Bypassing screen grid resistors???

[Chuck H]

Is this a good idea? It seems to be a modern trend with newer big bottle tubes to use higher screen grid resistors for safety. Though big caps are expensive, would there be any notable performance advantage to bypassing the screen grid resistors?

[soundguruman]

Performance advantage, like blowing the HT fuse, melting the power transformer, or blasting the tubes?
The screen grid resistors limit the current, and if they are removed, there is no more limit.

[NateS]

I think he meant adding a cap to ground to smooth out swings in the screen voltage with signal/load, not removing the screen resistors. Although it would tend to limit sag/compression, and as much as it does that, reduce the instantaneous limiting function of the screen resistors. I think what it accomplishes is basically a hi-fi goal, not a power tube distortion goal, though it might let you trade off for a different clip at the plates (by not limiting at the screens.)

[Chuck H]

I was suggesting a capacitor parallel to the screen grid resistors. Maintaining the same DC operating point but taking any AC to that circuits decoupling node.

There may be an application. There are plenty of amps where sag is introduced via other means. And it may sound a little different, as noted. Also, many amps now are using 1k rather than the 420r or 470r screen grid resistors typical of classic designs. So it may be that a circuit that partially bypasses the screen grids could get closer to the stock sound for vintage designs while still maintaining better DC characteristics. Just thinking out loud. I think this has been covered here a long time ago but I didn't search for it.

[Jazz P Bass]

The large value bypass capacitor on the screen resistor would make the amp:
1: Have less sustained power
2: Being an RC circuit, it would make the amp 'squishy'.
Simply my take on the subject.

[Chuck H]

The large value bypass capacitor on the screen resistor would make the amp:
1: Have less sustained power
2: Being an RC circuit, it would make the amp 'squishy'.
Simply my take on the subject.

Huh.?. I would have thought the opposite. By removing AC from the screen the electrons have a straight shot to the plate with no potential for phase errors. I know that the plate and screen operate somewhat analogous in a perfect world but I wonder how true that is.

[J M Fahey]

I tried it and soon ran away.
It makes the amp "Squishy", as JPB points out.
Exaggerated very unnatural sag and compression, in a bad way.
Unbearable.
And if you make caps small enough to minimize this effect, then you have bypass no more.

Well, 3 useful comments and a useless one, not bad for what MEF has become lately.
But on the other side, what would SGM understand of the word "bypass"?
He might even think this is "Cardiac Surgery Forum" and offer some of his wisdom
Which by the way would be something like: "don't waste time operating, that guy claims to be American but in fact is a fake copy made in Vietnam"
and:
"his heart failed because it was made to run on Japanese 100VAC, not Red White and Blue 120VAC"

[Chuck H]

Excellent! JPB, kudos for the accurate assessment! According to Juan you nailed it. And thanks Juan for the first hand experienced perspective (on two subjects ).

[Alan0354]

I just have question. What is the reason of putting in the screen resistor instead of straight connecting the power to the screen with a big filter cap?

For a small value cap parallel with the screen resistor is like removing the screen resistor at higher frequency. So back to the question why people put the resistor and what effect does the resistor to the sound?

I don't know the meaning of "Squishy", what does that mean?

I did experiment with the value of the screen resistor by increasing it from 470 up to 20K. It does not improve the sound to say the least. I think the sound loss the punch when increasing the resistance.

[Chuck H]

The resistor doesn't have to be there for the tube to function. In fact many amplifiers have been built without them, though not many guitar amps. So it's not going to be obvious why it's there. But the screen grid resistor limits current. An issue that becomes increasingly more important when the tubes are clipped and seems to be even more critical with newer production tubes.

[Alan0354]

The resistor doesn't have to be there for the tube to function. In fact many amplifiers have been built without them, though not many guitar amps. So it's not going to be obvious why it's there. But the screen grid resistor limits current. An issue that becomes increasingly more important when the tubes are clipped and seems to be even more critical with newer production tubes.

If it is for protection, then you have to worry about this bypassing the screen resistor as it is kind of shorting out the resistor.......but not quite as instead, the parallel cap should charge up and produce a DC. It will have a response time or something. I am just joining in, just guessing.

I think all talks are cheap, get an old pair of tube that you don't mind smoking it. Put it in, listen to the sound. More important, scope the screen and look at the voltage drop when you hit it full tilt and back calculate the power dissipation.

I am interested in your finding. I think there is a sagging effect that change the sound......hopefully for the better.

[Jazz P Bass]

I had a fella wanting to run a 1963 Fender Bassman (at 10 on the volume) into a half assed power soak.
The amp consistantly 'ate' power tubes.
The one on the left.
A 1K/ 5 watt screen resistor on each tube (factory 470/ 3 watt) did the trick.

[J M Fahey]

Alan, the bypass capacitor does not go in parallel with the screen resistor but from screen to ground.
2 things happen:

1) screen current is not constant , it raises with tube drive and *a lot* when the tube clips hard (situation not usually shown in curves because designers don't expect a sane designer to go that far).
So screen voltage drops with signal, and more when overdriven.
That's why in guitar amps screen resistors are a necessity, doubly so because tubes are used way beyond rated voltage.

2) screen voltage varies tube gain (to be more precise, transconductance).
If you drive a tube hard, gain/transconductance falls at the higher current, *following the driving waveform in real time* , so you have some tasty curve tops bending.
Not bad, it's part of tube sound.
*BUT* if you add a bypass cap from screen to ground, for the first second voltage will stay high, gain will be high.
As soon as it empties, voltage and gain die.
Time constant is way too long to follow the waveform, so instead of curve bending, volume goes down.
So pick attack sounds louder than normal, then all sound falls.
Then goes up as soon the screen bypass cap charges again and so on.
Like a poorly adjusted tremolo or compressor which tries to follow your picking.
Squishy indeed.

[J M Fahey]

Alan, the bypass capacitor does not go in parallel with the screen resistor but from screen to ground.
2 things happen:

1) screen current is not constant , it raises with tube drive and *a lot* when the tube clips hard (situation not usually shown in curves because designers don't expect a sane designer to go that far).
So screen voltage drops with signal, and more when overdriven.
That's why in guitar amps screen resistors are a necessity, doubly so because tubes are used way beyond rated voltage.

2) screen voltage varies tube gain (to be more precise, transconductance).
If you drive a tube hard, gain/transconductance falls at the higher current, *following the driving waveform in real time* , so you have some tasty curve tops bending.
Not bad, it's part of tube sound.
*BUT* if you add a bypass cap from screen to ground, for the first second voltage will stay high, gain will be high.
As soon as it empties, voltage and gain die.
Time constant is way too long to follow the waveform, so instead of curve bending, volume goes down.
So pick attack sounds louder than normal, then all sound falls.
Then goes up as soon the screen bypass cap charges again and so on.
Like a poorly adjusted tremolo or compressor which tries to follow your picking.
Squishy indeed.

[printer2]

Does this help?