Neutron Detection

Collected and mostly unedited posts from the Fusor Discussion Board (find it on Farnsworth Chronicles )

Absolute vs Relative counts


On Aug 04, 2000 10:23 am, Richard Hull wrote:

More info. from more reading and discussions with an NRL radiation metrologist.

What I am trying to do is make "absolute" neutron counts from the fusor. It turns out that this is the most difficult neutron measurement to make with accuracy. In low flux neutron fields, it becomes even more difficult. I was told, and have read, that absolute measurments with normal counters to within 10% are considered superb efforts.

Most normal measurements are relative neutron measurments in performance tests, etc. Accuracies of less than 1% error with a given instrument are relatively easy to make.

It turns out that it is almost impossible to arrive at a stable absolute efficiency of a neutron detector due to the nature of neutron interactions with matter both inside and outside of the device, altered conditions of measurment from calibration conditions, etc It is just something we have to live with.

So, once again we must take all absolute measurements of neutron output from our fusors with a +/- 10% window as a virtual given.

This oddly enough is just about the difference in instrument spread within the 3 different BF3 units I observed. Yet, all are capable of relative accuracies of a very high order. Hmmmm.....

As an old engineer I would hate to note that the bridge I am working on is absolutely within 10% of being the proper specified length based on work done with a ruler which is only approximately calibrated, while in the same breath proudly announcing all the individual pieces of the bridge are relatively measured against one another to 0.1% precision.

But, of course, neutron counting ain't bridge work.

On Jul 20,2000 9:17 am, Richard Hull wrote:

I have now amassed about 10 tomes dealing entirely or in a very thorough manner with neutron detectors. After reading and rereading them for the last year or so, it appears there doesn't exist a good, high efficiency neutron detector for small flux, fast neutrons, PERIOD!

The absolute best detector for our purposes is the good old BF3 tube. ( I cover this and the issue of counters in the tape #2). This is the oldest of the neutron detectors and is still used.

A special arrangement called a long counter is the absolute best fast neutron counter and with calibrated and fixed moderation can approach 15-22% efficiencies for a specific fast neutron energy. In a way this is good. We are dealing with what is effect a point source of mono-energetic, isotropically emitted neutrons, albeit a weak one.

The long counter relys on a longer than normal BF3 tube (up to 18 inches) and the neuts are allowed to shine in on the tube axis paralllel to the tube walls. Around the tube is a moderator of polyethylene or such and the fast neutrons are moderated and scattered, hopefully, uniformly over a long distance along the tube. Many more thermalized neuts now have an opportunity to enter the tube and cause a count.

I have a long conter tube and will fiddle with the thing as time permits.

The BC-720 proton recoil system (also seen on tape #2) is fine, but is in the .5% efficiency range and only comes out of the noise floor at about 1000n/sec area and only 5,000n/sec will give a sloppy person a definite and obvious smack in the face signal.

I currently rely heavily on the eberline rem ball counter and nuclear associates neutron ratemeter I have for low level neutron work. with the larger 5" PMT BC-720 still in construction.

Those who are really serious about fusion and hands on work WILL face these issues! you WILL need to readup and start looking for something to count with. My original indium detector idea is just a distant memory but might be usable to a limited degree at 10e7n/sec in the future.

Always thinking, doing, and trying to share the data. (often hard won)


On Apr 05, 2000 12:33 pm, Richard Hull wrote:

>I was looking at buying my first neutron counter, because without it I can't really tell if fusion is happening or not in my fusor.
>So, what do you guys recommend I should buy?

Neutron counters, at least all I have ever seen as finished or surplus units were found and purchased by me on E-bay. BUYER BEWARE! I have never seen one offered to be known as working or tested! All have said , "taken out of service working" (years ago) or UNABLE TO TEST! Assume they are having problems. IF the guy claims it works, you will need to contact him ask him just how he tested it??!! If he says he used a radioactive rock or a radium dial, he is looney. Intense radiation sources that can kill a man will have a zero count on the neutron counters. Beware.

Otherwise $5,000 is the normal new price for the bottom of the line survey meter that I have encountered.

You can make your own neutron counter as mentioned over many older posts on this board using the Bicron BC-720 and a photomulitplier tube if you are a skilled electronics guy.

On Jan 25, 2000 3:58 pm, Richard Hull wrote:

As we are totally snowed in here in the south with 18" of th' white stuff, I poured over and organized more paper work today. I found two more references related to neutron dose rate and "fluence".

