Bizzare? anti correlation between gammas and neutrons

[Doug Coulter]

I don't have much to say about this yet, but I took some interesting data the other day that might inspire some people to think about it and comment. Basically, I'd been thinking about trying to look for the super hot gammas from the rare DD->He reaction, to see if that wouldn't be a good way to detect fusion, since heck, they should be so "loud" as to be hard to miss, or be faked out by something else -- running a fusor makes tons of X rays in the 0-power supply kv range, we know, but what if there's something else coming out we could use with a fairly simple detector and threshold? In re that -- this could go on the metrology thread, but until we figure out what this means, it's here since it IS fusor data as well, taken during a real run.

In the first picture, I have a Cs-137 source taped over the end of the NaI head, which is pointing at the fusor grid from pretty close in, with no lead intervening between it and the fusor - only the stainless steel tank walls.

Calibrating...

b10gamma1.gif (6.69 KiB) Viewed 5787 times

Although I'm scanning slow (this is a roll mode that looks like the medical heart monitors you see on TV scrolling by) I have the scope set up to peak detect while sampling at a GHz, so it's catching all the peaks (and noise). Top trace is the NaI head, negative going pulses, and as you can see, the big 511 kev line from the source is about 1 major division negative going. The purple line is the output from the B10 tube preamp I'm working on (linear output) and the bottom line is an AC coupled faraday probe in the tank, far from the action. This setup is the same for the other pictures as well. Then I turn the fusor on, running not all that stably at this point (that's on purpose this time) to get kind of a sweep on the fusor conditions while we watch.

Running the fusor

b10gamma2.gif (9.65 KiB) Viewed 5787 times

b10gamma3.gif (9.82 KiB) Viewed 5787 times

b10gamma4.gif (9.63 KiB) Viewed 5787 times

b10gamma5.gif (8.84 KiB) Viewed 5787 times

Personally, I don't think this is conclusive of much so far. There could be pulse pileup going on and we couldn't tell at this sweep rate, but when I look at it with a fast sweep (in other runs) I don't see much -- the NaI head is very short pulses normally. I will be taking more data on that at any rate. But notice that there are gamma outputs well above the 511 kev calibration here,
but not so big as to indicate much if any 16 megavolt either, unless this head is very squashed output in the higher energy region. We do see some stuff that, assuming linearity, might be well into the 1 mev region though. Protons from the other common pathway hitting the tank walls? At any rate, it's interesting that this seems to happen most often when the neutron counter is NOT counting, hmmm. Could it be that varying the conditions varies which reaction pathway is preferred? In general, it's drawing less power during the high gamma/low neutron periods...a lot less. If I turn my hope-o-meter way up, I could try and convince myself that looking for neutrons, instead of total energy is the wrong thing to do, and that there's a possibility of a higher Q mode that favors the non neutron pathway in certain conditions. Obviously there's more data needed here, but this is a start. Thank heavens for this nice Ghz multichannel scope and the linux tools (which don't exist for windows) that let me do screen captures quick! You linux people go look up GDS-2000 tools or gdsh and feel like the lucky kids on the block for awhile (also integrates into octave and gnuplot if you want fancier things).

b10gamma6.gif (10.18 KiB) Viewed 5787 times

b10gamma7.gif (9.19 KiB) Viewed 5787 times

[chrismb]

Doug,

I think one has to suspect EM interference here, in the first instance, and try to rule that out first.

The way to test is the usual; use a dummy gas, see what it all measures then.

The D+D->4He+hv reaction is mediated by the electromagnetic force (as opposed to the strong force for the other reactions). It is simply which one shuffles itself into action within the nucleus, once it is fused. There are specific reasons why the EM reactions don't present themselves with the same frequency as strong mediated reactions. I think I'm right in suggesting all strong mediated fusion reactions have an EM branch, and they are generally 2 to 4 orders of mag lower in probability outcome. Some fusion reactions don't have enough exothermic energy to disgorge themselves of nucleons, so they are EM only.

The percentage (/crosssections) for the reactions is not determined for thermonuclear plasmas, it is determined by beam-target experiments. In fact, as far as I understand it, the inverse (to your suggestion above) is the case - that the thermonuclear guys assume theirs is the same as beam-reactions and have never gone about intentionally measuring anything to show that it is so. (Thermonuc data is 'consistent' with the assumption that it is the same as in beam reactions.)

