B-10 Tube Testing

[Tyler Christensen]

I stumbled upon a very old B-10 tube (tested ok in '55). I'm trying to get it working and am having some difficulties. As seen in the first picture, a makeshift moderator for testing. I'll build a nice 6" PVC moderator if it works, but for now this should suffice to test it.

I put a piece of ore on the tube and ran the voltage up until I saw gamma fuzz on my scope (second picture) that registered on the SCA set at 2V. This is with the pathetically simple pre-amp. I then pushed the voltage down a few volts and the counts stopped (around 650V). Then I ran the fusor up, there were no counts with 45kV applied and no plasma so no corona RF issues. I then put some deuterium in and ran it at about 35kV, 8mA for a while, got between 20 and 40cpm. But, then I removed the moderator and got the same count rate.

This indicates that it's not really detecting neutrons, perhaps x-rays or high energy gammas?

Any suggestions? I have no way of verifying I'm generating neutrons right now, but it was last tested 2.5 months ago on a small BF3 tube and not much has really changed, pretty sure it's still working great. When I was testing on that BF3 tube it got about 500,000n/s with peaks up to 1.5 million.

[Doug Coulter]

Tyler, I have two of these now, not well tested, and my very first forum question was "how do you make these work" -- funny.

I believe they want in the 850 volt region, but make so little signal you may not see a neutron signal well with what you call a "pathetically simple" preamp, you might be seeing preamp noise? Depends, I've seen some simple looking designs perform really well (prefer them myself). The currents here are really low, though.

Did you test with a "real" gamma source to see the gammas? You should see them (raw out of the preamp on the scope) at something like correct tube voltage settings. The neutron signal at those settings should be much louder, due to the Q of the boron capture reaction. So you set your threshold to ignore the gammas and let the neutron counts happen. You don't do this adjustment with tube volts, I think -- they just make it work at all when right, not if not, they need to be in a range that gives a little "gas gain" due to avalanches. These work better end-on(?), but I hear they are quite numb indeed, compared to the other tube types, so you may just be actually getting a correct count if you're getting the number you say! I have some old papers that show junk like a few of these shoved into a piece of wax moderator about the size of a compact car -- at a reactor, where things are loud indeed. So they can't possibly count all that fast on our level of N fluxes or they'd saturate dead on a fission reactor.

[Tyler Christensen]

All I have for a source is a fairly hot rock of ore that I got at HEAS, is that "real"?

When I have the tube voltage at 650V the gamma fuzz is nearly nonexistent, and at 670V they peak the 10V capacity of my pre-amp. I think the pre-amp is fine with very small signals, the schematic is on fusor.net. It works fine on BF3 tubes and I believe Scintillators.

I just tried at 660V where gamma peak is in the 4-6V range, set the SCA to cut that out (now bypassing the NIM shaping amp, pre-amp straight into the SCA).

Same results, a few counts per minute, but same non-moderated.

[Carl Willis]

Hi Tyler,

I apologize for my delay in sending the promised BF3 tube to you. That is something I must do soon, I can tell. It is still in my car. I am finding it hard to get out at lunch these days...at work we have a customer "supervising" a project and cracking the whip.

There is an old thread about these tubes here:

viewtopic.php?f=13&t=5577#p34021

Brian McDermott used one also I believe.

They are very sensitive to neutrons; the spectrum in the first post shows their response to neutrons from an AmBe source. They do not have the selectivity for neutrons over gamma rays that the BF3 and He-3 tubes do. Setup is a little more dicey. Your procedure is basically fine, however. Their fill gas is probably argon with an organic quenchant. 400-600V should be appropriate; 900V is excessive.

-Carl

[Doug Coulter]

Thanks, Carl, I was kinda hoping you'd show up on this one. I needed the info myself...

[Tyler Christensen]

I just ran a few more tests, this time I didn't discriminate out gamma at all and gave the SCA very low cutout. The bias was set so that gammas gave about 5V pulses and the SCA was set to 0.4V.

Background: 350CPM
Unmoderated Fusion at 33kV 6mA: 900CPM
Moderated Fusion at 33kV 6mA: 1750CPM

So it definitely does detect neutrons, it's just insanely poor at discriminating anything out.

