Ultralinear operation

[Frank Berry]

I've always wondered about push-pull ultralinear operation of Tetrode, Beam Power and Pentode vacuum tubes.
I hope that someone can give me the information that I seek.
These are my thoughts.

Under static conditions, the screens have higher voltage than the plates (if only by 10 or 20 volts).
I have always understood that the screen voltage should be lower than the plate voltage on most tubes.
As the amplifier is pushed and the plate voltage is pulled down on one of the tubes, the screen on that tube is even more positive than the plate due to the ultralinear connection.
Perhaps it's a matter of the duty cycle. Damage to the screen occurs due to heat.
The increase in screen voltage and current during half of the waveform may be offset by the reduction in screen voltage and current during the other half of the waveform as the plate voltage "soars" on that tube when the other push-pull tube is drawing the most current.

If I'm wrong, please correct my thinking.
Thanks for whatever information that you can provide.

[CharleyW]

The answer you seek may be here: Adjustable distributed load discussion

[tinitus]

in penthode, is there something about the screen grid resistor to drag voltage down below plate voltage is causing serious side effects

well, I know its more complicated than that, but in 'layman terms', maybe why the use of the transformer(UL OPT) is more reliable ?

[tubelab.com]

Damage to the screen occurs due to heat.

The screen grid is not capable of dissipating much heat. If it gets too hot it will start to glow red. When it starts to glow red it can act like a cathode and begin to emit electrons. If the plate goes more positive than the screen it can attract those electrons causing current to flow from the screen to the plate. This warms the screen up even more leading to eventual runaway followed by a bright blue flash and an erie silence from the speakers. Ask me how I know this....well it was a 98 cent tube anyway, and I was extracting far too much power from it.

In triode operation current flow from the screen to the plate can't happen....they are connected together. This is why you can often abuse the screen grid rating a bit in triode mode. The plate has far more surface area and attracts the most electrons.

In UL mode the screen can be slightly above the plate in the conducting tube. As the tube conducts heavier the voltage across it drops so the dissipation across the tube goes down near peak current. The other tube is usually cut off (P-P class AB amp) or at minimal current while the voltage is at max (P-P class A) so the dissipation is also low. The voltage difference between the screen and plate isn't very high at a time when any serious current is flowing.

Pure pentode is usually worse case especially in the typical audio tubes that operate with several hundred volts on the screen. During the time when the plate voltage is in the 50 to 75 volt range the screen is still at 400 volts, and gets a lot of current. It can get hot. a few milliseconds later (next half of the audio cycle) the plate can see 1000 volts while the screen is still at 400 volts. BANG happens.

As long as the screen grid is kept below its dissipation rating (averaged across the audio cycle) it won't glow. Some tubes also impose a peak screen dissipation rating to prevent arcing, and it must be obeyed. Yes, I blew up an E130L by ignoring it.

If you were to study the dissipation in the output tubes versus amplifier output power you will find that maximum tube dissipation almost never coincides with maximum output power.

In a class A amp maximum tube dissipation ALWAYS occurs at zero output and it the usually lowest at maximum undistorted power. It can be lower into clipping if the amplifier clips cleanly.

A P-P class AB amp usually dissipates the most heat somewhere around mid power. The tubes will cool down as the amp is cranked harder since the signal spends less time in the area where the voltage across the tube, and the current through it are both fairly high.

An amplifier thar operates deep into AB, near class B, may dissipate the most heat at max power output. Petes red board when cranked to 120 WPC and beyond exhibits this type of response which makes the limitation on output power directly dependent on the size of the output tubes.

[Wavebourn]

I know that tube manufacturers hated the fashion of UL.

[Frank Berry]

Thanks Tubelab. Probably a good idea to use current limiting screen resistors when operating in Ultralinear or pure Pentode mode.

[Jeb-D.]

G3 (suppressor grid) is what allows a pentodes plate to swing below the screen. In beam tetrodes, the "beam" serves the same purpose as the suppressor grid does in a Pentode. In a true tetrode, all hell breaks loose if the plate swings below the screen grid. But it's unlikely to find a true tetrode in audio applications.

With that said, there is something seemingly un-natural about allowing the screen to idle at a higher voltage than the plate. When running UL, I typically size the screen stoppers (resistors) so that the screen voltage is at, or just below the plate voltage. They also help current limit under overload conditions.

When running pentode mode, I prefer to set it up so that the screens run at a substantially lower voltage. Using a full wave voltage doubler for the plate supply and taking the screen supply off the half way point works well.

[Frank Berry]

Jeb-D
Thank you so much. That completely answers my question.

[Wavebourn]

Originally Posted by Frank Berry

Thanks Tubelab. Probably a good idea to use current limiting screen resistors when operating in Ultralinear or pure Pentode mode.

Current limiting resistors cause voltage on screen grid to depend on current it draws. It results in more distortions on lower power.


For lower distortions on higher power screen grid voltage must be stable and below the value that can result in excess power dissipated by screen grid. It is possible in pentode mode. In triode mode it can be of significantly higher value than in pentode mode. In UL actually tubes are abused. Manufacturers had to follow the fashion redesigning construction of tubes so they would work in UL for a long time, when fashion demanded it.

[Frank Berry]

I can certainly understand how ultralinear operation can be hard on the tubes.