building a scintillator

This area is for discussions involving any fusion related radiation metrology issues. Neutrons are the key signature of fusion, but other radiations are of interest to the amateur fusioneer as well.
Q
Posts: 200
Joined: Mon Jan 12, 2004 5:45 pm
Real name:

building a scintillator

Post by Q »

i've searched the archives and come up with squat, though i'm sure i've seen this question before.
anyway, in making my scintillator i've come across a couple of problems that i'm trying to figure out. i'll get to those in a moment.

but first, as i'm working on the detection portion of this thing, i've connected the pmt (through the blocking cap) to my trusty (crusty) old vectorscope. i can see the signal coming from the pmt, but the pulses are of varying heights. my check source is an ore sample. is this differance in pulse height due to multiple events happening simultaniously or is it related to the energy of the incident gamma?

next, in making my power supply, i cant seem to get a clean output. i've tried filtering it with various bypass caps, but with little change. out of despiration, i've also made a notch filter with the notch frequency to match the operational frequency of the power supply. it helped, but not much. any other ideas?

thirdly, on most scintillation detectors, is the anode resistor usually at the pmt end or the detector end of the system. in other words, if i plan to use the pmt assymbly with other detectors, does the anode resistor need to be integrated into the pmt housing or is it normally part of the power supply/counter circuit? i would assume that due to the variations in pmt's and differance in dynode srting resistances the anode resistor should be part of the pmt asymbly. is this the case?

some basic questions from your basic Q
Richard Hester
Posts: 1519
Joined: Thu Jul 05, 2001 12:07 am
Real name:

Re: builting a scintillator

Post by Richard Hester »

Gamma rays of varying energies will give varying pulse heights.

To clean up your power supply a bit, you can try running it through a couple of cascaded RC filters. For high frequency ripple (switching power supply), I'd try something like 100k and 0.01uF.

The anode resistor and output blocking/coupling cap should most definitely be at the PMT end of the system. You really don't want to have to deal with a BNC signal cable charged up to high voltage - they bite. This is assumng you have a separate connector for HV and output signal, which is the case for most PMT base assemblies.
Q
Posts: 200
Joined: Mon Jan 12, 2004 5:45 pm
Real name:

Re: builting a scintillator

Post by Q »

thank you, richard!

soon, i plan to post some pictures of my progress in the images section. this project is coming along quite nicely- even though it's at a snail's pace.

Q
Jon Rosenstiel
Posts: 1494
Joined: Thu Jun 28, 2001 1:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: builting a scintillator

Post by Jon Rosenstiel »

Q,

Just in case you haven't run across this site...

http://home.austin.rr.com/cthompson15/R ... Page2.html

Jon Rosenstiel
Q
Posts: 200
Joined: Mon Jan 12, 2004 5:45 pm
Real name:

Re: builting a scintillator

Post by Q »

thanks jon,
that's a great site. actually, the ccfl inverter based hv supply is what i'm using. it works well, but i cant get a very clean output from it. richard's cascade rc filter suggestion seems to be working well though.

Q
Q
Posts: 200
Joined: Mon Jan 12, 2004 5:45 pm
Real name:

Re: building a scintillator

Post by Q »

ever have one of those "duh!" moments where you realize you've beaten that dead horse into pulp for no good reason?
well, my hv supply for the scintillator project still puts out some ripple, but you know what? it doesn't matter! the reason i was having problems earlier is because my ripple was larger than the signal pulses coming from the pmt. i was trying everything that i could to clean up the 0.5 mv (p-p) noise. then it hit me... use a bigger anode resistor!!
now, the pmt signal comes in clear and nearly "drowns out" the power supply noise.
whew! now i can get on with this thing.

on a different note, i'm seeing some interesting pulses on the o'scope. at a somewhat regular interval (about once every 10 seconds or so) there is a massive pulse (at least 4x as large) that shows up. since i'm using a peice of u ore as my check source, is it possible that i am picking up a spontanious fission event? the scintillation material is general purpose plastic, 3 inches thick. would that be enough to catch the occational neutron? it just seems like a awful lot of events to be neutrons...

