DH>>If you wanted to get fancy, you could make a variable supply.  You
DH>>would want one based on a regulator rather than just using a variable
DH>>resistor since you do want the power supply to have as low an
DH>>impedance as possible.  If you use a high impedance, the supply will
DH>>"dip" every time you run a signal through it.  This -may- be musical
DH>>but it doesn't sound like it would be from just thinking about it.
DH>>More like the compressor from hell...
>     I've heard of people leaving off the cathode cap, to get compression.
>     (And occasionally, ear-bleeding feedback .)
>     And, yeah, I don't want to strangle the poor tube!  I've built
>     power supplies and such with 15V regulators, but I didn't know there
>     even was such a thing as a 45V regulator.
Yeah - these are not a simple 3-terminal deal but they are almost as 
easy.  It is an IC, you need a couple resistors, a pass-transistor and 
some caps but you can get regulation up to couple hundred volts very 
easy.  There is an internal voltage reference built into the regulator 
chip of 5 volts or so.  You use a resistave voltage dividor to take 
your plate voltage and reduce it to this level.  The chip will then 
adjust the pass-transistor to match the two and keep your output 
voltage constant.  You vary the leg of this voltage dividor to change 
voltages.  Any good electronics mail-order place will have them.
Mouser or DigiKey are two excellent places to start.
DH>>>     This solves a problem I've been mulling over for some years.
DH>>>     I built a larger experimental guitar amp with the passive tone
DH>>>     controls removed.  Yet, the amp still has the tube stage that
DH>>>     recovers the signal loss from the tone controls -- so the preamp
DH>>>     gain was way too hot.  Without that gain stage, the gain was much
DH>>>     too low, but with it the gain was too hot.  I fixed it with a 3 meg
DH>>>     resistor, but it didn't sound right.  Now it's obvious that the 
DH>>> thing
DH>>>     to do is reduce the plate voltages drastically, so the overall gain
DH>>>     is nominal but I'll get this effect you've mentioned.  Gotta try 

DH>>Actually, the gain is fixed by the components in the circuit and not
DH>>the voltage supplied to the circuit.  If you reduce the voltage, you
DH>>will still have the signal be too hot, it will just bottom out when it
DH>>runs out of plate voltage.
>     Ooops....  I guess I gotta use a lower value plate resistor and lower
>     the gain.
NO!
Think for a moment.
Lower value resistor for the same voltage will pass more current.
No more tube...
You must not touch the operating parameters of the tube if you only 
want to change the gain of the stage.  You need to look at the signal 
stages - what is happening at the grid.   Get or trace out the 
schematic.
--- Via Silver Xpress V4.4 [Reg]
---------------