Separate names with a comma.
Discussion in 'Amp Tech Center' started by peteb, Apr 15, 2018.
peteb said: "Elpico, most of what you wrote is about the screen. I did serious research and some testing. In fender amps, when the plate voltage drops, so does the screen, by the same amount. This plan for the screen to be unnafected by the plate drop needs the screen to have its own regulated power supply. This is never done in guitar amps."
My reply: "This should not be true of non-overdriven pentode power tubes. The B+1 node for the output transformer and power tube plates, and B+2 for the screens should decouple the screens and plates so the screen voltage should be fairly stable, at least until screen current begins to flow heavily during overdrive.
The behavior you describe is how a pentode/triode (half power) switch operates. In triode mode the plate and screen are tied together."
Were your original screen voltage readings in error?
Did you mean to post that in the plate load thread?
It's OK I think they all relate.
I'm not sure what the question is.
I've got good readings for AC and DC on the screens for a couple different amps during idle and playing conditions.
I feel fairly well informed.
Are there any questions what to expect on the screens?
I even highlighted the important line. You said plate and screen moved in unison in Fender amps. Now you have this thread posting the exact opposite. I just wanted to know what happened.
the screen voltage sags with the plate voltage, along with the voltages in the entire DC power supply.
There are no significant AC voltages on the screen throughout all playing conditions, screen resistor or no screen resistor.
The point of this thread is the screen being right there in the middle of all of the tube action, unexpectedly, develops no significant Signal voltage. Understanding how the AC signal develops on the plate and cathode is usefull for understanding why the screen always has a near zero AC voltage.
But why would anyone expect the screen to have AC on it? It has its own power supply node with a big filter cap?
Some designs operate under sliding screen conditions, where the screen is allowed to sag under signal load. It's often done to limit screen dissipation under high power and overdrive conditions, using a modest resistor value.
At one end of the range is fixed screen, where it's completely decoupled. This gives you maximum power tube gain and power. At the other end is triode mode, where the screen is tied directly to plate voltage. That gives you a significant drop in power and efficiency, and is what the half power switch in many amps does.
In between, you get some amount of screen sag. Designers can use this as a tool to change the feel of the amp since it tamps down power and adds compression. If you have a test bed amp, temporarily swap in a couple of quite large screen resistors - much larger than typical - and see what the effect on the feel of the amp is. Don't worry about the numbers and math, just play under low and high volume and see if you can tell a difference.
A pentode with a screen resistor doesn't get much screen voltage sag during non-overdrive conditions because the screen current average stays pretty close to idle screen current.
When the grid goes positive, flows more current through the tube and causes a drop in the plate current the screen with B+2 voltage attracts more electrons and voltage drops a little.
When the second half of the audio wave hits the grid and decreases current through the tube, the plate voltage rises and the screen gets fewer electrons and less screen current so it averages out to around idle current and is why we don't see an AC signal on the screen.
In overdrive conditions when massive amounts of screen current flow is when noticeable screen voltage sag and compression occur.
Well, let's add a simpler answer here. Tube design says screen current will vary under signal. The voltage swing seen, however, depends on the resistance between the screen and its supply node, and the node's impedance to ground - IOW the total impedance to ground it sees.
if there is no screen resistor, the impedance to ground is nil.
if there is a screen resistor it is usually very small.