relationship between high voltage supply and plate voltage

[studioit]

Is the relationship between the ac hi voltage supply and the plate voltage linear - In other words if you drop your voltage in the PT by ten percent, can you expect a 10 percent drop in plate voltages if all other things remain the same?
Thanks
Ian

[tbryanh]

I have extensive transistor technician background, and I am a little new to tubes, but I will give this a try anyway.

It sounds like your concerned with the plate voltage on the push-pull output stage.

To determine if the response is linear, you need more than two points. Here you are comparing the effect 100% on the PT has on the plate with the effect 90% on the PT has on the plate. You only have two points, so you cannot determine if the response is linear or not.

If you are wondering if the response is the same; i.e. does a 10% drop on the PT cause a 10% drop on the plate?

I think the answer to that is no.

I think there are two things that work in opposite directions when you lower the PT voltage.

Lowering the PT voltage by 10% tends to
1. lower the plate voltage by 10%.
2. lower the screen grid voltage by 10% too.

“2” tends to offset “1” a little; i.e. “2” reduces the gain of the stage which puts less demand on the power supply which in turn causes “1” to have less effect.

A guestimate is that lowering the PT voltage by 10% will lower the plate voltage by 7%.

The preamp and driver stages have not been considered yet. I think they would also tend to counteract “1” the same way “2” tends to counteract “1” and for the same reason; i.e. the preamp and drive stage’s plate voltages would drop which would reduce their gains. The lower gains would cause the output stage to not work as hard which would reduce the load on the power supply….

Including all of that, a guestimate is that lowering the PT voltage by 10% will lower the plate voltage by 5%.

[studioit]

Thanks, the second answer is what I was getting at. So there is no hard and fast formula? I guess not as there are so many variables.
Ian

[kg]

the answer to your question is yes.

proportionality of voltage is a function of turns ratio. if you don't change the turns ratio you don't change the voltage ratio.

if you increase line voltage by 10%, from say 110vac to 121vac, Vsec will similarly increase by 10%. if your B+ rail used to be 300vdc it will now be 330vdc.

[Steve Conner]

KG's answer is correct as long as you don't think about the problem too much ;-)

If you do think about it, tbryanh's answer is technically the right one: changing the line voltage will change all the voltages in the amp in proportion. This will then affect the currents drawn by the tubes, so the sag characteristic of the power supply will cause a further change in some of the voltages.

In math speak, this further change is an example of a second-order effect. KG's ignoring of it was an example of making a non-linear problem easier to solve by using a first-order approximation.

[kg]

KG's answer is correct as long as you don't think about the problem too much ;-)

If you do think about it, tbryanh's answer is technically the right one: changing the line voltage will change all the voltages in the amp in proportion. This will then affect the currents drawn by the tubes, so the sag characteristic of the power supply will cause a further change in some of the voltages.

In math speak, this further change is an example of a second-order effect. KG's ignoring of it was an example of making a non-linear problem easier to solve by using a first-order approximation.

well he DID say "ac high voltage supply."

as long as we're throwing higher order effects into the mix, how about the decrease in cathode current with very low heater voltage? are we up to 3rd order now???

i guess the real answer is, how technical do you want to get?

[Enzo]

Well, the original question was not will the B+ voltages drop 10%, though I bet that was what was in his mind, the question was will the plate voltages so shift. And there are the p-p output plates, and then there are the preamp plates with their large resistance loads. Two different stories.

If we want a physics lesson, then all the extra junk is valid, but if this is a "how does this work, daddy?" thing, then yes, it reduces the same percentage.

Did we include the resistance of the house wiring?

[tbryanh]

Well, the original question was not will the B+ voltages drop 10%, though I bet that was what was in his mind, the question was will the plate voltages so shift. And there are the p-p output plates, and then there are the preamp plates with their large resistance loads. Two different stories.

If we want a physics lesson, then all the extra junk is valid, but if this is a "how does this work, daddy?" thing, then yes, it reduces the same percentage.

Did we include the resistance of the house wiring?

Well said. Short summaries are always best.

[studioit]

The original question was, - in more understandable form this time I hope -
Suppose amp "A" has a plate voltage (output stage) of 350 V
The Ac high voltage winding of the PT in Amp"A' is 640 volts .
If... we put a different PT in amp "A" that has a high voltage winding that is
say 10% less or 576 volts AC, and, All Other things remain the Same -
(including wall voltage)
What percentage reduction in Output plate voltage can be expected.
Thanks
Ian

[Steve Conner]

"Roughly" 10%...

[kg]

The original question was, - in more understandable form this time I hope -
Suppose amp "A" has a plate voltage (output stage) of 350 V
The Ac high voltage winding of the PT in Amp"A' is 640 volts .
If... we put a different PT in amp "A" that has a high voltage winding that is
say 10% less or 576 volts AC, and, All Other things remain the Same -
(including wall voltage)
What percentage reduction in Output plate voltage can be expected.
Thanks
Ian

if, when you say "all other things remain the same," you really mean all you're touching is JUST the power tranny, then "something less than 10%" is correct.

on the other hand, if by "all other things remain the same," you really mean that the total B+ rail current draw remains the same, then 10% is correct.

the point that others were trying to make is that the total B+ rail current draw is itself dependent on the B+ rail voltage which is in itself dependent on the B+ rail current, which is dependent on the B+ rail voltage, and so on. as a result after changing either voltage OR current, a "new equilibrium" has to be found. the ultimate change will be somewhat smaller in magnitude than that predicted by the secondary winding output voltage difference.

so there is a simple answer, and a not so simple answer. unfortunately it's ridiculously complicated to try to calculate exactly how much all the operating points in the amp will shift, so you've basically just got to try it out. in your case you could string up some 65v zeners on your HV secondary winding to simulate the decreased output voltage and see what you get before buying another tranny.

hth
ken