NST, stacks, Spellman, Glassman etc.???

This forum is for specialized infomation important to the construction and safe operation of the high voltage electrical supplies and related circuitry needed for fusor operation.
Post Reply
scinut
Posts: 21
Joined: Tue Mar 01, 2011 3:42 am
Real name:

NST, stacks, Spellman, Glassman etc.???

Post by scinut » Sat Mar 26, 2011 1:56 am

With very little of my experience with voltages above 1kV, this question may seem a bit naive. What is it about Spellman, Glassman, et al power supplies that makes them so much more expensive than what you could build yourself? I can see things such as sensors or detectors that may have a very expensive substance involved like boron, or platinum, or..., but as far as I can tell, most support electronics such as power supplies and drivers do not normally have anything special like that. What about these expensive supplies is out of reach for a DIYer building one from scratch with regard to performance? Duty cycle?
Many here start with a NST, CW, a flyback transformer or MOT, but always seem to find them inadequate for one reason or another when they attempt to progress beyond plasma.
Anyone with some thoughts on this please?
Jerry

John Futter
Posts: 1624
Joined: Thu Apr 22, 2004 2:29 am
Real name: John Futter
Contact:

Re: NST, stacks, Spellman, Glassman etc.???

Post by John Futter » Sat Mar 26, 2011 2:38 am

Jerry
for the most part a lot of the parts are hard to get in low numbers

I have designed high voltage power supplies and the usual minimum number is 100 pieces with some being 1000

For instance the diodes i wanted 2CFM2 come from china min buy 1000 @ US1:50 each
200mA 20kV and fast single cycle surge capacity of 10A @ 60Hz

Murata HV ceramic caps min order 100 it took 8 months for them to arrive --another words I was piggybacking on a Glassman or Spellman order.

Large 3C90 U cores I had to buy 100 halfs min order--- not available anywhere else


Now a clever lad may want to strip a uWave oven you will get the ferrite cores from the invertor (note not U cores but usable) type but the diodes are only 6kV and theres only two per oven so plenty of ovens needed.

Yes Digikey and Voltage Multipliers stock some of the bits but at a greatly inflated price

The last thing is that you have to be able to control very closely dV/dT and dI/dT otherwise there will be alot of silicon that the smoke has escaped from.

Some here have used secondhand x-ray transformers but these are designed to run flatout for a very short time and will draw enormous current if left to idle without load.
NST's are self protected and therefore are very soggy in regards to output volts at any reasonable current.

If it were easy companies like Spellman /Glassman etc wouldnt make any money as everybody would roll their own

scinut
Posts: 21
Joined: Tue Mar 01, 2011 3:42 am
Real name:

Re: NST, stacks, Spellman, Glassman etc.???

Post by scinut » Sat Mar 26, 2011 3:29 am

That makes sense. I don't like it, but that does explain what happens with some of the components. I have seen posts where some have tried to hook up cheaper, easier to obtain components in series or parallel to achieve their desired results. In fact, it was your reply post about connecting MOTs in series that inspired my post. There is another thread about simply coiling up a large high ratio turns transformer with some rectifying diodes on the output and just going straight to the disired output. Some have called this brute force, but maybe it is pure and simple. I do, however, appreciate the alternate solutions that accomplish the end result with multipliers, etc. Perhaps, some here may be interested in pooling their purchasing power in order to buy some of these expensive minimum order items, like I have seen done with bubble dosimeters. Are the schematics for good "factory like" practical power supplies, that meet the fusor's requirements, readily available?
Thanks-Jerry

scinut
Posts: 21
Joined: Tue Mar 01, 2011 3:42 am
Real name:

Re: NST, stacks, Spellman, Glassman etc.???

Post by scinut » Sat Mar 26, 2011 7:59 am

You gave me quite a few options that sound like you have confidence in. Thank you! I think I have gotten the impression from many of the posts that using an NST, MOT, or flyback xfmr with a multiplier (especially with too many stages) is like wearing brown shoes with your black tuxedo, even though the high priced factory supplies use essentially the same ckts. I have a variac, a 15kV NST, one flyback xfmr, and a fair number of 12kV to 15kV diodes. I have seen CWs with disc caps, and door knob caps. I have some of both. The area that somewhat spooks me is the current limiting ckt on the output with regard to the kind of resistor to use such as wire wound, carbon, etc., and the proper wiring of ammeters where the meter is not actually part of the ckt. Current xfmr to ammeter? What is the mininum frequency for a pulsing ckt.? Thanks again-Jerry

Post Reply