Q
Jon Rosenstiel
Posts: 1494
Joined: Thu Jun 28, 2001 1:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: building a scintillator

Post by Jon Rosenstiel »

I would guess you're seeing cosmic ray muons, they produce huge pulses. As I recall, the average muon flux at sea level is one muon/cm^2/minute.

Jon Rosenstiel
Q
Posts: 200
Joined: Mon Jan 12, 2004 5:45 pm
Real name:

Re: building a scintillator

Post by Q »

hmmm, interesting. i had forgotten about cosmic sources... (another "duh" moment)
well, i am at approximately 2000 feet above sea level and the detector is 3 inches square. that might account for it, especially since the much larger pulses seem to be constant wether the check source is present or not. (i hadn't thought to try removing the source to see if that made a differance until this morning)

i was under the impression that muons were difficult to detect though.

Q
Jon Rosenstiel
Posts: 1494
Joined: Thu Jun 28, 2001 1:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: building a scintillator

Post by Jon Rosenstiel »

Muons pack quite the punch, about 200 times the mass of an electron and with an average energy (at sea level) of around 2 geV! They really tear up the scintillator.

Jon Rosenstiel
AnGuy
Posts: 208
Joined: Tue Jul 20, 2004 7:49 pm
Real name:

Re: building a scintillator

Post by AnGuy »

>i'm seeing some interesting pulses on the o'scope. at a somewhat regular interval (about once every 10 seconds or so) there is a massive pulse (at least 4x as large) that shows up.

Did you check to see if there is any AC devices that might be kicking on which would be causing these periodic pulses? A high current device such as a refridgerator or a furnace could be triggering your detector. Muons or other external sources would not occur at regular intervals. They would occur at random intervals.

BTW: The best way I found to reduce noise in a DC supply is with a coupled inductor filter. This device is like a Pi filter but with an extra low resistance coil wound on the inductor core that is attached to a capacitor. See Link:
http://www.orc.ru/~igorg/SMPS/zero_rip. ... %20Feature

You can order Pot cores from this site if you wish to roll your own:
http://www.cwsbytemark.com/

A source for very low noise HV power supplies is emco:
http://www.emcohighvoltage.com/
Their CA series devices are designed for PMTs and they will cost about $250-$300 per unit.

Hope this helps!
Q
Posts: 200
Joined: Mon Jan 12, 2004 5:45 pm
Real name:

Re: building a scintillator

Post by Q »

thanks anonymous,
i have checked to see if there was anything around that might cause these pulses, so far, i've come up with nothing. the timing isn't exactly 10 seconds apart, but roughly some where between 6 and 15 seconds, sometimes as often as often as 2 in one second, but this is rare. the 10 second figure was a rough average.

as for the power supply, i have contacted emco, but they will only sell to oem's, government, or educational organizations... since i'm none of those, they refused to communicate with me beyond "try searching the internet for high voltage power supplies"- and they weren't very nice about it either. oh well, not to worry. i solved the issue. ebay also came through for me, giving me another supply for the next pmt project!

Q
AnGuy
Posts: 208
Joined: Tue Jul 20, 2004 7:49 pm
Real name:

Re: building a scintillator

Post by AnGuy »

>have contacted emco, but they will only sell to oem's, government, or educational organizations..

Just them that your an engineer developing or repairing a detector, and give them a company name (you can always make up one) and PO number. They did not ask for a Federal Employer ID number. I placed a order for three modules about 5 or 6 months ago. The sales rep asked me what they will be used for, I told her that they were for a detector. I think their minium order is 2 units.
AnGuy
Posts: 208
Joined: Tue Jul 20, 2004 7:49 pm
Real name:

Re: building a scintillator

Post by AnGuy »

Here is another supplier of HV power supplies for detectors
http://www.picoelectronics.com/dcdclow/peHVP.htm
AnGuy
Posts: 208
Joined: Tue Jul 20, 2004 7:49 pm
Real name:

Re: building a scintillator

Post by AnGuy »

In order properly us an detector such as a PMT, and HE3 or a BF3 tube the noise level must be very low > 0.005% The power supplies from Geiger counters operate with noise levels above 1%.

