# Cockroft-Walton question.

RGClark
Why couldn't you have the output of a CW voltage doubler lead into the
input of a another doubler? It seems to me that instead of the voltages
being additive with additional stages as done now, with this method you
could double the voltage each time.
So with 10 repetitions you could multiply the voltage by 2^10 = 1024.

Bob Clark

Mentor
I think you can have as many stages as you want, as long as the ladder method is an efficient enough voltage multiplier for your needs. Where did you read that you can't cascade the ladder?

http://www.blazelabs.com/e-exp15.asp

Random
Yes, you may have as many stages as you can feed - at some point however the effective lumped series resistance will start to defeat your efforts.. so you can't start withl 12 v and multiply your way to 12 MV.

Antiphon
What you want to do doesn't work because the output of the newtork is DC, not AC.

If you had a voltage doubler with an AC output, the stages could cascade exponentially as you describe.

Mentor
Antiphon said:
What you want to do doesn't work because the output of the newtork is DC, not AC.

If you had a voltage doubler with an AC output, the stages could cascade exponentially as you describe.
I don't think that's right Antiphon. The voltage multiplier that the OP is asking about is a half-wave rectifier/multiplier:

http://www.blazelabs.com/e-exp15.asp

I suppose if you follow the last AC stage with a single diode and then a capacitor to ground, that will give you a DC output. But if you keep following the ladder structure of diodes and caps, the AC pulses get passed up the ladder.

Antiphon
Actually, no. If you look at the polarities of the caps, they are all charging with the
positive side on the right. The output will be DC.

Mentor
Antiphon said:
Actually, no. If you look at the polarities of the caps, they are all charging with the
positive side on the right. The output will be DC.
Oh, I see where I was confused. There are AC currents through the caps, but the result is indeed pretty much DC at the output. I misinterpreted the OP's question -- he was asking why it wasn't more efficient to series connect separate doublers instead of just adding more stages. Yeah, the C-W multiplier only grows by adding stages, not by trying to cascade separate ones somehow. Thanks Antiphon. -Mike-

CharlesP
Need 3 phase

OK now what is the circuit for a three phase machine? Powered by three phase RF hopefully.