In regards the EM issue, in my experiment I have magnetic fields and some hot electrons flying around, so I bought myself one of these;

http://cgi.ebay.co.uk/ws/eBayISAPI.dll? ... 0648159107

This gives a sum total of EM radiation in the area from 1 MHz to 8 GHz. Even though my B-field is not very high, it still has the potential to pump out 5 GHz EM. This isn't an easy region of EM to buy 'cheap' kit to measure with much precision, so I was happy with this. I didn't want to find out later I'm staring straight into what could easily turn out to be a microwave tube!! But a new RF-detection chip came out to 8 GHz, mainly driven by newer wireless networks running up at 5.2 GHz.

Anyhow, my suggestion would be for anyone in this kind of work to look at getting one of these. Very handy, instant compact readings of local sub-uW/m^2 to unit u/W/m^2 EM radiations across a broad spectrum.

[Doug Coulter]

Yes, EM is always a suspect, and why you see that blue line there -- it's an antenna (about 8" unshielded), and it's not that much louder when the fusor is running, see? Further, there's no apparent correlation between it and anything else. I've been checking that right along and making sure that wasn't it. This is all very well shielded (the whole HV outside the tank is in a copper faraday cage) and doesn't "go" for arcs and other insults in non-fusing gases, that has been tested many times on every detector - it's about the first thing we try, because yes, it can be a problem even when everything looks good. There could however be pulse pileups on the gamma head (which gives about a volt with the 511 kv gammas, no preamp involved here), and I will check for that next try.

I don't think this shows even a single very high energy photon as should be produced in the DD->He reaction, or if it did, I missed it (or the 4" deep NaI xtal couldn't absorb it), and I was watching the scope the entire time, not just these snapshots, looking for it. But what's interesting is that it's almost always showing some bigger than the 511 kev, or nearly always, and how that goes up and down vs the neutron tube which is very tight emi wise and a long way off. On the scale shown, 50kev power supply gammas would be 1/10 a major division. Yet we see two major divisions sometimes.

But we do see a bunch that, assuming linearity, would be in the 1-2 megavolt range. I'll be trying this again soon with both neutron tubes and a couple of gamma heads, since I have that nice 4 ch scope to show it on. I just wanted to see if anyone had a take on what it might be under the assumption I got the electrical engineering right and it's not EMI. I'm pretty sure I did.

That looks like a fun toy you bought. I find a few inches of wire on a scope probe does pretty well, and if it is all going into the same multichannel scope, you get the benefit of seeing if it's all time aligned in the bargain. Wire is cheap and I already have the scopes.

We do also always run with a cheap microwave oven leakage detector, and it doesn't read....you can strip one of those for a scope output and do pretty well too, for under 10 bucks. You just need that fast diode. No need to read EMI in the microwatts for checking if you're making dangerous amounts, or enough to make other detectors read false -- that takes real power.

[chrismb]

Yes, EM is always a suspect, and why you see that blue line there -- it's an antenna (about 8" unshielded), and it's not that much louder when the fusor is running, see?

I think you might find that your piece of wire struggles to pick much up outside of VHF. That's why I felt the gadget was a worthwhile cost - albeit a bit more than I'd liked to have paid for a little plastic box with a chip inside it! But it is a sophisticated chip in there doing wide-band normalisation, rather than just a diode.

Could you explain the lines a little more for me? I see a blue line that appears to have a range on the vertical of around 20V with the fusor off, and then a line which goes off the scale when the fusor is running.

I also see a yellow line that is nice and steady - excepting for high energy responses - when this blue line is 20V indicating a very tight spectrum excepting for the transients. Then I see a yellow line that is not at all steady and contains a spectrum of response when the blue line goes off the scale.

[Doug Coulter]

There are two vertical blue lines that are cursors also (hard to turn them off in this scope w/o the book in hand), that may be fooling you -- note they are in the same place every capture, and one is dashed. In this very slow sweep, all the lines are "thick" and contain the highest and lowest voltages that occurred during that pixel time (peak detect mode), measured at a 1 ghz sample rate more or less (scope goes to 2.5 ghz). The yellow trace is the NaI head output, negative going -- lower is higher energy from that detector. The purple line is the output of the preamp on the B10 tube with no thresholding, and the little stuff there is gammas/betasm, or neutrons that didn't happen to produce a full scale output (some fraction of the total do that). The blue line is an 8" or so antenna to pick up electrical noise (up to about 1 ghz) so I can check for EM -- half inside the tank, half outside. That's way better than good enough for this application, as nothing else in the detectors can see much above the single or two digit mhz range at all. And in this case, I'm not worried about high Ghz being created by anything.

