screen voltage for output stage

[ReadyTeddy]

First, let me apologize if this question has already been answered somewhere. In some push-pull output stages the screen is connected so that its voltage is some fraction of the plate voltage---that fraction varies in different designs, but the basic idea is that the screen voltage tracks the plate voltage. In other designs, the screens are connected to a fixed, filtered voltage. In this case, the screen voltage remains constant while the plate voltage varies due to input signal. In this type of design, the plate voltage can end up substantially lower than the screen voltage when the input signal drives one or the other of the plates to a somewhat low voltage. This latter condition has always baffled me. One might think that if the screen voltage is substantially higher than the plate voltage that during those times, the screen would draw lots of current, effectively stealing current from the plate. I'm obviously misunderstanding something.

Thanks

[loudthud]

When the screen is connected to a tap on the output transformer, that is called an Ultra-Linear circuit. The tube acts more like a triode and has a lower output impedance. When the screen is connected to a tap off the main B+ the tube has a much higher output impedance. The plate current will increase and the plate voltage will decrease as the tube supplys more signal to the output transformer up to a certain point where the plate voltage cannot decrease (about 50V to 80V) unless the current also decreases. At this point the screen current can go very high, if the plate votage decreases enough, the screen current can exceed the plate current. This is bad for the tube. It usually doesn't happen unless the speaker is disconnected.

[R.G.]

You're not misunderstanding.

The screen grid was inserted to "screen" the cathode and grid from the plate voltage going low and affecting the electron flow. The screen grid is positive to attract electrons toward the plate when the plate itself might not. It also screened the plate fro varying grid voltage and increased high frequency ability.

With the screen pulling electrons toward the plate (physically closer), the changes of plate voltage had much less effect on the current flow from the cathode/grid, and much less feedback to lower frequency response. So when the grid let electrons through, they went rushing for the screen grid and plate. The screen grid was relatively open and close to the plate, so most of them rushed right through the screen grid and hit the plate.

Some bounced off the plate, and it was the function of the suppressor grid, the fifth component in a pentode, to force the bouncers back into the plate, but that's another story.

As the plate voltage goes down, the percent of electrons going through the screen and hitting the plate (and staying there~!) goes down, and more of them hit the screen grid and stick. This DOES cause the screen grid to eat a lot more current and heat up. A lot of musical amps run the screen grids very high and have lots of problems with tube failure due to the screen grid melting.

But what you're missing is that the screen grid has a lot less surface area for electrons to hit than the plate, and they fly through due to inertia.

[ReadyTeddy]

You're not misunderstanding.

The screen grid was inserted to "screen" the cathode and grid from the plate voltage going low and affecting the electron flow. The screen grid is positive to attract electrons toward the plate when the plate itself might not. It also screened the plate fro varying grid voltage and increased high frequency ability.

With the screen pulling electrons toward the plate (physically closer), the changes of plate voltage had much less effect on the current flow from the cathode/grid, and much less feedback to lower frequency response. So when the grid let electrons through, they went rushing for the screen grid and plate. The screen grid was relatively open and close to the plate, so most of them rushed right through the screen grid and hit the plate.

Some bounced off the plate, and it was the function of the suppressor grid, the fifth component in a pentode, to force the bouncers back into the plate, but that's another story.

As the plate voltage goes down, the percent of electrons going through the screen and hitting the plate (and staying there~!) goes down, and more of them hit the screen grid and stick. This DOES cause the screen grid to eat a lot more current and heat up. A lot of musical amps run the screen grids very high and have lots of problems with tube failure due to the screen grid melting.

But what you're missing is that the screen grid has a lot less surface area for electrons to hit than the plate, and they fly through due to inertia.

Thanks, your explanation is very clear. Given that many guitar amps seem to run the screens very close to the B+ supply, the plates spend lots of their time with their voltages below the screen voltage. It still seems odd, but, I guess it works.

[bb5000]

Thanks, your explanation is very clear. Given that many guitar amps seem to run the screens very close to the B+ supply, the plates spend lots of their time with their voltages below the screen voltage. It still seems odd, but, I guess it works.

That's why you have screen resistors, they make the screen voltage drop and (hopefully) always stay below plate voltage.

[frus]

That's why you have screen resistors, they make the screen voltage drop and (hopefully) always stay below plate voltage.

no, the plate voltage is below the screen voltage 50% of the time (if there is a signal of course)
when plate voltage drops to some low value like 100V or 50V (depends on the tube), screen current rises abruptly. screen resistors help lower the screen voltage in that case, which in turn lowers the screen current (attracts less electrons)

[R.G.]

That's why you have screen resistors, they make the screen voltage drop and (hopefully) always stay below plate voltage.

As frus notes, the "off" side plate is way above the screens by transformer action.

Resistors cannot drop voltage by themselves. They only generate a voltage equal to their resistance times the current through them. So screen resistors cannot drop the screen voltage unless there is screen current flowing through them. The drop in screen voltage with screen resistors is not so much to make them stay below the plates as to try to limit their current and stop them from melting.