Page 1 of 2

Scintillator tests

Posted: Mon Apr 01, 2002 6:30 am
by Richard Hull

I have been preparing for and conducting tests on the real Bicron BC-720 fast neutron scintillator and its work-a-like fabricated by Tom Dressel. For the purposes of this report the BC-720 will be called BC-FNS. The Dressel Copy , work-a-like will be called TD-FNS. (Fast Neutron Scintillator).

Inorder to rule out other factors, I will test both scintillators in the same housing using the same EMI PMT and preamp connected in the same manner throughout all tests.

Each scintillator when installed in the housing, using the same optical coupling fluid, will be given 24 hours in the sealed, dark, housing prior to taking any measurements.

In each case, (5) 10 minute background runs will be made and averaged following the establishment of a gamma ray baseline and setting the appropriate SCA discriminator levels to avoid counting the Gamma rays from the sealed RaBe source.

Every reasonable effort has been made to insure a fair and even handed test.



The setup consisted of my original Bicron based BC-FNS 2" scintillator system. This consists of an EMI PMT with mated preamp and dynode string. The voltage applied was +900 volts from a precision Tennelec NIM regulated HV supply. The preamp +/-12 volts was supplied by the pulse amplifier NIM modules 9 pin standard connector. The signal from the preamp was fed to an Ortec NIM pulse amp and the amp's gain setup to yield an output ampltude of about .5 volt for detection pulses from a Cs137 source (660kev gamma).

** Note: As the FNS plastics are not meant to be a nuclear plastic, the pulse amplitude varies slightly. At least, more so than in a scintillator made for gamma detection, but no more than 0.1 volt difference was observed.

The pulse amp output was split and sent to the Tektronix TDS 210 Digital storage scope's channel #1 and to the input of a Canberra Single Channel Amplifier (SCA) nim module. The output of the SCA was sent to Channel #2 of the scope where it was split off to a Tennelec NIM timer/counter module and there, properly terminated.

The SCA was setup to offer the largest possible neutron detection window above the Gamma levels discovered during setup. For the BC-FSN this proved to be between 2-12 volts (10 volt window).


The Neutron Source:

I have made up a simple neutron source using Ra salts left insitu on old, cutoff aircraft instrument needles and Be metal powder. It is strong enough to give definitive long term results, but not strong by any standards of the industry. It is permenantly sealed in an aluminum cigar tube end nipple and stored in a lead vessel. The actual radiation from the lead vessel is under 1.5 mr/hr, which is less than a small piece of uranophane mineral in a mineral collection.

***WARNING****It is highly recommended that you NOT try this on your own. I have all the materials and the knownledge to construct such a device. Most of all, the safety concerns are severe and slipups by rank amateurs can result in crippling, debilitating and painful berylliosis and untimely death. There is no cure for Berylliosis save an agonizing, but blessed death. The radium salts on the needles were NOT disturbed by me in any way! NEVER, Ever scrape radium dial paint for any reason. A slip up here means a one way trip down cancer lane as radium lodges in the bone and is a promised lukemia death with as little a .1uc absorbed in bone. Just don't do it!


Background and Gamma Baseline

I spent a lot of time on this daunting task. I used a Cs137 source pressed against the end of the detector assembly to observe the output pulse amplitude. This hovered between a maximum of the 0.6 volts to a low of about 0.4 volts most pulses were 0.5 volts. This gave an indication of what 660 kev gamma would do and their approximate levels. I realized that my source, being a RaBe source, would throw out gamma up to and including, 1.6 mev. I next placed a large aircraft instrument from WWII against the face of the instrument. (The face plate on the end of the detector housing is a 2mm aluminum cover plate and the FNS was located about 1" on the other side.) This large Radium source gave a gamma reading on a Victoreen ion chamber meter of .75 R/hr! It was noted that pulses from the amplifier were now, on occasion, topping 1 to 1.2 volts in amplitude. Rarely, 1.5 volts would be seen, but often enough to force me to call the gamma discrimination SCA lower level set point to 2.0 volts for counting purposes.

Most of the Gamma rays from Radium are below 500kev with a large clustering at 150 - 350 kev. I estimated that that would be about 200mv on the amp's output.

I set the SCA lower point to 200mv and counted with the counter set to this low level with the large radium instrument pressed against the instrument.

The results of (2) 10 minute counts was an average of 186,231 counts for the ten minute interval.

I next raised the discriminator level on the SCA to 500mv and counted above most of radium's gammas. the results was 3096 counts averaged over the same two test periods.

