Opening an .mca file in ARTAX, is it possible?

[rmbordalo]

Hello,

I usually work with a Bruker Tracer III-SD and process the data using ARTAX and S1PXRF with no problems. I use it mostly to analyse pigments on easel paintings.
However, I am currently collaborating with someone who made some analyses (which can not be made again) and they used an Amptek XRF and sent me the results on .mca files. I can open them using PyMCA but I'm not an expert using it and the program is not really user friendly as ARTAX is. Ideally, I would like to convert the .mca files and open them in ARTAX. I tried doing so several ways but ARTAX does not accept any other files than .pdz. It does not accept .csv and it complains that .txt files don't have a header.
Any ideas of what can I do?

Thanks a lot!

Rui

[Lee Drake]

Hi all,

I have been working with Rui on this - the good news is that .mca files can be opened up as .txt files (just change .mca to .txt). What follows is a detailed look at the contents of the file types. The overall story is yes, you can open a .mca file in Artax, but there are some steps you'll need to follow. 

First, a general rule, every computer file is either one of two things, a .txt file, or a folder. If you click on the artax project filetype (.rtx) and open in a text editor like Word, you will find that an Artax project is really just a big text file with lots of code (technically an .xml file. Likewise, the .exe file extension, if you change it to .zip and uncompress it, gives you access to (often encrypted) code. In either case, don't hesitate to take a peak at a file by opening it as a .zip folder or as a .txt folder because most everything is one or the other. By they way, you can do this with music too - every .mp3 file can be opened in text edit, and if you bought your song from iTunes, you can even find that Apple buried your apple account information to track the file. 

In any case, I digress. Here is the beginning of a typical .mca file:

<<PMCA SPECTRUM>>
TAG - live_data
DESCRIPTION - 
GAIN - 4
THRESHOLD - 0
LIVE_MODE - 0
PRESET_TIME - 180
LIVE_TIME - 77.616000
REAL_TIME - 180.000000
START_TIME - 01/21/2013 10:25:28
SERIAL_NUMBER - 10360
<<CALIBRATION>>
LABEL - Channel
0 0
631 2.981
1362 6.4
1835 8.627
2242 10.549
2678 12.62
<<DATA>>
0
0
0
0
0

What you essentially have is lots of metadata, with a <<DATA>> followed by channel counts. Next, let's take a look at a .txt file that is opened by Artax:

BeginHeader
Elin=0.0199812 Eabs=0
Fano=0.11 FWHM=150
EndHeader
3813
2985
1971
1177

It is the same essential format, with channel counts following EndHeader. The Eabs and Fano data will always be 0 and 0.11 respectively with the Tracer. Elin (0.01998) revers to the energy calibration - the detector is a set of channels, which are then converted to energies. So if you multiply the channel count (2048 for most Tracers, 1024 for older models) by the Elin, you get the energy calibration. For example, 2048 X 0.0199812 = 40.92 keV. FWHM refers to the Full Width Height Maximum of of the Manganese K-alpha line (a mostly arbitrary standard). Basically, this revers to how many eV wide that peak is. So 150 eV would mean that the MnKa1 peak is .15 keV wide. A Tracer IIISD should be closer to 142 eV (.142 keV), an older Tracer can be as high as 220 eV (.22 keV). Sometimes you have to modify this a little bit to get the right match - different samples can have divergent FWHM values even with the same XRF unit. 

So basically, we need to find the Elin in the .mca file and then put 142 eV in for the FHMW, then we are done. The problem is that, for reasons I don't understand, the .mca file doesn't just give the slope of the line, it is instead giving points for the curve:

<<CALIBRATION>>
LABEL - Channel
0 0
631 2.981
1362 6.4
1835 8.627
2242 10.549
2678 12.62

What we need to do is run a linear regression and get the slope of this line (rise over run). This is the energy calibration curve:

​This line has a slope of 0.004708, with an r2 of 1. Alas, if only all calibrations were as good. This is the number we put into our new header for our .mca file:

BeginHeader
Elin= 0.004708 Eabs=0
Fano=0.11 FWHM=142
EndHeader

Make absolutely sure there are no extra spaces - Artax is very sensitive to spaces near BeginHeader and EndHeader. Sometimes, you need to repaste the original text a couple times. There is one extra step though, an Artax .txt file ends with the channel counts, while a .mca file ends with a lot more information, which we will need to delete. It is as follows. 

27
17
<<END>>
<<DPP CONFIGURATION>>
COM Port: USB
Rise: 11.2uS
Top: 0.4uS
Fast Threshold: 19
PUR Enable: PUROn
RTD ON/OFF: RTDOff
RTD Ratio: 25.0%
RTD Slow Thresh: 19% FS
AutoBaseline: Off
BLR: BLR:ON DN:64 UP:16 
Acquisition Mode: MCA
MCS Timebase: 10mS/channel
MCA Channels: 4096
Slow Threshold: 1.27% FS
Buffer Select: Buffer A
Gate Input (TTL): GateOff
Preset: 180.0 sec
Coarse Gain: 66.26x
Fine Gain: 0.9895
Input Polarity: Positive
Input Offset: -2.048V
Pole Zero: OFF
Det Rst Lockout: 102uS
TEC: 225.0K
HV: 119.4V
Preamp Power: 8.5V
Analog Out: Decimated Input
Offset: 0mV
Aux: ICR
Audio: Off
<<DPP CONFIGURATION END>>
<<DPP STATUS>>
Device Type: DP5
Serial Number: 10360
Firmware: 5.09
FPGA: 5.05
Fast Count: 5514689
Slow Count: 2377664
Accumulation Time: 180
Dead Time: 56.88%
HV Volt: -118V
TEC Temp: 226K
Board Temp: 27∞C
<<DPP STATUS END>>
<<DP5 BOOT FLAGS>>
Boot device mode: PX4
Boot configuration state: Use nonvolatile configuration
RS232 baud rate: 115,200 baud
FPGA clock: 80MHz
HV Polarity: Negative
ADC invert: Inverting
Spectrum offset: None
Fast channel shaping: Normal
MCA channel: Slow channel (normal operation)
Peak detect: Max only (normal operation)
<<DP5 BOOT FLAGS END>>

 Artax files, as you can see, are much simpler, which means we essentially have to delete lots of information from the .mca file. When we have done all that, this is what we get:

​Voila! We have opened a processed .mca file from pymca into Artax. Is it labor intensive? Yes. But it works great, and we are on our way to getting the best of both worlds. I've attached the raw .mca file (converted to .txt by renaming it; mcafile.txt) and a converted Artax-compatable text file (mcaforartax.txt). Also included is an Artax .txt file (artax.txt) for reference. In case anyone is curious, it is of a dinosaur bone, so if you ever wanted to see what trace elements can be in those, take a look. 

Thanks,

---

B. Lee Drake

Department of Anthropology
University of New Mexico
(505) 510.1518
b.lee...@gmail.com

    

--
Visit the pXRF for Cultural Heritage website at:
https://sites.google.com/site/pxrfgroup/
---
You received this message because you are subscribed to the Google Groups "pXRF for Cultural Heritage" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pxrf+uns...@googlegroups.com.
To post to this group, send email to px...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.