power supply suggestions and advice

[aardvarkash10]

Hi guys - I need to derive a +50/-50v supply for the source followers on my 1625 pp amp (vapourware but getting there). Bias on g1 of the 1625s is -22.5v and I'm aiming to run around 10ma through the FETs. Not a lot of current needed obviously, but rock solid V would be good.

Power supply is shown in the attachment - the 6.3v windings are already spoken for as all the tubes I'm using have 12v heaters. The 5v supply is good for up to 3A at least, and the B+ winding has a bit of capacity left.

That leaves me with the following options:

1. fit a separate 30-0-30 ptx
2. voltage multiplier on the 5v winding
3. create suitable rails on the main winding

#1 has space issues and is just, well untidy! I'd like to avoid it if possible but...
#2 is possible, but is it going to deliver enough voltage without noise and impedance issues?
#3 I have a gut feeling is begging for hum and stuff from mismatched impedance between the + and the - rails.

Comments?

 

[hesener]

hi, my preference would be #3 (if space is small) or #1 (if space is no problem). voltage multipliers can create noise due to the many diodes switching in there, and the efficiency is not that good. the noise can be filtered easily - in fact, the gate node of the mosfets are so high-ohmic that a smallish RC filter should do the job.

how do you intend to change the -22.5V to adjust the bias?

 

[aardvarkash10]

how do you intend to change the -22.5V to adjust the bias?

Balanced at the amp - see the pot between the two phases.

 

[wakibaki]

I like #1 because:

• It'll do the job as well as can be done.
• It's simple.
• There are no unknowns.

...except you won't get +/-50V from 30-0-30.

w