Archived - Carl's Jr. "first light"

[Carl Willis]

I often get around to doing things long after I say I will. This past weekend, I intended to finish Carl's Jr. and maybe even feed it some deuterium.

That didn't quite happen, but I did run it a bit in air to check things out. The photo shows the stable air discharge at 5 - 10 mtorr, 45 kV, and 2 mA. Under these conditions, the grid does not glow at all but the exposed portions of the feedthrough insulator fluoresce a bright bluish color tinged with pink at the edges. X-ray exposure rate measured with a "cutie-pie" is about 8 R / hr at the viewport and 150 mR / hr around the chamber. I apologize for the quality of the pic--I had to de-speckle it and adjust brightness and contrast since the glow is naturally quite dim and I didn't use the long exposure feature on the camera. I'll get some more eye candy shortly, hopefully with a deuterium discharge instead.

Here's a summary of what I've learned so far:

-The "oil socket" idea works great. At ultimate vacuum in the low 10-6 range, I can turn my power supply up as far as it will go, 60 kV, and still get no breakdown or threats of breakdown anywhere in or outside the feedthrough (which as mentioned previously is one of those Navy BeO feedthroughs that are up on eBay for about $25). In the absence of oil, spark breakdown usually comes around 40 kV. I have no doubt that I could put 75 kV on this insulator, and it would probably handle 100 kV or more.

-The ion source kinda works, but needs revisions. Namely, the conductance of the 18 ga. needle in it right now is way too high, and the glass tube for HV standoff / gas entry goes too low in pressure and starts acting like a neon sign. I'm going to move up to the 30 ga. neighborhood. On the rare occasions when the ion source is working properly, it does a remarkably good job at keeping the main discharge stable when otherwise it would be stuttering. The magnets seem to not be too important to it, although I have to play around some more with this when the pressure issue is resolved.

-Stability (in the absence of ion source operation) is excellent at currents of 2-3 mA and voltages of ~ 40 kV if a ballast (250 watt light bulb) is placed in the primary of the power supply. The ion source seems to extend stable operation to under 1 mA at 50+ kV. Without the ballast, operation at 50 kV and 5 mA is possible but does not seem very stable. In any event, I will need to get the water cooling working if I want to run long-term at this level.

Some specs on this fusor:
______________________________________________________

Chamber: 6 inch Wagner hemispheres w / 8 inch ConFlat flanges, water-cooled by 1/4 inch copper pipe

Grid: 5 loops 1.5 inch dia. 20 mil 316 stainless steel

Vacuum: diffusion pumped, estimated 5 l / s in air, base pressure around 3E-6 torr.

High voltage: Hipotronics 50 kV / 5 mA with switchable option of a 70 W power-limiting resistive ballast. Subject to improvement.

Ion source supply: Plastic Capacitors 15 kV / 5 mA module

Gas feed: two Hoke metering valves and one shutoff valve in series. Still to come is a single-stage, high-purity regulator for deuterium, just obtained.
________________________________________________________

-Carl

[Richard Hull]

Fantastic! Looking forward to more reports on this novel system of ideas. This is Archived, naturally.

Richard Hull

[DaveC]

Carl - Nice image. Your HV improvements seem to be working. Just a thought on the gas supply and metering. I would suggest making the gas supply a low pressure tank... say about 500 microns or so. Then use it to feed your metering/supply valves and such. You will get a much better flow control. But... you wil probably find even the 30 gage hypodermic tubing is pretty large... a coil of gas chromatograph capillary would probably give you more conductance limiting. Most all of the HiVac leak valves I've used, are very nasty at the initial opening, and become more predictable as they are closed down. Usually two in series work fairly well, once a flow is established.


FWIW....

Dave Cooper

[Carl Willis]

Well, the Willis "nuclear program" is finally back on line. This fusor turned out its first neutrons tonight, and while some issues need to be ironed out mainly with regard to the gas feed, I'm quite pleased overall with the performance. The first photo shows the deuterium discharge at 50 kV / 4 mA. A BF3 detector nearby was happily chirping away at this point. Obviously a much sharper, redder star than with air. Second photo is a shot of the whole system. A capacitance manometer will be added to the blanked-off QF25 tee on the fusor's left port as soon as I have the QF25-to-QF16 adaptor for it. Ion source is on the right port. Last photo is a closeup of the "oil socket."