At any rate, there's always some noise when the fusor is running, which is why that line gets "fatter" vertically when it's running -- but no huge offscale spikes on this run. I placed it on the bottom because the fusor noise is always positive-going. When it's not running, that's how much the setup picks up from local radio stations and so forth. This run was pretty quiet, so it doesn't really get a lot bigger at any time during the captures I'm showing. At any rate EM on this level is far below what it takes to make any of these detectors "false". When I start seeing noise that will make detectors give false answers it goes WAY off scale and for a long time, for unshielded detectors. These detectors are all fully contained inside tight metal boxes with no exposed antennas (other than the inevitable ground loop noise pickup). I don't see anything suspicious on the noise pickup here for the situation, it's actually pretty quiet compared to what I see sometimes. But I do watch that, being sensitive to the issue, which is why it's there at all.
I normally use that signal as an inverse confidence for my other detectors, and to look for things like big groups of charged particles going past the part of the probe inside the tank, in a bunch, which I see during the pulsed modes. I actually run several inside and outside EM noise pickup probes to keep all that sorted out, most of the time.

If the big output on the NaI turns out to be other than pulses piling up (and I will check for that soonest) then we have some truly new science here, something not reported anywhere in the existing literature -- I delayed reporting this so I could check. It's not that we see bigger than 511 kev that would be new -- it's the time variability of it -- mostly happening when the neutron count is low, that's new. If in fact the 1-2 mev stuff is something like charged particles hitting the tank walls and becoming gammas -- then that seems to mean that we see a lot of one reaction pathway sometimes, and the other some other times....for which there is no precedent. In other words it's an "extraordinary claim" which is why I'm not claiming it without some qualifications, yet -- I have to take more data to try and "break" this observation and see if it's some artifact, or truly the "real thing" before I can even think about claiming that.

So at this point, I'm merely suggesting that "something interesting is happening", with a possible explanation of it. There could be others, and I'm all ears.

Now, when I check this at a real high sweep speed, we won't be able to see this slow variation at all -- we'll only be able to see a bunch of skinny NaI pulses and maybe the odd (much more rare) neutron counter pulse if I get lucky and catch one at all at a sweep speed that will even see the very skinny (ns) NaI pulses individually. So that won't show this effect, just whether or not I'm getting overlapping pulses from the scintillator that tend to add to higher amplitudes or not. I'll do that next time I power up the rig and post it here.

BTW, a few years back, I did a design for a vendor to homeland security that had that broadband logging (logarithmic output) chip for EM in it -- it's not all that new. They wanted it to use to find people under rubble as their cell phones would still be trying to contact a tower. The trick there was the antenna design more than anything -- broadband antennas aren't easy to make and be directional too. They wanted it for search/rescue in case another event like 9/11 should happen. I've got some of those chips on boards still -- they were kind of expensive then, about $7 ea. But calibration with a user supplied antenna was totally "on you". For that use, it didn't matter.

[chrismb]

Doug, I'm still unclear why I'm seeing 'less' background signal on the yellow plot when the fusor is in operation?

[Doug Coulter]

Because as I said, that first one wasn't background; it was a Cs-137 source taped to the NaI head. A pretty loud one (brand new). So the thickness of that line is 511 kev at a real fast count rate. It gets a lot quieter with neither that or the fusor going, but it's still fairly noisy -- just more random pulse heights. I of course took that off during the fusor runs, which don't make nearly as many gammas as that hot cal source does. And sometimes the fusor makes no gammas as hot as 511 kev.

Or in other words, the height of the yellow lines are the energy of the quanta, and the density horizontally is the rate of them, which was high. This shows that even very high rates show very few pileups or pulses that are higher than any one quanta.

[chrismb]

OK. Thanks. I didn't understand that you were getting that many counts... That seems to be a pretty hot source!

