My completed 14-stage Cockroft-Walton voltage multiplier

[304er]

Hi Steve,

Nice little convenient mod. Great clean integration in to the multiplier. Just as I would expect from you 👍

Chris Reeland
Ladd Illinois USA

[MRMILSTAR]

Thanks Chris!

[davekni]

I was concerned that the intense electric field might damage the circuitry or affect its operation but everything seems fine. I still have the option of using the hard-wired remote switch.

Similarly, I was amazed that wireless microphones worked for Tesla coil and Marx generator demonstrations.  My friend brought one to our Maker Faire exhibit and we each used it for our respective presentations.  In our use, the receiver was farther from interference and the transmitter closer, so that likely helped.

[MRMILSTAR]

I made a short video of an ion motor being powered by my voltage multiplier.  It took 3 attempts to make this video because I kept getting shocked by the intense electric field! I was successful on the third try because I moved back about 6 feet.

[MRMILSTAR]

I added a new improvement to my voltage multiplier. It is a protection circuit that prevents damage caused by the accidental application of reverse-polarity power. If someone accidentally applies DC power from the external power supply with the incorrect polarity this circuit will not allow the current to flow. Without this my ZVS driver and RF receiver controller would be damaged.

You can do this with a diode but a regular silicon junction diode drops about 0.7 volts. With 10 amps of current this would be a 7 watt loss. In addition to wasting power, you would have to use a heat sink to cool the diode.

The next best method is to use a Schottky diode. These only drop about 0.5 volt. This is better but it still represents a loss of about 5 watts at 10 amps of current. I initially used this method.

The best method is to use 2 MOSFETs and some logic. This circuit is sometimes called an "ideal diode". This circuit drops less than 0.1 volts at 10 amps. This is a loss of less than 1 watt. This is what I currently use. It uses 2 MOSFETs and a LM74700-Q1 ideal diode controller chip. I could have built it myself but you can buy a complete circuit for less than $4 which is probably less than I would have spent if I bought the parts and made a circuit board.

I have attached a picture of the small circuit board as installed in my voltage multiplier. This protection circuit would be useful for any project that relies on an external DC power supply. This particular circuit can handle up to at least 30 volts and 15 amps.

[davekni]

The best method is to use 2 MOSFETs and some logic. This circuit is sometimes called an "ideal diode".

Nice find!  Doesn't appear to be any particular need for 2 FETs.  Looks like they are just paralleled for higher current capability.

I also see two-terminal versions on EBay.  Would be interesting to analyze how they work, measure the reverse current needed to cause turn-off, etc.  Claim is 2us turn-off time, but that is likely with high current such as connecting a reversed supply input.
https://www.ebay.com/itm/325423537248

[MRMILSTAR]

The 2 MOSFETs are for greater current capacity. This one is rated for up to 15 amps. I didn't think the 2-terminal designs would have worked for my application. I also bought another 4-terminal one from E-Bay that only had the 2 MOSFETs but without the ideal diode controller IC. I measured its forward voltage drop and it was about 0.5 volts which wasn't any better than a Schottky diode. I didn't investigate why it performed so poorly. I just ditched it for the one with the controller IC which performed much better. The controller IC is a tiny 6-pin device and is hidden behind one of the blue wires in the picture.

[IonWind]

After 6 months of work I have finally finished my Cockroft-Walton voltage multiplier. It is a half-wave design with 14-stages. The specifications are: ...

Wow !!! Absolutely lovely to look at too

[Alberto]

Amazing work!

[pete_dl]

* Each diode string consists of 3 fast-recovery (100 ns) diodes in series, with each diode being rated for 20 KV and 30 ma

Its an impressive multiplier Steve.

What was your reasoning for three 20KV diodes ? Was it to aim for a 100% safety factor ?

Also, the electric field strength around the wires connecting the caps and diodes could exceed the dielectric strength of the oil. Did you see any evidence of ionization of the oil ? I am not sure what that would look like

Peter.

[MRMILSTAR]

* Each diode string consists of 3 fast-recovery (100 ns) diodes in series, with each diode being rated for 20 KV and 30 ma

Its an impressive multiplier Steve.

What was your reasoning for three 20KV diodes ? Was it to aim for a 100% safety factor ?

Also, the electric field strength around the wires connecting the caps and diodes could exceed the dielectric strength of the oil. Did you see any evidence of ionization of the oil ? I am not sure what that would look like

Peter.

Thanks. The peak input voltage from the flyback transformers is about 15 KV as I recall. I used 3 diodes in series just for a generous de-rating. I haven't seen any signs of ionization or breakdown in the oil. The multiplier has been completely reliable since I completed it.

[pete_dl]

Thanks for your reply.

I am starting to test a 120kv voltage multiplier not unlike yours, with vertical acrylic tube filled with transformer oil. When switched on, I can see movement in the oil (slight shimmer and small bits of dust moving around).

What I suspect is happening is the field strength around the fine wires is above the dialetric strenth of the oil and the oil is obtaining a charge, then flowing to lower voltage components to drop the charge off. Not the end of the world, but an inefficiency that will reduce the output power.

I might switch to wax or epoxy. The disadvantage is no easy maintenance, but the ionized solid sections will stay in place and not drain the charge.

Peter.