Brütal neutron counts from tetrode fusor at higher voltage

[Carl Willis]

You read that right: brütal with a ü.

Earlier thread about this project is here:
viewtopic.php?f=6&t=3231#p20281

Tonight, the tetrode fusor got tested at 65 kV, 11.5 mA (748 W), and 33 mtorr, thanks to its new oil-insulated / cooled socket. The chamber is directly water-cooled by copper tubing coils. One of the photos below shows the neutron detector meter reading 4000 CPS under those operating conditions. Back-of-the-envelope calculations for this detector suggest that to be in the ballpark of 5 million n / s.

In low-voltage, air-insulated testing, the maximum voltage I reached was about 25 kV and the maximum current at that voltage, about 6 mA (150 W). The neutron count rate there was a whopping 7 CPS. I also had enough data to extrapolate a relationship for neutron yield at higher voltages and currents. That relationship predicted a count rate from today of 1340 CPS. I ended up a full factor of three higher. That's obviously a good thing for me, and a not-so-good thing for the model--at least this far beyond the range of physical observations that informed it.

I don't yet have automated data collection on the new assembly. I am replacing my horsecrap FuTek DAQ unit with a proper LabJack U6 acquired from Doug Coulter, and my LabVIEW interface for the LabJack isn't finished yet. When that is in business, I can expand and refine my earlier models for performance.

Another observation: the x-ray exposure rate in contact with my 2-3/4" viewport at 65 kV / 11.5 mA was 30-40 roentgen / hr. Typical x-ray artifacts (speckling) can be seen in the viewport photo. In the new configuration of this chamber, the viewport points toward the ceiling and can only be observed by mirror or camera. That's probably a good thing, too!

A quick word about some of the photos. Some show the oil socket, which is of acrylic construction. The 100-kV silicone noodle cable passes through a 3/4" compression fitting on a 2-3/4" ConFlat flange, which is itself face-sealed to the acrylic with a buna o-ring. Oil in the container is a mix of 5W-30 automotive oil and the rest of my Walgreens USP-grade mineral oil before they raised the price on this bowels-loosening commodity. The VTVM is set to read 15V per 75kV, via the Ross HV divider. The current meter is 25 mA full-scale on the setting shown. The Baratron pressure indicator needs re-zeroing and 3.4 mtorr must be subtracted from its reading as shown.

Good times. I can't wait to get this baby worked up to the power supply's max rating of 100 kV, at which point it would seem that no "metal umlaut" superlative will adequately describe the performance.

-Carl

[Chris Bradley]

I wonder how important having a regular geometry helps in fusion rate?

Doug has reported similarly that he swears by accurate geometry.

I don't see a particular obvious reason for that, in regards the process of the fusion reactions. But perhaps it is significant in *avoiding* the 'unhelpful' discharge processes that may also go on, and therefore deliver a 'cleaner' and more controllable set of discharge conditions so that the discharge can be held at higher volts.

Not sure how you're going to shield well at 100kV. Distance, and remote operation?

[Richard Hester]

If you don't mind something perhaps intended for a constipated horse, I've picked up gallon quantities of mineral oil on e-pay. My intended application is a combination moderator/fast neutron scintillator, but I suspect the oil would insulate just fine.

[JakeJHecla]

Carl- Have you formed a hypothesis as to why the neutron numbers are so "brütal"? If I recall correctly, this is very near the all time (amateur) record set by Jon Rosensteil (sp?) a few years ago! In addition, I notice the chamber pressure seems to be quite high in comparison to most successful fusors. Is this something that has been observed before? If so, do you have a hypothesis as to why this particular machine thrives in this pressure regime?

[Carl Willis]

Hi Jake,

Jon's record is 1.2E+07/sec. at 80 kV / 21 mA. Right now, I'm in the vicinity of 0.5E+07/sec at 65 kV / 11.5 mA. If I ran at the same current, it's likely I'd be seeing about 75% of his record. The additional 15 kV might put me beyond his record, but I'm not sure. I don't think there's any doubt that I can beat his record, and with lower power, at 100 kV.

So this device is a good performer, but it is not other-worldly in its efficiency. It is more efficient than my earlier fusors by a noticeable margin, and indeed more efficient than a lot of other fusors described here. I have some ideas about why that is, although they're not exactly revolutionary. First of all, this system is very effectively cooled. The grids are massively heat-sunk to conical copper sheet supports that go to the socket. At 750 watts in, there is NO color on that grid. That means there is insignificant thermionic emission. Electrons emitted at the grid represent a major inefficiency in fusors. So there is that.

The higher pressure has to do with the scaling laws that apply to gas discharges. This system has a larger grid relative to the chamber than most, so inter-electrode distance is relatively small. A certain number of gas particles need to occupy this space in order for the discharge to be able to sustain a certain current. When the space gets smaller, the density of gas must increase to maintain that number. In essence that is why the pressures are high. Also, it's a small fusor to begin with, fitting inside a 4" pipe.

-Carl

[Steven Sesselmann]

Carl,

These are impressive numbers...

