screen voltage higher than plate

[ohrider]

i've built an amp without follow an precise schematic ,but pasting various parts of other shematics , like the fender deluxe reverb blackface for the power amp but with 30 uf caps and without feedback ,and soldano atom 16 for pre .......the sound is good, but the screen voltage is little higher ,5 volt, than plate ,may be a problem ?.......sorry for my bad english
thanks

[Jazz P Bass]

What value screen resistors did you use?
1K/ 5 watt is sufficient.

[Bruce / Mission Amps]

I think that is too little info.... you didn't say exactly what you are doing nor what power tubes you are using....
A pair of 1K@5W screen resistors (on the sockets) is probably over kill for 6V6s... probably 6L6s too, but more correct for EL34s. etc etc
6V6s and 6L6s don't draw a lot of screen current.
If you are using a choke, the choke you are using has too little of DC resistance to drop the screen voltage down from the node you are taking it.
Nothing wrong with that either.
The plates of the power tubes are drawing so much current, with respect to the screens, that the DC resistance of the output transformer is dropping more voltage then the choke can, even with the load of the preamp tubes added on.
So, try putting a new 510 ohm to 1K2, 5 watt resistor after your choke lead and before the common point of the two screen connections... that way all (and only) the screen current has to flow through the new resistor. But remember, it's after the choke and nothing else but the screens are being supplied by this resistor.
That method will loose a few more V of DC easily enough.

[ohrider]

ok ,you have right,i using 2 6v6gt,and 5ug for rectifire but i want change with 5u4gb, the power traf is hammond 290axe 650 vct 70ma,0T is hammond1750h 6600 ct to 8 ohm,hammond 4h choke , i have 349 volt on plate and 353 volt on sceen , 26 volt on pin 8 cathode bias with 470r 5w resistor ,in the begining i have used 2 470r resistor 2w on screen pin and after i tried with 1k ,but nothing is changed
thanks

[ohrider]

OK this is the schematic that i use for the power amp, but the 2 first caps is 32 + 32 uf multicaps , the cap after the choke is 30 uf ,then one 30 uf and the last 22uf, but with cathode bias

[g1]

I'm not sure if you understood what Bruce was saying so here's your schematic with the resistor added in series with the screen supply.

[ohrider]

ok thanks now is very clear, i try this way

[Wilder Amplification]

Amps that use 6V6, 6L6 and KT66 with a 470R screen resistor oftentimes will have the screens a few volts higher than the plates will. This is because the screens on these valves don't draw enough idle current for a 470R resistor to drop the voltage to the screens as much as the DC resistance of the OT primary drops the voltage to the plates, which are drawing much more current than the screens.

If you have 2mA of idle screen current through the 470R resistor, you will only drop -

2mA x 470R = 0.94V

So the screen voltage will only be about 0.94V lower than the B+ rail voltage under quiescent conditions. The DC resistance of the OT primary will drop a couple more volts more than that to the plates. Even though the DCR is lower, you have considerably more current flowing through it, so again Ohm's Law takes over.

[Bruce / Mission Amps]

OK insert the resistor as I pointed out in your artwork... use a 510 ohm, 680 or 1K, 1k2, etc ... I'd go with at least a 3-5 watt resistor to be safe.

Oops... just noticed everyone else already correctly commented and my previous response... sorry.

[ohrider]

ok thank you all !!! tomorrow o after tomorrow i try in this way

[ohrider]

ok ,i try before 820r resistor , now i have put on a 1k2 resistor .i have the same tension between plate snd screen, 359 volt ,28 volt on cathode accross 470r bias resistor,28 milliampere for each 6v6 ,you think is good ? or i must put on a 1k5 instead of 1k2 ?
thanks

[Wilder Amplification]

ok ,i try before 820r resistor , now i have put on a 1k2 resistor .i have the same tension between plate snd screen, 359 volt ,28 volt on cathode accross 470r bias resistor,28 milliampere for each 6v6 ,you think is good ? or i must put on a 1k5 instead of 1k2 ?
thanks

I don't see why you're so concerned with the screen voltage being a few volts higher than the plate. It's not like it stays there. It's only that much higher when there's no signal present. As soon as you hit it with signal, screen current increases, which causes the screen resistor to drop more voltage off of the screens.

Increasing the screen resistor value will cause the screen voltage to drop more under signal, which reduces the transconductance of the valve and will have more of a "sag" effect. If you like a "tight" amp, this would be counterproductive to that.

[ohrider]

it' only paranoia ,it's my first "free" schematic
ok .i like this sound, i leave it
thanks again for all

[pdf64]

Now that ohrider is happy, I'll make a minor hijack and ask people here to review this thread over at the amp garage, on this same topic
The Amp Garage :: View topic - G2 power testing and conduction angle
The thread originator seems confident that bringing the quiescent screen voltage even slightly negative WRT plate voltage has significant benefits. My thinking is that he's basically inducing sag to the screen supplies but my theory on the fundementals is a bit sketchy and he may on to something. Pete.

[Steve Conner]

I agree. I think quiescent screen voltage is about as relevant as a car engine's idle speed is to its full power operation. What's important is how the screen voltage varies under load. The simplest example of this is, the power output is determined by the instantaneous screen voltage at signal peaks.

The curves on the tube datasheets are made with a regulated screen supply and no screen resistor. Adding a resistor changes the shape of them in a different way to altering the voltage.

I've never seen those formulas for conduction angle and I'm not sure what it means. I would certainly be wary of working out anything involving vacuum tubes to 4 decimal places, when they can generate 10% distortion. But I do know that if you alter the screen voltage, you have to rebias. If he forgot to do that, that would explain the move towards Class B.