Mirion PDS 100G, Spectrometric Personal Radiation Detector

Scintillation crystals, PMTs, voltage dividers etc...
User avatar
Go-Figure
Posts: 97
Joined: 04 May 2019, 22:24
Contact:

Re: Mirion PDS 100G, Spectrometric Personal Radiation Detector

Post by Go-Figure » 08 Sep 2019, 21:46

I post mine again since the images above for some reason are not clearly readable
I see my cps per µSv/h are higher than Ciro's for Cesium, lower for Americium and the same for Cobalt. So there are two diferences going in opposite directions, hence partially compensating each other.
I am not keen in changing any of those parameters, but since I have my default settings saved in this image I would see no particular risk in changing them to Ciro's values, just in order to see what difference does it make in terms of doserate, testing the same sample with the same geometry before and after. I can always change them back.

Question, are these values supposed to be exacly the same or they are calibrated for each device according to small differences that can exist in the crystal itself? Maybe luuk can help us with that.

Massimo
Attachments
Calibration+Factory Parameters-Default.jpg

User avatar
Go-Figure
Posts: 97
Joined: 04 May 2019, 22:24
Contact:

Re: Mirion PDS 100G, Spectrometric Personal Radiation Detector

Post by Go-Figure » 09 Sep 2019, 02:57

Ok guys I did the comparative test.
I wanted a hot sample for this one in order to have a lot of counts and have significant results with a 999 seconds measurement.
So I used the hottest sample I have, not an uranium or thorium ore, but a camera lens (you can read about the lens here , it was one of my first spectra viewtopic.php?f=5&t=519).
You can read in the original post the doserate on the crystal I calculated from that test was 2.85 µSv/h (sample + background) but of course the sample was not as close to the crystal as I could get it here, I placed the crystal right on top of it. so (nt of the inevitable margin of error) no surprise the doserate here was higher.
DSC06793R.jpg
Anyway, let's get to the results. First spectrum with my default settings gave me an average doserate of 4.51 µSv/h. Calibration good across the board. Here the spectrum in linear view. 1,205,234 counts in 999 second (1206 CPS).
PENTAX-Default Settings-999Secs-080919-LIN.png
PENTAX-Default Settings-999Secs-080919-LIN.png (91.86 KiB) Viewed 166 times
Then I changed the device calibration settings and entered Ciro's parameters and repeated the test. I tried to repeat the exact same geometry. The total number of counts this time was 1,205,627. The difference is 0.03% so I guess geometry was pretty close. With Ciro's parameter doserate was higher, 4.87 µSv/h, the difference is roughly 8%.
PENTAX-Cicastol Settings-999Secs-080919-LIN.png
PENTAX-Cicastol Settings-999Secs-080919-LIN.png (84.44 KiB) Viewed 166 times
Then I changed back the settings to my default ones and repeated the measurement and the result was exactly the same as the first test, less than 1% difference but probably I wasn't as careful as before in recreating the exact same geometry (I have the same 0.9% difference in terms of counts as well) there's a good consistency there I would say.

So my conclusion is Ciro's settings give a slightly higher doserate than mine. 8% higher on this particular sample which has Th232 progeny in it, so it has a wide variery of energy levels.
In order to know which setting is the best you would need to compare results with those from a properly calibrated one. I don't really have the chance to do so, Ciro probably can and I seem to understand Janni can as well. So if you have the chance let me know.
Of course a single digit percentage difference is not that bad anyway, being well inside the intrinsic margin of error.

P.S. As you can see the different settings didn't impact in the spectrum at all, another confirmation the two sets of calibration parameters are totally indipendent.

Massimo

Janni
Posts: 19
Joined: 22 Oct 2017, 04:36
Location: Hamburg
Contact:

Re: Mirion PDS 100G, Spectrometric Personal Radiation Detector

Post by Janni » 09 Sep 2019, 22:59

Hi!
So thanks to Massimo I got the software yesterday and found some time to play with it. So first of all I have to report that my settings were again different from the ones from Massimo and Ciro and my spectrum from my small piece of uranium ore was completely shifted (for example the Bi-214 Peak that should be at 609 keV was at 430 keV). I calculated the Gain and the Offset according to the formula in the manual and changed these parameters, now the spectrum looks ok.
So for first I changed only these two parameters. Also dose rate increased since then, but still shows not enough. But until now I did not change the sensitivity parameters (cps per µSv/h). I will try step by step, I just wanted to update you and document for myself, otherwise I will be confused because of too many data... ;-)
Attachments
spectrum before Gain and Offset calibration.PNG
spectrum before Gain and Offset calibration
spectrum after Gain and Offset calibration.PNG
spectrum after Gain and Offset calibration
Settings before calibration.PNG
Settings before calibration.PNG (96.81 KiB) Viewed 132 times
settings after calibration.PNG
settings after calibration.PNG (83.47 KiB) Viewed 132 times
factory parameters.PNG
factory parameters.PNG (149.01 KiB) Viewed 132 times