Can I ask you, how many minutes run time did you do today, and how far from the fusor are you standing when you are operating it?

What are you measuring 40 rem at the viewport with?

This run would put you next to Jon on the fusion reactors by energy quotient list.

PS: What is the internal architecture of that tetrode, how does the grid look?

Steven

[Jon Rosenstiel]

Awesome results Carl, keep pouring the coal to her and you have a pretty good chance of breaking the record. I'll clean and polish the trophy!

For a round-about comparison my recently upgraded / recalibrated Ludlum 12-4 (had the original BF3 tube replaced with an He3 tube) recorded the following from my fusor...
30 kV @ 10 mA: 3.0E+05 n/s
40 kV @ 10 mA: 1.0E+06 n/s
50 kV @ 10 mA: 2.1E+06 n/s
60 kV @ 10 mA: 3.2E+06 n/s
70 kV @ 10 mA: 5.0E+06 n/s

Jon Rosenstiel

[Carl Willis]

Even brütaler numbers from tonight: 15000 CPS at 74kV / 8.8mA / 651W. Just shy of 2E+07 n / s by my calculation.

I am having my doubts about one issue in particular, x-ray pileup. Quadrupling the count rate with 10 kV added is suspicious. My moderator-removal experiments on the detector were done at low voltages where I feel it is safe to handle the tube near the socket, but there is the significant possibility that at these higher voltages, x-rays are causing pileup at the detector. My means of answering that question will be to check the pulse-height spectrum from the tube and determine if it is still shaped appropriately for neutron-capture events. Adjustment of the LLD threshold may be necessary, or alternatively, a lead sleeve for the tube and its moderator. I also want to check the neutron yield on a secondary instrument that is actually dosimetric in response, the remball we have at work. I'll hope to bring that home tomorrow night.

Steven, you asked about the dose rates and I attached two photos from today, showing dose rate at the viewport (now partially shielded with a lead tube), with 32 R / hr; and near the socket assembly, showing about 2 R / hr. The instrument is a thin-walled ion chamber. My runs so far have been brief, not lasting more than about 5 minutes, during which time I have been mostly standing about 5 feet away.

I thought I had an earlier photo of the tetrode grid assembly. If not, I'll make a photo of the spare tomorrow.

-Carl

[Carl Willis]

Update: I can confirm there is a problem with x-ray pileup at voltages above about 60 kV. I have made some tests with the MCA and verified this problem exists and I can also solve the issue with a piece of sheet lead between the detector and the socket assembly. I'll try to get a more accurate neutron number tomorrow after a more permanent effort to shield the detector.

So all bets are off until that can be fixed.

-Carl

[Doug Coulter]

Great work, as usual, Carl. Once you get your instrumentation sorted out (and good of you to take that care) - I'd love to see what happens if you play a little with having your two grids at different potentials (in either direction). Focus, anyone?

Looks like you've got a great setup for activating silver there too - and there is enough experience to get pretty decent neutron numbers from 5 min runs doing that, which kind of sidesteps pileup and EMI kinds of issues with other instruments.

Very roughly, with a 5 min run, and a 2" sq piece of silver sheet .01" thick and 2" pancake geiger, you get (approximately) 1100 cpm per million neuts/second at the instant of fusor turnoff. This is not super hard to measure and back-extrapolate if the silver activation counter was already on, and shows the X rays (so you can see on the graph the exact instant of fusor turn-off). Using a log plot, you can simply back extrapolate a straight line to that instant. We use a counter that returns counts once a second - then multiply by 60 to get the cpm number.

Due to variations in moderators and gear, this won't get you 1% class results, but you'll never be off an order of magnitude either. When my counters and the silver disagree, I know which one to trust.

[Richard Hull]

Carl,

Fabulous results! I came late to the party as I have been gone a week on vacation.

That is a fearsome X-ray blast and while the 3He tubes are relatively insensitive to Gamma, the moderate energy x-rays might just be an issue, as you note, bouncing electrons off the SS walls of the tube into the gas environment. I would certainly believe that a combo lead shield and maybe a tweek on the discriminator level would take the tube back to the real world.

Meanwhile, a good silver run would also be in order as x-rays "be damned" with that method. If not highly qauntitative, until you can get at a Bubble detector, it would be the ultimate in qualitative, run-to-run neutron results.

All the best on this fabulous effort. For now, "lead up" or remote view/operate and watch those x-rays.

Richard Hull

[Carl Willis]

Update with corrected neutron numbers. I solved the suspected problem with x-ray pileup on the He-3 detector at higher voltages by making a 0.1" lead sheet enclosure for the He-3 tube and its moderator (shown below) and dropping the shaping time from 3 microseconds to 2 microseconds, producing a much more sober neutron count rate as verified by an MCA (spectrum is in the images below).

I also brought in a calibrated Bonner ball (Ludlum 12-4) and made a measurement with that.