I next took the neutron source and placed it againt the end plate of the detector. The count at 200mv lower window SCA setting was 66,515 counts.

A second count over two periods with the SCA's lower window set to 500mv yielded an average count of 1109 counts over the 10 minute period.

These were nice tests in that the meter contains about 3 times the Ra that the RaBe Neutron source has. at both discriminator levels.

I now re-ran the tests with all sources removed from the test area. Again, over (2) 10 minute periods

Setting of 200mv on the SCA - 1134 counts
Setting of 500mv on the SCA- 174 counts

Finally, a base background level for neutrons was set into the SCA as the lower window was chased up to 2.00 volts. (5) 10 minute count runs were made with no sources around and the average was 4.2 counts over 10 minutes. (.42 cpm).

*NOTE** A rare super event occurred maybe 1 or 2 times over a 10 minute period. This event produced pulses so large (>12 volts) that the SCA window could not include or detect it for counting! This is an old friend from the 1999 time frame and was noted by both Scott Little and I when originally working with the BC-FNS. It is probably due to a huge internal shower of electrons from the PMT internal elements due to a direct hit by a large gamma ray, cosmic ray or neutron, and not related to the scintillator crystal at all.

Once I aced out all gamma in the 2mev class and under by the high setting on the SCA, I personally felt that most all of the terrestrially based NORM, (Naturally Occuring Radioactive Materials), was excluded and all background counts were truly cosmic in origin. (only .4 cpm in this case.)


Counting Neutrons: BC-FNS
The first test was with the BC-FNS by Bicron with the SCA set for 2.00 volts.

Inorder to establish just how immune to a hot gamma source the instrument now was, I ran (3) 10minute tests with the large radium meter placed against the face of the detector head.

Results: 5.1 counts over 10 minutes (.51 cpm). This proved that the discriminator levels were set correctly as the huge blast of .75 roentgen Gamma hardly impacted the counting.

*note** All normal Gamma activity should have been suppressed by the discriminator setting on the SCA. The extra 0.09 cpm might have been due to neutrons actually generated within the old meter instrument itself!!

The meter has an aluminum body and the RaBr doped paint is applied directly to an aluminum dial within. It is well sealed, (gasketed), so 4 times the actual elemental Ra alpha activity is there due to accumulated daughters. It is well known that alpha bombardment of aluminum can produce neutrons. All light elements do this. The upshot is that the 0.09 cpm is not particularly significant and might be within the drift noise range over time. But I feel the extra is, indeed, significant and due to meter generated neutrons.


Next, the big test............

With the neutron source placed about 1/4" away from the instrument detector head, the Bicron BC-720 averaged 18.1 counts over the same 10 minute period in the same (3) 10 minute runs. Thus, at 1.8 cpm, this is 4.3 times background! This is proof of the neutron counting of the BC-720 or BC-FNS! While the pitiably weak neutron source didn't run up an impressive total, it is so far out of the noise that the tests can be considered definitive and significant.

Side Note: For those sticklers for detail.... We might want to assume the absolute worse case scenario and say that the 0.09 cpm additional over background with the aircraft instrument was totally caused by Gamma radiation. Given this assumption and the fact that the meter was three times hotter than the neutron source, (noted above), the neutron source could have bumped the background in this run by 1/3 of that .09 cpm or, .03 cpm. This would subtract from the 1.81 cpm to become 1.78 cpm or 3.96 times a high gamma field background! Still stunning by any standard.

ABSOLUTE BOTTOM LINE........1.33 real neutrons were detected each minute after subtracting background (worse case). As the BC-FNS (BC-720) is noted to be only about 0.6% efficient in absolute counting scenarios that meant that 1.33 X 166.6 = 221.6 neutrons zipped through the mass of the scintillator each minute or 3.7 n/s

Further extrapalation and assumptions such as a 1.5 inch separation of the source and scintillator and viewing the source as a point, yields a total isotropic source emission of about 11 n/s.

As it is known that about 10,000 alpha disintegrations in an intimate mix of Ra and Be produces only a single neutron, we can also get a crude figure that about 3uc of alpha emitters are present in the mix. As radium is only one fourth of the alpha emitters in a sealed source in equilibrium, I have only 0.75uc of actual Ra in my source. Not much at all. You would need over 4 uc of americium 241 to equal this or about 40 smoke detectors worth!

The math and several assumptions, by now, have tortured the limited data to give only order of magnitude data about the source. Still, it is valid, to a point.


