Lo scintillatore tascabile Mirion-MGP PDS100G

[eliocor]

BTW, today I completely reverse engineered the communication protocol of the PDS-100G(N).
I wrote some python scripts to test it: attached to this message is the output of one ¹³⁷Cs (TG77) 1024 points / 150seconds spectrum*, automatically created by my utility (no human intervention).
There are lots of things to do before the completition of a decent application, but this looks promising!

I have also added another file (chart_x2.xlsx) with two acquisitions:
- blue line is some radium (old WWII airplane dial), 90k pulses (237 seconds)
- red line is some ¹³⁷Cs (TG77), 90k pulses (606 seconds)
Acquisitions are not calibrated, but it seems to me that they are rather spot on, even if I discovered some shift due to temperature: this is not a problem because temperature is memorized during spectrum acquisition and it will not be so difficult to correct for the shift.

*) the spectrum was not calibrated but you can play with the two values 'Gain' and 'Offset'. Once done you can set those two calibration values in the PDS-100G parameters.

[cicastol]

[BOPOH]

BTW, today I completely reverse engineered the communication protocol of the PDS-100G(N).

WOW!!!
I was spying on PDS100 <-> PDSMass communications, but don't undertood all.
I found some limitations in spectrum transfer protocol:
- indicated acquisition time <= 999sec (PDS can collect spectrum longer if we select "continue" after first 999 seconds, but transferred spectrum indicates 999seconds)
- channel counters are 16bit so max value is 65335. I don't try to overflow it (it is not hard with Am231-based smoke detector plus several 999seconds interval)

[eliocor]

I found some limitations in spectrum transfer protocol:
- indicated acquisition time <= 999sec (PDS can collect spectrum longer if we select "continue" after first 999 seconds, but transferred spectrum indicates 999seconds)
- channel counters are 16bit so max value is 65335. I don't try to overflow it (it is not hard with Am231-based smoke detector plus several 999seconds interval)

You are correct, but please take care that the PDS-100G is not a "rea'l" instrument, so it can occur in some problems during a long term spectrum survey, especially drift/shift due to temperature.

During the reverse engineering phase I got only some doubts about some fields in '15' packet:

R15.png

They are/were never used by PDSmass

Do you know their meaning?

BTW, except for those fields, everything else was identified in all packets.

[BOPOH]

You are correct, but please take care that the PDS-100G is not a "rea'l" instrument, so it can occur in some problems during a long term spectrum survey, especially drift/shift due to temperature.

999seconds < 17minutes isn't loooooong time But we need to know the limitations and range of applicability of the instruments used to draw the correct conclusions from the measured quantities.

During the reverse engineering phase I got only some doubts about some fields in '15' packet:

My PDS eats too many energy then BT connected, so I interested only in "connect - grab spectrum - disconnect'. So I interested only in reverse engineering only spectrum transmission (G22000/G22002) and memory cleaning (S22000).

R22 format:
OFFSET,LEN DESCR
+0,1 0x02
+1,3: 'R22'
+4,8: PDS serial NUM, with leading spaces
+12,3: gain * 100 ('184' -> 1.84)
+15,3: offset
+18,3: temperature C
+21,3: acquisition time (sec)
+24,5: dose rate mSv/h
+29,3: ??? '000'
+32,12 time/date HHMMSSddmmyy
+44,4: channelsNUM
+48,3: PDS_specrum_memory_ID
+51,2*channelsNUM: channel counts

[eliocor]

here the results of my analysis of '22' packet:

R22.png

If you are able to install Python and some of its libraries*, in few days I will publish a small program which will exactly do what you need (will create a .XLSX file like the ones I have posted).


*) it is a very easy operation

[cicastol]

Elio, your's work is truly amazing, hope to give it a try ASAP !!

[eliocor]

Here it is the first draft for a PDS-100G spectra downloader: it was written in python and at this date it is a very rough utility... Try it at your own risk!
Some quick instructions:

• - create the following directory: "C:\PDS"
  - download the 'spectrum.zip' file and decompress it in "C:\PDS"
  
  - go to [External Link Removed for Guests]" onclick="window.open(this.href);return false; and download the Python installer v2.7.13
  - install it
  - press [Next] until you'll reach the 'Customize Python 2.7.13' window
  - in the feature list, ENABLE the 'Add python.exe to Path' option
  - press [Next] or [OK] until setup has been terminated
  - once done, open a command prompt window: press the 'Windows' key, followed by 'r' and digit 'cmd' followed by [OK]
  - in the command prompt window digit:
  pip install pyserial
  pip install xlsxwriter
  - you are done with the Python installation/configuration! (not to be repeated again)

if you have closed the command prompt window, reopen it (Win-R/cmd/[OK]), otherwise return to the command prompt window:

