JEOL stage shift issue

[Rom]

Hi everyone,
may be somebody heard about uncontrolled stage position shift when next unknown or standard measurement starts.
For instance, I set table positions   X,Y,Z = A,B,C. Sometimes, absolutely randomly my system use constant shift (mainly 0.5mm) X or Y or both of them in all points of unknown or standard. As result, measurements are in points X,Y,Z = A+0.5mm,B,C or A,B+0.5mm,C etc.
This shift happens randomly and only at the moment when system go to new group of point.
Is this problem source:
-soft Probe for EPMA
-soft Jeol
-softs connecting
-other

Thank you
 

[Probeman]

I've heard of stage problems like this on JEOL instruments because they are not using linear encoders for positioning. Maybe hardware/electronics problem?

What does your JEOL service engineer think?

[Rom]

Service engineer gave us just a suggestion to disconnect the internet during the measurements. We done it. The likelihood of issue is decreased but problem is not disappeared.   

[Probeman]

This is an old JEOL 8200, yes? It is very likely the stage is quite worn and needs servicing.

[Rom]

Yes, this is an old JEOL 8200. But everyone who use it hasn't this issue because use it through JEOl soft. Just several persons who use Probe for EPMA met the shift problem. It is why I ask to your team.

[John Donovan]

It makes no sense that this would be a software issue, first because the code for the 8200/8500 hasn't changed in 10 years and also because any stage problems like this have always turned out to have problems in the hardware.

Let's take this discussion off-line and if we obtain a solution we can summarize it here once we learn something.  I'll contact you by email.

[Paul Carpenter]

With our JEOL 8200 we have had several isolated incidents with stage issues.  Any behavior that appears and is variable in magnitude and direction is likely to be mechanical and/or electrical.

Because the Y-axis stage drive gets used extensively for stage mapping, there is more wear on the mechanism. You should check the following in general but especially the Y mechanism:

Wear or possible looseness of the feedthrough, gearing, and bearings for the stage.

Lubricant contamination or particles on any of the mechanism.

Because the stage uses LED encoders for stage movement, there can be problems from contamination on the LED sensor. There is a procedure for inspecting the signal from the LED system to ensure good signal to noise level, and monitoring of the system with drives to different positions in the stage range.

We use the JEOL stage and spectrometer jog settings so that the JEOL and PFE software can be run interchangeably. I have discovered that if the JEOL 8200 X and Y stage jog values are set to zero, then the JEOL stage mapping routine will acquire one mapping strip, then stop. So it is best to use the hardware jog settings for the JEOL and to not use software jogs in PFE.

Cheers,

Paul

[Brian Joy]

With our JEOL 8200 we have had several isolated incidents with stage issues.  Any behavior that appears and is variable in magnitude and direction is likely to be mechanical and/or electrical.

Because the Y-axis stage drive gets used extensively for stage mapping, there is more wear on the mechanism. You should check the following in general but especially the Y mechanism:

Wear or possible looseness of the feedthrough, gearing, and bearings for the stage.

Lubricant contamination or particles on any of the mechanism.

Because the stage uses LED encoders for stage movement, there can be problems from contamination on the LED sensor. There is a procedure for inspecting the signal from the LED system to ensure good signal to noise level, and monitoring of the system with drives to different positions in the stage range.

We use the JEOL stage and spectrometer jog settings so that the JEOL and PFE software can be run interchangeably. I have discovered that if the JEOL 8200 X and Y stage jog values are set to zero, then the JEOL stage mapping routine will acquire one mapping strip, then stop. So it is best to use the hardware jog settings for the JEOL and to not use software jogs in PFE.

Cheers,

Paul

A shift of 0.5 mm seems much too large to be caused by wear on the y-axis stage gear and/or bearing unless the wear is extreme.  When I've seen this problem occur during X-ray mapping (and then had the gear replaced), it was essentially invisible if the pixel size was about 2 microns or greater.  Since we haven't been on a service contract for years, I do much less X-ray mapping than I used to.  This makes me sad , but I know it is for the good of the instrument.

Wear on the y-axis stage gear should be easy to detect by collecting a stage scan BSE image using the X-ray mapping function in PC-EPMA on a material (such as brass with scratches in it) that has very fine linear features parallel to the x-axis.  Smooth linear or curvilinear features will appear "jittery" on a ~1-micron scale if the gear is worn excessively (through "normal" use).

[John Donovan]

I've had reports of all sorts of different size shifts from worn out stages. Not Cameca stages though which use optical linear encoders. The most common stage error you see on Cameca stages is that it fails to "home" on a specified position when the friction adjustment is too loose, and then times out giving a "motion not completed" error.

Brian and Paul are correct that it's the JEOL Y stage axis that takes a beating because of mapping.  See this topic which has some replicate stage scan images showing the sorts of reproducibility issues:

https://probesoftware.com/smf/index.php?topic=44.0

Be sure to log in to see the animated GIF attachments.

