Page 1 of 2

Criticize my neutron detection setup

Posted: Fri Oct 20, 2017 1:04 pm
by Silviu Tamasdan
I'm nowhere near having a fusion-producing fusor, but getting there little by little. I've found it more productive to concentrate on one particular subsystem at a time, so lately I've been working on the neutron detection.
I have now what I consider a reasonable setup for detecting neutrons, and I'd love to have comments or criticisms. It's not in the final configuration yet (still need some stuff to make a more convenient and mobile moderator because right now it's a pile of things that need to be taken apart and rebuilt when I make a change).
I did not build the system, I purchased it. I have a pretty good idea how it's built, but can't really answer very detailed questions about its components.

1. Detector. It's a Russian SNM-32 3He corona tube. AFAIK it's one of the biggest they made, at 332x18mm. Datasheet says 700V for initiation of corona discharge, plateau at 1500V and above. In practice however I've found that the tube cannot be used at any voltage above 1275V. At higher voltages the tube oscillates giving high frequency/high amplitude pulses (over -100mV at 1000000-2000000cpm). The builder of the system says that is typical of many of the SNM-32 tubes he has worked with and recommended to keep the voltage below 1200V. I will attach some test results below, but essentially I have found that the tube can be used as a counter at voltages between 700 and 1200V.

2. High voltage source. It is USB-powered, software-controlled (has no physical controls - I would have preferred it otherwise, but oh well). Can deliver up to 3000V, and have tested it up to 1500V with various other detectors (photomultiplier-based scintillation gamma detectors, Geiger tubes). One disadvantage is that I cannot modify the voltage while a count is running. I have to stop the running count, change the voltage, and then restart.

3. Detection electronics. Biasing resistor, coupling capacitor and probably a charge-sensitive amplifier based on Bob Higgins' design, but I don't know the details. The acquisition part is a USB oscilloscope (Picoscope 2000 series). The front-end is a custom program that combines acquisition and control of the voltage source. I can set the voltage, and the trigger amplitude of the detected pulses. I can present the output in cpm or cps, it does some basic statistics automatically, and can record and save the shape of every single pulse it counts. I have found that latter feature very useful for tweaking voltages and trigger values. It cannot filter pulses by pulse duration.

4. Home-made moderator. As I said above it's not complete (I have some HDPE blocks on order) but I've improvised something that works already fairly well. I had a 15x20 block of HDPE, 1 inch thick. I cut it in 4 pieces 5x15x1inch and stacked them together to make a 4x5x15inch block. By itself that worked to some extent, either with the neutron source (see below) on one side of it and the detector on the other, or with both on the same side. But it definitely could use improvement. I don't have any other HDPE yet, but I thought about moderators and found that I did have something that in practice I found suitable. Both HDPE and paraffin are comprised of carbon and hydrogen at approx 1:2 molar ratio. Another thing that has carbon and hydrogen at the same ratio is mineral oil. More specifically, motor oil. Which conveniently comes in brick-shaped bottles, 2x4x8inches. I took 12 of these bottles (unopened, so no leaks), laid them flat in layers of 2x2 bottles, 3 layers high. I put the HDPE block at the top. The detector tube is inserted between the 2nd and 3rd layer. The check source, if used for a particular experiment, is inserted in the same place, close to the middle of the tube and in direct contact with it. This way both detector and source are surrounded by at least 10cm of moderator in all directions. I will aattach results below.

5. Neutron source. It's made with a 5mCi polonium static eliminator, less than 2 weeks old so it should be still close to its stated initial activity. The beryllium came from some irregular cuts from a 1mm thick sheet. I have broken them in pieces and managed to arrange the pieces in the source window above the grate to get >90% coverage of the window, and held in place with tape. Picture attached.

6. Gamma rejection. Was checked using a 10uCi cesium disc source.

I have done many tests, with various voltages, trigger levels and moderator configurations. I'm attaching the most recent and relevant ones. They each comprise a background check with no source nearby, a gamma rejection check with the cesium source in contact with the detector tube, and a test source check with the neutron generator in contact with the tube. All tests were done with moderator in the configuration I described above surrounding the detector. Picture of the working configuration attached.

Tests were done counting pulses for 10 minutes in the given configuration. All the tests were done last night in the same location, in a contiguous 90-minute period. The voltages and trigger values are noted in the name of the file. Basically, one test was done at 1200V and a trigger value of -14mV, one at 700V and trigger of -24mV, and one at 700V and -22mV.

Each file contains the raw output from the software for the given experiment. They contain text files with parameters and basic statistics, and .png images with the shape of every single pulse counted, some graphs and temporal distribution of pulses.

Re: Criticize my neutron detection setup

Posted: Fri Oct 20, 2017 6:16 pm
by Nnnnnnn
I can say a few things about moderators. Moderators are used to turn fast neutrons into thermal (slow) neutrons. This means that you only want a moderator if you expect to have more fast neutrons than thermal neutrons, like in a fusor. For your source you can look at this report ... sertations
I only skimmed it but it seems that your source emits just as many thermal neutrons as fast neutrons. So I am not sure if you need a moderator. Keep in mind that thermal neutrons are more likely to be absorbed than fast neutrons.

