Plateauing a 44-2 with Cs-137

Scintillation crystals, PMTs, voltage dividers etc...
Post Reply
glhansen
Posts: 19
Joined: 05 Feb 2022, 08:50
Contact:

Plateauing a 44-2 with Cs-137

Post by glhansen » 07 Feb 2022, 01:53

I'm new here. Hello. One of my duties in a nuclear pharmacy is to calibrate survey meters, and sometimes I have questions about them, and nobody local to turn to. You people seem to know exciting things about that stuff.

My reference point for scintillators is the Ludlum 44-2. So, NaI, 1" diameter by 1" thick. It is standard to plateau it with Am-241, and that graph shows a basically null count rate at low voltages, a steep rise to a plateau, and then a high-voltage increase. I wondered why we weren't plateauing it with Cs-137, since we have so many check sources in that environment, so I tried it. I didn't see a flat plateau or a low-voltage cut-off, just a sloped line. Although Cs-137 is much spicier, I thought it couldn't be exciting more energetic photons in the scintillator, I understand them to have a characteristic spectrum. And I didn't think a higher quantity of photons could explain it because quantum mechanics -- if one photon doesn't eject a photoelectron, then ten photons won't, either. So I thought the gammas were going right through the crystal and ejecting photoelectrons in the PMT itself. But I tried hiding the probe behind a lead brick with about half the crystal exposed, and ran another plateau, and got the same sloped line (with a lower count rate; I used a stronger check source but it was a few inches away). Thinking back, I didn't take the voltage all the way down to zero, just to the region where the Am-241 dropped off.

My question is, what's going on in there? How to explain what I saw?
Greg Hansen

User avatar
Sesselmann
Posts: 928
Joined: 27 Apr 2015, 11:40
Location: Sydney
Contact:

Re: Plateauing a 44-2 with Cs-137

Post by Sesselmann » 07 Feb 2022, 10:38

Greg,

Thats an interesting observation and I don't know the immadiate answer. I guess it would be easy to test a PMT without a crystal to see if there is any detectable counts from Cs137 just from the dynode. I would have thought the photons from NaI(Tl) with 415 nm wavelength would have a constant effect on the dynode regardless or source.

Can anyone else answer greg's question?

glhansen
Posts: 19
Joined: 05 Feb 2022, 08:50
Contact:

Re: Plateauing a 44-2 with Cs-137

Post by glhansen » 08 Feb 2022, 04:24

Thanks, Steven. I was hoping someone would immediately know what happened, or tell me what I did wrong. I figure there's probably a reason that Am-241 is what the manufacturer recommends, just want to understand it a little better.

If I have time and opportunity one day, maybe I'll rummage around in the sources drawer and see how many isotopes I can try.
Greg Hansen

User avatar
Sesselmann
Posts: 928
Joined: 27 Apr 2015, 11:40
Location: Sydney
Contact:

Re: Plateauing a 44-2 with Cs-137

Post by Sesselmann » 09 Feb 2022, 08:10

Greg,

For what it's worth, I tested an old PMT without a crystal, over a range of voltages, with and without a source, what I saw was little or no counts up top around 850 Volts and then a rapid increase in the self count rate over this voltage. There was very little difference in the count rate when I introduced a Cs137 source.

Maybe your Cs137 source is too spicy, try the same plateau test with your source further away, maybe you will get a different result.

Remember the 1" crystal is 100% efficient to 59 keV gamma and only 50% efficient at 662 keV.

Steven

glhansen
Posts: 19
Joined: 05 Feb 2022, 08:50
Contact:

Re: Plateauing a 44-2 with Cs-137

Post by glhansen » 10 Feb 2022, 04:26

Screenshot plateaus.png
Screenshot plateaus.png (35.88 KiB) Viewed 131 times
That seems like interesting and useful information.

