Effect of lowered screen voltages on beam power tubes

[octal]

Interesting stuff, Merlin. The only two guitar amps I could find with wide spreads between the screen and plate voltages (not achieved through high value screen resistors) were the aforementioned Gibson GA 86 (it's anyone's idea what one of those would sound like- it's a rare bird) and the Silvertone 1484. The 1484 runs a pair of 6L6's at 475V plates, 340V screens with no stopper resistors into a 4K load. I don't think anyone would accuse that amp of sounding soft and triode-like! Admittedly, I haven't plotted the load line, so I don't know it they might lie in relation to the knee, but that seems like a typical load and plate voltage for 6L6s- it's not like they bumped the load impedance up to compensate. Any thoughts with what might be going on here?

Nathan

[GibsonLover]

Here is an article that claims a "boosted" mode by putting a regulator between the screen and plate for a specific response. The article is the third one in the PDF, so scroll down. I haven't posted attachments before, so if this doesn't work you can google "Boosted Triode Mode" from WWW.EDN.COM. Interesting application for a screen circuit, if nothing else.

[loudthud]

The RCA tube manual page for the 7027A shows a family of curves where grid #2 is the variable parameter and grid #1 volts=0. Basically they are showing you where the top line would be in a normal set of curves where grid #1 is the variable parameter. Lowering the screen voltage is going to reduce transconductance and voltage gain and more or less set a current limit for the tube. Seems like I remember a Gibson bass amp that had 3 6L6's. One was a screen grid regulator.

http://www.nj7p.org/Tube4.php?tube=7027a

[Enzo]

I remember that article. Read closely. They didn;t add a regulator, they added a regulated 100v power supply. Between plate and screen, it adds 100v to the plate B+. The screen is running 100v MORE positive than the plates.

The point of the article is that when running the 6L6 or other power tube as a triode, you lose power capability. This boosting idea increases the power capability back up to more pentode-like levels. But the tube is still functioning as a triode.

[octal]

Well, I built a demo amp (in two days from scrap parts!) It's a standard 5F6-A Bassman pre going into a Silvertone 1484 cathodyne PI and output stage with the stacked voltage doubler power supply. My build has about 470V on the plates, 330 on the screens derived from a tap on the voltage doubler circuit. It has a pretty cutting overdriven sound- not really to my taste. I will next try running the screens off the main B+ supply with a normal size dropping resistor, rebias and see what happens.

[black_labb]

very interested in your findings octal. i can imagine that it would have a softer overdriven tone, but we'll see.

[octal]

OK- I tried my experiment today- 445 on plates, 321 screens. -30v to bias @ 70mA. It made 11.8v into 8 ohms @ the onset of clipping. Running the screens off filtered b+ with an unbypassed 1.5k going into an individual 100 ohm screen resistor resulted in 443 plates, 440 screens, -45v bias for 70mA. It made 13.2v at clip. Clipped waveform looks pretty much the same- maybe a bit less bias shift crossover distortion when pushed harder? I have not actually listened to it- baby's sleeping and my tame guitarist isn't availible. (I'm not really a guitar player- just dick around a bit.)

[black_labb]

i dont expect the differences to be notable on the scope with a test singal, as i feel the differences are going to be related to the voltage drop across the screen resistor. when the amp is putting out a sin wave then there is not going to be a changing voltage on the screens unless you change the amplitude.

i guess it would be related to the dynamic response as well.

[octal]

Well, there is definitely screen sag with a sine wave signal too... as you approach clipping, the screen current goes up, and the screen voltage goes down (assuming it is driven off of a resistor as opposed to a low impedance fixed supply.) The screen grid current is actually cycling up and down as the operating point of the tube changes as the sinewave modulates the current in the tube.

Mind you, I'm not disagreeing that there might be some dynamic effects with a musical signal. On the other hand, Randall Aiken claimed that he could hear no difference in various values of screen resistor and he seems like a fellow who knows whereof he speaks.

[black_labb]

interesting. had a play with it yet?

[mooreamps]

I imagine this would be very good for tube reliability, but how about output power and the overdriven sound?

It tends to reduce the total output power down to about a 1/4, but makes it just so much easier to saturate the power tubes because it reduces the bias. Therefore, when the input signal begins to exceed the bias, the tube naturally begins to self saturate, even at a lower plate current.


-g