There are 3 things that can happen to a neutron in a moderator. (1) It can scatter (probability related to scattering x-section) (2) be absorbed (absorption x-section) (3) Nothing. Depending on the scattering and absorption x-section you can calculate the distance till absorption for your material.

You can now hopefully see that designing a moderator is task dependent. Look up Neutron X-Sections the calculations you need to do are very simple and look a lot like the calculations you do for radioactive decay.

Also I don't know why so many people like HDPE. What's wrong with water? Easily available and there is a lot of documentation on it.

Re: Criticize my neutron detection setup

Posted: Fri Oct 20, 2017 9:12 pm
by Silviu Tamasdan
Thanks for the comment.
My task is not to determine how many fast vs thermal neutrons the source emits, but rather to determine the appropriateness of my system for detecting neutrons (thermal, or fast after moderation). As long as it can detect neutrons from a relatively weak source like the one I made, I'm happy. In my opinion it does that.

I did run a few count series without moderator (not shown here). The counts I got this way were practicably indistinguishible from the background over 10 minutes. Perhaps with counting times of a few hours I could have seen a difference. I do not think that this source by itself generates many thermal neutrons.

I do plan on running a few more experiments, in particular counting with the neutron source shielded vs unshielded by lead to eliminate any contribution from possible intrinsic gammas.

Regarding water: apart from its undesirable interaction with high voltages in case of spills, nothing wrong with it. In fact one of my first ideas about an improvised moderator involved empty half-gallon cartons of juice or milk, rinsed and filled with tap water. Then I had the idea about mineral oil and it seemed much better. This way in case of a spill all I have to worry about is cleaning up the oil, rather than replacing the damaged HV source. (other fantasies involved surrounding the source and detector with cans of Coke...). The oil does contribute significantly to the moderation and detectability of neutrons. When using only the HDPE block as moderator, the counts are on the order of 4-5 cpm only.

(edit) I also have a SMN-11 boron-lined corona tube; I have just finished attaching a BNC connector to it and did some quick preliminary tests. Seems to work pretty well, doesn't balk at the voltage until 1650V. The baseline appears cleaner, but the sensitivity is somewhat less.

Re: Criticize my neutron detection setup

Posted: Sat Oct 21, 2017 4:23 pm
by Dennis P Brown
Randomly opened one of your data sets - rather a (very!) large set of data per file. This isn't something I'd think anyone here would want to go through - especially since there is no clear information on what you think the given file is displaying (or did I miss that?) You might want to find relevant data sets, smooth them (remove noise compared to signals), process as needed, analyse and present a few selected (relevant) sets that both show a clear neutron signal from your source with and without moderator with enough support information so we can offer an informed opinion. Until that is done, I'd not worry at all about testing with and without lead; one knows that without lead, one will have a lot of false signals. That is best a run after you prove you are seeing some neutrons via the moderator/non-moderator test.

If you are getting clear signals using that small source, a properly operating fusor will be a cake-walk for you.

You are following a very good methodology for your fusor effort. Building the required and separate components is an outstanding procedure. Getting a working detector is one of THE difficulties that cause far too many people from proving they have a working fusor (fusion.) Looks like you are well on your way in this aspect.

Re: Criticize my neutron detection setup

Posted: Sat Oct 21, 2017 5:23 pm
by Silviu Tamasdan
Thanks for the comment.
To keep track of things I tend to put the experimental conditions in the name of the file (that is set in the GUI before running an experiment).
For example the file you opened is part of my 10th series of experiments (that's irrelevant for anyone but me, helps me organize data), using the test neutron source, 700V on the tube, trigger -22mV and using the "enhanced" moderator pictured above.
Yes there is a lot of extra data in each file. You can safely ignore the pulse shapes if you want, the counts for each minute in text form etc. The two most relevant files in each set are stats.txt and the counts vs time image (called History.png).
At some point I will do some real data processing in Sigmaplot probably, but for now since all I'm doing is exploring the best detector parameters using my two small sources (not real fusor work) I didn't feel it necessary.
Perhaps a better way to present these test results in a simple way would be a screenshot of my GUI at the end of each run. It shows me most of the relevant information (counts vs time, basic stats and pulse shapes) together in one place.

BTW as an addendum, since yesterday I've been concentrating on using the boron-lined tube SNM-11in this setup instead of the helium one. I have had this morning even better results than in the files attached above. Very clean baseline with almost no noise, allowing me to decrease the trigger value considerably, and not much worse sensitivity to neutrons compared with the SNM-32. All combined with very low background counts and parasitic gamma pulses from the check source. I've also ran some counts with the neutron source and no moderator, and they're the same basically as the background counts (if you want numbers, 30cpm with moderated source vs 32cpm for the SNM-32; the background/gamma/unmoderated source is between 0.2 and 0.5cpm)

Re: Criticize my neutron detection setup

Posted: Sat Oct 21, 2017 11:09 pm
by Silviu Tamasdan
I'm attaching the results from the best run with the SNM-11 tube. Just the relevant files, no pulse shapes etc. I did not run a plain background count, the counts with the gamma source and the unmoderated Po/Be source were so low. Overall, I got 4 counts in 10 minutes with the 10uCi 137-Cs source, 5 counts in 10 minutes with the fast neutrons from the Po/Be source without moderator, and 282 counts in 10 minutes with the Po/Be source plus moderator. The operating parameters were: tube voltage 700V, pulse trigger value -12mV.