It occurs to me that I could have shared my data. Let's see if I can figure that out. 44-2 on a 2241-3, six-second counts. Looking at the red line, I could almost believe there would be a knee there if I had taken the voltage lower. But the Am-241 had cut off, and I figured at the time that cut-off was cut-off. The greenish line, "Hidden", is multiplied by a factor of 10 to scale it up, otherwise it would be laying on the bottom.

Your last comment triggered a question about the 44-3, which I actually haven't been working with. Made for low-energy measurements. The metallized polyester window makes sense. But that 1mm thick wafer of scintillator... My understanding is that to detect a gamma you have to stop it, and the high-energy stuff is more likely to go right through. I'm looking at the product manual from ludlums.com, it looks like the efficiency drops off a lot faster at higher energies than it does for the 44-2, but it must be thick enough to stop ~60 keV radiation.

It is standard, according to Ludlum and others I've talked to, to plateau the 44-3 with I-129. That's around 40 keV, isn't? That makes sense. I thought it was much higher, but I poked around to check. That's a little harder to find on the search engine, it's not as popular as some of the other isotopes, like I-131. I had wondered if it could make a difference because, I had thought, a plateau is a plateau, and when the PMT stops counting it has stopped counting. But after the Cs-137 I'm not sure.
Greg Hansen

glhansen
Posts: 19
Joined: 05 Feb 2022, 08:50
Contact:

Re: Plateauing a 44-2 with Cs-137

Post by glhansen » 10 Feb 2022, 04:28

What about thresholds? Whenever I've worked on a 2241 the threshold has always been around 34 mV on every channel. But if you're hitting the scintillator with gammas of ten times the energy, you would get ten times as many photons hitting the PMT and ten times the current?
Greg Hansen

User avatar
Sesselmann
Posts: 928
Joined: 27 Apr 2015, 11:40
Location: Sydney
Contact:

Re: Plateauing a 44-2 with Cs-137

Post by Sesselmann » 11 Feb 2022, 14:05

Greg,

The pulse height threshold is usually set to a level just above electronic noise. All amplifiers and especially radiation detectors with inverters will have some kind of ripple on the baseline and for obvious reasons you don't want to count the ripples coming from the inverter.

Old school survey meters with analogue dials convert pulses to a small current which then drives the panel meter, but digital circuits work differently and count the number of pulses, but the result should be the same.

Pretty sure the discrepancy you observe has something to do with the efficiency of the crystal.
Attachments
Efficiency-Calculations.pdf
(1.54 MiB) Downloaded 17 times

Mike S
Posts: 73
Joined: 25 Apr 2020, 10:24
Contact:

Re: Plateauing a 44-2 with Cs-137

Post by Mike S » 06 Jun 2022, 04:21

glhansen wrote:
10 Feb 2022, 04:28
What about thresholds? Whenever I've worked on a 2241 the threshold has always been around 34 mV on every channel. But if you're hitting the scintillator with gammas of ten times the energy, you would get ten times as many photons hitting the PMT and ten times the current?
Greg,

You asked this a few months back so I'm not sure if you got this sorted out, but I'll give you my theory.

I think if you lowered the voltage further you would see the roll-off for Cs-137. My theory is that because you are using a fixed discriminator/threshold but the pulse height for Cs-137 is about 10 times higher that AM-241, just below the knee for Am-241 you would still have plenty of pulse height with Cs-137 to be over your threshold. I believe the reason the HV is typically set using Am-241 is specifically because of its low energy gammas are a good way to ensure the detector voltage is high enough so those (lower) pulses are high enough to be over the threshold and be counted. If you set your HV using Cs-137, then lower energy gammas might not generate pulses high enough to be counted.

It seems pretty easy to test, just run the basic HV curve with Cs-137 but take the voltage lower than you have before. I looked through my data and didn't immediately find any examples to share where I ran curves with different sources.

Mike S.
Mike Sullivan
Central Coast of California, USA

Post Reply

Who is online

Users browsing this forum: No registered users and 1 guest