Testing the Dressel TD-FNS

After removing the BC-FNS from the face of the PMT following the above tests, the TD-FNS was put in its place and properly coupled to the PMT with optical coupling compound. The instrument was resealed and light proof tape used to cover all joints. 24 hours were allowed to pass before the following tests were made.



Same as above.


Gamma baseline and background readings.

Just for comparison, the instrument was just simply plugged back in and all of the old settings left intact.
A quick check with the Cs137 source showed that only about 10% more gain was needed in the pulse amplifier and the system was virtually ready to go as if nothing had happened!! This was stunning and a real coup for Tom Dressel. I could not tell the new scintillator from the Bicron original once this small gain boost was brought into play. The SCA lower level remained set at 2 volts with a 10 volt window, as above.

(5) 10 minute background checks recorded 8.2 counts averaged of the group of 10 minute runs. (.82 cpm)

(3) 10minute runs swith the .75 R meter in place averaged to 9.0 counts per 10 minute period. (.9 cpm)


The neutron count

With the neutron source placed against the end plate of the detector head, (5) 10 minute count periods averaged 22.7 counts over the period. (2.27cpm)

While the absolute counts in this test don't equal those of the first test, the ratios do and the actual detected neutron events are virtually identical. (1.45cpm) So with different absolute numbers, we arrived at the same basic number of detected neutrons after background is accounted for. Another stunning success, I think.

Tom can be proud of his efforts to make a lower cost scintillator available for those needing a neutron counter on the cheap by building there own.

I have put forth quite a bit of effort to setup and perform these tests and about 15 hours of labor total went into the above data collection and written report. In retrospect, for me, the effort was highly rewarding. However, I will let you folks be the judge as to the value of this work to those on the forum.

Tip**** The setpoint of the discriminator SCA is critical. Access to a digital storage scope is almost a must if one is to set the discriminator lower level based on a test source of gamma radiation.

One can easily see that too low a set point and the apparent absolute count will soar, but so will the background levels. Too high a set point and the count and background differentials grow vanishingly small. The ideal point is where the biggest delta exists between background and real neutron counts detected. I suppose this could be done tediously in multiple count periods, each time altering the SCA setting until the ideal is found, but it would be no fun.

Good luck all with your neutron counting efforts.

Richard Hull


The third night, I was about ready to tear down the above setup and decided to try my best BF3 counter calibrated by Ludlum on the neutron source.

Three 10 minute runs with no source yielded an average for the period of 0 counts!! AMAZING!

With the really hot, ( 0.75 R) meter pressed against the moderator, I averaged, again, 0 counts over the ten minute periods. Even more amazing! Total Gamma insensitivity.

With the neutron source against the moderator jacket I averaged 17 counts over the (3) 10 minute periods.

This was why I sorta' ditched the BC-720's for the Eberline and Victoreen BF3 neutron counters. (little or no background, almost total immunity to Gamma fields and positive neutron counts. The BF3 is a bit more sensitive to high energy pulse noise in the test area than a well shielded and internally preamped scintillation unit.

This afterthought test also reaffirms the value of the BC-720 type scintillator as a superb poor man's neutron detector. The differential is amazingly small between all three instruments looking at the same source. The BF3 has more detection volume and area of intercept which explains the fact that the BF3 read a little higher than either of the BC-720 based systems.


Re: Scintillator tests

Posted: Wed Apr 03, 2002 4:46 am
by Tom Dressel
Great job Richard and done like a true scientist.

Anyone wanting a scintillator or a roll-your-own BC720 scintillator core see the Trading Post.

Tom Dressel

Re: Scintillator tests

Posted: Wed Apr 03, 2002 7:12 pm
by Richard Hull
Hey folks, I just added more to the original posting here!!
I ran a cross check with the same NIM gear on the Eberline BF3 neutron counter just calibrated by Ludlum and it agreed very closely with the BC-720 tests. This speaks well of all systems.

Check out the postscript to the original post.

By the way, I have now edited the original posting 48 times!! It would be out of character for me to get it right on the first pass.

Richard Hull

Re: Scintillator tests

Posted: Wed Apr 03, 2002 10:32 pm
by guest
I'm Glad you warned about the radium paint.... That stuff has had it's share of casualties. Another item that was proposed on another post to use Polonium 210 as a source has most of the problems of the radium salts with a curious action all it's own. It self heats and will spread and contaminate any area you have it in. When the feds use the stuff it comes hermetically sealed. So never .... never ever open a destatic wand under any conditions. Same with the smoke detector items or self luminating signs... all will be with you as long as you live. The high energy alpha's of polonium will make sure that life will be a short one.

