System maintenance/fix-it tips and tricks

[DavidAdams]

Hi everyone,

So here's my scenario:

There's something going on with my Spectrometer 2. Every year since the instrument was installed in 2019 at about the 11-12 month mark the spectrometer drive wire breaks. This spectrometer doesn't get a bigger workout than any of the other 4 spectrometers on my instrument, in fact it probably moves less overall during an average analysis than the other spectrometers. This year's even happened in October and the service engineer didn't come out to fix it for 6 weeks leaving a big hole in the capability of my instrument. The service engineer is also not interested in diagnosing or investigating why this particular spectrometer is experiencing this problem and 6 weeks is way too long to wait for a repair service.

My questions are:

Do people here have experience personally replacing the drive wire on JEOL spectrometers? If so, could you share the process with the rest of us so that we can perform this bit of maintenance competently?

Thanks!!

[Scott B.]

Greetings,

Didn't hear about your issue till the Newsletter just showed up. Hopefully it's still helpful for the future.

I have an 8500F, but I'll assume it has the same configuration as your machine.

Are you referring to the belt that actually drives the spectrometers (flat gray band) or the braided silver wire with loops at each end that tensions the assembly, and runs around a bunch of pulleys all over the place?



If it's the drive belt, that's something you can certainly do for yourself, though it's pretty annoying, as there's a sort of Mobius Strip folding loop trick to get it installed correctly, and you have to get the loop the right way or it rubs.

The wire, on the other hand, requires removal of the baseplate assembly, which is a pretty hard core procedure that also requires the skill to do a baseplate alignment procedure after reinstall.

If it's the belt that's failing, I wonder if it has been repeatedly installed incorrectly? Believe me, it's easy to do, and I've spent hours staring at the tricky fold/loop trying to decide if I got it right, and even seen technicians second guess themselves. It's also possible that you're just having bad luck. The JEOL Rube Goldberg design is prone to belt/wire failure. I've probably had an average of 1 a year since I've been in the driver seat, but one per year on the same spectro seems excessive even for a JEOL...

Feel free to contact me offline if you'd like more information on the process.

Cheers,

Scott B.

boroughs@probesoftware.com

[DavidAdams]

Hi!

Thanks for the reply. Sorry, I missed it from so long ago!!

It is indeed the braided wire and not the belt that continuously fails. It's possible that the JEOL service engineer is missing something but it also could be a manufacturing defect with this spectrometer. The wire is the piece that fails most frequently on it but I have had to replace the motor O-ring multiple times, the belt has gone twice and the detector has had to be replaced. I think I just got a lemon SP2!

I might send you an email to discuss getting more info about the replacement procedure!

Best,
Dave

[John Donovan]

With time, there are various communication/networking issues that seem to accumulate (at least the JEOL probes I worked with) that may benefit from a restart. They can range from PFE being frozen or having erratic communication to problems on the instrument ("jog" doesn't work, spectrometer problems).

Depending on the issue at hand, there are different levels of escalation that one can go through. It may not always required to go as deep as Level 4 but occasionally I find that a level 4 reset, brings "Aprilfrische" back to the instrument.

Here are the different steps for a reset.

Level 1:
Close PFE, PI and whatever would be connected to the instrument
Close the PC-SEM/PC-EPMA and exit the JEOL software. Run the "kill" batch script for good measure.

Level 2:
Level 1 + Restart of both computers

Level 3 (This is what I would recommend for communications problems)
Level 1+ cycle the OPE power. On the front of the probe, behind the plastic window, it is the third button from the top, labeled OPE Power (obviously do NOT touch the main power or vacuum power). Just press the OPE Power button "off", count to ten or so, turn on again. Restart the JEOL software. When it is up, restart the PFE etc. The JEOL software will want to initialize the stage and you will likely have to reset your points unless you established fiducials in PFE.

Level 4 (likely needed for stage and spectrometer problems)
All the way to Level 3 but while the OPE power is off, also cycle the power on the intelligent unit. This is the big box behind the instrument. The power button is in a recess in the lower right corner close to the floor. Turn off, count to something, turn on again. Turn on your OPE power. Restart JEOL software etc..
The JEOL software will not come on fully before the electronics are all back, so give it enough time between each step too.

A colleague recently has had some issues with their instrument, specifically some firmware (1003 and 1004) errors with a spectrometer initialization, but it is always worth trying Anette's suggestions above, because as she put's it:

If she has not already I would recommend for her to do a full restart that includes the intelligent unit (See Level 4 restart in the description above).

Cycling the OPE power on the 8230 only deals with the electron optical system while the spectrometer boards are in the intelligent unit. In the newer JEOL models, they incorporated the intelligent unit and just cycling the OPE might be enough.

She also adds:

The out of bounds motor position error message might either require an initialization of the spectrometer limits (in the JEOL software) or might be a result of an initialization of the spectrometer and now the limits don't agree anymore. Easy fix to update the new limits in motors.dat.

There might be a real problem with the spectrometer, but I would start with the above.