The NRC likes the figure 2.9X10e7 n/sec/ for 2.45mev Neutrons.

A table from and ICRU report #20, 1971 give a nice ready to use value of: 8.0 neutrons/sec/ for 2.45 mev neuts.

These are not far from the reduced values given here recently and I like the 8.0 neutron figure for the mrem and will use it exclusively in the future.

Also some corrections.....

Some of the bonner sphere units and other neutron survey meters deliberately over read the actual dose in an effort to play it safe and not allow accidental overdoses on "their watch" I found one manufacturer that is so frightened of the 1MEV RBE 11 neutron that it gives a correction ratio on a supplied graph of 8:1 for the measured dose versus the theoretical dose based on the rpg curve! I guess they supply that graph for those poor slobs trying to do flux measurements with their survey meters. Some of this is due to the type of thermalization done with a mix of inner borons shielding, cadmium and gadolinium moderator screens, in addition to multiple Polyethylene moderator cylinders.

The thought here is to snag all the manufacturer's data on any rate meter you get or you will probably end up overestimating your neutron production. This is why a proton recoil system is just a better deal albeit requiring a number of skills not often encountered in a single person and equipment found only in some of the more advanced hobbyist's labs.



On Jan 26, 10:51 am, Jones Beene wrote:

Measuring neutrons accurately is not an easy task even with fairly expensive equipment. As much art as science, it seems.

With that in mind, has anyone constructed or even thought about a homemade neutron detector that would employ, say a normal gamma detector placed adjacent to a large container full of perhaps boric acid? In effect you could have a giant BF3 tube. Alternatively a plate of cadmium steel might work as well.

One of the reasons that the BF tubes are so inefficient is that they contain only milligrams of actual boron (maybe micrograms? ). With few neutrons being produced and few atoms of boron to intercept them, then, of course statistically there can be only a few hits.

But you could multiply the boron inventory of a detector by billions of times with a pound of boric acid from the drug store dissolved in a few quarts of water. As a bonus you have your moderator built in. The drug store product is not pure stuff, of course, but it is unlikely that anything else in the mix would detract from the ability of the boron to react with neutrons from a fusor.

The n + Boron 10 ---> He4 + Li7 + 2.78 Mev may not produce a gamma signature itself but all of those resultant Mev alphas and Li ions zipping through the water would definitely give your gamma counter plenty to look at.

Richard responds

Good thinking, but there are some serious problems here. (or it would have been done years ago)

The alpha reaction is very short ranged in air about 2cm at max energy and the alpha is not necessarily at the full energy you note making the bulk of them even shorter ranged. The gamma thoughts are equally good, but the efficiency is way, way down below a common BF3 tube.

There are boron tubes using a thin film of pure B10 lining the entire walls of a normal counter tube. The tube is then filled with a much more useful detection gas (argon and quencher) at a much lower pressure and then the alpha reaction within the boron film which proceeds in that gas is very strong due to gas amplification. Unfortunately, they are rare and very expense. (As if BF3 tubes aren't expensive enough!)

The better BF3 tubes are about 25 percent efficient provided the moderator is calculated for the best response of a particular energy neutron.

The proton recoil system of the Bicron BC-720 has abysmal efficiency (~0.6%), but will not detect virtually any neutrons below 1mev and is virtually gamma insensitive. This allows for a much cleaner playing field albeit with bottomed out detection levels of at least 10,000neuts/sec virtually point blank in the 2" diameter.

The BF3 tube requires the unhandy large poly thermalizer which is averaged for a full spectrum of neuts. This places you much farther away from the chamber where the neuts are made and thus intercepting a smaller "solid angle".

Test sources using Ra/Be or Am/Be

In the early days, neutrons were produced by using alpha particles from Radium hitting a Beryllium target. Initiators for atomic weapons have used Po210 and Beryllium to produce neutrons to make sure the fission starts in a timely fashion (rather than waiting for spontaneous decay). Companies such as Isotope Products ( , see also resources page) sell a variety of neutron sources based on Am241 or Cf252.

It might be nice to have a reference source of neutrons to verify that your neutron counter is working, or to calibrate it. On Jan 25, 6:11 pm, Jim Lux suggested::

Could one generate test or calibration neutrons by using the 5 MeV alpha particles from a AM241 source on a suitable target (Be, Al, etc)?. The 1 microcurie source in a smoke detector produces about 27000 particles per second, over a 4pi sphere. If you put the foil target right next to the source, you'll intercept, say, 1/4 of the particles. I haven't looked up the cross sections, but, off hand, it seems that you could generate at least a few neutrons for testing your neutron detector.

