screen resistor

[catalin gramada]

Hallo. Can someone explain what tone diferences to expect using a shared screen resistor vs individual ones
? Supposing I intend to use a common 1k/5w instead two pcs on each tube for 2 el34 . I read somewhere to half the value and double the power but I think it is not a valid choice.This should be right just for iddling when the average screen current is small.for hard driving condition have to calculate keeping în mind the max admisibile screen power dissipation for one tube not for both cause they dont draw the same ammount of current in the same time . Am I wrong please? Over that my question regard diferences in tone for the same power dissipate by screens. please. It is hard for me to understand into electrical way how screens works relative to the plate in hard driving conditions.Thanks.

[Gingertube]

Screen resistors do 2 jobs, they act as grid stops to prevent parasitic oscilation, and they provide some protection of the screen grid against over dissipation.
The grid stop function requirement means that EACH tube should always have a screen grid stop resistor of say 100 to 150 Ohms minimum.

This is because just as you get gain when you apply a signal to grid 1, so too do you get gain when you apply a signal to g2 (the screen).

The common screen resistor is a Mullard idea. In a push pull amp, as the drive signal swings in one direction, one side of the push pull draws more screen current the other side draws less, and then they swap for the signal swing in the opposite direction.
The net result is that the total screen current draw is reasonably constant. A single screen resistor will therefore give reasonably constant screen voltage, a desirable thing for HIFi Amps. In this case there is no screen signal voltage for that gain to respond too. It will sound "tighter" with better "attack".

When changing to individual screen resistors of larger sizes, the screen voltage drops a little as the screen current goes up and so you get a compression effect introduced. There is now a signal at the screen caused by the voltage drop across the screen resistor in response to the screen current. AS the tube tries to conduct more anode current the screen current goes up and the screen voltage drops. That drop in screen voltage will be acted upon by the gain at the screen grid to reduce the anode and screen currents. It is Negative feedback. This introduces a compression effect. That can be very nice for instrument amps PLUS that compression tends to protect the screen from over dissipation.

Pure pentodes such as the EL34 have higher screen currents than beam power tetrodes (like 6L6 etc). That has resulted in the EL34 having a reputation of having fragile screens. THis is not really the case, but it is true that they are more susceptible to damage from screen abuse.

For EL34 Push Pull you will usually see the following recommendations:
1) Triode Mode - 150 Ohms between screen and anode on each tube.
2) HiFi Pentode Mode - 470 Ohms common (PLUS 150 Ohm individual is recommended)
3) HiFI Ultralinear - 1K individual (has to be individual to implement the ultralinear connections)
4) Instrument Pentode Mode - 1K (minimum) individual

Cheers,
Ian

[catalin gramada]

Thanks for explanation Ian. I know there are compulsory small values over the socket to prevent posible oscillations but in the same time there are a lot of vintage circuits which works without it. Have to consider the risk of osscilation is related by gain level .input impedance or tube type ? I saw the mullard sheets and they have very precise direction în respect with application used. But did not found any description what.s happening with screen when hard driven. I know it goes into severe conduction and start to draw a lot of current but have no ideea how much relative to a momentary plate voltage. What I know it is what was allready said. 1k it is safe enough to be used with contemporany tubes as time we not exceed a decent voltage level. said max.400v when dimmed. My question was just for tonal diferences. Thanks for You answer. Catalin

[Jazz P Bass]

Not too sure that I can help on the "Tone" question.

Consider the fact that the higher the screen resistor, the lower the screen voltage.

The lower the screen voltage, the less the tube conducts.

An extreme example is the Univox amps that run 700 some volts on the plates.
The screen circuit feeds the screens with 350 some volts.
(which helps the poor, modern 6L6 tubes survive)

Remove the screen voltage & the tube will not conduct.

[loudthud]

I've experimented with the common screen resistor on some Traynor amps. I never noticed any big shift in the tone, but I never set up an experiment with a switch for rapid comparison.

[catalin gramada]

Yeah.I saw traynor used shared soluțion even with a small 47 ohm individual resistors over sockets. I have to wait for a oportunity to receive a Marshall to test it without individual resistors at all. There are a lot of jtm jmp 67-72 era which not used for 50w versions. Maybe keeping the power tubes into a border with instability add a rich content of over tone in some way.but screen protection should be a must still

[teemuk]

Gain compression effect naturally happens only if modulation of screen voltage is allowed by the circuit design.

If screen voltage is derived from plate voltage supply you have modulation from plate voltage dropping under load. Screen supply filtering will counteract this modulation to some extent and basically the effects of this modulation are not instantenious since screen supply voltage has a finite speed of shifting, determind by the RC filter circuit involved.

The individual screen resistors will also modulate, since screen current naturally develops a voltage drop over them. Since there are no RC time delays involved the compressive effect of screen stopper resistors is instantenious. Compression is naturally asymmetric in nature, since screen current draw is high during amplification of one halfwave and correspondinhly low at the other. Combined effect of push-pull operation is of course symmetric. All screen stopper resistors will modulate screen voltage to some extent, but in my experience you need to go really overboard with them to make the effect distinct. e.g. If screen resistors were 470R - 1K before change them to something like 10K. Basically, the higher the resistance the more voltage will be dropped across it, the lower the screen voltage and lower the gain of the tube.

You may not like the effect, though. While it may sound counterintuitive, in my experience many people do not like the gain compressive effect of tubes when the effect is exaggerated enough to be distinct. An interesting side note in that regard is that certain "tweed" Fender amps didn't have screen stoppers at all and hence no gain compression at all, only plain "clipping". Peavey has introduced amps that artifically mimic grid conduction with a solid-state diode. In effect you get similar "grid clipping" but since no actual grid current draw is involved the screen current draw is also minimized to good extent. The common theory is that such scheme provides more "graceful" overdrive characteristics, not to mention it will not kill tubes with excessive screen current draw and screen dissipation.

Common screen resistor values are practically too low to have distinct effects to gain compression, but they serve purpose as current limiters or oscillation killers.

And of course you need to crank the amp to hear any effect of the screen circut. The screens won't be drawing much of current unless you do and thus the voltage modulation effect remains minimal. You really want to get those power tube grids to draw current because only then will screen current draw skyrocket. Oh, bear in mind that this is one method to kill tube power amps, unless the screen supply is heavily current limited / regulated to remain at safely low voltage (which has the same effect). So, there's a reason for the designs where plate voltage is close to 600V - 700V while screen voltage is probably only about half of that (a.k.a. safely low). Note that often in these cases the screen voltage happens to be regulated to some extent as well, perhaps derived from an individual supply completely.

Good point to make distinction between pentode and beam tetrode tubes. All of them too often get lumped to same "pentode" category while there are obvious differences in grid drive requirements, in operation of the screen grid, in characteristic plate curves and in overall harmonic distortion they produce. These will cause distinct differences in how such tubes perform under overdrive conditions.

[catalin gramada]

Thanks you teemuk . Maybe I will decide to blow up a pair of used tubes into a test. Dont know how to draw a right plot but Can record everything and debate here.Have a dual scope but need a bunch of multimeters to record simultaneusly ac and dc condition still.I expect some sugestion how to organise a execution bench test.please. Thanks