Adjusting Screen Grid resistors

[EETStudent]

Isnt adjusting your screen grid resistors a way to increase or decrease your maximum power tube output/distortion? Well if so, Im having trouble locating this resistor. I know where to screen grid is on the power tubes, but there are two different resistors running into them. One is a 3000 ohm power resistor going into a 30MFD, 500V filter cap. The other is a 10K resistor going across the other three filter caps. which one should i toggle with?

[Chuck H]

It sounds, by your description, like niether one of those resistors are screen grid resistors. They are both most likely dropping resistors in the B+ rail. Screen grid resistors are placed between the B+ rail and the screens as two equal value resistors (one for each screen grid). Screen grid resistors aren't used on every amp, so you may not find them (not being there and all). Often they are mounted on the tube socket itself with one end to the screen grid and the other soldered to an unused pin. In this case the screen supply lead is soldered to that junction at the otherwise unused pin.

Chuck

[EETStudent]

thanks for your reply Chuck H. I guess screen grid resistors dont exist in this amp. I dont have a schematic for this amp, but after looking at other schematics of popular amps I tended to fine common cathode resistors between the two power tubes. In my case, the two cathodes are noding at ground with no resistance between them. I dont have enough knowledge to know how this amp would otherwise be biased, but i do know its voltages are fairly low so I would like to figure out how to adjust this for hotter output.

[Old Tele man]

...the "formal" purpose of the screen resistors with "beam tubes" is to ensure the screen voltage remains below the plate voltage (~50Vdc) during normal operation, so that the 'virtual cathode' stays in front of the plate.

...when the plate voltage falls below about 50V above the screen voltage, the 'virtual cathode' that's normally created and positioned in front of the plate begins to fade away...and, begins to re-form in front of the screen grid (BAD,BAD,BAD). When THAT occurs, the screen grid is acting as "electron catcher, ie: meaning WATTS dissipation!) instead of just "electron accelerator" and the screen grid wattage is VERY soon exceeded and rapidly causes the screen grid to distort or melt...ie: physical failure...then things start "shorting" together within the tube, resulting in electrical failures...tube, OT windings, PT-fuse, etc.

[Chuck H]

OTM,

But there are a great many amps using beam tubes with the screens tied to the same B+ as the OT CTR. And the plates are indeed lower than the screens. Old Gibsons come to mind as an example. Perhaps what you say is true in theory, or for some applications, but in practice I don't know any of those amps to suffer melted screens, lost efficiency or blown transformers.

I don't know enough about theory to say why or even why not the screen grids should or should not be there. but now I am truely interested.

EETS,

Most popular amp designs use screen grid resistors. Perhaps we will find out why. For now, I can say from your description that your amp is most likely fixed bias. Between the phase inverter coupling caps and the power tube grids (schematically speaking) you should find two resistors tied together. Probably valued somewhere between 100k and 330k. Where these two resistors meet they go to a negative voltage supply. A power tubes output and performance is tuned by the voltage relationship between the cathode and grid. For now, suffice it to say that the grid should remain at a negative potential to the cathode. In a "fixed biased" amp the cathodes are at ground (0v) and the grids are tuned in to a negative supply via another circuit in the amp. The bias arrangement you described in your post is a "cathode biased" arrangement. In this case the grids are referenced to ground (0v) via a resistor, and the cathode is elevated from ground via a resistor. This places a positive voltage on the cathode. So in a cathode biased amp the grid is still at a negative potential to the cathode.

You can add screen grid resistors to reduce a little voltage at the screens. Try a pair of 1k 2watt resistors. One to each screen from the B+ rail. Then you must rebias your fixed bias amp. To do this you must locate the bias voltage supply circuit. Let me know when you have found it.

Chuck

[Old Tele man]

"...the "formal" purpose of the screen resistors..."

...both Gibson and Fender routinely ran Vs at or greater than Vp, but that's more the result of NOS tubes having published 'specs' far below their actual limitations.

