AB167 bias balance

[Helmholtz]

My grid voltages are -51V.

I should have mentioned that measuring bias voltage directly at power tube grids only gives accurate readings with a meter having high input resistance (e.g. 10M).
With only 1M meter resistance, readings will be low by about 20%.

It'a always best to measure at the other ends of the (220k) grid resistors. As these resistors don't drop voltage this will give real grid voltages.

[vintagekiki]

The value of bias voltage is determined by quiet current at a given HV anode voltage.
Quiet current is depends by the current emission of the cathode, and therefore it is necessary the bias voltage adjusts each time when the power tubes are replaced.

http://web.archive.org/web/20121118023641/http://www.diyguitarist.com/GuitarAmps/PT-Biasing.htm
Biasing Power Tubes

http://web.archive.org/web/20120623045707/http://www.diyguitarist.com/GuitarAmps/PowerTubeBias.htm
Common Power Tube Bias Charts

https://music-electronics-forum.com/showthread.php?t=8291
About biasing tube guitar amplifiers

[Randall]

"I'm thoroughly confused by this. The 100k, being a lower value, should have reduced bias voltage more (more 'effect") than 470k. Did you mean 47k?"

Yes, I meant 47K. It seems I can't type anything without a typo these days. Sorry.

[pdf64]

...I wouldn't have expected such a simple and general relation. Where did you find this?...

I read it, somewhere
If it had been anyone else asking this, I'd be searching your posts for it
Perhaps I've overstated it, eg rather than being precise / direct proportionality, it's more of a general trend.

...Consequently, when mains voltage rises by 10%, bias as well as screen voltage can be expected to rise by 10% and both effects on plate current should compensate.

My experience is that is seems to hold fairly true (what's your's?).
Of course, even if this leads to plate current staying fairly steady as mains Vac rises, plate dissipation will still increase, due to the increase in plate voltage.
But it demonstrates why it would be a bad idea to regulate the bias supply voltage without also regulating the HT (or at least screen grid) voltage.

[vintagekiki]

Adjusting the bias with grabble method is a painstaking and uncertain job, which is always repeated with each replacement of output tubes.
With trimmer connected in a series with 15k resistor adjust the quiet current (bias), measure value the series connected resistors, and instead of 15k install the newly created value.

The rest is known.

[pdf64]

...https://music-electronics-forum.com/showthread.php?t=8291[/FONT]
About biasing tube guitar amplifiers

As I see it, a major failing of so many of those info sources is that where a particular idle plate dissipation % of the limit is suggested, no clarification / distinction is made regarding whether the limit is derived under the design centre or design max rating system (or even whether it's an absolute max rating)!; they're all just lumped together.
The point being that design max limits are intended to have an appropriate derating factor applied by the equipment designer (and generally gets devolved down to us), whereas with the design centre system, such a derating factor is already built in.
http://www.one-electron.com/Archives...on%20Tubes.pdf

[Helmholtz]

If it had been anyone else asking this, I'd be searching your posts for it.

Thanks, I feel honored.
But maybe McCaul or Terman?
Might be the result of a first-order Taylor series approximation of Ip(Vg, Vs, Vp).

My experience is that is seems to hold fairly true (what's your's?).

I haven't done specific measurements on a variac yet.

Of course, even if this leads to plate current staying fairly steady as mains Vac rises, plate dissipation will still increase, due to the increase in plate voltage.
But it demonstrates why it would be a bad idea to regulate the bias supply voltage without also regulating the HT (or at least screen grid) voltage.

Completely agree. And there's also the effect of different heater voltage.

[vintagekiki]

... when mains voltage rises by 10%, bias as well as screen voltage can be expected to rise by 10% and both effects on plate current should compensate ...

Yes, but 10% rise voltage does not have to mean 10% rise current (the changes are not linear)

[Helmholtz]

Yes, but 10% rise voltage does not have to mean 10% rise current (the changes are not linear)

Doesn't matter as long as increased grid bias and increased screen voltage have exactly inverse (mutually compensating) effects on plate current.

[vintagekiki]

As I see it, a major failing of so many of those info sources is that where a particular idle plate dissipation % of the limit is suggested, no clarification / distinction is made regarding whether the limit is derived under the design centre or design max rating system (or even whether it's an absolute max rating)!; they're all just lumped together.

