Gamma Spectrometry Forum

Hi All,
After a while of inactivity due to lack of time for the hobby, I need your comments and advices again.
In my last post I showed my first CS-137 spectrum with no shielding, now lets show the second attempt with what I lately constructed.
This is maybe for the Shielding forum, but I post it here since the spectra are the main topic.
The detector:

https://imgur.com/a/CacbX

The castle (a compromise between effectiveness, size, weight, price, efforts):

https://imgur.com/a/dIptX
https://imgur.com/a/TGeLA

3mm copper + 25mm lead rolled around 75mm plastic tube, same for the bottom, sealed with epoxy resin. The upper part could get another thinner stage, that's why I left it this way for now. Weight is 19.5 Kg. Reduces the background counts with a factor of 10 (from ~100 to ~10/s).

Here is what I got from a quite weak source of CS-137 - a spark gap tube. Please note the settings used, around 10 hours of recording, software gain boost 1, in inTune the peaks are at about 60-70 units, 1050V supply to the detector, calibration is roughly done only with this same probe, so can't be 100% trusted:
https://imgur.com/a/yaIAL

And here, about 4 hours, gain boost 2, 1100V (gave in 10-15 min a recognizable spectrum):
https://imgur.com/a/1kI5y

Since I am new and inexperienced any comments and advices for improvements are very welcome. Do I do it right at all? :)
I wanted to ask you also: What are the peaks after the 662Kv one? (around 980Kv and 1115Kv on the first picture and around 950KV on the second).
Should I better use the gain boost with this detector, as it gave an acceptable result faster, or the quality suffers? Which of the two measurements is better performed (not taking the longer time of course)? How do you find this performance overall? Thank you in advance, any comments will be really helpful since at the moment I can't really judge for myself.

Hi Svilen,

Nice detector and nice job on the shielding!

Your spectrum and resolution looks pretty good, but I too wonder about the bumps to the right of your Cs137 peak, very strange indeed.

Have you tried recording a spectrum in your shield without a source? maybe worth trying, in case you have bought some contaminated lead.

Regarding the boost gain, you should not need to change this when using a NaI(Tl) detector, it's something Marek added for proportional detectors with very weak pulses. For normal PMT type detectors you should have enough gain adjustment between the bias voltage and the computer line in volume. (Worst case some extra gain control on the GS-PRO circuit board trimmer potentiometer). Optimally your Cs137 peak should sit on or around 25 arb.u (arbitrary units).

Yes, increasing the gain by changing the boost gain will reduce accuracy.

Keep us posted on the mystery humps..

Hi Steve,
Thank you very much for your comment. I read in the PRA's manual that the gain boost is for the weak signal of other detectors, but just was a bit unsure in the signal strength of mine. I have just to play a bit more with the software options and hardware to find the optimal settings. Have to try an external sound card too, this integrated one is really a poor cheap stuff.
But the other problem now seems worse... Somebody wrote here "There is a raD in the castle" and I may have two..
Here are the spectra of the castle and the room:

https://imgur.com/a/kc16S
Castle with the rough calibration from the CS-containing tube

Same in logarithmic mode:
https://imgur.com/a/RswAb

With au scale:
https://imgur.com/a/IzE0X

Background in the room (without lead castle):
https://imgur.com/a/Vmu8Z

Same logarithmic:
https://imgur.com/a/njF1i

AU instead of the calibration:
https://imgur.com/a/LckuG

I really need your help analyzing this problem... Thanks in advance.

Svilen,

Contaminated lead is definitely a possibility, through scrap metal recycling, radioactive elements have made it into batches of lead, so it it quite common. I had to stop using a particular brand of solder after I discovered (purely by accident) it was radioactive. Now I am only using lead free solder.

Just to triple check you could do a background away from the shield and see if there are any signs of a peak there.

Steven

Just bad for the money wasted. Could be the copper maybe also..or could it be the epoxy?
So these two peaks definitely shouldn't be there or? And how you find the background spectrum from the room, out of the shield? There are also some small peaks of activity between 700 and 1000KeV (approximately, have to mention again), is this normal?

Svilen,

It is normal to see a peak at 1460 keV this is Potassium or K40 it exists in all organic matter including people, and it's almost impossible to eliminate it from your spectrum.

Do you think your linearity could be that far out, that you are seeing K40 at 960 keV ?

http://www.gammaspectacular.com/gamma_s ... a-spectrum

Note: Can you select the pulse height histogram and use the up arrow key to compress the spectrum, then we can see further out what you have there. maybe K40 becomes obvious.

I know this about the K40 and also thought it may be it, but just too far away. The calibration I could only make using the 32Kev X-rays peak and the 662Kev peak from the Cs-137. I selected linear instead of slope, which is better to use in this case? I have also radium watch hands and thorium mantle, but the energies there are even lower then 662Kev so not very useful for calibration when we look above 1400Kev.
There is nothing more to the right of the room spectrum, than what you see on the images above. And this is a compressed spectrum of the castle:
https://imgur.com/a/OaCtJ

Something isn't right, because if you just record background radiation with a 2" detector in a non contaminated area you should see K40 as an obvious bump at around 1460 keV.

Just saw this video from Tom (antiprotons), doesn't it look like he have the same problem with his shielding here?
https://imgur.com/a/1fBTs

https://www.youtube.com/watch?v=gcNzV0jWf7I

Hi Steven,
I made today some recording with Theremino and superimposed the graphs. It is easy to see that the peaks I have inside the castle are present also outside. Looks like it is not the lead, but more likely the environment or contamination on the surface of the detector. The second is maybe more possible since I recorded same peaks from different room, there are no (known) sources around and I guess even if it was the case, it wouldn't be that obvious shielded with 2.5 cm of lead then. On the other hand, the detector is carefully cleaned few times with a cloth damped with water + a bit of detergent, than just pure water, than alcohol. Any suggestions?

Brown is the background in the room
Blue is the spectrum inside the castle
Green is CS-137 spark gap tube for reference
https://imgur.com/a/OzqQk