Another update: Carl's Jr.

[Carl Willis]

Carl's Jr. is in the neutron business, if only barely, with the anemic HV power supply currently on hand. Tonight I made the first quantitative neutron test, and installed some upgrades mentioned below.

-NEW CAPILLARY GAS FEED: To improve my gas feed and ion source performance, I ordered a 30" piece of the finest capillary available at Small Parts (0.005" bore) quite a long time ago. According to Poiseuille's Law it will restrict flow to a couple sccm. It got back-ordered, it got lost by the Postal Service, and finally a piece got here a couple days ago. After some challenges trying to cut it, solder it and clean it without plugging it up (very frustrating, trust me!), it is now in business. And it works great. The metering valve is almost pointless. To get the right flow, it's easier to change the line pressure at the regulator.

-NEW LB GAS REGULATOR: I am also sporting a new, sweet little single-stage lecture-bottle regulator, distributed through Joseph Norrie of TeslaToys.com and made for hydrogen service. The Series 2500, whose brass body is machined out of bar stock and has low internal volume, strikes me as being very economical at only $85 (which includes the option of a stainless steel diaphragm, basically what distinguishes a high-purity hydrogen regulator). The big boys like Matheson make these regulators also, but they cost in the hundreds of dollars and rarely show up on eBay. The regulator ships with the CGA-350 flammable gas fitting. I had to trade it for a CGA-170 obtained elsewhere.

-AND FINALLY, SOME NEUTRON NUMBERS: Courtesy of my BD100R bubble dosimeter. Running stable at 47.5 kV and 2 mA with no cooling (only 95 watts), the dosimeter picked up 34 bubbles in 10 minutes next to the equatorial flanges. That's 6 mrem / hr, and the emission rate is probably in the range of 150,000 - 200,000 / s based on this test.

I was hoping to do a little better honestly, but there are some issues that might be standing in the way. First of all, that capillary and the associated plumbing is very difficult to outgas and right now is still full of water and air and other junk. I can tell based on the not-so-red color of the discharge, and because when heated, the capillary evolves a lot of contaminants that also change the discharge. I have a change in mind that will clean up the plumbing: I need to connect a vacuum purging line into the plumbing behind the capillary so I can pump it out quickly and get rid of the condensation, air, and junk. The other issue I think relates to power supply and cross-section. My power supply is very nicely capacitively filtered. What shows up on the meter is an accurate representation of the voltage at all times. The peak voltage is 50 kV, the average is 50 kV. Now on an unfiltered x-ray trannie with rectifier, the effective voltage seen on the meter is not the peak voltage. Most of our x-ray trannies are probably good for 60+ kV peak, though the meter might only show an effective 45-50 kV. The cross section shows quite a bit of curvature in this low-energy realm, and is quite more sensitive to the peak voltages.

-Carl

[Richard Hull]

Thanks for the update. I have a couple of feet of SS capillary as well as a 10ccm MFC. I ultimately hope, with Fusor V, to install this critical upgrade.

If you figure on 250,000 n/sec then that is not whimpy. It is about the best I have ever done! Still it would be nice to hit the 10e6 mark for easy neutron acrtivation experiments. This is my ultimate goal. I have a ways to go yet and little free time now with all my other activities.

Carl Jr. will easily glide to the mega mark, I am sure. Keep those updates coming.

Richard Hull

[Richard Hester]

Carl - I wouldn't complain. Those are the best efficiency numbers I've seen on any fusor here.

[Starfire]

Carl & Others - did you use the reservoir idea to reduce the flow or just the long tube?
I have several MFC but all above 20 sccm { much to much for my design }. I am going to try to pulse flow a one bar supply into a gas reservoir before feeding into the Ion gun, to reduce delivery. Was thinking to include a hypodermic line as I have several one metre lengths ? (donated by a local company who make medical kit ) The problem is to calibrate. Electronic pulse timing a doddle but not so gas control. -- your thinking will be appreciated.

