Cockcroft-Walton

something like this?

-Lasse

Yikes, that 433 volt circuit starts to make inductors look good again.

this?

yeh I guess that is one of those "just because you can"

-Lasse

Sure, we have a full wave 2 phase WC rectifier system in a vessel of SF6 gas that generates ~3M volts.

Up to 10kV RMS @ 100Khz from a litz wire Pie transformer and by the time it gets to the other end it can get to a few Million volts.

Oh, btw, since there is also a ripple of 100khz at the other end, we use this ripple into a RF transformer where we can generate the heater voltage for the filament. All of this is control via motors at the cold end coupled to rheostats at the other end via delwin rod material.

since the first stage coupling and many of them afterwards are capacitive coupled, the 74HC04 can serve to work the same way using cap decoupling.

Replace the transformer output here with the output of the 74HC04

Jamie

Jim, possibly something like this is what you are looking for:

peter

This one? ;~)

Beat me to it. Art

message

Thanks, Peter and Art! I'd lose track of my @#$$, except it's bolted on ;-) ...Jim Thompson

Must be an RF engineer, they do things like that.

Cheers

Interesting article...but I don't understand it.

If the output of every other CMOS gate is the same (ignoring prop delay), why do I need more than two CMOS driver gates for ALL the caps???

with_CMOS_gates.pdf

Each gate charges a cap (then switches a diode). Each cap eats all the current a gate can provide. Hence one gate per cap.

Mark L. Fergerson

this?

Ok, suppose that makes sense if you've got a lot of gates laying around. Doesn't take too many stages before a dual 555 or gate driver chip starts to look attractive.

Also noticed that the caps have increasing voltage as you move along the chain. C-W doesn't require high voltage caps.

Yes, they could be in parallel all the same.

True, but CW also accumulates diode drops. After maybe 20 stages, between diode drops and a losing battle with series resistance, things start to look futile -- well, this is true of any multiplier, really.

Tim

this?

Sure it does. It has to have a cap rated at least twice the voltage of the first stage's drive signal. so, if you have a 750 volt drive, each cap in the multiplier should be at least a 1500 volt cap. Every cap will see the exact same voltage. C-W DOES require HV caps, idiot. What a stupid thing to say.

And, then also note that the voltage from stage-to-stage DOES increase, and EACH gate chip MUST be discreet from ANY others in other stages, which means your single chip idea has about as much validity as your affect in this group does. Which is nil.