Page 1 of 1

BF3 tube energy spectrum

Posted: Thu Oct 06, 2005 12:15 am
by Carl Willis
The following data were collected with a Nancy Wood G-10-2 BF3 tube housed in a 3" Cd-lined, HDPE moderator (Ludlum 42-9), exposed to a very weak AmBe neutron source. This is a pretty nice modern BF3 tube, in contrast to the antique tube I showed on 2005-02-03, and so gives more of a "textbook perfect" response.

This graph illustrates nicely the physics of the B-10 neutron detection reaction.

The response has three distinct plateaus of sorts, which I've labelled A, B, and C.

Region A is counts that come from full absorption in the active volume of the 0.84 MeV Li-7 product and only partial absorption of the energy of the 1.47 MeV alpha particle (the rest of whose energy is expended in the tube walls). In Region B, which lies above the 1.47 MeV alpha energy, we have events possible in which the full energy of the alpha and less than full energy of the Li-7 nucleus are absorbed in the tube gas, in addition to the same type of events in Region A. Hence the plateau in this region has more counts. At the high-energy end of Region B is the "full energy peak" which results when the energy of both reaction products are fully absorbed in the tube gas (the detection event was probably close to the center of the tube). Finally, Region C represents the situation in which the B-10 reaction goes to the ground state and produces higher-energy charged particles (occurs in about 6% of events). In the other 94% of reactions, a 479 keV gamma is released whose energy escapes the tube.

Below 0.84 MeV, counts from neutrons are possible but very unlikely. This is because the kinematics of the B-10 reaction causes the products to fly off essentially in opposite directions and if one is lost to the tube wall, it is very likely that the other will deposit all its energy in the active gas volume. So to count neutrons and not pick up any gamma response, the LLD should be set near where the red dotted line is seen, somewhere below the 0.84 MeV Li-7 cutoff.

Also noteworthy is that the gamma / noise tail is very steep. This suggests that if you don't have an MCA and are setting up a BF3-based detector, one way to get the LLD set about right is to stick the bare tube next to a strong gamma source (like a piece of uraninite or a aircraft radium dial) and crank the LLD up until the rapid countrate just stops. This tube normally chirps about once every 10-15 minutes from neutron background, but with the neutron source nearby it goes off about once every 15 seconds.

The other equipment used was my Tracor-Northern MCA, Ortec 571 amp, 1000 kV bias voltage, and an anonymous charge-sensitive Ortec preamp of '70s vintage.

-Carl

Re: BF3 tube energy spectrum

Posted: Sat Oct 08, 2005 11:30 am
by Frank Sanns
Very nice Carl! I vote for an archive.

Interesting how the energy barriers' transitions are quite sharp. What are the chances of you doing a similar spectrum with your large He3 tube? Do you think the resolution would be good enough for neutron spectroscopy?

Frank S.

Re: BF3 tube energy spectrum

Posted: Sun Oct 09, 2005 3:22 pm
by Carl Willis
Hi Frank,

I've taken a spectrum with the He-3 and it is up here...somewhere. I'm certain it is in this forum and not Img. du Jour (by the way, everything not on Image du Jour is automatically "archived"). I also have a comparison of He-3 neutron detection spectra for different amplifier shaping times. Maybe I didn't put that graph up.

The He-3 tubes can do neutron spectroscopy over a limited energy range. However, the cross-section of the reaction falls off as the neutron energy goes up. I haven't really tried spectroscopy with my tube since I don't have any source of neutrons that would give a distinct spectrum. BF3 tubes could conceivably be used for neutron spectroscopy but only with some kind of complicated unfolding program.