They say that in producing the V1 weapons that the Nazis killed more people producing them than were ever killed by the missile.

During the rise of the strategic arment ,the future generations will probably find that the production of the atomic weapons killed more people than the actual weapons.. (assuming no use)

I say we as an intelligent group should internallise these leasons and not repeat the same old silly mistakes.

Larry Leins
Physics Teacher

Re: Scintillator tests

Posted: Thu Apr 04, 2002 3:43 am
by Jon Rosenstiel
Thanks for the great report; I just finished printing it out. I would imagine that my BC-720 replica would mirror the results of Tom's replica as they're quite similar. (Same size, same casting resin to ZnS ratio, etc, etc).

I also want to mention that in my experience a freshly assembled scintillator/pmt combination takes about 6 days to calm down completely. In the first couple of days after I assembled my replica I was getting 13cmp. That slowly tapered off to around 9cpm after 6 days. It's held at the 9cpm count ever since.

Jon Rosenstiel

Re: Scintillator tests

Posted: Thu Apr 04, 2002 5:48 am
by Richard Hull
Thanks for the input Jon. I remembered you made one first and am glad Tom decided to go public and offer them up to others.

Have you had an opportunity to check yours against any kind of neutron source yet? I would be interested in hearing about them when you do. I would suspect that the results will be similar to the real thing just like Tom's are.

It is good to know your observations regarding the high noise "activity period" of a light shocked scintillator of this type.

Richard Hull

Re: Scintillator tests

Posted: Thu Apr 04, 2002 6:58 am
by Jon Rosenstiel
No neutrons yet. My x-ray power supply is coming along, just have to finish the controller portion of it. I have a pair of 6" ss hemispheres. I'll order the flanges and viewport when my checking account allows! I picked up a 9" hi-res b&w monitor at the TRW swapmeet last weekend. ($10.00) After I re-soldered the voltage regulator to the circuit board it works like a champ! Something else I picked up at the swapmeet that I've not seen before is a 0 to 20uA meter. ($3.00) Is that puppy ever sensitive!

My job requires me to work many weekends, so progress is slow. Hopefully things will mellow out at work and I'll be able to take a few days off here and there and make some real fusor progress.

Jon Rosenstiel

Re: Scintillator tests

Posted: Tue Apr 09, 2002 4:13 am
by guest
I was pleased to read the report on the comparison between the Bircron and the "TD" neutron detectors, especially how similar they are. It is definitely a vote of confidence for the use of polyester casting resin as a matrix. I was a liitle worried about the possibility of energy storage in the resin, but the gamma tests have put that fear to rest.
I would be interested to know the preamp used with the detectors, how many stages were in the PMT, and if any subsequent amplification was used. I can use this information to estimate the approximate currrent per pulse for pulses at the lower limit of the SCA setting. This information can then be used to set the gain for the preamp I'm getting ready to unleash on the world...

Re: Scintillator tests

Posted: Tue Apr 09, 2002 7:47 pm
by Richard Hull
The PMT was an 8 stage EMI bi-alkali tube with a mated dynode string and 3 transistor preamp on socket. (shown on my 2nd fusor video) It is not as good as the Hamamatsu preamp, but probably sports a gain of 10 or 20 and is mainly a low impedance matcher to the line out, (50 ohm). The average height of the pulse out of the preamp was between 20 and 40mv. This was amplified by the NIM spectroscopy amp set to a gain of about 50-65. (Output seen in channel #1 of image du jour section test image #3).

I had about 30 of these tubes and preamp/socket combos in 1997 and sold all of them to old list memebers and hamfesters by 1999. I then obtained about 15 of the Hamamatsu tubes and preamps and then sold all those to list members and others by late 2000. Scott little bought three sets and his scope pics of the preamp output can be seen on his report of fusion at

Tom or Jon or someone on this list recently picked up one of my old combos with schematics from someone at a hamfest. Other than for myself and Scott Little, I am unaware of anyone out of all those buyers who ever turned one of those nice setups into finished hardware.

Hope this helps...

Richard Hull

Re: Scintillator tests

Posted: Tue Apr 09, 2002 8:50 pm
by Tom Dressel
I was the one who picked up one of your EMI PMT-preamp combos on E-bay a few months ago. I have tested it and the NIM components with a LED input and it works. I just have to mount the scintillator in the PMT enclosure, set the discriminator levels, fire up the fusor and bleed in some deuterium. But I may have to build a beefed up power supply.

Tom Dressel