Capacitor multiplier circuits

[Zozobra]

I've recently started thinking about making my own PCBs for the main power supply boards, mainly as I use snap in caps and they dont look so great for turret builds and also because I can make the board much more compact and fairly generic as well as include elevated heaters and multiple bias feeds. Anyway if I'm going to that effort I may as well use a capacitor multiplier for the screen supply and save myself some money and real estate over using a regular RC filter as I'm mainly interested in using higher voltages. I had a look through Merlins power supply book and looked at the cascaded BJT multiplier which allows for up to 700V input and 100mA current draw without the use of a heatsink (looking at a pair of KT88s and 3-4 ecc83s so I should be safe there) and I got a little confused about the voltage rating of the capacitor to be used as it isn't explicitly stated. This is my proposed circuit. Shooting for a plate voltage of around 550V.



My silly question is do I need to double up C3 and balance it as I've done with the reservoir caps so that they're comfortably rated for the voltage (say 2x22µF, 350V in series) or have I missed a trick and any old cap will do? I suspect I need to double up but its not too clear in the book. If so what would be a sensible choice of resistor to balance the caps?

Any advice gratefully received
Ian

[loudthud]

C3 needs to be rated for the expected maximum voltage it will see. Two 22uF 350V caps in series with equalizing resistors would do the job. I like to use the snap-in caps with eyelets and put the equalizing resistors on the back of the board. The board will have to be removed when the time comes to change the caps.

[printer2]

On a related note, I was thinking about a future build with some 25W6's I have. Thought of taking a 24V transformer for the heaters and a voltage multiplier to get the supply voltage in the 150V range. Then I realized I could not elevate the heaters. Kind of a disappointment but for some reason I thought to go negative with the multiplier rather than positive. Heaters automatically elevated.

[Gingertube]

Zozobra (Ian)
,
That schematic doesnt look right to me for a capacitance multiplier. There is'nt enough resistance in series with the 10uf to form a decent filter time constant. I would expect to see something like 47K minimum for those resistors

Look up Elliot Sound Practices description of a capacitance multiplier to see how they work and then tweak the design here:
http://sound.westhost.com/project15.htm


For the voltage rating I would probably use a single MOSFET rather than stacked bipolar transistors.

Cheers,
Ian

[Gingertube]

Circuit for consideration / discussion

Here is a little circuit I posted to another thread for your consideration. This one is capable of around 500mA and can be scaled down.

With the link in the Time OFF position you have a simulated inductor (sometimes called a gyrator).

With the link in the Time ON position the gyrator action is compromised (reduced) by the 100nF/1uF capacitance divider BUT that 1uF now contributes a lot of capacitance multiplier action

Cheers,
Ian

[Zozobra]

Thanks for all the comments and suggestions, I'll do a bit more homework on this