Its pretty difficult to make a low noise power supply. Basically you need to build DC-DC converter that operates at a very high frequency ( > 1 Mhz) so that you don't need giant caps and inductors to filter out the switching noise. The HV power supplies from EMCO are also very small, about 1" by 2" by 0.75" and are shielded to prevent EMI.
Richard Hester
Posts: 1519
Joined: Thu Jul 05, 2001 12:07 am
Real name:

Re: building a scintillator

Post by Richard Hester »

You don't necessarily need extremely high frequency operation for low noise. Sometimes this can get in the way. A proper regulator design is important. In the files section, you will find a schematic for a HV supply based on a CCFL inverter, a pretty dirty source. With some shielding and a good regulator, I got 40 millivolts ripple for 900V out. This is 4.4 X 10^-3 %. If I set up the dropping resistors on the shunt regulator for lower quiescent current, no doubt I would get lower ripple, with lower power consumption to boot.
AnGuy
Posts: 208
Joined: Tue Jul 20, 2004 7:49 pm
Real name:

Re: building a scintillator

Post by AnGuy »

>I got 40 millivolts ripple for 900V out. This is 4.4 X 10^-3 %. If I set up the dropping resistors on the shunt regulator for lower quiescent current, no doubt I would get lower ripple, with lower power consumption to boot.

I don't believe 40 millivolts isn't too hard, its getting it below 4 millivolts that becomes the challege. The EMCO supplies have ripple below 1 mv. You may also need to consider noise between your input supply connections, since there might be common mode noise between the two HV terminals, that isn't evident until you check against your DC inputs. Also what happens when you add a load? Does the noise level remain at 40 mv?
Richard Hester
Posts: 1519
Joined: Thu Jul 05, 2001 12:07 am
Real name:

Re: building a scintillator

Post by Richard Hester »

The supply I cited uses an output shunt regulator, so it is always loaded at a constant value, whether or not anything is attached to the output terminals. Also, if I remember correctly, I used an unmodified scope probe to measure ripple, so some of that ripple shown could have been pickup, always a problem when measuring switchers. Not too bad, anyway, for a 5 dollar investment for the inverter and some parts from the junk box.
Next time around, I'll pop off the end cap and ground lead from the probe, take my life in my hands, and try to measure ripple with a real short loop area. At any rate, the supply shown powered the PMT on a plastic scintillator fast neutron detector/SCA with no problems (yes, it detected neutrons, too). These days, I would use MPSW42 transistors instead of the TIP50s (higher gain, less space consumed), and something like 100-150k instead of 68k for the shunt regulator dropping resistors. One of these days, I'll try and post another circuit with less power consumption. If I did a blocking oscilllator with a hand-wound transformer, I don't doubt I could do better, but I wanted to post a circuit where people could just go and buy the inverter, as not everyone is set up to do their own transformers.
BTW, there's no real reason to be putting lots of money in the hands of folks like Pico or Emco unless you have it to burn. The power supplies used in the standard REM ball detectors aren't all that sophisticated. I have one of the Ludlum REM balls and a "Snoopy", both with schematics, so I know first hand. The manual for the Snoopy is posted on this site in "Files", so you can have a look and see what I'm talking about.
User avatar
Richard Hull
Moderator
Posts: 14991
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: building a scintillator

Post by Richard Hull »

I guess I am just too used to NIM steups for neutron instrumentation. I still retain but rarely use any of my portable neutron detectors anymore. I haven't used a scintillation based neutron counter since 2000 and haven't worked with my venerable Eberline rem ball since 2002. I did use my snoopy and my PNC-1 in 2004, but now I only have one portable (PNC-1) with batteries even in it! All the other systems are in storage with their batteries removed.

