BF3 tube pulse height spectrum

[Carl Willis]

If you have an MCA and access to a few neutrons, setting up a unfamiliar BF3 proportional tube system is pretty straightforward. (In fact, without using an MCA, it's likely to be hard to get it just right).

The first image is a pulse-height spectrum from the "antique" N. Wood BF3 tube shown in the second picture, a G1500S. It came from a dysfunctional Nuclear-Chicago instrument that had a late '50s look to it, and N. Wood never answered my questions about the tube's intended operating voltage. I do know that it has an enriched (B-10)F3 fill at 200 Torr and an active length between 3 and 4 inches, and is probably much like the modern G-5-3.

Anyway, the anode was biased at 1700 V for this data, and the counting time for both the "neutron source" data and the background data was 100k seconds. For the former case, a very low-level, moderated neutron source was positioned near the tube. The charge-sensitive preamp in use was homebrew, with a 2N3819 front end.

The resulting pulse-height spectrum shows the salient features that are common in all BF3 spectra: low-energy counts that represent gamma ray detection, a very broad and non-uniform high-energy peak representing neutron detection, and an intermediate valley in which very few counts are registered. The actual shape of the neutron peak is a signature of the tube's construction (fill pressure and diameter mainly) and has nothing to do with the neutron energy. If the tube tends toward low-pressure fill and a narrow diameter, the peak will be skewed to the low-energy end. If the tube is ideal, i.e. very large, the peak will be skewed toward the high-energy end and might show several distinct plateaus that are correlated with one or the other of the B-10(n, a) reaction products. A BF3 tube is bad if the neutron peak is smeared all the way back into the gamma region such that the two regions are not effectively separated.

I put a red line on the spectrum showing where one would want to set a LLD (low level discriminator) to cut out the gamma events in this tube. The LLD should be set right in the middle of the valley to get the widest counting plateau voltage range. The total counts right of the LLD were 1119 with the source, and only 69 without the source. This latter figure corresponds with a background of only one count every 24 minutes.

-Carl

[Richard Hull]

Very, very good work Carl!!! I'll have to review my work here and use my multichannel analyzer in future. This is first rate stuff.

I always started out with an intense, high energy gamma source near a tube and waltz immediately out of the gamma noise using the LLD (viewing the output of the SCA on an O'scope). This puts me well out of the intense gamma peak seen to the far left of your chart. I next use a weak gamma source that is also rather intense like Am241 and do a careful timed back ground versus source and finally move the LLD clear of this zone to background. This, I feel, gets me pretty darned close to your valley so cleanly presented in your chart. Once here, I usually roll the LLD just a tad beyond this for security.

This method has never failed me.

I next take my neutron source and painstakingly make constant 10-20 minute runs with a background check every hour, resetting the LLD if needed, (very rare), and go for yet another count.

The process I have used is tedious and long, but arrives probably at about the same zone and perhaps just a little "inside" the neutron region, but gives no false detections beyond normal background.

The problem with all these methods, yours and mine, is that they tend to require either equipment, skills and or sources that are just not in the hands of the average neophyte fusioneer. Of course, very few fusioneers roll their own counters.

We, nonetheless, post our work which we regard as our duty and as a result of the "hands-on" imperative.

Thanks again for this fine addition to the neutron metrology tool box.

Richard Hull

[Frank Sanns]

Thanks for the data Carl. Good work!

Now if only I can get my tube running properly. I ran the fusor tonight with disapointing results with the BF3 tube. The remball clicked away but nada on the big BF3 tube. Did multiple 15 minute runs with no neutrons detected on the big tube. I even ran the threshold down to 0.23 mv and started to pick up noise. It was zero counts in 15 minutes or random clicks of noise on the order of 200 per 15 minutes. At 0.2 mv it was runaway counts.

Carl, what is the level of signal on your MCA? I have never run down at this low a signal so I do not know if I am just too low in voltage or is there something else that I have missed? With my PMT tubes I normally run a threshold of around 1.0 mv.

Frank S.

[Richard Hull]

With most of the tubes I have, I require a gain of 50-100 using a superlative Ortec charge preamp to pull the tube charge up to a neutron signal level of 60-100mv. This would infer sub-mv levels. This is a charge device anyway. You can't expect a voltage without a very high impedance.

Are you using a known functional charge preamp? I assume you are. Finally, what is the tube voltage you are using? Did you inch the voltage up higher and higher?

If you have no preamp you are shooting in the dark for the most part. It is a matter of impedance matching and a little bit of gain to see neutron pulses out of a He3 or a BF3 tube.

I don't think I have ever seen a pulse right off a a tube directly.

Richard Hull

[Frank Sanns]

A storage scope is definately the tool to have. I have tried to watch my scope and run the fusor and look for the pulses but they are just fleeting futility.

Also the input impedance of one of my Ludums is only 1 megohm while the other is 100 megohm. I am going to switch and see how the higher impedance works. It has to be tons better. Higher voltage and a preamp at the tube is most likely going to be the solution. I just was trying to get by with what I had on hand.

Frank S.