Corrected measurement from the He-3: 417 cps @ 75 kV / 11 mA. The new tube position is 13.6 cm from the center of the tetrode fusor, the assumed intrinsic detector efficiency (established from previous uses) is 0.2 with this moderator; and the calculated neutron source rate is about 540,000 / sec. The pulse height spectrum from the shielded tube shows all the expected features of neutron capture, and the window includes no x-ray tail at the low-energy end. The lower cusp of the wall-effect continuum can be seen distinctly, again confirming that with the shielding installed, discrimination against x-rays is excellent.

Remball measurement: dose rate is 7.2 mrem / hr; ball center to fusor center distance is 28 cm; calculated yield is 570,000 / sec.

Those numbers are not impressive, and in fact at the power levels involved, this machine turns out to be a poorer performer than my two previous fusors and several low-power beam-target contraptions. Of course the question is why, and what can be done about it. I'll be working on getting my data logging software in order first, and then onto some ideas to try to salvage a decent record of success from this tetrode grid.

-Carl

[Steven Sesselmann]

Carl,

Thanks for the update, and a good example for the newbies, that electronic neutron detectors can fool the most experienced users. The pulses are so tiny and the environment so noisy, detecting neutrons is a real art in itself.

The other sobering message here, is that fusion is harder than it looks, the darn Deuterons hate each other more than we love fusing them, and it is going to take quite some ingenuity to come up with a device that works better than the humble Farnsworth fusor, but we keep trying...

I imagine your tetrode fusor will eventually burn itself in and produce some better numbers as the components outgas and become saturated with deuterium. I wouldn't mind seeing one of those 3D plots, now that you have sorted out the detection.

Nice work with the lead around the detector, but how about wrapping the lead around the fusor instead?


Steven

[Richard Hull]

Thanks for the usual full featured update Carl. Nice work, as always. Sorry for the neutron reduction, but wasn't it sort of expected? It remains interesting that the good old spherical fusor is still a great performer in good hands. Is it geometry that attends this? If so can recirculation be part of the differential or is it size? If your tetrode were, say, 10 times its current volume, how would it perform? Interesting questions abound.

I feel that it is a number of spherical/cylindrical fusor specific processes that might make the difference. Large absorptive wall areas might play a key role.

Richard Hull

[Mike Beauford]

Richard, are you saying a larger surface area on the wall of a fusor would produce more neutrons?

[Richard Hull]

No and Maybe. I can't say definitely, but we have recently discussed at some length the probable multitude of processes that might take place in a simple fusor capable of producing fusion. We are positive that no one advantageous process in the fusor produces all the fusion.

Electrons bombarding the shell would knock out buried deuterium atoms and possibly ionize them in the very zones where you really want deuterons to be formed. More surface area would tend to produce more deuterons as fast neutrals buried themselves in the inner shell walls.

Richard Hull

[Mike Beauford]

That leads me to my next question, given the number of fusor's that have been created and that have shown to generate neutrons, does it directly correlate the larger the surface area the higher the average number of neutron's or does this also mean you need to worry about it's shape also at the same time?

I'm wondering if I have a large surface area, could I create a set number of neutrons say X at a lower voltage then a fusor that has a smaller surface area, but at a higher voltage?

Mike Beauford

[Frank Sanns]

This has been posted in past threads but I have seen enhanced data in my fusor and in my analysis of other people and organizations fusors. My fusor is a large 6 way CF cross. The outer grid is often only half the distance to the shell. I believe that there is a recirculation scheme that can be established that enhances neutron output up to a point. It manifests itself as a leveling off of neutron rates at a given voltage and a given current based on the geometry. It is in every single piece of good data that I have looked at. The results are larger than large scale electrostatic or magnetic focusing of the inner grid can account for.

Frank Sanns

[Richard Hull]

Frank is certainly correct. Circulation and wall surface area are possible players, but even these are not the whole scenario. We live and we learn. I am loathe to lay claim to some of my earliest, naive postings on the old "songs" site in the 90's, when we all thought this was 100% recirculation. Those old thoughts were just too simplistic for this ultra simple device with its many possible modes of doing fusion.

As for Mike's question.... There is no real data to suggest that spherical shape is more advantageous over other shapes or that wall area is a major factor. Thus far, the anecdotal evidence is that fusion proceeds nicely in most any well operated environment. Controls, if ever applied, would involve, shape, volume and wall surface area. At least 6 fusors would have to be made with 12 being more definitive to answer some of these questionsin a scientific manner. Here we are talking about a good deal of money in amateur circles.

Nonetheless, Mike's questions are well considered opportunities for research. This is research that demands tight controls and multiple iterations of identically assembled devices all operated exactly at the same input energy, voltages, pressures, etc. Such a piece of work, as noted, above would involve thousands of the money on hand and a year or more of careful effort. The result would be a great piece of work and perhaps point to what among many theoretical operational mechanisms might be the greatest causitive fusing agent in the simple fusor.

I wonder if we would all be shocked at the result?

We are drifting a bit far afield of commenting on Carl's fabulous effort that started this, but his refinement of his neutron measurement process amidst the hail of x-radiation and subsequent reduction in numbers, seems to have brought up this old discussion, which, if it proceeds, might best be moved in a specific posting to the general fusor fusion theory forum.

Richard Hull