Is this Detector Too Big, Too Old, or Just Normal?

[Mike S]

I bought a used 5"x4" detector with an end well, and it seems to at least partially work, but I've noticed some differences compared to the two other detectors I have (an old very used 2x2" and a recently acquired, but also used, 1.5x2.25").

I'm trying to determine if it's normal behavior for the size and end well configuration of the detector, if it's degrading due to age (water absorption?), or maybe something else is wrong. In comparison to my other detectors, this big one:

1. Needs a lot of voltage. It seems to operate best about 1050V
2. Seems to have a lot of noise if I go any higher than 1050V (a peak starts growing around 100 keV)
3. Needs a lot of gain. In Theremino MCA, my other detectors use a gain of 1. This one needs a gain of 4 to bring it up enough so the energy trimmer can be used effectively.
4. Doesn't seem to have much response at all at the low end. I'll include some comparison plots, but below 200 keV, there's significantly less response than the other detectors.

The detector does appear to be older, but I'm not sure how old it really is. It uses a 3" PMT even though it's 5" diameter, so there's a light collector of some sort. This is also the first time I've used the 2-wire input on the GS-USB-PRO.

With all that, there's a lot of variables going on here, so I'm wondering if someone has experience using a larger detector with a well and can share if some of what I'm seeing is inherent in the design or a symptom of some sort of degradation of the detector. I could potentially return it if it is no good, but it does seem to work well above 200 keV.

In the three attached plots, the green traces are with the 5x4" detector and the red traces are with my 2x2" detector. Both use some shielding.

Thanks in advance for any advice.

Mike S.

Ps. Sorry if I get mixed up and type 4x5" sometimes, that was a an old film format that I played with in the past, and it comes more natural than 5x4.

[iRad]

Hi Mike - Next thing to do is to take the PMT housing off of the crystal can (remove the 4 allen head machine screws). The PMT housing should lift up a little bit easy enough, but the PMT may be stuck to the window of the crystal can and might need some persuasion. If they don't come apart easy enough, I use a shot of compressed air to separate them. Be careful not to pull the wires off the end of the PMT while lifting the PMT and its housing away from the crystal. Then it should be easy enough to use a flashlight to inspect the crystal. You will probably have to clean the crystal's window of gunk to get a good peak. From your symptoms, I expect that it may be a bit wet, but a visual inspection is always a good idea. If the crystal is cracked, you'll want to know so you can base your determination to return it or not based on what you find.

If the crystal is in good shape, you may want to update the PMT with something newer and better quality.

Nice workbench...

[Mike S]

Hi Mike - Next thing to do is to take the PMT housing off of the crystal can (remove the 4 allen head machine screws). The PMT housing should lift up a little bit easy enough, but the PMT may be stuck to the window of the crystal can and might need some persuasion. If they don't come apart easy enough, I use a shot of compressed air to separate them. Be careful not to pull the wires off the end of the PMT while lifting the PMT and its housing away from the crystal. Then it should be easy enough to use a flashlight to inspect the crystal. You will probably have to clean the crystal's window of gunk to get a good peak. From your symptoms, I expect that it may be a bit wet, but a visual inspection is always a good idea. If the crystal is cracked, you'll want to know so you can base your determination to return it or not based on what you find.

If the crystal is in good shape, you may want to update the PMT with something newer and better quality.

Nice workbench...

The PMT seems to be bedded in substance that behaves like very clear RTV silicone. It flexes quite a bit when I put force on the PMT and has good adhesion. I stopped there, since I was expecting something with a consistency more like grease or hardened grease. Is this normal? It doesn't seem like it would be reversible if I removed the PMT.

Thanks on the bench. I've had multiple incarnations over the last 15 years or so and this time I decided I was going all-out to make it just like I wanted. Maybe if I ever get it cleaned off, I'll take pictures and post in the Lounge section.

[iRad]

Yes, what you are seeing is normal, as the stuff can get very thick over time. Although if you are careful, and the coupling compound is still clean and pliable, you could save and reuse it, but I would recommend replacing it. You'll want to inspect that crystal, and the only way to do that is to separate the PMT get that old gunk out of the way...
Clean workbench... that's nearly an oxymoron?

