also new here

All new members, your first post should be an introduction here;
Forum rules
Welcome, these forums are for professional and amateur scientists to discuss matters relating to radiation detection, and to facilitate friendly and honest collaboration. It is a requirement for all members to register using their real names and location. If your name has not already been used, feel free to use it as a login handle, otherwise add your name and location to the bottom of your posts as a default signature. Thanks for cooperating.

BUSINESS REGISTRATIONS
If you have a rad related business and want to post information about your products you may do so in the trading post, providing you first introduce your business formally in the introductions forum and add your business details in the signature sections so it is clear to everyone that your post has a commercial purpose.
User avatar
Sesselmann
Posts: 1142
Joined: 27 Apr 2015, 11:40
Location: Sydney
Contact:

Re: also new here

Post by Sesselmann » 05 Dec 2016, 14:26

Kolja,

Nice first try...

Settings look okay at first glance, and I can see a little pulse in the "Audio input" window, which looks quite well formed.

What stands out as a problem is the large peak on the left in your Pulse height histogram, this is most likely noise in your system being larger than normal (<0.5 arb.u) for some reason.

If you saved this spectrum, try opening it up and activating the 'Advanced pulse filter', then discriminate the pulse height so the peak on the left disappears. I suspect your spectrum will suddenly appear.

The most likely reason for the noise threshold to extend out to several arbitrary units is that your volume setting is too high, so you are amplifying the ripple too much. Try changing the combination of voltage and volume.

Steven

keuleeule
Posts: 67
Joined: 31 Oct 2016, 09:09
Location: Hamburg/Germany
Contact:

Re: also new here

Post by keuleeule » 06 Dec 2016, 13:21

Ok, playing right now with the PRA and the detector...
you are right, there are changes possible.

So first I will state my observations.
connected detector to GS2000, 900V, noise is audible, a white noise.
@1300V - 10 pulses will be detected each second

If I put my Cs137 source to the detector, the background noise increases rapidly, but the sharp "clicks" which are only counted within the "pulse view vs time" as total pulses and filtered pulses just occour randomly at 900V lets say a pulse in 2 seconds.
The audio input show one pulse triggered in the middle of the window, just half that size then yesterday.

If I increase the voltage to 1100V, sound is just louder, but maybe 1,5 pulses each second is detected.
"count rate" window show up and downs during 100sec of measurement - so no stable countrate.

increasing the voltage to 1200V brings approx 50pulses each second, "audio input" window changes to several not synced or triggered pulses.
1300V will count more than 300 pulses each second, still multiple audio peaks - seems at least one stable pulse triggered. stable count rate

1400V counts nearly 600 pulses a second. stable countrate, audio window show multiple peaks completely unsynced not triggered in the middle. stable count rate

1500V - the audio sounds overloaded. a little more than 620 counts are detected in a second. Audio window show sharper peaks, double that hight from the beginning and still not synced.
Sounds 1500V is too much.

The blue was done in PULSE SHAPE ACQUISITION


I keep it on 1300, perform the pulse shape acquisition followed by a data acquisition.
OK, that will take time now... 64min for the milion pulses -much longer than before.
I do have a different spectrum now, playing around with the advanced filter brings it not in a better shape lets say,
The picture show the last value until it changes its behaviour, just to see where the edge is.

There is no low keV visible :(

choosing other nuclids...
Ra226 show indeed now a peak at approx. 380keV and a big peak at 610 maybe. These values dosnt fit :)

to resume my observations,
I can see in the window "counting rate vs time" that I do have a source placed infront of my detector.
This count rate should be quite strait as my source dont move. and as long I do not have nuclids with a short halflife.
This window just counts without corelation to the energy of the nuclid, except that a detector should have a more or less good or bad resolution to a defined energy I think.
If I play around with the applied voltage this countrate will in or decrease

The pulse view vs time/s
just tell me how many pulses were detected in which time.
So far the total pulses are the same as the filtered pulses

Both, the countrate and maybe the pulse counter should be relevant to identify the best operating voltage of my PMT.
Normally I know that I did that with logarythmic paper and choosing different voltages copy down the rate and adjust to 1/3 of the resulting plateau, as far I remember.

the audio input window needs to receive positive (or threshold needs to be reversed) negative pulses.
I really dont know how high they need to be and if it is necessary that these need to be triggered in the middle of the window.
during background measurement @1300V they are very clear stable in the middle with a nice peak.

the pulseheight histogram should read a large amount of low energy gamma as far the detector isnt shielded at all, as mine is in its own old cover, with approx. 5mm lead, it might be better then nothing, but I can´t see any keV lower then 100 lets say - and as far I got it that is unrealistic.
I was even not able to receive any pulses over 1800keV with backgound measurement. I do not have high energy isotopes which I could measure.
if I place Uranium infront of the detector, the energy goes not more than 600keV for example, even if the background showed me some peaks at 1800keV before... that I dont understand.