...I just state "why" the screen grid is needed...and, think about it: how much "control" is a 100- or 470-ohm screen grid resistor gonna exert? Obviously not a hell of a lot! And, unlike the 100-ohm resistor that's typically inserted in the screen grid to plate connection of TRIODE-wired beam/pentodes, they're (100-470 ohm) not providing any "voltage-feedback" *when* the screen voltage is from a separate, well regulated (ie: post choke) voltage point along the high-voltage string.

...such small value resistors are basically acting as "short-circuit" protection devices in the unfortunate case of screen-grid failure and "shorting" to plate. They're certainly not there as major "voltage-controlling" resistors for the screens, for instance (6L6 at idle / 55W max power):

6L6 Vs = 400Vdc at 2.8mA screen current, Vr(screen) = 470*2.8mA = 1.32Vdc
6L6 Vs = 400Vdc at 22mA screen current, Vr(screen) = 470*22mA = 10.34Vdc

...so, obviously, they're also NOT doing a lot of screen-grid voltage "regulation" either.

...thus, the REAL reason they're there is to "limit" screen currents during maximum output periods when the power tubes are being momentarily driven "positive" (and thus grid draws current, meaning it's no longer AB1 but now AB2 operation) and the screens begin to draw heavy current on the SIGNAL PEAKS (not at the signal-RMS level).

[EETStudent]

Chuck, thanks for your replies. i have indeed located the two grid resistors you mentioned, both equaling 100k going to the negative supply. The voltages where the leads meet the sockets are -31V, and the plate voltages are 425V. As for the screen grid resistors, what would lowering the voltage do performance wise? From my understanding those previous voltages mentioned are good for a lower output/distortion and a colder bias. I woud like to bring this up if that is safe and acheivable. what do you think?

[Steve Conner]

I agree with OTM. Screen resistors are basically there to stop the screen grids from melting when the amp is cranked hard and the power stage is clipping. A lower screen voltage has the same protective effect, but it's usually easier and cheaper for the designer to just hook the screens to the same B+ as everything else, and add the resistor.

I'm not sure if 6L6s can survive OK without any screen resistors, but I doubt EL34s could. Even with a 1.5k screen resistor you can see their screens glowing red on signal peaks. With 1k, and higher plate voltages, they light up like lamp filaments!

[EETStudent]

Steve, the amp Ive been posting on is indeed an EL34 based amp, and from what Chuck explained to me there are no screen grids. Another thing, how can you locate the B+ rail? Does it usually follow the same spots throughout most amps? thanks fellas

[Chuck H]

"how can you locate the B+ rail? Does it usually follow the same spots throughout most amps?"

Schematically, yes. But physically it can be anywhere the builder wants to put it. Follow the power supply from the power transformer, to the rectifier (tube or diode type) then to the B+ rail. Your bias supply should be located there, between the rectifier and the power tube grids (again, schematically).

Or you could locate the bias supply by following the path from the bias resistors on the power tube grids (remember, the two that each go to a grid and are tied together at the other end). Follow the circuit from the junction of those two resistors moving away from the power tubes and you will locate the bias supply.

Chuck

[Steve Conner]

From the look of it, and the voltages you posted somewhere else, that 3000 ohm resistor is your screen resistor, and the designer hung that 30uF capacitor off the screen node too. This is a quite valid way to do it, though it might make some funky-sounding compression when you overdrive it with your guitar. You can change it to a more conventional setup, as long as you make sure that other parts of the amp aren't also powered off this node.

No, the B+ rail does not usually appear in the same spots, except in so far as you commonly find it connected to things like the + terminal of large 500V capacitors, the center taps of output transformers, the ends of preamp tubes' plate resistors, and the cathodes of rectifier tubes or rectifier diodes, and if you somehow get it connected to yourself, a trip to the emergency room may be in your future.

Enjoy the practical side of your EE education

[EETStudent]

Steve, thanks for your comment. When you say more conventional setup, what could I do to achieve that? Come to think of it, the compression on this PA is a bit unconventional. All I know so far is that screen grid is going to that filter cap mentioned, the cathodes are going to ground sharing no resistors, and the negative voltage supply is -31V with respect to grounded cathodes(thanks to Chuck I could identify that). The plate voltages are 437V, and the screen grid voltage is 376V. I just feel like there is something I can do to improve the overall performance for guitar use. any thoughts?