Has anyone tried converting data obtained with a Bruker Tracer III-V+ pXRF to be comparable to data obtained with the new Tracer III-SD? I'm working on obsidian and I have tried both compressing and expanding the spectrum but the numbers do not even come close to being comparable.
Have you tried contacting Bruker themselves? I attended a workshop on the Tracer III-V by Bruker in 2011 and they assured us that they were always ready to answer any questions we might have about the use of the pXRF and about the resulting datasets. I think they might at least be able to tell you what differences you might expect when comparing the datasets. Also are you at least detecting the same elements with both devices?
Dear Simone, on both instruments the acquisition conditions were the same - the x-ray tube was operated at a 40 kV, 14µA, using a Cu-Ti primary beam filter and each sample was scanned for 200 seconds. I had analyzed some samples on the Tracer III-V+, then my department upgraded their instrument to Tracer III-SD. I also run two standards (AGV-1 and SCo-1) on both machines to see how they compare. I am working with obsidian, so I used an obsidian calib file to get quantitative data (counts). Under the same parameters used, the elemental counts are different (as expected) for the two instruments. However, I was told I can in S1PXRF I can either compress the new data from a 2048 channel spectrum to a channel to make it comparable to the old data acquired or expand the 1024 to a 2048 channel. I tried both ways and the counts don't come close to being comparable.
I have attached a small file to illustrate.
Thank you SO very much for responding! I have been trying to figure this out for a few days now without much success.
Dear Veerle,
I have not yet contacted Bruker, but that's my next step if the RG community is also unable to help me with this problem :). Thanks for your suggestion.
The issue, as I detailed above, is not that they instruments are yielding different results in terms of elemental counts (same elements are being detected), this is to be expected. But that I was told that you can either expand the channels or compress them to make data collected on two different instruments comparable. But this is just not working out. I have tried making the old data comparable to the new by expanding its spectrum channel from 1024 to 2048 and have tried the reverse and neither work.
Again, thanks for your suggestion!
Dear Simone,
Again thank you for your response. Right, it absolutely makes sense and I knew that the data collected on the two different Bruker machines would not be the same, but I am wondering if anyone knows how they can be made to be comparable. According to Bruker, in the S1PXRF software we should be able to either compress or expand the channels to make the data comparable but it doesn't work. Just wondering if anyone else has run into this problem and may have a solution.
Many thanks!
- Badges
- Science topic
- Similar topics
- Computer Communications (Networks)
- Simulators