Moderator: Ellery Frahm, [log in to unmask], Electron Microprobe Lab, University of Minnesota

Hi Everybody,

Thankyou for your kind help. To quickly address a few issues.

Spectrometer Odometer - we don't have them. Sorry this is an 8800 EPMA, I was answering Ritchie's question about the 733 (which some how did not make it through) when I thought out loud about my old 8800.

Beam Position and charging in OL pole piece - good point, I have seen this before in other machines, but I do not see any sign of it now. Image is stable even under high beam currents, and the optical image of my spot on a fluorescence specimen does not jump about or shift.

WDS belt linearity/tension/length - we have a very experienced factory engineer and this is always checked with belt replacement, plus the new probe with the same engineer is doing great. Also, I have used unadjusted belts before and they do not give me this degree of trouble.

Backlash issues - Both probes old and new use the same software and same backlash correction, only the old probe gives trouble, plus the old probe worked fine with this backlash correction when new. Also, when I did my reproducibility tests I tried starting numerous times from above the peak and below the peak, no real difference. Having said that, maybe it would be a good idea to increase this backlash correction on the older machine, thanks!

I've also tried these various tests with open slits and closed slits, again no real improvement.

Thanks again for all your help.

Best Regards,

Ron

JEOL Probe Users Listserver

Moderator: Ellery Frahm, [log in to unmask], Electron Microprobe Lab, University of Minnesota

Post a message: send your message to [log in to unmask]

Unsubscribe: send "SIGNOFF PROBEUSERS" to [log in to unmask]

On-line help and FAQ: http://probelab.geo.umn.edu/listserver.html

* Ron,

First, your software should be setup so the spectrometer approaches the peak from the same direction
to avoid backlash issues.

There is a spectrometer adjustment that has a large impact on reproducibility called the belt linearity adjust. This usually is not a problem unless the stainless steel belt on the spectrometer has been replaced. I know some service guys will avoid doing it because it is a little tricky and very time consuming. When the adjustment is out it is most difficult to reproduce at the higher L values (>170mm).

You can do a quick check of the belt linearity by defocusing your sample and re-peaking the spectrometer (at a high spectrometer L-value). First, defocus the optical image (Z) at +/-20um intervals and re-peak the spectrometer. If the count rate goes up and is considerably higher with the sample defocused vs focused, you have a belt linearity problem. The specification is to have maximum count rate within 40um of sample focus. If you plot this over the range of the spectrometer it can also tell you if your baseplate and crystal tilt are aligned correctly as well. Note: this check and/or adjustment should be done with at least the 500um slit in place.

Bill

On Thu, May 21, 2009 at 8:19 AM, Ellery Frahm <[log in to unmask]> wrote:

JEOL Probe Users Listserver

Moderator: Ellery Frahm, [log in to unmask],
Electron Microprobe Lab, University of Minnesota

Post a message: send your message to [log in to unmask]

Unsubscribe: send "SIGNOFF PROBEUSERS" to [log in to unmask]

On-line help and FAQ: http://probelab.geo.umn.edu/listserver.html

*

Hi Ron,

What about the beam position on the sample? Is that stable? If, for example, there is charging in the objective lens pole piece or somewhere else in the column, the beam could wonder a lot across the sample, constantly changing the Bragg angles (and therefore L values) and affecting all the spectrometers.

Best,
Ellery

On May 20, 2009, at 11:49 PM, Ron Rasch wrote:

Hi Ellery,

Yes all good questions. To go into more detail:

This probe has seen a lot of work, 15 years of 24 hrs a day, 7 days week, excluding down time etc.
Doing a peak search during Qnt analysis, does not help anymore, but it use to when it started "going off". For most of the probe's life I only did a peak search during Standard analysis.
All four WDS spectrometers are giving bad results but to a greater or lesser extent.
My FCS is probably the best and my H type is probably the worst.
The beam current looks very stable.
The standards are good and have been tested in my new probe. Plus they were good at the start (15 years ago).
There seems to be no strong connection between L value and reproducibility, some days the high values some days the low, but mostly the low L_values are better.
I've had the factory check the base plate, align the crystal tilt and do the SCA conditions, twice.
I do not believe it's the x-ray detectors (see below), but just in case I've replaced one Xe detector and the window for one GPC, no great improvement.

My usual test (apart from just doing a typical analysis) is to do a peak search on a standard, then take a 10 second count three times, then move off the peak then repeat.

On any given peak search, the three count rates are quite consistent with each other, just not consistent with the count rates after the next peak search. Two very close peak positions may not yield similar count rates.

Next on my list is to try swapping a stepper motor and gear box with my new probe, but I'm also looking for any other ideas or things to check.

Cheers,
Ron