Faraday erroneously low current on unknown halting automation (8200 W)

[orlandin]

Hello all! First, I am somewhat of a newcomer here and it will be a while before I succeed in absorbing the feast of helpful information in this forum; apologies if this issue has been reported elsewhere and I've missed it.

My problem is that within the first dozen or two unknown points of an automated run (I think usually/always at the end of a point), PFE will give an error stating that the measured current is below 0.1 nA and that the filament may have blown. This halts the automation. The beam current is a modest 10 nA for apatite analyses.

It seems possibly related to the FaradayWaitInTime issue reported here: https://probesoftware.com/smf/index.php?topic=80.msg6823;topicseen#msg6823

However, two things make me think that maybe this is a different issue: first, that this problem does not seem to ever occur during the lengthy standardizations that precede these unknown points; and second, that when my admittedly human eyes are recording the beam current via the UNIX software, the current never seems to significantly vary. Certainly when I posthumously discover that this error has occurred in PFE and I check the offending current in UNIX or GunAlign, there it is sitting innocently at the 10 nA that I thought we had all agreed upon.

I have tried increasing my FaradayWaitInTime from 1.0 to 2.0, and this does not seem to help. I will try increasing it further tomorrow (Teamviewer has finally realized that I am an millenial moral degenerate and I have yet to upgrade to the Windows 10 remote access alternative). There is a possibly related issue with being able to set the beam current to high values on the JEOL side (also being addressed tomorrow) and so I'm sure that my Set Beam Current Calibration is not ideal. I tried to bypass complications from that by selecting the 'always use current instrument conditions' along with the 'Read Current Instrument Conditions' in Analytical Options and elsehwhere - my hope being that this would cause PFE to not attempt any manipulation of the current and we would all just chug along passively having faith in our new filament's stability.

I suppose one mildly irresponsible workaround would be to disable the kind-hearted feature which halts the analysis upon reading a blown-filament level of low current? That may just create more problems down the road, though.

Thank you all in advance for any advice that you might have!

Phil

[John Donovan]

Hi Phil,
I should defer to the many JEOL experts on this forum, but I have observed that since the JEOL 8900, each new JEOL model EPMA instrument seems to require a longer "settling time".  For the 8230/8530 the Faraday picoammeter "wait in" time needs to be over 2 seconds, even 2.4 seconds or so.

But 2 seconds should be fine for the 8200, so there must be some other hardware issue I would guess. I would have an engineer look at the Faraday cup and wiring. Maybe be it's sticking mechanically. I've seen that before. 

Also, did you re-start PFE after you edited the FaradayWaitInTime value in the Probewin.ini file?  That is required to obtain the new value.

By the way, the problem with PFE saying the beam current is too low, is completely independent from setting the beam current. Reading the beam current is strictly a picoammeter issue.  The "Always Use Current Conditions" flag won't help.

The basic issue is that after the Faraday cup is inserted, it takes time for the JEOL picoammeter to "settle" in order to get a good reading.  You might also try setting the number of "Beam Averages" to greater than 1 in the Count Times dialog.

JEOL 8200 users: what value do you have for the FaradayWaitInTime in your Probewin.ini files?

[Probeman]

On the WSU 8500, FaradayWaitInTime=3.5.

Really?  Did you test that this length of time was actually necessary for reliable beam current readings?

Any other 8200/8500 users have values for the FaradayWaitInTime in the Probewin.ini file?

[orlandin]

Hi all! Thank you so much for this feedback - I think that John's suggestion is what did the trick. The successful run of the full automated queue thanks to setting the beam averages to 6 (the number of points per grain, although I'm not sure that matters), has empowered me to discover other problems with my analyses. The halting of automation due to erroneous beam readings is no longer one of them, though!

[John Donovan]

Hi all! Thank you so much for this feedback - I think that John's suggestion is what did the trick. The successful run of the full automated queue thanks to setting the beam averages to 6 (the number of points per grain, although I'm not sure that matters), has empowered me to discover other problems with my analyses. The halting of automation due to erroneous beam readings is no longer one of them, though!

Hi Phil,
I suggest another experiment also: try setting the beam averages back to 1, but change the FaradayWaitInTime in the Probewin.ini file to 3.4. Owen says that is what he needed on his 8500.

Be sure to re-start the software after editing the Probewin.ini.  And if you are so inclined it would be interesting to see how small a time delay you need to read the beam current reliably.