-The water cooling is in business now, and wow, does that REALLY improve the discharge stability at 250 W. I bought a 1/10 HP pump at Surplus City here in Albuquerque, and I think it can probably support several gal / min at 20' of head. Anyway, it's overkill for this job but still small and quiet. The pump takes suction from a plastic Tupperware container that holds about 4 gallons of water. Most of the pump discharge is routed through a coil of 1/2" copper pipe to act as a heat exchanger to the ambient surroundings, and then returned to the reservoir. A small fraction (controlled by valve) is routed through the 1/4" cooling coils on the Fusor and returns to the reservoir. I'll get some more data on the flow rates and pressures later. The fusor shell used to get too hot to touch at 250 watts before cooling was added, and the discharge had to be carefully baby-sat at this power level. I think rising temperature would increase outgassing, raising the pressure and the current: a positive feedback mechanism. The water cooling keeps the fusor shell cool to the touch, and more surprising to me was that there is not much temperature differential on the surface between the copper coils and the empty areas. SS is a poor conductor of heat, but evidently it's good enough for this design. I can run at 250 watts with no ballast, and the fusor is rock-solid.

-Replaced the ballast in the power supply (250 watt light bulb) with two 200 watt light bulbs. Gives a better range of protected power output.

[bwsparxz]

Nice work Carl! Glad to see you are back in the neutron business. Hope to be there sometime soon. Your cooling system is working alot better than air cooling, I will have to use that when my chamber is contructed. I will try not to steal all your innovations . I will be taking ideas from many fusors built here and use what I like and can afford. Good luck .

[Jon Rosenstiel]

Good going, Carl, welcome back to neutronville. Sounds like your H2O cooling setup is the way to go, and it has to be nice to not have to listen to an air cooling blower screaming away.

Also, nice pix, particularly your star photo.

Jon Rosenstiel

[Starfire]

Ground-breaking Carl, first class work, congrat's.

[Carl Willis]

Hi Dave,

The gas feed is a challenge, particularly since I want to conserve deuterium a little better on this fusor than on the last.

I bought a 3 ft. length of 0.005" capillary from Small Parts (the longest length of the finest capillary they stock). The entire length should limit the flow rate to about 0.5 sccm, although there is naturally a great deal of uncertainty linked to the variations in bore diameter, surface roughness, molecular flow at the downstream end, etc. Unfortunately, I'm going to have to connect the coil of capillary directly to the HV feedthrough, putting it downstream of the glass-tube gas feed / HV standoff in order to keep the pressure in this standoff high enough to prevent it lighting up like a neon sign and shorting the HV. Ideally I would have preferred to have the capillary upstream of a cutoff valve, but oh well. The metering valves and cutoff will necessarily be on the grounded copper conduit leading to the glass standoff so I'm not tempted to shock the bejeezus outta myself.

I hope this will solve my ion-source problems!

-Carl

[DaveC]

Hi Carl - Think I got the picture of what you have in mind. Seems like it should work.

I don't have a source to suggest, but I have seen ceramic capillaries, that were used for real time Mass Spectrometer analysis. Heated too, as I recall.

If you can keep the voltage gradient low around the low pressure lines, by appropriate shielding they won't light up, but there is still the matter of getting into the HV region, so as not to have valves at high potential. An insulating capillary laid out so as to have most of its pressure drop in the HV terminal, might get you past the problem. Based on my experience, you need to be below about 10-4 Torr for potentials above approximately 10kV.

Very interesting and creative work. Thanks for sharing!!

Dave Cooper

[Todd Massure]

I must agree with everyone else: nice work! I'm really enjoying watching the progress - especially the new innovations you've come up with.

[bwsparxz]

Carl, a good source for mass spectrometer capiliary and other items you might find usefull is Scientific Instrument Services. There site is:

http://www.sisweb.com/

Not sure if you have seen them or not.

[Richard Hester]

Sisweb were the folks I came up with last night as well when I was searching for capillary tubing. They also have needle valves that work with the tubing, and a host of other interesting items and tidbits of information - a good place to bookmark.

[Richard Hull]

Lots of food for though and future copycating possibilities here.

A lot of us commenting here will get together live in front of my sad little fusor IV this weekend for some real campfire discussions.

Richard Hull

[bwsparxz]

Richard, Fusor IV is far from sad! The only 2 fusors I have seen in operation is yours and Carl's old fusor. I look forward to seeing yours again this weekend. It keeps me going till I finally collect all the pieces for mine. Yes, I see some copycatting from me , possibly from the cooling system. I expected to see more of Carl's ideas being used.