Janni
Posts: 19
Joined: 22 Oct 2017, 04:36
Location: Hamburg
Contact:

Re: Mirion PDS 100G, Spectrometric Personal Radiation Detector

Post by Janni » 09 Sep 2019, 23:06

Here is the photo with cps and µSv/h with exactly the same piece of uranium ore at same point of detector. Cps are about the same, but dose rate is bit higher now.
Attachments
cps and dose rate after changing gain and offset.PNG

User avatar
Go-Figure
Posts: 97
Joined: 04 May 2019, 22:24
Contact:

Re: Mirion PDS 100G, Spectrometric Personal Radiation Detector

Post by Go-Figure » 10 Sep 2019, 00:20

You are welcome Janni,
In my experience changing the spectrum calibration has no impact at all on the dose rate, just as changing the calibration which determins the doserate doesn't have any effect on the spectrum.
Anyway in order to compare CPS and doserate of two measurements the best way is to record a long enough spectrum, not for the spectrum itself, but because among the results you will also have total number of counts and average dose rate over the acquisition time, and fluctuations (which are inevitable) will be averaged out.

Massimo

Janni
Posts: 19
Joined: 22 Oct 2017, 04:36
Location: Hamburg
Contact:

Re: Mirion PDS 100G, Spectrometric Personal Radiation Detector

Post by Janni » 10 Sep 2019, 01:28

Hi Massimo,

yes you are right, anyway I need to make a longer spectrum to specify better and maybe it is just fluctuation about dose rate change. Definitly it still shows too less, so I will try to adjust these other parameters, I did not want to change all at the same time but now I will change them. Also for checking better other peaks I want to make longer spectrum that will be more representative.

I will update the next results :-)

User avatar
Svilen
Posts: 66
Joined: 23 Sep 2016, 04:25
Location: Germany
Contact:

Re: Mirion PDS 100G, Spectrometric Personal Radiation Detector

Post by Svilen » 12 Sep 2019, 09:05

I can happily announce that I also became a part of the PDS-100G owners group few days ago. It's quite nice device, fast, sensitive and portable. Not cheap, but a good deal, especially having in mind the price for a new one of that type. Connected to a tablet it's really a very portable spectrometer (logging counter, dose rate meter) that can be great company for any trip.

My device also almost doesn't have traces of use, but one small malfunction is still present - the charging doesn't work. Which is not a big deal, since the 2 standard AA batteries are easily charged with an external charger. Actually, as Massimo already mentioned above, the battery indicator is very inaccurate, and charging this way is even preferred. By the way - thank you Massimo for the help!
Another thing that I personally consider bad design is batteries placed directly on the uncovered PCB, which also contains the charging circuit and the BT module (even when on the opposite side of the board). This way, if the batteries leak due to improper long time storage or malfunction, the board under gets destroyed (as can be actually seen on two other devices, the same seller listed as non functional). This is maybe the main reason for the welcome message, reminding the user to check the type of batteries inside, every time he starts the device.

My calibration was also a bit off, but after slight adjustment of the gain and offset looks good now, at least for the lower energies. Here are some pictures of the spectrum taken from two radium painted clock hands in linear and logarithmic view. Also I show the difference using 512 and 1024 channels. Since I recorded only 15 minutes, the source is quite active and higher resolution here is not a must, 512 channels can be even preferred.
1_Ra_512ch_Lin.png
1_Ra_512ch_Lin.png (32.11 KiB) Viewed 105 times
2_Ra_1024ch_Lin.png
2_Ra_1024ch_Lin.png (33.13 KiB) Viewed 105 times
3_Ra_512ch_Log.png
3_Ra_512ch_Log.png (46.91 KiB) Viewed 105 times
4_Ra_1024ch_Log.png
4_Ra_1024ch_Log.png (40.14 KiB) Viewed 105 times
If the peak around 1130KeV is from Protactinium-234 (1001 KeV), then there is nonlinearity and only the first half of the scale fits the calibration. Longer recording would be of coarse better, but what is your opinion on that? I also see that the HV supply is higher than the others posted above, but don't want to play with that, not to affect the dose rate calibration (if correct, since I can not compare it with a known correct device to be sure). Any comments here will be highly appreciated.

Here are my settings:
PDS_Settings_1024.png
PDS_Settings_1024.png (15.09 KiB) Viewed 105 times
Svilen

Janni
Posts: 19
Joined: 22 Oct 2017, 04:36
Location: Hamburg
Contact:

Re: Mirion PDS 100G, Spectrometric Personal Radiation Detector

Post by Janni » 12 Sep 2019, 19:39

I have again some news about calibration. I think my device is now calibrated nearly perfect.

So I made a longer more accurate spectrum and found out that the peak at 609 keV was just a few keV shifted (almost good but not perfect). Besides in the higher energy area I could not allocate exactly the peaks and I wanted to take one of this area for a better calibration, so I looked for one that I can exactly allocate and compared the spectrum with another spectrum that I once made of the same uranium ore sample with a HPGe detector because in this spectrum the peaks fit exactly to a spectrum I found in the internet. I then decided to take the peaks at 609 keV and at 1760 keV for the calibration.
After calculating and changing the Gain and Offset again a little bit the spectrum fits completely fine consistent the HPGe Spectrum. I marked just four significant peaks, but also the other fit perfect now.