-Carl

Re: BF3 tube energy spectrum

Posted: Mon Oct 10, 2005 9:06 pm
by AnGuy
>Also noteworthy is that the gamma / noise tail is very steep. This suggests that if you don't have an MCA and are setting up a BF3-based detector, one way to get the LLD set about right is to stick the bare tube next to a strong gamma source (like a piece of uraninite or a aircraft radium dial) and crank the LLD up until the rapid countrate just stops

Question, with my BF3 tube (1 inch by 12 inch) I get no noticable hits from a unranium ore. I get about 1 to 1.5 hits per minute from background with or without the ore sample. It was my understanding that BF3 did not respond to gamma radiation. Is this assumption incorrect? Now the HE3 tube on the other hand is very sensitive to gamma.

How are you translating hits into MeV with the BF3 tube? Pulse height? From my observations with background, all the pulse heights appear to be the same.

The only interesting thing I noticed, is that occationing the there will be a double pulse hit. I assume that this is the same neutron (or cosmic ray) striking twice. I though one might be able to roughly measure the neutron's speed by measuring the time between the two pulses, and use the tube length as the error margin. Of course this isn't very precise, since the two pulse might occur less than a millimeter apart and we don't know how much energy was lost during the first strike.

Re: BF3 tube energy spectrum

Posted: Mon Oct 10, 2005 11:12 pm
by Carl Willis
Hi AG,

First of all, nice tube. That's a big one. I suppose it's an N. Wood?

If your detector gain is high enough, you will indeed see the gamma counts if the tube works. Any tube will see gammas to some degree because of liberated electrons and ions formed by the rays' passage. The insensitivity of the BF3 that is commonly referenced has to do with the large Q-value of the boron reaction with neutrons, which makes it easy to separate these events from gamma events. We don't mean that the tube itself will not detect gamma events. You probably just have to increase the gain. Gain is a function of both the gain pot setting on the counting instrument AND the gas amplification factor of the tube, basically corresponding to the bias voltage. If you cannot get gamma counts to register on the highest gain setting / lowest LLD setting, turn the voltage up some. You might also want to use a shorter cable (loading the detector with a large cable capacitance degrades the signal) or a higher load resistance (100M for instance).

The channel-to-energy calibration is done by noting that the "stair-steps" of the differential energy spectrum correspond to the kinetic energy of the lithium nucleus and alpha particle, respectively, and that the peak is at the sum of E(Li) and E(alpha). Seeing this clean response requires that you have low noise, linear pulse-processing electronics behind the detector.

I haven't taken a long-term background spectrum with my BF3, so I don't know what a background response looks like. I'd guess that you'd still see the neutron-related spectrum from natural neutrons and that anything else is noise or comes from natural alpha contamination in the tube.

The double-pulse issue isn't a neutron "striking twice," it's probably an electronic problem. (Once a neutron is absorbed by boron, it is gone.)
Double-pulsing can come from insufficient quenching or too high a voltage, maybe also from a very-badly-set pole-zero control, or even a long coaxial cable between some elements of your setup that is not terminated in the right impedance.

I'm curious to know more about your setup. Maybe you've posted something on it before. It sounds like you have the right elements in place.

-Carl

Re: BF3 tube energy spectrum

Posted: Tue Oct 11, 2005 10:41 am
by Richard Hull
I am assuming annoymous guy has a digital scope and is looking at the pulses. If so, then the gammas are obvious with the approach of a mineral or radium source. I use only this method for setting my discriminator after the preamp stage. Without a scope, you are just going to have to tough it out by dialing in the lower level pot until all source gamma detection disappears. A digital scope usually lets you accurately set an intelligent, lower level, "noise margin window" quantitatively related to an observed maximum gamma pulse height on the scope.

Richard Hull

Re: BF3 tube energy spectrum

Posted: Wed Oct 12, 2005 12:25 am
by AnGuy
>First of all, nice tube. That's a big one. I suppose it's an N. Wood?

No, its a brand new Centonic BF3 tube (supplier in the UK)

>If your detector gain is high enough, you will indeed see the gamma counts if the tube works.