Richard Hull responded with some experimental data, and a little analysis:

I have four nice very old industrial smoke detector Am241 strips of 15uc each! I put all four together and then covered them with finely powered elemental Be .9999 pure. The counters (both the survey meters) went from about 1 count every ten minutes to 1 count every 3 minutes with the source taped to the PE moderator!!! This makes sense, too.

Lets do the math.

Am241-Be sources made by Isotope Products produce 2.2X10e5 neuts per second per Curie of Am.(data sheet, page 85 of the online catalog) This is about twice that of the best fusor III run to date. Now lets figure out what we might expect from my homemade 60uc source. It would be reasonable to assume that an amateur construction would not equal and certainly do no better than a pro source. Given this, and the best case scenario, we might hope for 12-13 neutrons per second isotropic emission from my "HOT" source at 4cm which is the closest I can get to the detector (BF3 tube) or about .06 neutrons per second hiting the 1sqcm tube end (optimum angle in a long BF3 tube for ~20% effeciency) From all this 1 neutron might enter the tube every 16 seconds with near perfect effeciency (20%) in a tube moderated for a specifc energy neutron we might expect a single recorded count every 5 X 16 or 80 seconds. My detection every 180 seconds speaks of the less than ideal detection conditions of my particular survey meter and, or, the skill of my source assembly, accuracy to the assumed (4) 15uc Am241 sources, etc. The math works out close enough for government work. (lotsa' slop in there)

Sooooo.... No! An Americium-Beryllium source of the homemade type, even a rather rich one, 60 times hotter than Jim proposed, is of little value.


Neutron Rate Meters

On Jan 23, 2000 2:03 pm, Richard Hull wrote:

I have two neutron ratemeters. These are specifically designed to give an averaged dose of neutrons from thermal energies in the .01ev to fast neutrons in the 5mev range. As such, tradeoffs are necessary in the moderator surrounding the BF3 tube contained in each. The BF3 tube relies on a THERMAL neutron hitting a Boron 10 gas atom and producing an alpha particle in the gas which can then be detected by gas amplification. We are hunting only fast neutrons (2.45mev) and for the BF3 tube to detect these, they must be "thermalized" or "moderated". This being said, I regard these two instruments as being more qualitative than quantitative.

The general upshot is that someone, at sometime in the past defined a curve called an RPG curve (radiation protection guide) based on the RBE (relative biological effectiveness) of neutrons from thermal to fast neutron energies. The object of the moderator is to mime the "inverse" RPG curve over a broad range of energies to give an averaged energy independent dose in millirem/hr of neutrons. WHEW! If you followed that, you were better off than I was when I first read it!!! Stuff like this slowly settles into place in my old brain by study and use in calculation and empirical study.

T.W. Bonner of Rice Institute was one of the first to utilize a detector within a spherical polyethylene ball to mime this curve, and D.E. Hankins of Los Alamos National Lab was the first to codify and specify the idealized 9-10" sphereical polyethylene detector arrangement as a fully functional neutron ratemeter/dosemeter and general survey instrument. His paper, "New Methods of Neutron-Dose-Rate Evaluation", presented at the December 1962, Harwell England, International Conference on Neutron Dosimetry is a classic.

Eberline, Nuclear Chicago, Victoreen and others quickly produced a vast line of "Bonner Sphere" neutron survey meter dosimeters. Perhaps the most famous and infamous of these was the Eberline PNR-4 with its 9" poly sphere and log-linear twin needled meter single channel indicator. I recently acquired one of these classic instruments. About a month ago, I also picked up a Victoreen-Nuclear Associates more modern, non-spherical equivalent. Thus, I now have two professional neutron rate/survey meters calibrated in millirems/hr dose rate. Both are relatively well calibrated with the non-spherical Victoreen unit being orientation sensitive. In general, "Bonner sphere" based instruments are not orientation sensitive.

Remember, I still have a fine, self made, quantitative neutron counter which is well known and calibrated to detect the precise, characteristic 2.45 mev neutron from D-D fusion. This is my plus ultra instrument for neutron measurement.