Well, thanks Steven that you gave me some advices, I really need to work on it.
decreasing the Voltage might get a better result now, as well as discreminating.
But I´m still sure that I need to learn to work with the equipment, would be nice to have a good tested detector as well.
Maybe I will search for one.
regards, Kolja
Attachments
second_Cs137@1400V.png
second_Cs137@1400V.png (113.99 KiB) Viewed 6626 times

User avatar
Sesselmann
Posts: 1142
Joined: 27 Apr 2015, 11:40
Location: Sydney
Contact:

Re: also new here

Post by Sesselmann » 07 Dec 2016, 13:33

Kolja,

Scintillation detectors with PMT's should have a voltage plateau, a range of optimum voltage. You can find it by using the Count rate over time window.

Start at a low voltage (around 500V) and record the count rate in 10 second bins, increase the voltage by 50V avery minute and look for an interval where the counting rate is relatively flat. Then your best voltage setting is in the middle of this plateau.

Your Cs 137 spectrum doesn't look great, are you sure this is a NaI(Tl) detector?

Steven

keuleeule
Posts: 67
Joined: 31 Oct 2016, 09:09
Location: Hamburg/Germany
Contact:

Re: also new here

Post by keuleeule » 08 Dec 2016, 08:41

Steven Sesselmann wrote:Your Cs 137 spectrum doesn't look great, are you sure this is a NaI(Tl) detector?
No, I´m not sure. As I cant open the device. The PMT is in a black rough housing sealed in aluminum.
Luuk stated it could be as well a Harshaw detector. As it is in a portable lead shield it could also be just a plastic scint for a meter... but it comes from the 70ties, was plastic common there?
Would a bad crystal look like this spectrum?
or would a plastic scint "spectrum" look like this?

All I can do is to take some X-rays to it to maybe check it. As we have just medical X-ray, I´m in need to try 40kV at the minimum, maybe I can arrange tomorrow.
Kolja

User avatar
Sesselmann
Posts: 1142
Joined: 27 Apr 2015, 11:40
Location: Sydney
Contact:

Re: also new here

Post by Sesselmann » 08 Dec 2016, 09:58

Kolja,

Just by looking at your spectrum the FWHM (Full Width at Half Maximum) of your Cs137 peak is about 5 arbitrary units and the peak is at around 18, so this is 5/18 = 27% resolution. You can also see this directly by selecting the region of interest and a bin within the region.

A good NaI(Tl) detector should have a resolution better than 8%.

I can't say for sure, but it looks similar to spectra taken with plastic scintillator BC412

keuleeule
Posts: 67
Joined: 31 Oct 2016, 09:09
Location: Hamburg/Germany
Contact:

Re: also new here

Post by keuleeule » 10 Dec 2016, 03:04

Like promised, I just did some xrays on the fly of my detectors.
Here the results:

I cant see any cracks - but a molten Crystal would maybe look the same...
Attachments
4.jpg
4.jpg (161.37 KiB) Viewed 6600 times
3.jpg
3.jpg (276.11 KiB) Viewed 6600 times
2.jpg
2.jpg (320.93 KiB) Viewed 6600 times
1.jpg
1.jpg (232.12 KiB) Viewed 6600 times

keuleeule
Posts: 67
Joined: 31 Oct 2016, 09:09
Location: Hamburg/Germany
Contact:

Re: also new here

Post by keuleeule » 10 Dec 2016, 19:22

As both scints look the same under x-ray, and the large one is marked as NaI (Tl) i assume both are from same material.
I even would think it would be visible if the crystal starts to melt. Only possible case is that complete crystal is already molten. I even checked with lower kV but there was nothing different.
As my large one do not read anything in the under 1000V bias, I still think to adjust the divider again until calling it scrap.
Kolja

luuk
Posts: 308
Joined: 01 May 2015, 21:00
Location: the Netherlands
Contact:

Re: also new here

Post by luuk » 11 Dec 2016, 01:29

Hi Kolja,

Both are for 99% NaI(Tl) crystals the x-ray looks interesting and nice but you cannot see to much on it, the pmt is a from what I can see an old EMI if I look at the construction.
If the crystals are wet you can't see that on the x-ray even crack will be difficult to see, same for a bad coupling between crystal and pmt.
But you may expect that the pmt's are in very bad shape so the gain dropped dramatically because of the age and/or usage.

Luuk

keuleeule
Posts: 67
Joined: 31 Oct 2016, 09:09
Location: Hamburg/Germany
Contact:

Re: also new here

Post by keuleeule » 12 Dec 2016, 09:06

Ok Luuk, that is nice that you call both NaI(Tl)
You think the PMT´s are bad, so is there a possibility to open both ? I cant see any screw thread. Or is that only possible in your environment?
Maybe I will try to x-ray the crystals again from other directions and angles - just to check.

Thanks so far, Kolja

User avatar
Sesselmann
Posts: 1142
Joined: 27 Apr 2015, 11:40
Location: Sydney
Contact:

Re: also new here

Post by Sesselmann » 12 Dec 2016, 14:32

Kolja,

Love the x-rays, very cool indeed..

Steven

Post Reply

Who is online

Users browsing this forum: No registered users and 3 guests