are about the same.

Andy had suggested measuring them at the socket to see the difference in potential between them.

The same or about the same?

This is where a good accurate meter is helpful, one what displays a couple significant digits after the decimal point.

[markr14850]

This measurement seems odd to me. The plate supply will have a ton of ripple on it. There will be much less on the screen supply (and it will be slightly phase shifted). Any numbers that your meter shows you when comparing these two will totally depend on how your brand of meter integrates the ripple AC component.

[Structo]

Well I say about because the reading was jumping around.

Mark, what you say makes sense.

I have a Meterman 37R which is supposed to be equivalent to a Fluke.

[Andy Le Blanc]

I hadn't run into excessive ripple on plate voltage, yet...
Been thru a pile of different amp brands too...
Yes try measuring the difference, it was happenstance that I measured it there.
And I was surprised that it has been a stable reading in so many amps...
try the difference and don't assume the ripple will be an issue before you try it

crazy ripple would be a problem to itself, bad cap or poor parts value selection, ground on the power rail, not enough current flow or minimal voltage drop to make for an effective filter stage.

I use krapp meters, I'm cheap and and don't mind pitching one if it gets cooked.
I bet the sensitivity in mine is low, but it will measure o.xxx ma., its a guitar amp after all...

[jbrrrrr]

This old Mullard trick gives you one more design option. A common divider with individual screen resistors off the divider. To fit the discussion above that would mean a 500 Ohm / 220K divider with individual 1K screen resistors from the divider. That would give similar sound to the individual dividers with bypass caps option but without the need for any bypass cap (always nice to ditch electrolytic caps when ever possible).

Sorry for bringing up this necrotic post, but the content of the screen grid bleeder resistors has been on my mind lately and there's not a whole lot else touching on the idea that Andy has presented.

I'm particularly interested in this arrangement mentioned above (at least my understanding of it):

B+ screen supply node -> 500r
                                           |
                                        220k -> 1k screen grid resistors (per OP tube)
                                           |
                                         GND


So there's a common 500r series resistor with the supply, and the screen grid resistors are taken from the junction of that 550r - 220k to ground divider. Has anyone ever given that a try? The logic seems sound in that there's a shared source resistor connected to the B+ screen node to respond to both screens demands, but each also gets its own series resistor to the grid, and the voltage divider help mitigate the plate & screen voltage relationship based on what Andy was describing with this technique.

[martin manning]

A shared screen resistor is not much different from separate ones wrt current limiting since one side is in cut-off at high power anyway. The advantage of separate ones is that they can be on the sockets and therefore better at preventing oscillation. Adding the large resistance to ground will pull the screen voltage down a few volts, and maybe get it to be less than plate voltage if that's the goal, but I'm not sure there is any advantage to that. I've never tried it though, so it could be that Andy's report of smoother tone may be true.

[Helmholtz]

If e.g. the voltage across the 220k is 450V, it draws a current of 2mA and causes an additional voltage drop of 1V across the 500R.
So removing the 220k would increase the screen voltage by 1V only.

[jbrrrrr]

If e.g. the voltage across the 220k is 450V, it draws a current of 2mA and causes an additional voltage drop of 1V across the 500R.
So removing the 220k would increase the screen voltage by 1V only.

Thanks, Martin & Helmholtz. It always sounded a little odd to me but I'm game to try any weird arrangement that people find to be beneficial. I can kind of understand forcing the individual screen grid resistors to have to sip from the same 500r straw, but the voltage divider wasn't something I could wrap my head around wrt why it would make much difference. I'll give it a shot just for the hell of it, and because it seems like it did something positive for at least one person, consistently.

[sluckey]

but the voltage divider wasn't something I could wrap my head around wrt why it would make much difference.

That 220K was never meant to be part of a voltage divider intended to reduce the screen voltage. It functions as a bleeder for the screen node filter cap. If you look at the original post in this thread you'll see that the OP said so.

FYI, back in the good ole days voltage dividers were commonly used to drop B+. The rule of thumb was to have the current flowing through the divider resistors to be much greater than the current required by the load. We're talking about huge high wattage resistors. A 1K and 220K/1W just ain't realistic.