The acquistion of the two recent neutron survey meters was to give me an idea of the biological dosage that I am receiving and to act as a crude cross check on the proton recoil Bicron BC-720 scintillator instrument.

This weekend I setup the two survey meters near fusor III (~25cm from the fusor core, inner grid.)

I decided to adjust the fusion rate to indicate 1mrem/hr as close as possible on both instruments. This was achieved at about 20.5kv @ 10ma. The longer time constant on the Victoreen-Nuclear Associates instrument caused a slight slippage in the values such that the Eberline instrument was indicating about 1.25mrem as the slower reponse Victoreen unit was dead on 1mrem.

It is my guess that the Eberline was made in the 70s and that the Victoreen unit is an 80s creation. Both are transistorized, battery powered units.

Now to back figure from dose rate to isotropic, point source, neutron production is possible provided some key data is known. I have operators/service manuals for both instruments.

The key item is that 1 mrem for each is considered to be 30cpm or .5cps. Now how does this relate to flux and what are the efficiencies of the counters for the 2.45mev neutrons??!! There will be some hand waving and approximations here by the manufacturers as the system is not made for specific energy detection but quite the opposite. It is designed to smear the energies to arrive at a time ordered average dose based on RBE.

It just so happens that both The Hankins paper and another by I.O. Andersson and J Braun at the same conference give a relative efficientcy of the bonner sphere to 2.5mev neutrons as being about .5 counts/sec per neutron/square cm for 80mm polyehtylene moderation. Given the generalizations thus far, we can now do some crude order of magnitude calcs to get at isotropic emission from the fusor.

As I had the instruments rather close at 25cm, the spherical area of the measurement is 4 X pi X 25e2 or about 7850 sq. cm. As we were getting .5cps from the meters, there were on the order of 7850 neutrons being emitted per second, isotropically. Or, with a small stretch, 10e4n/sec. This is very close to me earlier reports in Jan/Feb 99 on this BBS using my Bicron detector which is highly accurate.

Nice, huh?

I am getting a bit more comfortable with neutron measurment now and while no expert, do have a handle on things, I feel. It is nice to be able to cross check, and even nicer to be able to get neutrons at will with a functional fusor.

I would never attempt to rely on the BF3 tube measurements in the rate/survey meters for precise results, but you can at least, (1) get an order of magnitude idea about what you are doing, (2) prove that you are producing neutrons to any doubting Thomas idiot observer, (3) watch out for dangerous levels of neutrons in order to prepare shielding and (4) check the biological effectivenss of that shielding.

Note these instruments were very lucky finds and relatively expensive ~$500 each. I am sure that someone will be able to beat these figures on a lucky strike in the surplus market.

On Jan 25, 10:27 pm, Tom Ligon wrote:


We have an old Eberline He3 counter. NIST said they don't calibrate neutron counters any more, and also said Eberline's calibration is flaky.

We sent it to Eberline anyway. The resulting data took considerable handwaving to interpret, and I have never trusted it. Just a warning to anyone with this brand of meter. They calibrate for mRem, and can't understand why anyone would actually want neutron flux instead. Tom

On Jan 26, 2000, 12:19 pm, Richard Hull wrote:

Thanks for the heads up on the Eberline He3 counter, I remember it from my visit to the EMC squared labs. It might yet be turned to a good purpose.

In th' biz of neutron counting there are only two really broadly accepted methods. Proton recoil and Boron counters (B10 and BF3 tubes). The boron method is very old and well understood, in Neut counting circles. Most folks have real fluxes to count, we don't. We have to count indivdual neutrons slowly. This is very tough. Efficiency of counting is very low and this mandates very long counts and detailed background count data for light to moderate neutron production levels. For very rare and extremely sparse neutron events the process sinks into the "nosie" and one becomes easily lost in the quagmire of extended handwaving and magical statistical incantations. I call it reading the noise scientifically......."they" used to call it "casting th' bones".

On Oct 08, 1999 2:12 pm, Richard Hull wrote:

I will limit my discussions to the proton recoil system I use - (Bicron BC-720 scintillator attached to a PMT). I will also focus on its use with the fusor system.

Proton recoil can occur when fast neutrons enter any hydrogenous material. In this case, it is the styrene plastic of the scintillator. These charged heavy particles can then hit the silver activated zinc sulfide inclusions in the special scintillator. This produces a flash of light (scintillation) which transmits through the clear plastic into the PMT causing an electron avalanche and electron multiplication. This, inturn, puts a pulse of current in the anode circuit which is amplified further for display and counting.