That already belongs in the domain of technological (factory) documentation which exists for each tube individually.
Based on factory documentation, for development engineers is foreseen handbooks, data manuals, datasheets with maximum and recommended values.
Each product (tubes, transistors, etc) is defined as a four-pole block (with input and output) and the necessary initial data needed for the designing of various assemblies.

1)
https://vacuumtubes.net/How_Vacuum_Tubes_Work.htm
2)
http://www.tubebooks.org/Books/Atwood/RCA%201940%20Vacuum%20Tube%20Design.pdf
3)
https://www.youtube.com/watch?v=EzyXMEpq4qw
^{Hand Made Vacuum Tubes by Claude Paillard}
4)
https://www.youtube.com/results?search_query=how+to+make+vacuum+tube+at+home
^{how to make vacuum tube at home}
5)
http://bee.mif.pg.gda.pl/ciasteczkowypotwor/%23Book/Elektronika/Ksiazki/Inside-the-Vacuum-Tube-John-F-Rider.pdf
Inside the Vacuum Tube
6)
http://bee.mif.pg.gda.pl/ciasteczkowypotwor/%23Book/Elektronika/Ksiazki/Electron%20Tube%20Design%20-%20RCA%201962.pdf
Electron Tube Design – RCA 1962

[pdf64]

Interesting.
I wouldn't have expected such a simple and general relation. Where did you find this?

Consequently, when mains voltage rises by 10%, bias as well as screen voltage can be expected to rise by 10% and both effects on plate current should compensate.

I just happened across RCA Application Note AN-38, which seems to support, within a limited range, that relationship / approach http://www.one-electron.com/Archives...cteristics.pdf

[Helmholtz]

I just happened across RCA Application Note AN-38, which seems to support, within a limited range, that relationship / approach http://www.one-electron.com/Archives...cteristics.pdf

Actually I'm not sure if I understand the method correctly.
Does it really mean, that changing the plate voltage by 20% would have the same effect on plate current as changing screen voltage by 20%?
If so, that would be in contradiction to what I learned about pentodes, namely that the screen voltage should have a considerably stronger influence than the plate voltage.

[pdf64]

Actually I'm not sure if I understand the method correctly...

Bugger, I was hoping you'd know what it all meant

My take is that the 'Fe' factor is intended to be applied to the anode, screen grid and control grid voltages, then from the chart, the resulting characteristics etc can be derived.

So that a set of 'typical operating conditions' from the published info can be adapted to be applied to, eg the voltages that the PT we've got puts out.

[Helmholtz]

Bugger, I was hoping you'd know what it all meant

My take is that the 'Fe' factor is intended to be applied to the anode, screen grid and control grid voltages, then from the chart, the resulting characteristics etc can be derived.

So that a set of 'typical operating conditions' from the published info can be adapted to be applied to, eg the voltages that the PT we've got puts out.

Had to look up "bugger". Hope you're not calling me that name .

I think we agree that the application of the method described in the document is not well explained.

If you're right, plate, screen and grid voltages always need to be varied by the same percentage.
That's way less exciting than what I was expecting. I was hoping to find the effect on plate current if I vary only one parameter at a time.

It also wouldn't really confirm your theory of a self-compensation with varying mains voltage.
E.g., from the table : Vp = 200V, Ip = 22.7mA => Pa = 4.54W and Vp = 250V, Ip = 32mA => Pa = 8W.

So still not sure.

[pdf64]

...It also wouldn't really confirm your theory of a self-compensation with varying mains voltage.
E.g., from the table : Vp = 200V, Ip = 22.7mA => Pa = 4.54W and Vp = 250V, Ip = 32mA => Pa = 8W...

Yes, I agree, you were right to be skeptical of my claim
Re-reading post #16, I went too far and overstated the effect I'd noticed. I should have phrased it as being more of a general trend of the anode current changing by a lower ratio than the electrode voltages.
I don't think the accompanying change in heater voltage would have much effect, as Tomer p18 notes that the electron cloud around the cathode operates with a massive surplus; hence a change to the magnitude of that surplus (ie due to a change, within tolerance limits, of heater voltage) should affect operating characteristics.