Vacuum system parts collection and my development plan.

[Bob Reite]

Please find attached a photo of the vacuum system parts that I have accumulated.

The major components are:

A Pfeiffier pumping station which includes the forepump, turbo pump and controller.
The eight inch Conflat vacuum chamber. The major attraction of this particular unit is that it has a water jacket. This could be used for cooling and or measurement of heat output.
It also has four Swagelok fittings, one of which will be used to admit the deuterium gas.

Other components:

Hornet IGM401 Hot Cathode Bayard-Alpert ionization vacuum gauge for 1.00E-09 to 5.00 E-02 Torr measurements.
2 3/4" Conflat elbow
2 3/4" Conflat tee
2 3/4" Confalt bellows
3 circuit high voltage feed through
UHV butterfly valve
Extra copper gaskets
Misc hardware

Although this project will be accomplished in two or more stages, I don't want to build things twice, so the initial "demo" fusor and power supply will be built in such a way that the project can be expanded to a fully functional neutron producing fusor without having to rebuild anything, just add to the existing system.

HV Power supply design goal: Variable from near zero to 50 KV with 20 mA current at 50 KV. No oil filled components. I have begun modeling an air core resonant transformer design similar to Off-Line Tesla Coil designs. However I will only try for 25 KV output and use either a Villard cascade or Cockcroft-Walton multiplier to get to 50 KV. Operating frequency will be between 50KHz and 350 KHz, depending on how far I can push the IGBT module that I will be using to drive the primary coil. I may even decide on a lower AC voltage and more multiplier stages, depending on how the modeled performance works out.

I may also consider an "old school" 833 triode tube driver, since I already have some of those parts and tubes can take more abuse than solid state devices. The major disadvantage of a tube design is the 10 volt 10 amp filament transformer and the even bigger plate transformer.

The transformer pictured was originally from a broadcast transmitter. The secondary is 2500-0-2500 V center tapped with 240V input. This would work fine for the 833 power oscillator. But I still may go the solid state route, depending on how difficult the IBGT driver turns out to be.

I discarded the idea of a 60 Hz voltage multiplier from 5 KV to 50 KV for a couple of reasons: 1. In theory it would require ten stages with large and expensive capacitors. 2. The energy stored in those capacitors could do serious damage in case of an arc over.

Layout Considerations: Number one is safety! Number two is serviceability. I've studied most of the photos here on fusor.net and the high voltage connection always seems to come from the top of the vacuum chamber and great care has to be taken to guard it from accidental contact, as well as getting 25-50 KV out from the power supply to the fusor. With my cylindrical chamber I'm seriously considering placing the high voltage feed through on the bottom of the chamber, which in turn would sit on top of the high voltage supply enclosure. That way the HV connection between the power supply output and the chamber input would be quite short. All the high voltage connections would be totally enclosed inside the power supply cabinet. X-ray emissions through the insulator would be aimed down where no humans or pets could wander. The main disadvantage is that if there is ever any water leak in the cooling jacket connections, it would go right into the power supply, unless the top of the enclosure is made watertight.

The vacuum system would be to the right of the power supply enclosure. The vacuum chamber could be serviced from the top. In fact, I would just have an 8 inch blank on the top. This would be a good place for the bubble detector when I get to actual fusion and the moderator when I get to activation experiments. The view port will go on the 6 inch chamber outlet. I have not yet decided if the Hornet vacuum gauge is going to go on one of the angled chamber ports or on the tee. I also haven't decided if I need to get a "high range" vacuum gauge, so as to know when I can safely turn on the Hornet without damaging it. Although the gauge is supposed to be interlocked to prevent operation at higher pressures, the manual suggest using a convection gauge to know for sure when it is safe to turn on the Hornet without risk of damage to the fragile filament.

Overall development plan:
1. Design and build power supply. Build grid, probably with tungsten wire, so that it does not have to be built again for higher power levels. Assemble vacuum chamber and associated components. Operate the system as a "demo" fusor with no more than 14 KV input to fusor using air to check vacuum system and grid geometry. Completion of this stage should qualify me for the plasma club.

2. Build electrolysis system. Add completed gas system to the equipment built in stage one. Test with ordinary distilled water to test for hydrogen leaks and gain experience operating the system at higher voltages, up to the 50 KV power supply maximum. Of course no detectible neutrons are expected at this point. The reason for this intermediate stage with ordinary hydrogen is to troubleshoot any vacuum, fuel supply or high voltage issues, before purchasing the bubble detectors which have a limited shelf life.