Assuming we have a functional counter system made up from the above, it is instructive to first just hook it up to a digital storage scope nd observe the pulses evolved from back ground for several hours.

I first take a Co60 of Cs137 source and tape it to the thin metal window at the end of the detector just .5 cm from the scintillator. Bicron professed a virtual gamma ray insensitivity for the scintillator and they are correct. The gamma rays are at least a full order of magnitude in amplitude below even the weakest neutron/baryon/meson signals found in the major background components. Likewise, they have a pulse signature (appearance) which also sets them appart from the proton recoil event. In short, they are easily discriminated by pulse height right out of the recording scenario.

Now we have a counter that is 100% proof against locally produced natural radioactive materials and eminations (alpha, beta, gamma).

It is part of the Bicron Literature to note that the counter can't be used as a neutron spectrometer, per se. Proton recoil energy is a function of incident particle angle as well as energy. Thus, no reducable data is carried in the photon amplitude! What we see are a vast range of background events caused entirely by "cosmic rays". The term is nebulous and should be as the mix is rather profuse. We may be assured that only those more massive items, secondary neutrons, mesons and other items in the mix of really large kinetic energies will be involved in making a pulse appear on the scope. The number of pulses seen per unit time will depend on our pulse height discrimination level.

Naturally, it is assumed that you are not co-jointly fiddling with diffeent PMT high voltage settings and your preamp is hooked to regulated supplies. These are usually "doped" out ahead of time before the terrestrial pulse height discrimination process mentioned above. get your PMT data sheet and never ever go too near the max allowed anode voltage. Most of the time you won't hit even half the maximum. I adjust my HV until i start to get noticable floor noise (zero line hash) on the scope and then back off a notch.

If we just ace out the Co60 and Cs137 sources mentioned above, we can often count 10 or more pusles per minute. A really long term trial over many different times of the day and night over a couple of weeks will reveal stunning variations which seem diurnal in nature. They are!

The chance of solar neutrons entering our counter are vanishingly small. The short lifetimes of neutrons, coupled with the required velocity to survive the trip plus with the attenuation factor of our moisture laden atmosphere and the fact that we will see only very fast neutrons anyway, militates against such events.

All the real neutrons counted in background will be secondary radiations created in and around earth's atmosphere mostly by incredibly energetic solar protons. A whole gang of ultra hot mesons and other nuclear debris will also trigger the counter.

An occassional count will be internal to the counter. Some electron avalanches will be triggered by radiations from the materials within the tube materials themselves. All PMT's are carefully constructed and materials selected to eliminate this effect, but there is always some radioactive material around even in the best constructed tubes.

You will be stunned by the amplitude of some pulses being again a full order of magnitude larger than 90% of all the other pulses. Normal fast neutrons, (2-5 mev), will fall about in the middle of your pulse systems detection pulse heights.

With a lot of looking and background data collecting under your belt, You may now start to implement the actual man made neutron data collection process.

I leave a scope attatched, always! To just trust in an electronic counter or scaler is a fool's paradise. Yes, I have a nice HP counter attached and set its level trigger discriminator mostly by the scope! A scope will show noise and other false signals which would otherwise be taken for a count by the neophyte. In most simple work involving small voltages in d-d fusors, the count rate will be just above background. Hopefully, it will be satistically meaningful. This is where a full understanding of your local background level and rigid pre-run and post-run background counts will payoff.

A quick lesson in solid angle interception and counter efficiency is in order. That, in the next installment

On Oct 07, 7:51 pm, Richard Hull wrote:

The absolute best book in existence for such things is Knoll's "Radiation Detection and Measurement", 1979, Wiley&Sons.

Neutrons are not detectable directly, at least in a quantitative manner. All detectors rely on secondary reactions. All the secondary reactions produce simple ionizing radiation which is then detected by normal means. (all in above ref.)

The real bitch is that there is no counter that will count every neutron! As a matter of fact, all such counters are graded in terms of efficiency. For most all neutron counters that efficiency is abysmal, at best. An efficiency of 0.6% is considered great. Thus, you need a hail of the suckers before you get a single count! The bigger the neutron hailstorm, the more statistically meaningful are the time averaged results.

I have said it before and I'll say it again, Neutron measurements, especially in the noise floor, is an art form! Self deception is easy and rampant among newcomers to the art. One doesn't just fetch a counter, place it next to a suspected neutron source, get a reading and believe it! A lot of background counting, diurnal effects, etc have to be factored in.