• - digit:
  cd c:\pds
  - now you can run the program: to see a small reminder of the available options, digit:
  python spectrum.py -h
  - you will get the following message:

Codice: [Local Link Removed for Guests]

usage: spectrum.py [-h] [--version] [-v] [-k] [-d] [-l] [-e] [-p PORT]
                   [-s SPEED] [-t TIMEOUT] [-o OUTPUT] [-m MAGNIFY]
                   [-r ROTATION]

download spectra files from MGP PDS-100G(N) and write .XLSX file

optional arguments:
  -h, --help            show this help message and exit
  --version             show program's version number and exit
  -v, --verbose
  -k, --kill            after retrieving, delete spectra from PDS-100G
  -d, --delete          delete spectra from PDS-100G
  -l, --last            save only last spectrum taken
  -e, --endian          change endianness of spectrum data
  -p PORT, --port PORT  serial port name
  -s SPEED, --speed SPEED
                        serial port speed [default: 115200]
  -t TIMEOUT, --timeout TIMEOUT
                        serial port timeout (ms) [default: 200]
  -o OUTPUT, --output OUTPUT
                        output file(s) name
  -m MAGNIFY, --magnify MAGNIFY
                        graph size (magnification), [default 3.0]
  -r ROTATION, --rotation ROTATION
                        graph X label rotation (degrees), [default -75]

• - the most important parameter is the Bluetooh serial port: on my PC is COM4, so to run the spectrum.py program I will have to use:
  python spectrum.py -p com4
  or (it is exactly the same)
  python spectrum.py --port com4
  if used on linux it could be:
  python spectrum.py --port /dev/stty01
  
  in general the '-t' and '-s' and '-e' parameters should not be modified.
  At this date, the '-k' and '-d' parameters are not enabled: spectra files will NOT be deleted from the PDS-100G device!
  
  the default output file name is: SerialNumber.xlsx ; it can be renamed using the '-o' option. eg:
  python spectrum.py --port com4 -o foo.xlsx
  
  You can also modify some graphical properties of the graph in the Excel file: size of it (I use a 4K monitor, so the x3 magnification (default) is OK...) maybe you can use a different value, the same is for the X axis label rotation: it can be changed; eg:
  python spectrum.py --port com4 -m 2.1 --rotation 30
  
  Once run, the program will wait for the Bluetooth pairing of the PDS-100G, printing every some seconds the string:
  Waiting for Bluetooth connection.... (press <Ctrl-C> to exit)
  
  As an example, this is the output of a download session between my PC and the PDS-100G device:
  
  C:\PDS>python spectrum.py --port com4
  Waiting for Bluetooth connection.... (press <Ctrl-C> to exit)
  Waiting for Bluetooth connection.... (press <Ctrl-C> to exit)
  Waiting for Bluetooth connection.... (press <Ctrl-C> to exit)
  serial port [com4] opened at 115200 baud
  Serial number: A11xxx
  Serial number: A11xxx
  Got spectrum #4: creating excel worksheet...
  Got spectrum #3: creating excel worksheet...
  Got spectrum #2: creating excel worksheet...
  Got spectrum #1: creating excel worksheet...
  
  [ A11xxx.xlsx] file created.
  You can see the contents of the attached .XLSX fie: the 4 spectra files were stored in 4 different worksheets called 'SpectrumXX'

I would be grateful if you will test the python script with your PDS-100G just to see if it works even for you.

_ Elio.

______________________________________________________________________________________
Veloce traduzione in italiano (fare riferimento al testo in inglese che ha la formattazione del testo)
:
Ho scritto un programmino in Python che permette lo scaricamento degli spettri acquisiti con un PDS-100G, i quali vengono visualizzati creando un file Excel con dati e grafico (vedi allegato).
Seguono istruzioni e maggiori dettagli su come installare/usare tale programmino:
- creare una directory C:\PDS in cui scompattare il file .zip qui presente
- andare sul sito di Python e scaricare la versione 2.7.13
- installarla, assicurandosi (una volta arrivati alla finestra "Customize Python 2.7.13" che il campo "Add python.exe to Path" sia abilitato.
- terminare l'installazione
- aprire un prompt di comandi (finestra DOS) digitando [Windows-R] seguito da 'CMD' e dare invio
- installare i due moduli Python richiesti dal programma digitando i comandi:
pip install pyserial
pip install xlsxwriter
- l'installazione di Python è terminata.