[Carl Willis]

Hi Richard,

Jon Rosenstiel still has the efficiency edge I think. He's putting in ~12 times my power and getting out 30+ times my neutron count. But he's in my cross-hairs and I do plan to clean up Jr. to the point that it is doing as well. Looking in the viewport, I noticed some drops of vacuum grease that must have fallen off an o-ring or something and splattered the walls. Yuck. And I need to replace my stainless grid with tungsten for kW-level operation. And I need to implement the gas feed system changes. I know what I need to do.

Of course, the power supply is a long-term improvement that I am working on. I believe some of these x-ray machines, like the one I had previously and the one Jon has now, can hit 65 kVp with 45-50 kVRMS (full-wave rectified unfiltered DC). That is a number to beat.

-Carl

[Carl Willis]

Hi John,

I don't really intend to have a reservoir. The 30" coil of capillary actually needs some pressure behind it to get decent flow, and in the interests of conserving deuterium I will keep the volume in the feed circuit low.

I could not find hypodermic needle stock small enough to be effective. It may indeed be made, but I don't know where to get it around here. Keep in mind that Poiseuille's Law for laminar-flow conductance depends on the fourth power of the diameter! The product I obtained from Small Parts Inc. is called "capillary" and has a thick wall. The OD is probably around 1/8." Cutting it is a bastard. It arrived with one end sort of crushed (probably after receiving the obligatory drop-kick from a Postal worker) and I had to trim it. My technique: make a circumferential cut about 1/4 the way through the metal, all the way around, and then fatigue off the scrap end with a sharp bend with pliers. Before installing this capillary, be sure to drive out the water with a torch (particularly if you blew through it into a glass of water in order to make sure it was unobstructed). Anything in there with more viscosity than air is effectively a plug.

-Carl

[Jon Rosenstiel]

Carl- I don’t follow your n / s calculation.

Jon Rosenstiel

[Carl Willis]

Hi Jon,

Always much room for an error when I actually do a calculation on the fly. Who knows what happened. Maybe I forgot to divide the diameter by 2. Maybe I multiplied inches by 2.54 twice. Anway, the number is definitely bad. The corrected number should be around 64,000. FYI I figure it thus:

(n / s) = (mrem / hr) * (n / cm^2 / s / mrem / hr) * (surface area at detector radius)

(mrem / hr) = 34 bubs / (33 bubs / mrem) / 0.167 hr = 6.2 mrem / hr

(n / cm^2 / s / mrem / hr) = 8 (approximately; from the USNRC, who takes recommendations from the ICRP)

Detector radius = ~ 4" (3" for the fusor + 1" for the flange). Surface area is about 4*PI*R^2 = 1300 cm^2

Just means I have all the more more work to do on improvements.

Thanks for the correction.

-Carl

[Jon Rosenstiel]

Carl- That sounds about right...I thought I was the one that was losin' it.

And of course, you have challenged me, so I gotta' do whatever I can to "beat the man down"!

Jon Rosenstiel

[Richard Hull]

That was a great call to task by Jon and a fine response from Carl. All done in the humor it deserved.

Goodness knows I have made my share of errors in math. This is all part of NOME as opposed to NORM. "Natural Occuring Math Errors". They happen in blasting of e-mails and like meson showers, are considered unavoidable. It is great we have folks who have a "Feel" for fusion based on a long history of the "doing". There is such a thing as intuition based on experience that something just doesn't "sound right"....It is hardly scientific, but is often right a bit more often than it is wrong.

I felt a bit stunned that Carl was getting big numbers at such low currents, but didn't poke into it or have the flags go up as Jon did as I figured he was at nearly double the voltage I run and thus, higher up on the cross section. I figured "I gotta' get more volts!"

Now I realize that the current was just too low, but I still gotta' get more volts.

The simplification is that, given any operationally fixed voltage, you are at a fixed point on the cross sectional curve and that current drives the number of POSSIBLE fusions at that point. Up to the point where the fusor chokes out with thermionic, grid, electron run away, more current will always supply more fusion opportunities.

This was a good thread and the humor of the competition is great. There is always something of a relief to see the best amoung any of us screw up. A ray of hope exists for us slower types.

Richard Hull