I suppose that the noise ( <4 mv) could include gamma hits. I assumed that anything under 24 mv is noise generated from the input power supply, the HV module, or general emf noise. The pulse hits that I assume are real are about 50 mv. I will have to take a closer look next time and see if I I can spot gamma pulses. BTW the 4 mvV and 50 mV is (raw output from the tube, before the JFET). I don't remember what the 50 mv pulses were after the JFET amp I think they were near 2 Volts. I haven't been working with the BF3 tube because I been playing with a HE3 I recently aquired.

>The double-pulse issue isn't a neutron "striking twice," it's probably an electronic problem. (Once a neutron is absorbed by boron, it is gone.)

Thats what I originally thought but these double pulses are distinct. For instance there are periods of minutes of flat line and then the double pulse occures with in less than a 1 usec with similar pulse heights. When I see noise (> 10 mV), I see dozens of pulses that are quite chaotic (random pulse heights and envelopes). This occures when a device (such as a refig) kicks on or off. I don't have a digital scope so I can't show the actual waveform but the way it appeared didn't see like electronic problem. I suppose that it could be a glitch in the oscilliscope, but that seems far fetched. these double pulse are also extremely rare. Perhaps once every few hours. I would guesstimate that one out 80 pulses is a double pulse. Could it be that when these double pulses occur, that the neutron is striking a non-boron atom (ie fluorine, containment, etc)?


>I'm curious to know more about your setup. Maybe you've posted something on it before. It sounds like you have the right elements in place.

Its pretty simple. HV+ 100Meg HV resistor (TRW) to MHV connector to Tube (HN connector) with a 2.5 RG6 cable (98+% shield copper cable). The MHV ground is connected to a 1Meg 1/4 resistor. I have a pair of 100 pf capacitors that dump into a JFET op-amp (with a 5 V zenor Diode for protection connecting the outputs of 100 pf caps). The pp noise is about 4 mV from the HV source. The pulse heights are about 50 mV +/- 10 mV (pre- JFET Op-Amp, direct from the output of the 100 pf caps.)

Re: BF3 tube energy spectrum

Posted: Fri Oct 14, 2005 11:11 am
by Alex Aitken
Just a quick check, but your preamp is charge integrating, right?

I'm wondering if you are running the tube voltage too close to the plataeu and getting pulses from failed quench events. Just a thought.

Pulse height (maximum) of a charge integrated output is the usual method. The decay time of the charge sensitive stage is usually set to be significantly larger than the time a pulse actually lasts coming from the device.

Re: BF3 tube energy spectrum

Posted: Wed Oct 19, 2005 10:19 pm
by AnGuy
>Just a quick check, but your preamp is charge integrating, right?

No, just a JFet Op-Amp configured as a inverting amp. I see the same waveform that the tube puts out, just amplified from millivolts to volts.

Question, what happens when Neutrons are slowed down? Boron is an effective moderator right?

Re: BF3 tube energy spectrum

Posted: Thu Oct 20, 2005 2:50 am
by Richard Hester
The moderating is done before the neutrons reach the BF3 tube. That's why commercial neutron detectors have a big hunk of polyethylene around the BF3 or He3 tube. The boron inside the detector tube is not really dense enough to do much moderating.

Re: BF3 tube energy spectrum

Posted: Thu Oct 20, 2005 9:05 am
by Alex Aitken
If you dont charge integrate the relative max pulse height will depend much more on other factors like where in the tube the fission happens, ion and electron diffusions and the way your electronics is designed and handles fast pulses. It will also contain a lot more noise as only the early charge stands a chance of registering. This may cause the neutron events and gamma events to overlap.

Re: BF3 tube energy spectrum

Posted: Thu Oct 20, 2005 10:23 am
by Richard Hull
Marvin is correct. You just gotta' have a charge sensitive amp behind a He3 or BF3 tube. Why get a nice tube and not give it the chance to do its stuff?

There are often, on e-bay, just the amps you need. even if they are dead, you can select a replacement FET if you are careful. I have bought and repaired several such (NIM) based amps. I use one of those amps on my He3 system for fusor IV. (princeton-gammatech)

These amps are made by Ortec, Tennelec, Canberra and others. They have been optimized for just this work and can often be had for under $50.00.

