Screen resistor as on/off switch

[lowell]

How big does a screen resistor need to be to shut off a tube? To be specific say we have a twin reverb with a 470ohm resistor, and a Va around 460v, fixed bias at 70%: 46ma. I may sub in a pot to test this if anyone hasn't already found the answer to this... Just a thought I had and I'm not at my bench.

[booj]

If you control the conductivety of the tube with the screen, yer messin with HV in whatever circuit you choose, but it would be possible to switch the screen connection from the 470 ohm resistors to ground which would shut off the tube. But then you'd have the ends of the 470 ohm resistors with HV and nothin to do....

[lowell]

Ok yeah, I guess it could be done by just disconnecting the screen connection altogether. Do you know if this creates a POP?

[Old Tele man]

...here's a worked mathematical example of what would happen to a 6L6GC operating at Vp = 450, Vs = 400, and Vg = -37 when you "remove" the screen voltage Vs:

Ik = G * (Vg + Vs/µ1 + Vp/µ2 )^(3/2)

where:
Ik = Cathode current, ie: Ik = (Ip+Is)
G = Tube perveance, ie: 6L6 = 0.0010 A/V^(3/2)
µ1 = Screen (triode) amplification factor; 6L6 = 7.99 : 1
µ2 = Plate amplification factor; 6L6 = 135 : 1
Vg = Control grid bias voltage; ie: -37 Vdc
Vs = Screen grid voltage; ie: 400 Vdc
Vp = Plate voltage; ie: 450 Vdc

Ik = 0.0010 * (-37Vg + 400Vs/7.99 + 450Vp/135 )^(3/2)
Ik = 0.066A; so, roughly Ip = 0.063A and Is = 0.003A

...now, set Vs = 0.0Vdc, and repeat:

Ik = 0.0010 * (-37Vg + 0.0Vs/7.99 + 450Vp/135 )^(3/2)
Ik = 0.0010 * (-33.67V)^(3/2) ...which cannot work.

...the "negative" voltage within the parenthesis (the "equivalent diode voltage") means there's NO "positive" plate voltage to attract electrons from the space-charge around the cathode...ie: the tube cannot "conduct."

...the tube can only "conduct" when the equivalent diode voltage is greater than zero volts.

[booj]

...here's a worked mathematical example of what would happen to a 6L6GC operating at Vp = 450, Vs = 400, and Vg = -37 when you "remove" the screen voltage Vs:

Ik = G * (Vg + Vs/µ1 + Vp/µ2 )^(3/2)

where:
Ik = Cathode current, ie: Ik = (Ip+Is)
G = Tube perveance, ie: 6L6 = 0.0010 A/V^(3/2)
µ1 = Screen (triode) amplification factor; 6L6 = 7.99 : 1
µ2 = Plate amplification factor; 6L6 = 135 : 1
Vg = Control grid bias voltage; ie: -37 Vdc
Vs = Screen grid voltage; ie: 400 Vdc
Vp = Plate voltage; ie: 450 Vdc

Ik = 0.0010 * (-37Vg + 400Vs/7.99 + 450Vp/135 )^(3/2)
Ik = 0.066A; so, roughly Ip = 0.063A and Is = 0.003A

...now, set Vs = 0.0Vdc, and repeat:

Ik = 0.0010 * (-37Vg + 0.0Vs/7.99 + 450Vp/135 )^(3/2)
Ik = 0.0010 * (-33.67V)^(3/2) ...which cannot work.

...the "negative" voltage within the parenthesis (the "equivalent diode voltage") means there's NO "positive" plate voltage to attract electrons from the space-charge around the cathode...ie: the tube cannot "conduct."

...the tube can only "conduct" when the equivalent diode voltage is greater than zero volts.

Ok...your point being?
I'm curious now. Do these numbers mean you can drop the screen to a short to the ground and it will shut the tube off without a pop? Or does it mean something bad will happen and the tube gets sucked into a negative universe?

[J M Fahey]

Hi Old Tele Man.
Impressive.
We need more stuff like that.
Hi Lowell: If you are looking for a "thump-less" standby switch, I guess you might connect your screens not in the "regular" way, but on their own side-chain with a small time delay as shown.
The time constant is only a few milliseconds, but it might turn a sharp "click" into a soft thump which besides would be almost cancelled by the push-pull action.
When "on" the screen voltage and tube current would "ramp up"; when off the screen consumption will probably do the same.
You try it and tell us.

[Steve Conner]

While I agree with OTM's math, I'd be a little worried about switching the screens directly. I think a quick cutoff of the plate current, and noise from switch contact bounce, might cause a high voltage kickback from the OT.

JM's idea of a RC circuit to "slug" the screen voltage rise and fall would cure this, too.

OTM: If the screen isn't grounded but left floating, does the math still apply? Or will the screen pick up some positive voltage by itself?

[Old Tele man]

While I agree with OTM's math, I'd be a little worried about switching the screens directly.

...agreed, but our USN APS-20E search radar (also used in your Shackleton AEW planes) actually did use Vs-control to "sleep" certain circuits, but using a series-path voltage regulating tube, not a mechanical "switch."

OTM: If the screen isn't grounded but left floating, does the math still apply? Or will the screen pick up some positive voltage by itself?

...excellent question, with multiple (time-wise) answers. At first, nothing untoward would occur as the screen would suddenly go to zero volts; but, eventually, the unconnected screen would slowly accumulate electrons (like the grid does with contact bias) and become more and more "negative," but without a discharge path (assuming it is the Vs circuit that's been "opened")...but, still nothing would happen as long as there was no discharge path for the Vs.

...now, if the Vs was sourced from a voltage-divider from B+ and the lower resistance were still "in the circuit," then things become "interesting" because now the screen can truly "act" like a plate and discharge the accumulating electrons collected by the screen...how much current flow would be determined by the "capture-ratio" (physical ratio of screen grid diameter to grid-to-grid spacing) of the screen itself and the resistance to ground in the screen circuit. Definitely would "stop" plate current.

...the screen grid controls tetrode and pentode operation, not the plate (it's strictly the 'electron catcher'). And, typically, plate voltage alone cannot overcome screen grid influence and control.

[Gregg]

A "screens standby" is implemented by Marshall in their JVM410:

http://homepages.slingshot.co.nz/~da...2-iss2-2PA.pdf

This way you can use the preamp only and run it through the simulated output or whatever you like.