Then I still was not satisfied with the dose rate, so I played with these factory pile sensitivity parameters. In the manual is written that cesium and cobalt thresholds should not be changed. Anyway I changed them because I wanted to see what happens, but nothing so much happened. So I changed them exactly back to the initial numbers and changed only the americium. I had to change it in a severe way from more than 1000 to 200 now. That is completely different to the initial and also to all other numbers of this parameter posted in this forum, but now I have a much better fitting dose rate... I compared with automess AD6 and also checked higher and lower dose rates with it from other uranium ore in different distances and all complied good with the automess that is even official calibrated (geeicht, look on the sticker :)).
So for whatever reason this parameter is completely different now, for my understanding everything is perfect now :-)
Attachments
1.Uranium ore sprectrum.png
Uranium ore energies from internet
1.Uranium ore sprectrum.png (23.56 KiB) Viewed 97 times
2.HPGe Spektrum.png
2.HPGe Spektrum.png (83.09 KiB) Viewed 97 times
3.Settings.png
Settings now
3.Settings.png (73.32 KiB) Viewed 97 times
4.Uranium ore PDS spectrum log.png
Logarithmic scale
5.Uranium ore PDS spectrum lin.png
Linear scale
6.Factory parameters.png
factory parameters now
7.cps and dose rate.PNG
cps and dose rate
8.dose rate AD6.jpg
dose rate compare with AD6
8.dose rate AD6.jpg (44.75 KiB) Viewed 97 times

User avatar
Go-Figure
Posts: 97
Joined: 04 May 2019, 22:24
Contact:

Re: Mirion PDS 100G, Spectrometric Personal Radiation Detector

Post by Go-Figure » 12 Sep 2019, 20:33

Svilen wrote:
12 Sep 2019, 09:05
If the peak around 1130KeV is from Protactinium-234 (1001 KeV), then there is nonlinearity and only the first half of the scale fits the calibration. Longer recording would be of coarse better, but what is your opinion on that? I also see that the HV supply is higher than the others posted above, but don't want to play with that, not to affect the dose rate calibration (if correct, since I can not compare it with a known correct device to be sure). Any comments here will be highly appreciated.
Hi Svilen, Good to see everything's working now!
As you probably noticed when you load the spectra from the PC or tablet (loading the file you previously saved), instead of from the device, the average doserate over the acquisition time doesn't show anymore. I don't know if that can be changed, but that's the reason why I always save a screenshot of the spectra from the device together with the file, so I save that information as well. Or even better a photo of the detector's display at the end of the measurement so actual acquisition time is showed. The software never shows acquisition times longer than 999 seconds, therefore if your acquisition time is anyware between 1 and 999 seconds the figure showed will be correct, but if it's longer the software will still show 999 seconds no matter what, and you won't be able to know how many CPS you actually got in average. It's better to save that information while available, because when it's no longer on the detector's display I don't know any way to retrive it.

As for the peaks, I think your 1120-1130 is from Bi214 at 1120.29, which is the most likely peak to be found there in an U238-Ra226 gamma spectrum, the peak from Pa234m has a far lower probabiity, it's there, but it should not be the "biggest" thing you see.
The rest doesn't look too bad to me, I can't check the details from here, but the one above 1200 should be Bi214 again at 1238.11 keV, then I see K40 and again Bi214 at 1764.49 keV.
That is completely different to the initial and also to all other numbers of this parameter posted in this forum, but now I have a much better fitting dose rate...
Hi Janni, yes, those calibration settings are very far away from the range I've seen so far. Being just a user of the detector I am in no position to comment on that.
The only thing I can report is that (as you can see in one of my last posts here) when I changed those settings by much smaller amounts I immediately found measurable difference in the result.

Massimo

User avatar
Svilen
Posts: 66
Joined: 23 Sep 2016, 04:25
Location: Germany
Contact:

Re: Mirion PDS 100G, Spectrometric Personal Radiation Detector

Post by Svilen » 13 Sep 2019, 03:43

Thank you for your reply and analysis, Massimo. Yes, I may be wrong about the Pa-234m and then is all good.

I noticed that about the saved spectra, but just didn't think about it when I took the pictures, since I was more concentrated in the energies, than the dose rate.
Maybe still a bit of fine tuning is needed, but not much, and I am quite satisfied with what I get.

@Janni: Did you try with 100 for the Americium threshold? Looking at mine, Massimo's and Ciro's settings - all of them 100 (which of course doesn't need to be the same, but still), makes me think that somebody just made a typo and wrote 1000 instead of 100. Could that be the case?

Greetings,
Svilen
Svilen

Post Reply

Who is online

Users browsing this forum: No registered users and 0 guests