voltage divider for screen grids

[Andy Le Blanc]

have a 72 twin on the bench, it came in with an eaten power tube and BBQ screen grid resistors.

I replaced the screen grid resistors with 1k 1W at the tube socket, they where what was on the bench.

I then took the opportunity to install a voltage divider with a bleeder to ground between the usual point on the fender component board before the screen grid resistors at the tube sockets.

So... on each leg of the push-pull circuit, the screens are supplied at the junction of a 1k and 220k resistor.
the 1k goes to the usual tap on the power rail and the 220k is a bleeder to ground.

This produced a consistent 7v difference between the plates and screens for the 6l6 in the power side of the amp, the plates being consistently positive relative to the screens when the difference between the two is measured at the tube socket, the plate voltage with a .045A static bias for each tube was riding around 441V for about 20W dissipation per tube.

I have been playing with this approach for some time now and it seems to produce a warmer smoother toned amp consistently, the character deepens in the nicest way.

if the screens are slightly positive they must first go to 0 before going negative relative to the plate as signal is applied to the circuit.

I beginning to see it as a harshness or a hardness in the tone of amps.
inner modulation and distortion maybe...

[Colossal]

Very cool Andy, thanks for posting! Nice there is an audible difference as well as a performance improvement. I like it as a means to keep the screens in check but the smoothing effect is an added bonus.

[M Fowler]

Thanks Andy I'm going to try this not sure which amp yet.

Some amps will have the screens at or near the same voltage as the plates I noticed this more in four EL84 cathode bias amps though.

Mark

[Andy Le Blanc]

the miss just come into shop asking which amp it was I was playing cause it was so pretty.
I've been surprised by it, some thing so simple, that little measurement of difference
between the plate and screen is so telling, and just that little measure of control of that voltage relationship can make such a change.

it not the value of screen grid resistors, its the measured difference between the plate and screens

[Andy Le Blanc]

I've set this up on el84 to kt88 and everything in between that I could lay my hands on in the shop, its up to you and how you want to set up the power side of your amp irregardless of tube type, like any other sweet spot for loading , bias, or operating point any where else in YOUR circuit

[rock_mumbles]

Just to clarify this ...

from the screen (power) supply you have installed a 1k/220k voltage divider that then connects to the 1k screen grid resistor(s)

and the 1k resistor in the voltage divider needs to be the same wattage rated as the screen grid resistor.

[Andy Le Blanc]

1k / 220k the 1k is the same as the screen resistors, you could use any value to "set" the voltage where you want it.

The bleeder can not be any smaller than 220k, or it will draw significant current and then the component wattage becomes less practical.

The 220k can be 1/2 watt however, which is nice, its just a bleeder

so its very practical... I've even placed 220k bleeders right from the screen grid terminal on the tube socket going to ground,
does the same thing but so far I've like the result of the voltage divider before the screen grid resistors
its subjective, they both achieve the same thing but having something to act as a grid stopper right on the tube socket sounds a little better maybe.

[surfsup]

Mathematically, a 1k/220k will produce 1.5v less with a 350v supply which is pretty minor. Surprised to hear it makes this much difference to your ear. Maybe the amp you did this with is just a nice sounding amp?

[Andy Le Blanc]

the result after screen grid resistors was a few "less" volts at the tube socket.

nope its just another crap silver face, filled with ceramics and blue and brown turd caps

its simply fine tuning screens voltage to insure that the plate is positive
I was surprised... get into an amp and measure the difference between the plate a screen at the tube socket.

I got into a textbook circuit, and when I played it I heard a distortion, that fender fizz, when I measured it the screens were slightly positive, I could hear the distortion, and everything was technically correct.

getting that extra step of control allows you to measure and set up that aspect of the circuit the same as you would the bias, it really seems to make a difference

if your not too busy cloning, get in there and play with it

[Bob Simpson]

Andy,

do you have a photo or drawing of where & what you did?

TIA

Bob Simpson

[ChrisM]

Andy,

do you have a photo or drawing of where & what you did?

TIA

Bob Simpson

Add a 220K 2W or bigger resistor from the screen grid to ground on each output tube.

I'd use 5W. Once the screens start drawing current that resistor is going to get hot!

[Structo]

Hey Chris, haven't seen you around these parts in a while.

Hope everything is going well for you.

I am still a bit confused by this voltage divider because so many posts have been made about it.

At one time I thought all that was needed was a single 220K 5w resistor from the screen node after the choke to ground.

Then it sounds like you have to have a 1K screen resistor on each tube and a 220K resistor from each screen pin on each tube to ground.

Can somebody draw me a picture?

Gaz had posted this schematic but I don't understand the need for the additional filtering there.

[img1131]http://i55.tinypic.com/28icnjl.png[/img]

Dazed and confuse.

[Andy Le Blanc]

its simply an extra resistor off the same point where the screens are usually feed from then a bleeder to ground

[John_P_WI]

Tom, I believe the attached file is what Andy is talking about.

Surf, is it possible your calculations are neglecting the current draw of the screens on the first 1k resistor? If you calculate the v drop across the 1k resistor with an assumed 5% current draw of two output tubes biased at 45ma being fed from this node and add to the v drop that you calculated, you will come up with a much closer value to 7 volts. I believe Andy suggested that he has two voltage dividers, one feeding each pair of tubes on "each side of the OT".

Andy, IF this is not correct I will remove the drawing....

John

[John_P_WI]

Damn Andy, you are too quick....

Take care,

John