JEOL Spectrometer Backlash

[John Donovan]

Scott Boroughs and I have been looking at situations where the emission line of interest is close to the lower limit of the spectrometer range on JEOL instruments.  Probe for EPMA has included an automatic adjustment to account for the built in JEOL jog that can exceed the lower limit in these cases.

But previously we had assumed a constant 0.6 mm spectrometer jog, which might not be what your instrument actually uses, so recently we modified the code to utilize the actual JEOL backlash value from the MOTORS.DAT file as seen here:

     "SP1"   "SP2"   "SP3"    "SP4"    "SP5"       "X"       "Y"       "Z"    "motor labels"
    60.4     82.4     60.4     60.4     60.4       1.5       1.8      9.0     "low limit spect/stage motors"
   254.0    254.0    254.0    254.0    254.0      99.2      99.2      12.9    "high limit spect/stage motors hilim"
742.8646 742.8646 742.8646 742.8646 742.8646 27593.425 27593.425 27593.425    "unit to steps conversion"
.014381  .014381  .014381  .014381  .014381    1000.0    1000.0    1000.0    "unit to LiF200 angstrom or micron conversion"
       5        5        5        5        5        20        20        20    "completion step tolerance for motors"
   -400.    -400.    -400.    -400.    -400.     -400.     -400.     -400.    "backlash size, (hilimit - lolimit)/backlash size"
     200      200      200      200      200       200       200       200    "derivative gain"
     750      750      750      750      750       700       700       700    "integral gain"
     700      700      700      700      700       700       700       700    "integration limit"
    2000     2000     2000     2000     2000      2000      2000      2000    "following error"
      20       20       20       20       20       300       300       300    "acceleration"
  200000   200000   200000   200000   200000    400000    400000    400000    "velocity"
     600      600      600      600      600       700       700       700    "gain"
      11       11       11       11       11        11        11        11    "hard limit mode (low xor high xor smooth)"
     1.0      1.0      1.0      1.0      1.0       1.0       1.0       1.0    "joystick sensitivity"
     1.4      1.4      1.4      1.4      1.4       1.4       1.4       1.4    "joystick acceleration"
    .002     .002     .002     .002     .002      .002      .002      .002    "backlash tolerance"
   180.0    180.0    180.0    180.0    180.0        12        12        10    "park positions"
  500000   500000   500000   500000   500000    400000    400000     40000    "JEOL velocities"
   50000    50000    50000    50000    50000     -5000     -5000      5000    "Jeol Backlash (1/100th micrometer units)"
       0        0        0        0        0         0         0         0    "SX100 velocities"
       0        0        0        0        0         0         0         0    "SX100 minimum speeds"
       0        0        0        0        0         0         0         0    "unused"

Note that the JEOL backlash values are specified in 1/100th micrometer units (this is what JEOL uses internally to avoid floating point math!), therefore 50000 equals 500 um or 0.5 mm.  The new code uses these JEOL backlash values from the MOTORS.DAT file whatever they may be.

On 8900/8200 and 8500 instruments these values are read in from the MOTORS.DAT and sent to the JEOL instrument so they will be utilized in the automatic backlash adjustments for the spectrometers and stage. 

Note also that the newer 8230/8530/iSP100 and iHP200F instruments do not have an API command to accept these JEOL backlash values sent through the instrument network, so the values in the JEOL PC-EPMA software will need to be edited so they match the values in the MOTORS.DAT file on these newer instruments.  Though to be honest we do not know where in the JEOL PC-EPMA software these backlash values are edited, so if anyone does know, please let us know here.

Note also that these backlash values from the MOTORS.DAT will be utilized based on these keywords in the Probewin.ini file and explained in the Probe for EPMA User's Reference Manual:

SpecBacklashFlag=1      ; spectro backlash
StageBacklashFlag=1      ; stage backlash
SpecBacklashType=1      ; spectro backlash type, 1 = software backlash, 2 = ROM based backlash
StageBacklashType=2      ; stage backlash type, 1 = software backlash, 2 = ROM based backlash

Therefore the backlash "flag" needs to be non-zero and the backlash "type" needs to be 2 for the automatic hardware based backlash method which uses these MOTORS.DAT values.   While the "smart" software backlash correction (backlash "type" = 1) (which applies to both JEOL and Cameca instruments), the "backlash factors" on line 7 in the MOTORS.DAT file are what are utilized.

Finally, we should note that although we had been checking the low off-peak, ROM scanning and peaking lower limits, we were not checking the multi-point background (MPB) limits for this JEOL backlash going beyond the lower spectrometer limit.  New code to check the lower MPB positions  has now been added in the latest version of Probe for EPMA (v. 13.6.9).