[Doug Coulter]

Yes, looks real hot at the slow sweep speed. It's a .25 uC source, so it says. Our brandy new Canberra military soldier monitor says it's between 600-700 uR/hr - it would probably give you a burn if left in your pocket, eventually. A geiger tube counts in the over 10k cpm range (I forget the exact number, but it was "wow"). The scint counts much more, since it's better at seeing the gammas, but I've not counted that yet -- that preamp is yet to come (and will be both linear and thresholded like the current one but have to be much faster and fancier to handle the real short pulses).

Henny penny, our frightened scintillator counts it at 370 cps -- pretty hot.

FWIW, that same canberra sitting on the couch 5 ft away is reading about 7 uR/hour. The hottest rad samples we have read about 2 mR/hr, and I don't keep those in the place I live!

I wanted to establish vertical scale and the relative lack of pileups that way. The trace looks real choppy with nothing there -- pulses from "can't see it" to around 1 div, maybe two if a cosmic hits and and you happen to see it happen. CS-137 sources are universally used for that because they have just two lines -- 511 kev (it's a positron annihilation) and ~30kv -- so you get both ends of the scale to calibrate your gamma spectrometer in one source, with nothing in the middle.

I think NaI heads have a limit on how big a peak they can make, due to not all the energy being absorbed in the finite crystal size before some scatters out. This one is 1" sq by 4" deep, and we of course use it long-ways for all this. It's not a super good one, it's from a PET scanner -- git it cheap on ebay.

Thus, what I'm saying above, is that if I crank up the hope-o-meter, and plug that straight into the guess-o-meter, I'm seeing the first reaction in the gamma spec head, and the second one in the neutron tube. And that we seem to be seeing either one, or the other, mostly, at different times which is intriguing to say the least (if true), because that in turn implies that it is possible to tune for one or the other reaction, perhaps on purpose, because it looks like nature is doing it by accident as the fusor parameters drift around.

For those who haven't seen this lately, this is the official chart, the second set of numbers is Q.

JustDD.gif (1.78 KiB) Viewed 5768 times

Since stainless steel, or any nearby lead won't (easily) make lines that high energy when hit by anything -- but rather make a lot more photons lower energy in a bunch of scattering events per charged particle strike...the ratio could be larger than indicated here.

The only other explanation I can offer is that this is close to a big wax moderator for another neutron detector tube, and the capture gamma in H is also in the same range.
Edit: This doesn't satisfy me in two ways. One is we see more when we see fewer neutrons. The other is, this would imply a scintillator near a moderator is a more sensitive neutron detector than a neutron detector. So the only explanation possible that would make this true is neutrons coming out one side, then the other, of the fusor. Possible, but, Occam's razor kinda argues against any such fancy explanation.

Or, something is just broken. We did see this a little when the NaI head was half across the room while the tech guys were here, just not conclusive enough in the time I had to look at it, so I set up this run with the thing close in to the fusor and put it all on one scope. That's when I noticed the fact that times of high neutron counts coincide with times of not many high energy gammas. I need to put more neutron tubes on the scope, and more NaI -- say two each, and repeat this to get a little more sure, and check at higher sweep speeds to be sure of pileup issues in the spectrometer heads too.

This run wasn't spectacular, about 1/3 or 1/4 our best from a neutron standpoint. I have some things to repair in there. But since it works at all, I'm using it now to get all the detectors working so I'll have them when it does work its best.

[chrismb]

Since stainless steel, or any nearby lead won't (easily) make lines that high energy when hit by anything -- but rather make a lot more photons lower energy in a bunch of scattering events per charged particle strike...the ratio could be larger than indicated here.

Doug, there's no ratio indicated there. Those numbers are the MeV exothermic potentials of the fusion reactions. For the top two strong-mediated reactions, those energies are purely kinetic energies of the fusion products, whereas the lower is a gamma. So they are all 'energy' numbers, but they are apples-and-pears.

Whether those kinetic fusion products strike the chamber wall and cause X-rays, I'm sure they do, but the number and energy levels of such photons is not something I have any knowledge of. But I'd hazard a guess, just like you say, that a 3MeV proton hitting the chamber won't produce a 3MeV photon but instead slow down gradually, sending out a stream of lower energy photons. I'm not sure how you'd filter those processes out from your data. Would 3MeV protons hammering steel lead to many 100keV photons? I think they probably would. So I'd make a guess that photons below a few 100keV are prob from that, and above 10MeV are fusion gammas, and in between, could be either.

Now, as to why you are getting the temporal discrimination as your instruments appear to be showing, I don't have an 'answer' for that.