I have posted rather extensively on my neutron counter which is about the best and cheapest the amateur will be able to get his or her hands on.

Good luck on your quest.

On Oct 08, 08:55 am, David Taylor wrote:

>The absolute best book in existence for such things is Knoll's "Radiation Detection and Measurement", 1979, Wiley&Sons.

I talked to Dr. Knoll this morning, and he says that his third edition will be coming out at the beginning of 2000, and that he has updated the neutron detection section. (His 2nd was published in 1989 and costs $90 on Amazon, ascompared to $20 for his 1979 version on abebooks. Ouch!) I will try to be patient - after all, "Ionized Gases" should take at least that long to absorb!


On Mar 15, 1999(?), 4:30 pm, Richard Hull wrote:

The key to good measurement and accurate data is related to the taking of good background count data with whatever nuclear instrumentation you have. In our case it is neutron background.

This is one of the more difficult measurments. Trying to mask out all the other stuff which makes counters click in a neutron detector is tough.

I have so shielded and prepared my instrument with discriminators that I can hold a seething hot source of Alpha, Beta, Gamma or X-radiation within inches of the counter's sensitive end and I get "Zippo" out on the counter beyond background. Ideally, the only thing which can trigger my counter is fast neutrons!! Fast Neuts of about 3.45mev or more should be the product of D2 fusion reactions. Some Gamma's can be expected, but according to the Bicron data with my Scintillator, they are much reduced in amplitude and unless multiple coincidences occur, (additive), they are easily removed by pulse height discrimination.

This was found to be correct by me. I removed all discrimination and looked at an O'scope output. On my system, the radium and Co60 sources caused many pulses in the .5 -1.5 volt range (thousands/minute!). Fast neutrons created huge 6-8 volt pulses. Thus, I could safely set the discriminator for 4 volts and safely be assured of never seeing rather intense gamma ray or X-ray readings.

There is no substitute for clean, long term background readings. I have not really done a gross study of where background fast neuts might come from, but primary cosmic radiation would be unlikely as the neutron only has a 15-20 minute lifetime when out strolling around on its own. most of our fast neuts are probably solar in origin with a few produced by secondary cosmic ray reactions in the upper atmosphere. Also, the counter can be activated by any Baryon or heavy charged particle which can cause proton recoil in the scintillator. All these contribute to a variable background count. This count will vary with discriminator setting as well as diurnally and with changing ionospheric conditions.

With this in mind, A background count is best taken just before and just after a fusor run with the discriminator level FIXED!

I have found that a minimum of 10 minutes is needed to acquire a reasonable background with 20 minutes being preferred. Remember, this is both before and after your run. Likewise, I try to make a fusor run at least 10 minutes long. This is a busy time filled with monitoring conditions and jockeying controls. Time slips away and time endings are missed and one must go to the next minute, or more. I was using a standard stop watch.

To avoid this, I purchased the neatest little toy which will, for $15.00, save a lot of hassle and let you pay attention to your work instead of the clock. It is Radio Shack's "talking timer". It has a great countdown mode where you set a fixed time to countdown (up to 24 hours) Let us say, 20 minutes. The time verbally announces the minutes under ten minutes with a pleasing woman's voice "nine minutes left"..., etc. Under one minute it announces every ten seconds. "forty seconds left"...etc., until it reaches ten seconds when the classic count down from 10 begins on the second and at zero a beeper pulses until you stop it. This allows you to forget the time and attend to business until the last moments when you can be prepared to take data, shut down the fusor or whatever.

Armed with the two neutron backgrounds, I find the average of the two readings over the two twenty minute periods and divide by 60 to get a back ground Count per minute. I then figure the count per minute rate during the fusor run and the difference in the fusor rate per minute and the background rate is what I term the Delta CPM for my equation especially developed for my particular counter efficiency and geometry.

While my count might not be close to yours, I find my particle background is in the 2-4 neut/minute range. Current fusor levels record about 23 neuts/minute. While statistically highly significant, this thing is not clickin' away like a geiger counter. An error of just couple of counts per minute will change the output data by 500 neutrons/second!!!!

Data collection must be scrupulously overseen and shepherded to get meaningful and truthful, results.

Why spend time and money, sweat bullets over dimensions and tolerances with making a fusor if you aren't going to collect data well? It's all part of good science and good sense.