I'm wondering what the addition of 550cpm is between background and unmoderated, just RF noise? With the voltage at 50kV with no gas, it is nearly background so it doesn't seem like it's noise (I've never had noise that was a function of current, only corona discharges).

Is it even possible to get this background out of the picture? Any tips on finding the right settings?

[Doug Coulter]

You can stop the gammas with lead....at least those up to the power supply volts will be stopped by 2 or 3 mm more or less completely (they are here along with the tank walls). When there's fusion, a background continuum of higher energy ones that would take real thick lead starts to come out too, and you'd count those -- but hey, do you really care? They mean fusion just as much as anything else does. They don't happen running the same conditions in say, argon, or helium, for example; they are purely a byproduct of real fusion happening, something with high Q for the reaction to make that higher energy happen.

I like this plan -- a tube detector for moment by moment feedback, so you can "tune for max output" pretty easily. I have some tubes for neutrons, but it really doesn't matter what I use for that as long as it's a "fusion only" detector, however you get there.

Alongside that, have a moderated oven for baking silver or indium in the neutron field, then count it with a geiger (a jig to make it so you always do this the same way is a good idea). This will give you a reading for the entire run (we usually use 5 minutes).

That's what I'm doing anyway, and it makes some of those issues just go away for me. And it gives a somewhat redundant "sanity check" -- if my counters were going nuts but the silver stayed cold, well, the counters have a problem! Open and shut, no fancy troubleshooting about which is wrong, the sliver or indium is NEVER wrong -- gammas and EMI don't affect it at all.

Remember too that this detector will have a little response to fast neutrons anyway, and more importantly that there is no way from keeping some of them from being moderated in the average lashup -- there's always water in the air, the air itself, the table and a bunch of other things that can slow down neutrons and bounce them around. So getting some counts without explicit moderation could mean some gammas (likely) and some moderated neutrons from junk nearby (likely) and some response to fast neutrons anyway (likely).

[myID]

Hi-

just a hint- from what I know PVC eats Neutrons (I think Chlorine has a relatively high cross section)...
So probably you do not want to use it for your moderator?

Greets

[Steven Sesselmann]

Tyler,

Lots of good advice in the posts above, I would just add that, your neutron counts would improve if your tube was moderated on all sides, and not just a slab of paraffin between the fusor and the tube.

Steven

[Doug Coulter]

PVC does eat some neutrons due to the Chlorine. My neutron oven was originally made of it, and when I went to HDPE things got better, but not hugely -- maybe 30%? I have my 3He tube in a 6" PVC pipe, but the moderator is wax cast in there, so not so much PVC is involved, and it does work fine when PVC's not the main/only thing there. The thing about PVC is that it's cheap and it's everywhere, wheras big chunks of UHMW HDPE etc are neither.

[Richard Hull]

I rang out one of these tubes in a past posting here and have a couple on hand. I found them "devilish" and rated them poor to moderate in my report. As Tyler noticed, they are a poor but workable GM tube more so than they are a neutron detector.

The ideal voltage range roams from 500-800 volts, depending on many factors. Among them is age, gas condition, prior use history etc. The key point is that all of these are very old tubes. Thus, your mileage may vary.

Carl and others have monkeyed them into a finicky functionality. I suspect there is a narrow sweet spot that, once locked in, would prove them somewhat serviceable or in the best instances, with a prime tube, quite useful. I'll bet you will need a first rate, charge sensitive preamp.

With the He3 hiatus and BF3 tubes being snapped up as replacements in the lurch, I would naturally expect these old tubes to be looked at by some hopefuls who have latched on to them.

I think that with a lot more effort than it takes to say, someone who is dedicated to the task and who can verify real operation once arrived at, these tubes might prove a valuable toy for budding fusioneers on a budget.

A good, thorough, operational study of these electronically and radiologically would be a good task for someone so disposed or backed into a "detection corner". What was once considered a collectable relic of the cold war neutron detection world might soon fill a gap.

Tyler, a good hot piece of ore is a great gamma only source if you put a piece of 1/4 thick plexiglasss between it and the detector. This is due to its trapped radium daughter content. An instrument dial in a poly baggie would be a lot better source.

