These are the tests we run during our evaluation of Cameca AND JEOL microprobe. Hopefully you can easily follow them. It is likely we will run similar test when we receive our instruments (prior to the final acceptance). Key one is the half-peak
spectrometer reproducibility test: instead of moving back and forth between two peak and measure them repeatedly, you measure the position at half the maximum of the peak height. This test was suggested by John Donovan, and it is a very difficult test (for
the instrument) and take some time… I believe John has documents about that (see email from Jeremy).
We managed to get most of these tests done in 3 days (and 2 nights), except for the long term standardization stability - but this is something you should check (i.e., standardizing and analyzing some secondary standards, then wait 1-2 weeks and
re-run the analysis without standardization - even more difficult for the instrument is to perform several crystal flip over this week (or 2) and re-run the analysis without standardization).
Another test I find necessary is the spectrometer and crystal alignment (geometry). You can vary the Z-position (similarly deflect the
beam along the X and the Y axis - not the stage!), and ideally the maximum intensity should be at Z = 0 (and X = 0, Y = 0).
Stage reproducibility test should definitely be performed, too, and I believe JEOL does NOT have a diagnostic tool for this, you will have to do it manually (locate a small feature, center it on the screen, move away from a few um to a few mm/cm,
then come back and measure the displacement between the center of the screen and the small feature).
Other tests described in the attached document. Of course, you might need to adjust this based on your samples / standards available and / or the main purpose of your probe.
On comment about your tests vs. JEOL’s tests: unless you have clearly stipulate this in your contract, some of these tests might not be “recognized” by JEOL and they may not want to delay acceptance of the instrument if their tests have passed.
Or it might be difficult for them to acknowledge a problem you might have seen (but they don’t according to their criteria).