Once you use a stable nim supply and nim modules including a good preamp to cob up a neutron counter around a nice sized He3 tube, you will just never use much of any thing else,.... as anything else will be inferior.

Likewise, once you use a BTI bubble detector you will laugh at terawatt pulse noise in the vacinity of you detector and not be concerned about 100 rad gamma or X-ray fields giving errors in your neutron count either. They won't even make a single bubble.

For me, at this stage of the game, there are only two real neutron counters for the amateur encountering low level, fast neutron counting problems. This, inspite of my have invested about $3,000 in "other" neutron counting systems and efforts since 1997.

The noise in any NIM supply, regardless of type, (and I have used about 8 different), is not enough to impact an He3 tube count rate. You will receive far more heartache keeping noise out of the system from the electrically noisy fusor environment rather than a silly ole supply.

Old post of my current detector.....
viewtopic.php?f=13&t=5528#p33972



Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
AnGuy
Posts: 208
Joined: Tue Jul 20, 2004 7:49 pm
Real name:

Re: building a scintillator

Post by AnGuy »

>The supply I cited uses an output shunt regulator, so it is always loaded at a constant value, whether or not anything is attached to the output terminals

What are you using as a Shunt regulator? I suspect that 900V is close to the max operating range. Can you operate it voltages near 2kv? For instance the BF3 tube I have requires an operating voltage near 1900 volts, the HE3 tube I have will operate at about 1600 volts.

> any rate, the supply shown powered the PMT on a plastic scintillator fast neutron detector/SCA with no problems (yes, it detected neutrons, too).

I don't think ripple is much of an issue if all you are doing is counting. As long as the ripple level is below the pulse level of the detector it can be managed with voltage discriminator or SCA. However I believe for MCA is a different story since the pulse height will be a summation of the ripple and the detector output, which will distort the results.

I always thought PMT scintillator detectors where even more sensitive to ripple than BF3/HE3 detectors but I've never worked with them.
AnGuy
Posts: 208
Joined: Tue Jul 20, 2004 7:49 pm
Real name:

Re: building a scintillator

Post by AnGuy »

>The noise in any NIM supply, regardless of type, (and I have used about 8 different), is not enough to impact an He3 tube count rate

What is the average noise level of your NIM supplies? I assume that your using just a SCA instead of an MCA?

>You will receive far more heartache keeping noise out of the system from the electrically noisy fusor environment rather than a silly ole supply.

With the ripple & EMI filtering on the outputs of my DC power supply I have been able to eliminate spikes created by power cycles of large appliances (refrig, Washer, etc). I don't have my fusor in operation yet, but I am hoping that the spikes generated from the HV supply will also be suppressed. (I know I am probably overly optimistic!)
AnGuy
Posts: 208
Joined: Tue Jul 20, 2004 7:49 pm
Real name:

Re: building a scintillator

Post by AnGuy »

>I'm guessing that in this hobby of neutron detection, most Home Labs are using discrete preamps and NIM modules to make the measurements?

FWIW: I am working on a all in one, including High voltage PS (EMCO), Charge-Sensitive preamp, JFet Amp, SCA/Discrimator, TTL count output, ADC with USB data output to feed a PC based MCA. Everything except for the PC and the AC-DC PS will go onto a single PCB.

>Gas proportional probes (BF3, He3) require even lower threshold settings and not all meters can accommodate those. Settings as low as 1 or 2 mV are called for, very difficult to achieve and retain stability.

As I recall, the BF Tube I have outputs in the 50mv with gammas in 3-8 mv range. the HE3 Tube outputs in the 100+ mv range with gammas in the 20mv to 70 mv range. My numbers might be a bit off a bit since I haven't worked with these tubes in a couple of months.
Richard Hester
Posts: 1519
Joined: Thu Jul 05, 2001 12:07 am
Real name:

Re: building a scintillator

Post by Richard Hester »

The shunt regulator is shown in the schematic I posted in files. The regulator uses the doubled output of a small CCFL inverter, about 1200 VDC. Since I wanted 900-1000V to run a 8 or 10 stage PMT, a doubler was sufficient for the job. Higher voltages would require a tripler or quadrupler.