22 Feb

Neutron measurement is no trivial matter and at the lower levels of fusion, the need for extreme care is in order as the measurements can get tangled up in noise and false sources.

I used the Bicron BC-720 fast neutron detector coupled to a Hamamatsu R-1306 bi-alkali extended blue range PMT with preamp. This was sealed in a light tight shell of aluminum lined with 1/2" of lead for radiation shielding. Electrostatic shielding was taken care of by the outer aluminim shell and an inner wrapping of copper around the PMT tube itself. Also wrapped around the PMT was a mu-metal shield to fend off any magnetic fields.

I used a Stanford Research photon counter system,(hamfest buy), as the power unit,and discriminator. Both a digital Tektronix scope and Hewlett Packard digital counter were used to set the discriminator levels and count the neutron pulses.

Extensive long term checks were used to establish a base level for background radiation. Large gamma and X-ray sources were placed near the system and the discriminator set to ignore them.

Bicron BC-720

For those that might be interested, I have posted the current prices of the Bicron BC-720 fast neutron scintillator material below.

I have purchased and used the 2" piece for my first Neutron counter which used a 2" Hamamatsu PMT and it worked out great. A 2" detector appears to have a low end noise free threshold window at about 5,000-10,000 neutrons/second when butted up against a 6" fusor.

My proposed 5" design might dip well below 1000 neuts per second.

The BC-720 scintillators are plastic (styrene) disks of the stated diameter and about 1" thick. Somehow, the Bicron folks make an "archery target" design of thin, silver activated, zinc sulfide vertical rings within the plastic piece. The whole thing is polished like glass, except one face of the disk is painted white. The PMT tube side is clear and open. The side which is snow white, faces away from the PMT.

The list price of all the ingredients is below $2, I am sure. The process of pulling it all together is the money maker.

5" PMT

I just recently picked up a nice, new, inbox, Hammatsu, 10 stage, 5" photomultiplier tube! This was a real find indeed.

I need a neutron counter with a bit more sensitivity which will defeat the low end barrier a bit more. This new system should allow a low end, minimum count of 1000 neutrons per second to come into view with statistical certainty.

The tube cost me $50.00, but the required 5" diameter Bicron BC-720 scintillator just flattened me for about $475.00! Ouch again!.....But, hey, you can't take money to the grave with you!

While it is on order, (about an 8 Week lead time as it is made up), I am working on the dynode string and preamp, and finally, the special housing that will be required.

For those interested, the 5" diameter detection volume placed against a fusor will intercept a great deal more neutrons sending the statistical data into new regions, where only noise existed before.

Of all nuclear radiations, fast neutrons are the most difficult and costly to detect. As I have said before, it is really an art.

Rolling one's own neutron counter is not for the beginner in electronics. Non-electronics types must usually do without. Knoll's book on radiation detection is one of the best references I have on the subject. Too bad it is out of print.

I had once held the illusion of using Indium foil and neutron activation to get at the count rate for a non-electronic solution, but the flux is so low from the low end amateur fusor that such methods would demand even more sophisticated and costly instruments than a full blown neutron detector. The indium foil, even then, might only offer a crude order of magnitude idea about the neutron yield.

Tom Ligon's words on indium activation

I did get a couple of runs on a 15 kV Fusor which did get statistically significant counts from neutron activation of indium. We used less than an ounce of indium (I know you have about 6 pounds of it or so) rolled to thin foil strips, moderated in water, and counted with a cheap "Alert 4" GM counter. The runs we detected were making short bursts at around an estimated 100,000 neutrons per second total reaction.

Indium has a HUGE resonance crossection for neutrons as the energy drops thru the epithermal region during moderation, something like 30,000 barns! The crossection is so high the indium must be thin foil or fine particles, because a neutron will only travel something like 0.002 inches before being absorbed. Thus, it is pretty efficient. The activation produces a half-life on the order of half an hour.

I would recommend this as a back-up method to an electronic counter (PMT and BiCron 720 as a case in point. The electronic methods are subject to false positives when the "pulses anomolous discharges" occur on the inner grid.

My preferred system would suspend fine indium powder in parafin wax or a comparable moderator, then put a GM tube on a computerized counter in the middle of this. I would do 5-minute counts for at least half an hour, and look for a statistically significant fall-off compared against a nice, long background count.

/nuc/ncount.htm - edited 28 Sep 2000 - Jim Lux