Richard Hull

Re: BF3 tube energy spectrum

Posted: Thu Oct 20, 2005 11:55 pm
by AnGuy
>moderating is done before the neutrons reach the BF3 tube.

Humor me. What would happen if neutron was moderated inside the tube?

Re: BF3 tube energy spectrum

Posted: Fri Oct 21, 2005 12:55 am
by AnGuy
>You just gotta' have a charge sensitive amp behind a He3 or BF3 tube. Why get a nice tube and not give it the chance to do its stuff?

The JFET works fine. The BF3 tube puts out a 50 mv signal which is easy for a JFET that is sensitive in the nV range. With the HE3 tube, the JFET was saturating the Neutron hits. Until Carl started this thread I thought BF3 and HE3 tubes were not capable of measuring energy spectrum. Now that I know this is possible I will consider rewiring as a charge integrating amp. For now, as it is it should be fine for counting hits.

>These amps are made by Ortec, Tennelec, Canberra and others. They have been optimized for just this work and can often be had for under $50.00.

Well I wanted to constuct my own detector circuit, with a "all in one" circuit board: HV supply + Detector Resistor + Amp + TTL output for counting. Because all of the components are on a single board with little distance between components and a solid ground plane, I have a very good noise ratio. With a seperate HV supply + preamp + HV resistor, Every cable, every extra power input adds noise. As I meationed, I was originally under the impression the HE3 and BF3 tubes were only capable of counting, not measuring energy levels.

Re: BF3 tube energy spectrum

Posted: Fri Oct 21, 2005 6:11 am
by Alex Aitken
I'm afraid you have misunderstood. This is not a neutron energy spectrum. These are the signals produced after charge integrating for monoenergetic neutrons (thermals) as well as tube background. Aside from illustrating the physics that makes the tube work it also shows where the discriminator should be set in order to get neerly just neutrons detected.

Boron is a lousy moderator and a good absorber of thermals. No moderating will happen in the tube to all intents and purposes but you might get very occasional recoil pulse if you have no other moderator. You will of course also get next to no thermals even in a good flux of high energy neutrons and thus next to no reading.

Re: BF3 tube energy spectrum

Posted: Sun Oct 23, 2005 12:56 am
by Carl Willis
Anonymous Guy wrote:
> >moderating is done before the neutrons reach the BF3 tube.
>
> Humor me. What would happen if neutron was moderated inside the tube?

You'd have recoil nuclei produced by inelastic scattering. Depending on the incident neutron energy and the target atom, there would be various amounts of energy deposited in the tube. Consider a 2.5 MeV neutron scattering, which would produce a recoil boron ion of up to about 420 keV; or a recoil fluorine ion of up to about 238 keV. The distribution of recoil energies (i.e. pulse-height spectrum to be expected from the detector) will approximate the shape of the angular differential elastic cross-sections for the nuclei in the tube provided the tube is "thick". In any case, you can see that the deposited energy is always going to lie far below the 2.8 MeV of the B-10(n,a) products, and in fact will most likely be lost in the "noise tail" of the tube's spectrum.

Another possibility is that you get a capture reaction in boron with a fast neutron. The cross-sections are fractions of a barn above an MeV, compared with 3840 b for thermal neutrons, but the reaction could still happen. Ideally the deposited energy in the tube would be E(n) + 2.8 MeV.

I agree with the others that having improved pulse processing is probably what you need to get a clean response from your tube. A charge-sensitive preamp eliminates pulse height fluctuations due to constantly changing detector and input capacitance, and has the added benefit of a long tail that counters "ballistic deficit." In a proportional tube, charge collection times differ because the events are dispersed throught the tube volume. Pulse charge is a more reliable measure of deposited energy than peak pulse voltage magnitude as a result Your main amplifier should produce a pulse output much longer than the preamp pulse rise time, about 10 us long is long enough for a BF3.

-Carl