Page 1 of 1

Images of 100 kV Voltage Multiplier

Posted: Sun Sep 18, 2016 5:40 pm
by Dennis P Brown
Well, put some more parts together and apparently, managed to build a simple multiplier (half wave) that appears it can reach 100 kV using a small NST; tested to 50 kV but that required only 42% of the variac range. Until I put it under oil, won't risk the diodes any higher and is the limit for my HV probe. Once I redesign the system I will add a voltage divider on the output to read the real voltage and what the system can do in the 50 - 100 kV range.

Also, will replace the first two cap/diode array stages with door knob/smaller diodes and see if this works as well. The first two stages just carry far too much charge and I would prefer not to have to that issue. Also, such an assembly would be significantly more compact with just the door knobs capacitors.

The caps have identical voltage ratings (40 kV) but slightly different charge capabilities. I just keep it balanced so the connecting hardware is easier to match between each cap. Since I got a deal on these cheap (7 of each type), I am using what I have on hand. So far, this has cost me under $165 total (caps, diodes, hardware, and NST.)

Re: Images of 100 kV Voltage Multiplier

Posted: Sat Sep 24, 2016 2:57 am
by prestonbarrows
That is some nice and clean work. If you are shooting for 100kV, you might begin to have corona issues on those ring terminals towards the HV end. Even under oil, the small radii of the terminals and the wires themselves can start to become a problem.

As you are learning, producing high voltage is not that hard, the trick is a high voltage supply that can produce high current also. You don't mention the values of your components, but you are reaching up into the area where ripple will become a real issue with this multiplier when it is under load, especially if driven by a NST.

It would also be nice to see all that enclosed within a grounded enclosure for safety though that understandably introduces a multitude of feedthrough issues.

You may want a cap to ground to smooth the output, a resistor divider across the output to ground to monitor the voltage and a series resistor on the ground side to monitor the current.

Re: Images of 100 kV Voltage Multiplier

Posted: Sun Sep 25, 2016 5:12 am
by Paul W Fontana
Looks really nice! Is that a five-stage Cockroft-Walton?

Let us know what happens when you put a load on it.

-- pwf

Re: Images of 100 kV Voltage Multiplier

Posted: Tue Sep 27, 2016 3:52 am
by Bob Reite
Looks to me like it's half wave, I'm afraid that it's going to sag terribly when you put a load on it.

Re: Images of 100 kV Voltage Multiplier

Posted: Tue Sep 27, 2016 1:56 pm
by Richard Hull
This is the type rectifier/multiplier used by many of the early gas He-Ne lasers in their power supply. Metrologic was famous for it in their long slender aluminum housings. It allowed a simple, small 200 volt 50ma transformer to start and run the He-Ne. They used a heptupler arrangement with a single 2ufd 600v electrolytic as the first cap and .001ufd 2kv discs for all the others. When turned on, the cold laser saw about 1500 volts DC across it. Once the tube fired and the gas ionized, the multiplier buckled under the 10ma load of the lit tube and it ran at a reduced voltage of a couple hundred volts due to the larger 2ufd cap. Brilliant and simple electronics design. (They relied on the failure of the multiplier) As the tube aged and outgassed, this supply proved a bit onery as it was just able to start a good tube. Often, on a weak tube, the tube would flicker on and off as the supply bounced bewteen HV and the lower run voltage.

Other companies used a more substantial, 1kv 30 ma xformer and a 1ufd cap with only a simple tripler using .001 discs at 3kv. While this supply was a bit bigger, but it never failed to lightoff the laser even when it would no long lase.

All used a 30k to 50k ohm 2 watt carbon resistor as a current limiter between the supply and the tube.

All such half wave systems are no better than the transformer that feeds them and the size of the caps in the multiplier. As mentioned in many other posts over the years, larger caps can supply current, but also store unforgiving, one touch and you are dead charges.

Richard

Re: Images of 100 kV Voltage Multiplier

Posted: Wed Sep 28, 2016 11:41 am
by Dennis P Brown
Thanks for all the advice. I will put this 5 stage system under oil in a plexi-glass tube (just have to add oil now.) Since it isn't going to act as a real power supply I hope saging will not be an issue. This unit is simply going to be used to charge a 12 inch diameter dome (so a few tens of micro-amps at worst current draw.) So I'd think being half wave shouldn't be an issue nor the use of a 60 Hz NST.

The corona issues even at 45 kV are terrible (things begin to get flaky above 25 kV) - the stack is humming like angry bee's. Arcing between places that just doesn't make sense (but never along the 5 stage stack.) Yes, five stages are in the photo but the whole system is really 7 stages. This 5 stage stack is fed by another two stage system as seen in the picture. That system can be very deadly since the ends are exposed in a dangerous manner; I did enclosed that feeder system in a special housing. Even then I now keep that system away from my general work area for safety for testing and use.

An aside: I was concerned that my high voltage probe was way off so I did pull out a digital DC supply that goes to +/- 6 kV and the probe was only off by 500 volts. It also agreed with my Glassmen at 40 kV unit but that is scale is a bit hard to read (can be +/- 2.5 kV.) So, I do believe the number I obtained (and the arc over of an inch past my insulator plate to the floor was impressive and the system was barely at 40 kV.)

However, my voltage divider that I had designed for use with this system was reading very different numbers (this is a 1G ohm and 1 M ohm divider. ) That simple device only read 0.8 volts for an applied 6000 volt. Wow, never believed a resistor could be that far off - of course, 1 G is not readable by my ohm meter but it claimed 1% accuracy - maybe it was damaged and conducting part way.

Re: Images of 100 kV Voltage Multiplier

Posted: Wed Sep 28, 2016 11:20 pm
by Richard Hull
If you need a gigohm do it with 10 mice long 100meg resistors. Single resistors at super HV are not the norm unless manufacturer for that purpose. A 1 gig divider resistor might be about 10 inches long, if measuring about 50kv.

Richard Hull

Re: Images of 100 kV Voltage Multiplier

Posted: Fri Sep 30, 2016 1:06 am
by Bob Reite
If you are only going to have a microamp load on it, the half wave multiplier should work as predicted by calculation.