[Mike S]

Yes, what you are seeing is normal, as the stuff can get very thick over time. Although if you are careful, and the coupling compound is still clean and pliable, you could save and reuse it, but I would recommend replacing it. You'll want to inspect that crystal, and the only way to do that is to separate the PMT get that old gunk out of the way...
Clean workbench... that's nearly an oxymoron?

That stuff was like clarified boogers, but I finally got the PMT off and the bulk of the coupling compound cleaned off the light pipe (which is permanently mounted to the crystal).

Any tips on how to get the residue off the light pipe without damage? My interpretation of the part number indicates the light pipe is undoped NaI, but I'm not sure about the actual surface.

I can't say for sure if it's yellow in there. I'm hoping I get a better view once the residue is removed.

[iRad]

The window should be made of quartz, with the doped crystal and undoped light pipe all safely inside the can, underneath and sealed to the window (hopefully not detached). I use alcohol and plenty of tissues to clean the window surface. From the pictures, it looks a bit cloudy and yellow, but may be hard to tell until better cleaning of the window.

[luuk]

Hi Mike,
Nice big detector you got there!
This detector has a stainless steel housing so it will be difficult to detect low energies they will be strongly attenuated..
It looks like the crystal is “brown” but that could be just the picture that is difficult to see.
Clean the surface with some alcohol so you can look inside, use a small strong flashlight to inspect the crystal, if you see darker spots inside than you can expect the crystal is wet.
If the complete crystal is one color so no spots than you may assume that it is not wet, also look at the interface between the glass window and the crystal that too should look the same all over, if you see lighter parts the crystal and glass are not connected anymore.
send some more pictures after cleaning if possible
Luuk

[Mike S]

Thanks for the input guys!

It's not looking too good for this crystal...

The light pipe looks clear, but things get very milky at the actual crystal.

In addition, it appears there a large separation between the light pipe and the crystal (about 15% of the area). This could be a crack but I don't think so based on the way the light was responding.

Note that the picture directly below is the closest to the actual color.

In addition, while it's very hard to see, it looks like there is an actual crack in the crystal on the the other side (across from the suspected area of separation). The pictures don't show it very well.

Thoughts?

--Mike

[luuk]

Hi Mike,
Don't worry about the crack that does not have to be a problem, I have repaired detectors with big and many cracks that still work great.
The color is difficult to see on the picture but it looks not so bad to be honest.
The interface between the crystal and glass looks good to me for an old detector.
Maybe a good pmt would help to get just a little bit better resolution and you probably will need less high voltage then.
In my opinion the crystal is still usable do you made a spectrum with Cs-137? or you can also use K-40 or Th-228 if so please post the measured resolutions, to made a better a diagnostic.
Luuk

[Mike S]

Hi Mike,
Don't worry about the crack that does not have to be a problem, I have repaired detectors with big and many cracks that still work great.
The color is difficult to see on the picture but it looks not so bad to be honest.
The interface between the crystal and glass looks good to me for an old detector.
Maybe a good pmt would help to get just a little bit better resolution and you probably will need less high voltage then.
In my opinion the crystal is still usable do you made a spectrum with Cs-137? or you can also use K-40 or Th-228 if so please post the measured resolutions, to made a better a diagnostic.
Luuk

Here's some additional plots:

This was not a pure Cs-137 source because a nominally 1uCi source generates a very high cps when put in the well of the detector. This is from a Trinitite sample with a strong Cs signature. This does use background subtraction (it says no background subtraction in the notes, but the notes are wrong).

****

Here's a plot showing the response to voltage. It's a little messy because I'm showing 5 plots at once, but you can see how the low end noise (~80 keV) becomes more prominent with higher voltage. I would have expected noise to have a broader impact instead of looking like a spike. Is it possible that the higher voltage allows the lead x-rays to be detected more? The shielding is mostly lead wrapped around the detector with only limited copper and pewter on the top. I would expect to have a large lead peak, but because the poor low energy response at 1050V, it doesn't really show up in my normal sample spectra.

Note that this is plotted at a fixed counts per second per bin, with no change in energy scaling, so the higher voltage causes fewer counts per bin because the pulses are spread over more bins. You can see how the relationship between the low end noise (~80 keV) and the k-40 peak changes while the relative height of the background peak (~250 keV) and K-40 peak remains relatively constant.

Here's a lower voltage comparison using a very week source (28 cpm net with a pancake probe, ~6 cps in the 5x4 detector well). At 950V, the response below 300 drops off sharply.