Rebuilding a Precipitator Supply

[Alan Sailer]

So, as the proud owner of three broken 30kV precipitator supplies I decided to take the flybacks from one and build a whole new
driver circuit. Since I spent a year building several tesla coils I had some knowledge and lots of spare parts. Amazingly I was able
to do this without buying anything new.

The rebuilt supply uses the same style half bridge driver as the original circuit. The major difference is the use of a gate driver
transformer. The original uses a IR2153 to directly drive the two gates. The new circuit uses a SG3525 to generate the variable
complementary gate drive signals. The circuit didn't have the poop to drive the IRFP460 MOSFETs so I used a set of two totem
pole drivers to up the available output current.

The schematic;

I have tested this circuit with a dummy load of 10Mohm. The load can only handle 15 watts so I am only able to run the circuit for
a short time. It was putting out 30kV @ 3ma while drawing about 120 watts from the wall.

The circuit has a few advantages over the original circuit. First, I can actually put a good ten turn pot on it without it blowing up.
Second, I am using a variac to set the input voltage so I can turn it on carefully. I can control the output voltage two ways, first by
changing the operating frequency of the SG3525 or second, keeping the frequency fixed and changing the input voltage using the variac.

The circuit is also operating on the low side of the flybacks LC curve, unlike the original supply that operated on the high side. I talked with
Carl Willis about this and he agreed that operating on the low side is the conventional approach.

Some photos of the set-up;

I am sorry about the quality of the pictures. The forum software seems to hate modern cameras. I had to print out the images and
photograph them with an old Canon G11.

If and when I try this with an actual fusor I will report my results.

Cheers.

[Mark Rowley]

Looks great. And it sure is nice when all the parts are on hand. Rare occurrence over here!

Mark Rowley

[Richard Hull]

Great hacking there my friend. Very interested in how it comes out.

Richard Hull

[Alan Sailer]

I did a short non-fusion attempting run with old deuterium which is probably contaminated with air.
The rebuilt power supply worked as well as the old design. Reached ~28kV at one point, current unknown
as my attention was monopolized but yet another meter being shut down from EMI.

It has become increasingly clear to me since the fusor technology dates back to the 50-60's any equipment used with
the fusor must also date from the same era. No integrated circuits allowed, no LCD displays and nothing but d'arsonval
meters.

If anyone on this forum ever reaches the energy breakeven point it will definitely EM pulse the planet back to the bronze age.

[Richard Hull]

Fabulous and a big thank you, Alan, for noting what I have pushed. I might be an old throwback in many ways, but today's miracle electronics demands an absolute clean room environment of ultimate stability, EMP wise. EMI, however, is just a modern norm that most manufacturers of hyper sensitive electronics working at light speed have to plan for as best they can. EMP is another matter. The right blast at the right time at the right level and millions will lose they life!!....(smart and cell phone death)....For many, these are life itself.

The old analog world is a bit more rough and tumble, as you note. A person undertaking fusion might use the latest technology where it is great like switching supplies, but put a few old clumsy analog instruments in the mix to avoid destruction of costly modern instruments by the very modern switchers. This is specifically directed to fusioneers, of course. To keep the modern EMI sensitive instruments alive, the fusioneer might have to develop an expert level of EMI engineering to save them due to his working in a tortured high energy, high voltage environment. Today's common electronics world just doesn't see stuff in the EMI range that we fart with. Take a baby step back to the 60's at least in some of the instrumentation.

Richard Hull