Richard Hull

[Carl Willis]

Hi Tyler,

My first suggestion is to lower the bias until the gamma counts are under 1V. With gamma counts coming in at 5V, most of the linear range of the NIM electronics is being wasted on the gamma and noise continuum.

Then, adjust the LLD on the SCA until the rush of gamma counts stops. At this point your count rate should be at most about 20 CPM. Any counts remaining are either neutrons, gamma pileup events, or electronic noise. The argon in the fill gas makes these tubes more susceptible to x-ray pileup than the modern tubes, but a thin piece of sheeting lead should solve that problem if it is suspected.

Note that the surroundings cause some moderation of neutrons, even if you take away your moderator. There will still be some neutron counts. Your moderator also looks less than effective. The tube should have a better hydrogenous reflector behind it. If those jugs do in fact contain a chromate solution, the capture cross-sections of the dissolved materials may be much higher than plain water. I recommend plain water, wax, or polyethylene.

I think this tube will work for you. Brian McDermott, Richard Hull, Jon Rosenstiel I think, and I have found their sweet spots and made them work. They used to be very inexpensive on the surplus market. But recently, Los Alamos has been buying them back from surplus outlets. No idea why, but they are not going to be as common as they once were. You've got a real piece of history there!

-Carl

[Doug Coulter]

I think I should fire up my two samples and try them real soon. If I find anything good, I'll post.

[Dustinit]

Jon Rosenstiel has a great primer here
http://www.rtftechnologies.org/physics/ ... tector.htm
Where he reports complete insensitivity to gamma and xrays.
He also covers moderators and showing counts vs orientation and sensitivity to various calibrated sources
Seems a very useful tube.
Dustin.

[Carl Willis]

Two corrections.

First, that is Andrew Seltzman's website.

Second, this is not information particular to the AEC B-lined tube under discussion here, but concerns an N. Wood BF3 tube of modern manufacture.

-Carl

[Dustinit]

Thanks for the correction Carl, my mistake(s).
Interesting that one is completely insensitive to gamma and the other not, presumably this is due to the gas fill or the agon ionising sensitivity to Xray/gamma.
I assumed they were the same type of tube and the "B" was the important part for alpha generation but clearly photo ionisation of the gas plays a part.
Seems my ignorance on neutron detection requires some work.

[nemesistech]

I have only been using a bubble detector so far, so I have a few question here.
First, I can get both UHMW and HDPE solid rods 6" in diameter. Is there any advantage to using one compared to the other with the B-10 tubes? Also, if I rounded one end off and drilled out the center to inset the tube (it would be shaped like the bubble detector), that would give me about 2" of solid polyethylene all the way around the tube. Opinions about if this is thick enough? And if not what do you think would be the optimal thickness for a tube like this.

[Doug Coulter]

Perhaps we can get Carl to chime in here, but here goes.

UHMW is more or less the same stuff as HDPE, but is slightly denser so you need a little less.
More often than not, it's cheaper too. So I'd go with the less expensive one. I know from experience that the UHMW isn't bad to machine at all -- and that's quite a hole to be drilling so it matters.
No need to round the end.

2" should do it. This is not like the neutrons just get slowed down while going straight -- every collision completely randomizes their direction, and a really good example was Carl's recent posting on YouTube with a link here. Just putting a slab of moderator *behind* the tube, with nothing between the source and tube created enough slow neutrons to make the tube count -- a fairly dramatic demo.

We've tried from 1.5" to 2.5" as the first layer of our neutron moderating activation oven, and both extremes work about the same (there's a lot more behind the sample) with perhaps a slight dependency for things that like neutrons a bit faster than thermal. But it's not an oh-wow amount of difference.

[Tyler Christensen]

I've found the problem, and that is that my moderator wasn't large enough. I just surrounded it by slabs of 1" paraffin on all sides effectively making it a 2" moderator, and my rates went way up. Now the unmoderated background is statistically insignificant for the accuracy of the detector (usually less than 10-15% total moderated count rate).