3. Acquire bubble detectors. Operate electrolysis system with deuterium oxide. At this stage operating with deuterium gas, I should be able to get statistically significant detection of neutrons and join the neutron club.

4.Neutron activation experiments and future research yet to be determined.

[Bob Reite]

PEM Fuel Cell from the "Hydrocar" kit I scored on Ebay.

Rooting through my junkbox turned up a high voltage probe that was damaged during my last move. However the 600 megohm multipler resistor survived intact and the 50 uA meter movement is OK, so I now have a means of measuring the power supply output to 30 KV which will more than enough for the demo fusor, and others have done fusion at 30 KV. In the future I will get another 400 megohms worth of high voltage resistors to give me 1 gigohm, which will indicate 50 KV full scale on a 50 uA meter.

A cockroft walton multiplier and a couple of transformers that I purchased arrived today. United Parcel Smashers lived up to their name, the anti-corona top terminal is just sitting there in the photo, the nylon screws got sheared off in transit but it is repairable. Next step is to see which of the two transformers I want to use. The red one is supposed to have been originally used with the multiplier. The one with the blue windings is a design by Carl.

[Bob Reite]

I disassembled the cockroft walton multiplier pictured above to reverse all of the diodes as it came configured for positive output. While I had the diodes out I tested them, using a techinque I used for testing high voltage diodes out of transmitters. HV diodes typically have a drop of 6-50 volts, depending on their rating, so an ordinary multimeter will show them as being "open". All the diodes tested good. Then I drove each module by itself with a function generator and the blue transformer and verifed that they all made negative DC. The multiplier was reassembled and tested with the red and blue transformers. I used a 400 watt "monoblock" audio amplifier of my own design driven by a function generator. The blue transformer gave the best results of the two, best efficiency was at a 30 KHz operating frequency. No load I can get 60 KV without a problem, although the corona becomes very noticeable much above 50 KV. Since 50 KV was my design goal, I'm going to "call it good". Next was to test under load. I could get to about 40 ma at 4KV (into a 100K wirewound resistor) before the rail fuses on the power amplifier blew. New fuses and the amp was good as new, but I'm going to have to come up with a suitable load for load testing at higher voltages. I will probably come up with a more "cost effective" drive circuit in the days ahead.

[richnormand]

Hi br2600,

(((( While I had the diodes out I tested them, using a techinque I used for testing high voltage diodes out of transmitters. HV diodes typically have a drop of 6-50 volts, depending on their rating, so an ordinary multimeter will show them as being "open". ))))

You got me intrigued here. Can you say more about this process/

Cheers and thanks.

[Richard Hull]

HV diodes will never test on a simple digital meter on the diode function. (unless they are shorted) Too high a drop.

Richard Hull

[Bob Reite]

I used a 48 volt DC power supply in series with a 4700 ohm 1/2 watt resistor in series with the diode under test connected so that it will conduct, then measure the voltage across the diode. A good HV diode out of the multiplier will read 6 volts. A shorted diode will read zero, an open diode will read 48 volts across it. Testing using this method goes real fast, since you don't have to reverse the diode. A 48 volt test supply will work for HV diodes up to 20 KV rating.

[mike-bushroe]

Hi br2600,
You got me intrigued here. Can you say more about this process/
Cheers and thanks.

If you want to see more details of the process, use a variable power supply and hook the volt meter across the diode. As you bring the voltage up from 0 to 6+ volts, you will see the entire voltage drop across the diode. There will be virtually zero current until the diode is fully conducting and zero amps through the 4700 ohm resistor means zero volts across it. Once the diode gets into forward conduction mode, all additional increases in the power supply output voltage will appear across the resistor and the voltage across the diode will remain constant. Then reverse the diode and start from zero again. This time the entire voltage will appear across the diode the whole way as it is in blocking mode. There are curve tracers that can do the whole process automatically and display the results on a screen, but you will have just demonstrated the functional properties of a diode. But once you get bored with the whole process, as he said above, straight to 48V with 4.7K and read the voltage off the diode. 0 volts -> shorted, 48V -> open, 6+V -> good.

[Bob Reite]

Pictured is a totally enclosed rack cabinet that I won on Ebay for about 1/4 of what it would cost new. The digital meters are a bonus, I was originally going to use analog meters, but since these came working with the rack, I'll use them.

Except for the odds and ends that you always seen to be short on when beginning a build, the acquisition phase of the project has ended and the construction phase is beginning.