[Brian Joy]

I've had reports of all sorts of different size shifts from worn out stages. Not Cameca stages though which use optical linear encoders. The most common stage error you see on Cameca stages is that it fails to "home" on a specified position when the friction adjustment is too loose, and then times out giving a "motion not completed" error.

Brian and Paul are correct that it's the JEOL Y stage axis that takes a beating because of mapping.  See this topic which has some replicate stage scan images showing the sorts of reproducibility issues:

https://probesoftware.com/smf/index.php?topic=44.0

Be sure to log in to see the animated GIF attachments.

Hi John,

But how much time have you actually spent using a JEOL microprobe?  I get amazing stage reproducibility in a ten-year-old instrument that has no optical encoders.  Furthermore, failure of an optical encoder within the Cameca stage assembly can thrust one into a world of misery (and this is from personal experience).  Ask Edgar if you don't believe me -- he'll know what I'm talking about.  A JEOL stage is pragmatically simple in design.

Brian

[John Donovan]

I've had reports of all sorts of different size shifts from worn out stages. Not Cameca stages though which use optical linear encoders. The most common stage error you see on Cameca stages is that it fails to "home" on a specified position when the friction adjustment is too loose, and then times out giving a "motion not completed" error.

Brian and Paul are correct that it's the JEOL Y stage axis that takes a beating because of mapping.  See this topic which has some replicate stage scan images showing the sorts of reproducibility issues:

https://probesoftware.com/smf/index.php?topic=44.0

Be sure to log in to see the animated GIF attachments.

Hi John,

But how much time have you actually spent using a JEOL microprobe?  I get amazing stage reproducibility in a ten-year-old instrument that has no optical encoders.  Furthermore, failure of an optical encoder within the Cameca stage assembly can thrust one into a world of misery (and this is from personal experience).  Ask Edgar if you don't believe me -- he'll know what I'm talking about.  A JEOL stage is pragmatically simple in design.

Brian

Hi Brian,
Yet you hesitate to run stage mapping...   

About 80% of our customers are JEOL instrument users, so I am mainly speaking of experience in our working with them.

I'll just say this: the Cameca stage is reproducible to 1 um *without* a stage backlash operation being performed with every move.

[Brian Joy]

I've had reports of all sorts of different size shifts from worn out stages. Not Cameca stages though which use optical linear encoders. The most common stage error you see on Cameca stages is that it fails to "home" on a specified position when the friction adjustment is too loose, and then times out giving a "motion not completed" error.

Brian and Paul are correct that it's the JEOL Y stage axis that takes a beating because of mapping.  See this topic which has some replicate stage scan images showing the sorts of reproducibility issues:

https://probesoftware.com/smf/index.php?topic=44.0

Be sure to log in to see the animated GIF attachments.

Hi John,

But how much time have you actually spent using a JEOL microprobe?  I get amazing stage reproducibility in a ten-year-old instrument that has no optical encoders.  Furthermore, failure of an optical encoder within the Cameca stage assembly can thrust one into a world of misery (and this is from personal experience).  Ask Edgar if you don't believe me -- he'll know what I'm talking about.  A JEOL stage is pragmatically simple in design.

Brian

Hi Brian,
Yet you hesitate to run stage mapping...   

About 80% of our customers are JEOL instrument users, so I am mainly speaking of experience in our working with them.

I'll just say this: the Cameca stage is reproducible to 1 um *without* a stage backlash operation being performed with every move.

Hi John,

That's a mischaracterization of what I wrote.  I do plenty of X-ray mapping, I just don't do it excessively (as I might have been accused of doing in the past, just because I like pretty pictures).  I have plenty of experience with both JEOL and Cameca stages, and I much prefer the former.

Brian

[Brian Joy]

Would you like me to unleash "other Brian" upon you?? 

[entoptics]

Greetings,

Thought I'd chime in with my experience with JEOL stage issues (8500F), though I'm not sure it's relevant to the OP's issue. Also, this happened ~1.5 years ago, so a little fuzzy on details.

I had intermittent reproducibility issues, with a somewhat "randomly occurring, but fixed magnitude" behavior. It was particularly weird, because it seemed to only occur in certain regions of the stage. We serviced the stage, and cleaned the mirrors, checked gears and bearings, etc without curing it.

We finally figured out it was an issue with the stage controller boards. I don't know the details or reason, but the fix involved an oscilloscope and lissajous figures. 

Cheers,

Scott B.

[Rom]

Greetings,
Our service engineer checked, cleaned etc. our JEOL, including stage.
The shift problem was not disappeared.
Engineer Sergiy discussed this issue with JEOL USA colleagues, but they could not confirm that the similar problem is experienced by many other 8200 instruments.
I hope to link our JEOL engineer with entoptics (Scott B.)'s JEOL engineer  (sent him a mail).

But may be somebody could recommend us some test for understanding the source our problem of huge scene shift which happens absolutely unpredictably.
Thank you!