I'm now getting 7500CPM at 40kV 5mA, and more than 20,000CPM at 40kV 12mA (can't keep it going for the minute count due to outgassing immediately driving the pressure down when I hit the higher current so I can't get an exact value on that).

But overall, this tube is really rocking now. No gamma rejection, but with proper moderation it's statistically irrelevant, and as Doug pointed out, the gamma function is based on fusion anyways so it's still an indicator.

[Richard Hester]

If you are using the pathetically simple preamp circuit from files, you may need to modify the charge storage capacitor and reset resistor to tailor it to the amount of charge released by a typical neutron event. I originally used values that were in line with a Bicron 3" NaI scintillator (with a 10-stage PMT, I think)., excited by gamma radiatation from a radium dial.

For example, if the preamp needs to be talored for a lower value of picocoulombs per event, one would reduce the value of charge storage capacitor and increase the value for the reset resistor.

[myID]

Dear Richard-

I asume this will be the same for a He3 and BF3 tube also?
I also built the pathetically simple pre amp from the files section and so far just tested it with a pulser (also home brew from the files section)- nice!
Could you please advise what values to use to detect a "typical Neutron event" from a "typical detector tube" or share guesstimated values?

Thanks for your help in the "Art of building homebrew Neutron detectors"!!!
Roman

[Tyler Christensen]

I have used the default schematic capacitor and resistor for BF3, B-10 at high rates from gamma, and GM tubes at up to 50kCPM, they all work fine. I don't think there's any real need to configure it specifically for a tube type. Unless, would this help with discrimination?

[Richard Hester]

If the value of the charge storage cap is too small for the amount of charge in a valid neutron event, you will saturate the preamp for good pulses, and amplify grass/noise that could otherwise be properly discriminated out. If it's way too large, the amplitude will be low for both good signals and bad. I brought a home made fast neutron detector and electronics to HEAS a few years ago, and found that valid neutron events from a small chemical neutron source saturated my preamp. Increasing the charge storage capacitor by a factor of two helped (decreasing the reset resistor by a factor of two at the same time to keep the same time constant). If I recall correctly, data sheets for proportional detectors give numbers for the amount of picocoulombs expected from an average signal-causing event. At any rate, it would be a good idea to use a moderator with the B10 tube and blanket it with some lead, then hang an scope at the preamp output and look at the pulses coming out.

[Carl Willis]

Roman,

In a charge-sensitive preamplifier, the output voltage is very nearly given by

Vout = -Q / Cf

where Cf is the feedback capacitance. (This approximation depends on the voltage gain of the input FET being large, and on the input capacitance being insignificant compared to the product of voltage gain and Cf. From Knoll, Radiation Detection and Meas., 3rd ed.)

For example, in the Cremat CR-110 preamplifier (I use this quite a bit with He-3 and BF3 tubes), Cf is 1.4 pF and the sensitivity would be expected to be 0.7V / pC. Ortec's 142PC preamp, designed for proportional tubes, has a listed sensitivity of 6.5V / pC. In Richard's "Pathetically Simple CSA", Cf is 27 pF and the sensitivity would be expected to be about 37V / nC. To make Richard's preamp as sensitive as the Cremat, you would change Cf to 1.4 pF. Unfortunately, that will make the decay time too short for use with most nuclear electronics, so R3 will have to be increased to several dozen MOhm. Also, you have to be careful not to overload or saturate the preamp. When you do this, the linearity of its response to detector charge is lost.

A proportional tube's gain varies enormously with bias voltage. You raise the high voltage until you obtain signals of the appropriate magnitude in the electronics. Based on the construction of many preamps made for proportional counters, it appears that they're usually run so as to obtain a charge on the order of a pC for detection events.

-Carl

[Richard Hester]

I did some snooping around, and the typical output of a B10 or He3 proportional tube is in the range of 0.1-0.2 pC for a typical event. This means one should really use a charge storage cap of 1-2 pF, and some amplification after the charge sensitive input stage. The simple circuit in files was meant for use with scintillator/PMT combinations. For the proportional tubes, I would look at devices with lower input capacitance for the input stage, like a jfet. I'll do some thinking about a circuit better optimized for this application. Too bad I couldn't get a tube to try with it...