Breaking the 1/2 meg barrier

[Bob Reite]

A bit of introduction for those not on fusor.net. The first successful version of my device uses a traditional spherical grid made out of three loops of tungsten wire crimped to a barrel which has 4-40 threads at the other end to fasten it to the HV feedthrough. The feedthrough stalk is not insulated on the vacuum side past where the stalk leaves the ceramic and enters the chamber proper. A sheet of lead is wrapped around the chamber and I can go to 50 KV without any detectable X-rays where people will be. However I blew out a grid trying to do 50 KV at 10 mA, so I now limit the input power to 400 watts or less, which means operating at 40 KV The bubble detector was positioned 20 centimeters away from the grid center. I also made a little stand for the bubble detector with my 3D printer. The exact position is marked on top of the cabinet so that measurements will be repeatable. The BTI bubble detector employed had a sensitivity of 33 bubbles per mrem.

I measure the operating pressure with a wide range ion gauge that will go up to 50 microns and is calibrated for D2. It has fussy electronics, so the "head" stays in an anti static bag on a shelf until the end of a run. After I shut off the HV, I plug in the ion gauge electronics module and measure, after verifying that the reading from the convection gauge did not change after turning off the HV.

So equipped, here are the results of my 40 KV runs.

October 12, 2014

Code: Select all

Eg 40 KV
Ig 7 mA
Pressure 12.5 microns
Flow rate 4.0 sccm
40 bubbles over 5 mins at 20 cm distance
calculated isotropic neutron emission 569,000 n/s


I did not have time to do any more runs that day. I shut off the electrolyzer and closed the mass flow contoller , but left the vacuum system running holding the system at 3 x 10-6 torr so I would not have to go though baking out again when I had a chance to resume testing.


Oct 22, 2014

As long as the system is properly purged, after "walking it in" from standby I am now getting fairly consistent results. Here are my 1/2 meg plus results for today. Between each measurement, I kept the fusor "idling" at 35 KV and 3.0 sccm gas flow for 30 minutes while waiting for the bubble detector to reset.

Run #1

Code: Select all

Eg 40 KV
Ig 7 mA
Pressure 13.4 microns
Flow rate 3.7 sccm
36 bubbles over 5 mins at 20 cm distance
calculated isotropic neutron emission 517,000 n/s


Run #2

Code: Select all

Eg 40 KV
Ig 7 mA
Pressure 13.1 microns
Flow rate 3.8 sccm
37 bubbles over 5 mins at 20 cm distance
calculated isotropic neutron emission 526,000 n/s


Run #3

Code: Select all

Eg 40 KV
Ig 7 mA
Pressure 13.1 microns
Flow rate 5.0 sccm
42 bubbles over 5 mins at 20 cm distance
calculated isotropic neutron emission 577,000 n/s


That's about a 10% difference from best to worst of the three.

[Doug Coulter]

Half a meg from those speeds and feeds is really decent performance.
We're using a pfeiffer pkr-251 gage and kind of guessing from its calibarion curves that claim it's reading about 2x high on deuterium. We also find that since at our pressures it's partly ion gage, and partly pirani - it reads off (or at least differently) when the fusor is running vs not, as some of what comes in is already ionized. I tend to report in millibars that the gage reports - and know that the real number is around 1/2 that. I'm not an enormous fan of metric, it's just what my gear tells me, and I'm lazy about constantly converting - in other words, it's the luck of the draw there.