Re: Criticize my neutron detection setup

Posted: Sun Oct 22, 2017 7:11 pm
by Richard Hull
First off, a millicurie Po source with Be is no small source of fast neutrons. It is a fabulous source! Second a good neutron source testing a well tuned neutron counter with its tube, outside of a moderator, should not detect hardly any neutrons beyond background of any significance! A Po-Be source is a vast source of fast neutrons only! This is seen in this report as Silviu noted, and that was reassuring in the reports here. Silviu, in my opinion, has done it all correctly and double checked his work. This is one of the finest reports on one these often mis-applied Russian Corona tubes that has been published here. I have no crticisms of this effort.

I was particularly impressed that it was noted that at higher voltages the noise picked up as it should. Past credible reports have noted this with these tubes. Refining of the circuit and other corrective actions have helped raise the voltage a bit, but these are corona tubes and the very word corona means electrical noise. This is why this tube and many boron lined neutron tubes are real bears to turn into good neutron counters. At some HV point they become good GM counters and leave neutron detection behind as they start to count everything.

Silviu recognizes the rather heroic effort and experiment needed to turn a corona tube into a neutron counter. Just as I realize that anyone using one of these tubes and claiming fusion is always a candidate for possible self-deception through ill-application of the tube.

As a general rule, anyone claiming fusion using a Corona tube or a Boron lined tube should have a neutron source to bring the tube into reliable adjustment showing it is counting only neutrons.

I will not have to worry about Silvui if he claims fusion due to his fine report issued here.

Richard Hull

Re: Criticize my neutron detection setup

Posted: Sun Oct 22, 2017 10:16 pm
by Silviu Tamasdan
Richard -

I see you found my thread. I saw your reply first in the new user forum, and was going to point here, but no need now.
Thanks for your confidence. I do have a long way to go though. I need a fusor chamber, and my voltage supply is at best in the planning stages. I have almost given up trying to find an Xray transformer, and have started thinking more often about making my own high frequency supply with ferrite transformer from scratch. But that's another story.

FWIW I'm attaching the full datasets from my "best run" with the SNM-11 tube at 700V and -12mV trigger. If you compare the waveforms with the ones from the SNM-32, you will see they are much cleaner.

I'm also attaching a picture of my 3 current detectors which I use. From top to bottom, SNM-32 with SHV connector (bought with the system), SNM-11 (from sovtube on ebay) with my homemade BNC coupling (made from 1mm copper sheet, originally intended to be roof shingles, insulated with several layers of teflon tape and electrical tape; the reason for the teflon tape is to act as moisture barrier; for the next tube I plan on 3D printing a proper adapter from PETG, but this one seems to work quite well) and a 3inch Bicron BC-412 scintillation detector. I have on order a 1.5inch photomultiplier to which I will attach a LYSO crystal matrix and hopefully have something usable for gamma spectroscopy and activation experiments.

One lesson I've learned in working with these corona tubes is that more voltage does not equal better signal. In fact I get the best signals at the lowest end of their usable voltage range, i.e. 700V. I have run tests (not shown here but I have the data if there's interest in it) with the SNM-11 up to 1600V, and the noise increases to a point where it becomes troublesome, and signal decreases proportionally. Above 1650V the tube will oscillate at high frequency as does the SNM-32 at over 1275V.

Re: Criticize my neutron detection setup

Posted: Mon Oct 23, 2017 12:45 pm
by Finn Hammer
Silviu Tamasdan wrote:
Sun Oct 22, 2017 10:16 pm
I have almost given up trying to find an Xray transformer, and have started thinking more often about making my own high frequency supply with ferrite transformer from scratch. But that's another story.


If you live in the states, you could get in contact with The Geek Group, they have offered X-ray transformers for sale recently. Chris Boden is their leader.

Hope this helps.

Cheers, Finn Hammer

Re: Criticize my neutron detection setup

Posted: Mon Oct 23, 2017 1:57 pm
by Silviu Tamasdan
Finn Hammer wrote:
Mon Oct 23, 2017 12:45 pm

If you live in the states, you could get in contact with The Geek Group, they have offered X-ray transformers for sale recently. Chris Boden is their leader.

Hope this helps.

Cheers, Finn Hammer
Thanks for the tip. I had never heard of The Geek Group. I don't see anything HV-related for sale on their web site or their ebay store, but I'm going to leave a message through their contact form and see what comes out of it. To be frank, all my recent buying of vacuum and detection parts has kinda depleted my finances for now but we'll see.