- da finestra di comandi digitare (posizionarsi nella directory in cui è stato scompattato il file spectrum.zip):
CD C:\PSD
per avere un "aiuto" su come utilizzare il programma, digitare:
python spectrum.py -h
che stamperà il testo riportato più sopra.
Per eseguire il programma bisogna ALMENO sapere quale è la porta seriale/Bluetooth a cui fa il 'pair' il PDS-100G. Ad esempio nel mio PC è COM4, quindi per far partire il programma dovrò digitare:
python spectrum.py -p com4
oppure (è esattamente la stessa cosa):
python spectrum.py --port com4
Se fossi sotto linux (il programma funziona anche lì) il comando potrebbe essere:
python spectrum.py --port /dev/stty04

I parametri '-t', '-e' e '-s' di norma non dovrebbero venire richiesti.
Al momento i parametri '-k' e '-d' , usati per cancellare gli spettri dal PDS-100G non sono attivi.

il nome standard del file Excel è [Numerodiseriedeldispositivo.xlsx] ma può venire modificato usando il parametro '-o', ad esempio se volessi generare un file di nome pippo.xlsx dovrei digitare:
python spectrum.py --port com4 -o pippo.xlsx
oppure (è esattamente la stessa cosa):
python spectrum.py --p com4 -output pippo.xlsx
oppure (è esattamente la stessa cosa):
python spectrum.py --p com4 -o pippo.xlsx

Esistono anche delle opzioin per modificare i grafici generati da Excel: dimensioni del grafico e rotazione delle etichette sull'asse X. Sul mio PC la scalatura standard del grafico è 3x, ma è anche vero che uso un monitor 4x (3840x2160), per cui su altri PC potrebbe essere più pratico ridurlo un po' (ad esempio a 1.5x); anche le etichette di testo saranno create orizzontali (0 gradi) invece che lo standard (-70 gradi):

python spectrum.py --port com4 -magnify 1.5 --rotation 0

Una volta fatto partire il programma, esso rimarrà in attesa del pairing tra il PDS-100G ed il PC, stampando ogni qualche secondo la stringa:
Waiting for Bluetooth connection.... (press <Ctrl-C> to exit)

riporto per completezza ciò che stampa una intera seesione del programma di acquisizione (nel PDS-100G ci sono 4 spettri):

C:\PDS>python spectrum.py --port com4
Waiting for Bluetooth connection.... (press <Ctrl-C> to exit)
Waiting for Bluetooth connection.... (press <Ctrl-C> to exit)
Waiting for Bluetooth connection.... (press <Ctrl-C> to exit)
serial port [com4] opened at 115200 baud
Serial number: A11xxx
Serial number: A11xxx
Got spectrum #4: creating excel worksheet...
Got spectrum #3: creating excel worksheet...
Got spectrum #2: creating excel worksheet...
Got spectrum #1: creating excel worksheet...

[ A11xxx.xlsx] file created.

è sufficiente aprire il file [A11xxxx.xlsx] per farsi una idea dei grafici che genera: ogni spettro è in una pagina separata del foglio elettronico (tab in basso)
______________________________________________________________________________________

[Enotria]

.

E' un vero peccato che un argomento così interessante come questo sia precluso a chi, come me, non conosce l'inglese.

Neppure è valida l'opportunità di utilizzare i vari Traslate su questi argomenti così marcatamente scientifici, perché alle volte saltano fuori delle affermazioni assolutamente ridicole.

Comprendo ed apprezzo la volontà di mettere a loro agio i nostri ospiti stranieri, ma ciò non dovrebbe avvenire a scapito dei locali, anche se colpevoli di ignoranza della lingua inglese.

Se non è troppo fastidio, suggerisco a chi ben conosce l'inglese, tanto da saper rispondere direttamente in questa lingua, di aggiungere in coda la traduzione in italiano, in modo da rendere maggiormente usufruibile a chiunque la discussione.


[attachment=0]147.gif[/attachment]

[eliocor]

Spero di averti soddisfatto: ho aggiunto anche una parvenza di traduzione al mio messaggio....
Nelle prossime risposte vedrò di aggiungere anche il testo in lingua madre