So now, in this latest version of PFE, if these spectrometer positions need to be adjusted, the software will provide a warning to the main log window that this adjustment was performed as seen here:

[Ben Buse]

Hi,

What's the difference between "Jeol Backlash (1/100th micrometer units)" and "backlash size, (hilimit - lolimit)/backlash size"?

What does "backlash size, (hilimit - lolimit)/backlash size" do?

Thanks

Ben

[John Donovan]

What's the difference between "Jeol Backlash (1/100th micrometer units)" and "backlash size, (hilimit - lolimit)/backlash size"?

What does "backlash size, (hilimit - lolimit)/backlash size" do?

The "Jeol Backlash" is a hardware (ROM based) backlash, the "backlash size" is a software backlash. We did this because some people prefer the hardware backlash method, some prefer the software backlash.

These are defined by these keywords in the Probewin.ini file:

SpecBacklashFlag=1      ; spectro backlash (0 = no, non-zero = yes)
StageBacklashFlag=1      ; stage backlash (0 = no, non-zero = yes)
SpecBacklashType=1      ; spectro backlash type, 1 = software backlash, 2 = ROM based backlash
StageBacklashType=2      ; stage backlash type, 1 = software backlash, 2 = ROM based backlash

To be honest, we worked on this with Paul Carpenter and others with JEOL instruments (because it doesn't apply to Cameca instruments), so we can't really say exactly how it is utilized in all circumstances. If you have a specific question we can look at the code and get back to you.

[Joe Boesenberg]

John

I had the JEOL engineers here yesterday trying to address the stage issues on my iSP-100. Several hours after saving a point the stage will return to the proper X and Y positions, but the Z is often visibly off by 5-10 microns - enough to screw around with your analysis. John Glass noticed the Z backlash in. the JEOL software was set to do a 50 micron backlash rather than the 100 micron backlash, so it has now been changed and I am testing it. One of the questions that came up was whether PfE was doing the backlash when it returned to the point for analysis. We determined that it was indeed, but we were unsure what the probewin.ini file meant concerning the backlash.  There are two lines:

StageBacklashFlag=0;  0=false no stage backlash PFW, and nonzero for true do PFW backlash (overridden by Move dialog box)

StageBacklashType=2;  1=software backlash 2=ROM based backlash   


Can you explain what these mean?

Also, does PFE do the same magnitude backlash as the JEOL software setup (now 100 microns) and in the same direction (approaching from under or over the Z value). John Glass mentioned the JEOL software approaches from under the Z value.

Thanks. Joe
   

[John Donovan]

Also, does PFE do the same magnitude backlash as the JEOL software setup (now 100 microns) and in the same direction (approaching from under or over the Z value). John Glass mentioned the JEOL software approaches from under the Z value.   

If you are using the so called "ROM" based stage (or spectrometer) backlash, PFE uses the backlash values specified in the MOTORS.DAT file in 1/100th micrometer units.

See above posts for details and let us know if you have further questions.

[Joe Boesenberg]

John

Just to be sure, and maybe I am a little dense, if the StageBacklashFlag=0, that means it is only reading the JEOL hardware, not anything in the PfE software, correct?

Thanks
Joe

[John Donovan]

John

Just to be sure, and maybe I am a little dense, if the StageBacklashFlag=0, that means it is only reading the JEOL hardware, not anything in the PfE software, correct?

Thanks
Joe

Most labs will set it to these flags:

StageBacklashFlag=0;  0=false no stage backlash PFW, and nonzero for true do PFW backlash (overridden by Move dialog box)
StageBacklashType=2;  1=software backlash 2=ROM based backlash   

That means it will use the ROM hardware based backlash (not the PFE software backlash), based on the values in the MOTORS.DAT file.

[Joe Boesenberg]

John

Thanks very much for the clarity. The value in my Motors.dat file is 5000 (which is 50 microns), which I will change to 10000. Does whether the value is positive or negative (10000 or -10000) indicate the direction of the backlash (positive backlashes in a higher direction than the Z value and negative the lower direction)?

Joe

[John Donovan]

Thanks very much for the clarity. The value in my Motors.dat file is 5000 (which is 50 microns), which I will change to 10000. Does whether the value is positive or negative (10000 or -10000) indicate the direction of the backlash (positive backlashes in a higher direction than the Z value and negative the lower direction)?

I took a look at the JEOL backlash code and as far as I can tell, for JEOL 8230/8530/iSP100/iHP200F instruments we rely completely on the JEOL instrument for stage and spectrometer backlash functions. So I don't think editing these parameters will have any effect.

[Joe Boesenberg]

Thanks John. Appreciate it. Makes things somewhat easier. I am going to leave the change I made in the Motors.dat file anyway to match the recommended JEOL setting. Joe