Fusion Run this weekend...

[bthoma]

I have finally found enough time away from work to get the fusor up and running. Preliminary results are posted below as well as pictures. I would like to confirm with a BTI detector, as soon as I can get one.
The chamber has been sitting under vacuum since the HV feedthrough was replaced last April. A “standard” pump down to the E-07 Torr range was followed by raising the pressure to 20 micron via the deuterium feed. First run was very anemic with an occasional neutron picked up on the detector. Subsequent runs were started by pump-down, then flooding the chamber all the way up to 200 micron before throttling to 20 micron and lighting the power supply. This produced better results. The chamber temperature hit 80+C during the runs.
The chamber is a 8” x 6” cross. Deuterium feed is a .005” capillary through a needle valve into the chamber top. Power supply is a Glassman 75KV/12mA supply.
Voltage was set for 50KV max and current all the way down as the chamber decreases down through 40 micron. At this point, current is increased to about 4 mA and plasma forms in the grid. The throttle valve is adjusted via voltage at this point as the density of D2 determined the voltage the power supply will provide. Once reaching 20kV (about 20 micron), the current is increased to 12mA. Neutrons on the probe start to record but are sporadic. The pressure is adjusted lower to bring the voltage up to about 27kV and neutron production starts to become consistent. Voltage was increased to 44 kV where peak neutrons were recorded. No higher voltages were attempted in these runs.
The best 3 runs were all short and I was concerned about the grid melting. The grid is tungsten and the vertical (??) grid wire was white hot while the other 3 grid wires were bright red when hitting good numbers. The grid stalk (304SS tubing) was also glowing red for 3cm from the grid. The 1/2+kW was mostly going to heating up the chamber, and I believe it was the electron stream hitting the chamber walls as a result of ionization. I used the Jon Rosenstiel method. Thanks Jon!
The neutron detector is a BF3 type (PNS-19 from Technical Associates Nuclear Instruments). Bias voltage was set at 900V. My pre-amp is blown so I used a Steve Sesselmann (Thanks Steve!) gamma spectacular hooked directly to the probe. All that investment in the NIM just sitting there powered down! I feel confident that the Gamma Spec works well. As mentioned before, I would like to back up the numbers with a BTI, though. The probe was recording during the pump downs with zero counts. The PRA software was tuned by eliminating all background by threshold level, regardless of shape. Once fusion started, the count level graph makes a nice upward sweep and was repeatable and consistent with expected results. The probe was calibrated with NIST traceability last October. The probe has a 25mm circular sensing element with a attached all housed in an aluminum cylinder. The outer OD of the probe is about 50mm. The probe is located 14.7cm from the center of the IEC grid.
Run #1 had an average count of 92 n/sec for a total average production of 16,317 n/sec. Peak on this run was total of 29,749 n/sec.
Run #2 had an average count of 118 n/sec for a total average production of 20,949 n/sec. Peak on this run was total of 48,810 n/sec.
Run #3 had an average count of 302 n/sec for a total average production of 53,709 n/sec. Peak on this run was total of 84,972 n/sec.
After run #3, I had to let the chamber cool down.
I think that the increase on each run was the condition of the chamber baking itself out with the high temperature of the chamber walls. The first run was preceded with a pump down bottoming out at 7e-07 Torr. By the 3rd run, it bottomed out at 3e-07 Torr. The chamber was hottest in the band surrounding the grid. The turbo only heated to 38C at it’s highest.
Anybody near NE Ohio have a BTI in warranty?

Pictures:
1. High pressure light up (5kV at 4mA in this one) to start plasma. Looks neat, but no fusion occurring at this point yet.
2. Hitting peak fusion on run #2. Note that vertical grid wire is cookin’! X-ray and Gamma on my Navy radiac set pegging the 1000 mRem scale at 10 ft.
3. Black probe is Technical Associates PNS-19 probe.
4. Granville-Philips 354 Ion gauge on left, Convectron on right for full vacuum range coverage.

[Doug Coulter]

Very nice! No doubt you made plenty of neutrons. You should try activating some silver in a moderator as well as the BTI verification. We use 1.5" thick HDPE between the chamber and the sample, then a bunch more HDPE (4") behind it for that -- works nicely and it's a very good verification for neutrons.

I'd bet that some slight adjustment of your grid would make it so different parts got hot instead -- I've observed that here. What I've mostly seen in other's sphere grids is that when everything else is right, it's the end that gets hot from all the ions hitting it where the wires are too close together.

I too had a ton of troubles with overheating when trying a sphere type grid in a cylinder type tank.
It works better for me with a very precisely built cylinder type grid, FWIW.

We find here that it takes sensitive gear to see neutrons at 16-20kv, but that the rate really goes up quick with increasing voltage. Though I can go up pretty high in current, I rarely see any improvement by going much over 10-15 ma (I can get 45 ma at 53kv here). Just more heat.

[Richard Hull]

Good report and little doubt about the neutrons. Welcome to the Neutron club. I will make the entry today. This is a fine looking system and the images are perfectly consistent with your results. Really nice work and I am sure this is just the start for increased numbers as you learn to work the system to peak performance.

Richard Hull

[DaveC]

Bob -

let me add my congrats to those above. Very nice test setup, and report. Easily followed.

The higher temperature of your vertical stalk, tells me it's the primary thermal path to ambient, - a combination of conductive and radiative heat transfer.

This suggests that you might want to increase the diameter, to decrease the temperature drop across it. If you could do a tungsten or Tantalum tube with a copper interior, this would give you the high melting temp ( and sputtering resistance) on the outside, while conducting more thermal watts away.

Of course the HV feedthrough is next in the thermal path, and its sealing area doesn't usually want to be too hot.

Nice work!!

Dave Cooper

[Steven Sesselmann]

Nice work Bob,

Welcome to the neutron club..!

Steven