For neutron scintllators, BF3, and He3 detectors (especially the latter two, as a moderator is used), the individual energy information from a fast neutron is lost, so counting schemes work just fine. An MCA is wasted for these types of detectors, except for diagnosing detector operation. It is possible using software to recover the average neutron energy information from a plastic scintillator excited by fast neutrons, but individual recoil events have a random energy distribution up to the max incident energy.
For higher output voltages to feed a proportional neutron detector, I would possibly think of closing the loop around the CCFL inverter and using a buck converter to modulate the average input current for the inverter input inductor. I came up with what should be a low power approach using a CMOS gate package and a couple of small MOSFETs this afternoon, which should work.This approach gets around using the big string of series transistors required by a shunt regulator. A simple Neanderthal approach would be to score the appropriate voltage Victoreen Corotron regulator tube on EBay and use that as a shunt regulator. With a stack of RC filters in front of the Corotron, the output would be pretty quiet. I have also seen stacks of small HV transistors operated in the breakdown mode as zeners.
Richard Hester
Posts: 1519
Joined: Thu Jul 05, 2001 12:07 am
Real name:

Re: building a scintillator

Post by Richard Hester »

A string of 100V zeners will work, but given the usual current you will want to pull in the regulator string, thay will be running "starved", so that regulation will not be all that tight,. You would definitely want to be using 500mW or smaller units that do not require as much current to get them up past the "knee" in their operating curve. Using HV transistors in breakdown mode may be better - That's what the design engineers used in the "Snoopy" HV supply. The manual and schematics are in "Files" here.
The voltage and current available from a CCFL inverter will of course vary from unit to unit. I'm using the small TDK 5V input, single-output inverters available from Electronic Goldmine and elsewhere for a pittance, and I'm leaving the output current limiting capacitor untouched.
It is possible to use a TL431 and pass transistor at the front end of a CCFL inverter, though it won't be as efficient as a switching approach. Also, there will be a limited range of operation set by the point at which the CCFL oscillator quits due to insufficient voltage. Modulating the average current in the input inductor is a more efficient and flexible approach, though a more complex. It takes a 1A Schottky rectifier like the 1N5818, a CD4093 quad Schmitt trigger, a small signal MOSFET (a 2N7000 may do the trick), and a larger device like an IRFD110 or a MPF960. Il'll post a circuit when I have the time to build it up and verify functionality. If someone wants to beat me to the punch, they are quite welcome to do so...
Richard Hester
Posts: 1519
Joined: Thu Jul 05, 2001 12:07 am
Real name:

Re: building a scintillator

Post by Richard Hester »

I haven't been all that concerned about input draw for my portable stuff, as I've been using a large linear wall wart with +5V and
+/- 15V outputs. If I were depending on batteries, I'd be singing another tune. The blocking oscillator should be the way to go for portable stuff operating on batteries. All the schematics I've seen for battery operated instruments use some variation on that theme. One alternative to a hand wound transformer would be to use a filament transformer backwards as a blocking oscillator - not the smallest solution, but still fairly readily available. I show a circuit in Files for a simple blocking oscillator. One could also use a 555 in astable mode with a bipolar or low threshold MOSFET.
Richard Hester
Posts: 1519
Joined: Thu Jul 05, 2001 12:07 am
Real name:

Re: building a scintillator

Post by Richard Hester »

One more thing I forgot to mention - when you run zeners in the "starved" mode (low bias current), they get really noisy. This is actually the basis for some noise generator circuits. I had a big output noise problem in an earlier version of my shunt regulator circuit that I traced to a 15V zener I was using in the bias string. It was running at only a couple of hundred microamps, and it didn't like it one bit. Replacing the zener with a resistor cleared up the noise right away.
This is no real concern for geiger aplications, but could be a real problem for more sensitive stuff.
Post Reply

Return to “Neutrons, Radiation, and Detection (& FAQs)”