Estimating Power Output VS Voltage

[guitarmike2107]

I am thinking about using voltage control on my next amp and will probably do it with different taps on the PT rather than a power mosfet.

I need to approximate the output power at different voltages to “guess” the required voltage taps.

Since P = V^2/R , and R is a constant resistance for measuring output power is it safe to assume the equation holds up for all operating conditions of the valve. i.e. as per the chart below




Cheers

[pdf64]

That should work, as a guideline rather than exact, eg you'd get 0% power at a voltage above 0%, but 50% voltage would likely give a little more than 25% power, as the voltage losses across the ouput devices may be less and the power supply may sag less.
In one project I put a 50% power switch using this idea, but the actual power in that setting was about 60%, due to the above. Pete.

[guitarmike2107]

Thanks pete

Thats what along the lines of what I was thinking, but just wasnt sure.

[J M Fahey]

I think that if you want to make a "stepped master volume" equivalent with many taps and a many poles switch, it may become complicated; but a simpler solution with, say, 50% voltage and 25% voltage will provide quite a usable range of output levels, say : "full", "small stage" , "garage/loud living room" or something like that..
If you want to be able to fine tune to get "bedroom", etc. then a continuous DC variation with a MosFet and a pot will provide higher resolution.

[guitarmike2107]

I was only going to do a few taps, maybe 3 or four, but the mosfet is cheap so I am looking at it as an alternative now.

thanks

[J M Fahey]

Well, at least it's easy to try with any amplifier and transformers you already have.
I'd only add that you cool it well, with a nice little heatsink properly placed, taking all due precautions with insulating the high voltage; I've seen more than a few just bolting it to the usual thin *iron* chassis, which is a very poor heatsink, or placing it carelessly close to very hot tubes..

[Steve Conner]

Or, you can use the MOSFET to adjust only the power tube screen voltage (and maybe the preamp supply to taste) which may let you get away with no heatsink. This is the method I used in my latest build.

If the power tubes are cathode biased, then the bias voltage will adapt as the screen voltage is changed, which makes it a very simple power scaling method. If fixed bias, you'll have to adjust the bias supply too.

The adjustment range available with voltage is limited, as far as subjective volume goes. If you lower it too much, the tone just gets weak and sputtery, like a transistor radio with a dead battery. That is why I added a 3-position switch that gives 10dB "coarse" steps by reconfiguring the whole circuit.

[guitarmike2107]

I have scaled the whole amp and it seems to work well till the lower regions, I will try scale just the power amp tomorrow to see how that sounds. I have mounted mine on the chassis, but will in the future probably use a heatsink, even if just for reliability.

I will also try and map out the power/voltage to see how closely it matches the graph above.

Steve, Did you try scale the anode and screens, before just scaling the screens, to see any tonal difference, if any. It would be easy for me to try so I may do that too.

Any thoughts on the best way for switching between two different pot settings, on two different pots. i.e. as switching master volumes on two channels. I have some ideas, to sketch up but am not sure about popping.