News:

:) Remember, you need to be logged in to see posted attachments!

Main Menu

Wish List for PFE Features

Started by jeb, October 07, 2013, 11:35:16 AM

Previous topic - Next topic

John Donovan

Quote from: dawncruth on February 24, 2023, 01:02:02 PM
Is it possible to include the beam 2 values in the Analyze window? Right now, you can only see them in the working window. See attached files for reference.

Almost everything is possible. But some things are more work than others!. 

Let us take a look at this.
John J. Donovan, Pres. 
(541) 343-3400

"Not Absolutely Certain, Yet Reliable"

John Donovan

Quote from: dawncruth on February 24, 2023, 01:02:02 PM
Is it possible to include the beam 2 values in the Analyze window? Right now, you can only see them in the working window. See attached files for reference.

Your wish is granted:



John J. Donovan, Pres. 
(541) 343-3400

"Not Absolutely Certain, Yet Reliable"

entoptics

Is it possible to change the frequency of beam current measurements during an analysis run? I poked around, and all I found was "on/off" in the Acquisition Options dialogue, and didn't see anything relevant in the Probewin.ini file.

I currently have an incredibly stable gun, with typical drift of 0.05% per hour, and as you know, due to the Faraday delays on JEOL probes, the current measurement protocol actually eats up 5-10 seconds per analysis. Would be super cool if I could set it to read the current every X number of analyses, or every X number of minutes.

I can see how that might be tricky to code into the current normalization calculations in Analyze!, but thought I'd mention it and see what you thought.

John Donovan

#588
Quote from: entoptics on April 02, 2023, 02:27:14 PM
Is it possible to change the frequency of beam current measurements during an analysis run? I poked around, and all I found was "on/off" in the Acquisition Options dialogue, and didn't see anything relevant in the Probewin.ini file.

I currently have an incredibly stable gun, with typical drift of 0.05% per hour, and as you know, due to the Faraday delays on JEOL probes, the current measurement protocol actually eats up 5-10 seconds per analysis. Would be super cool if I could set it to read the current every X number of analyses, or every X number of minutes.

Your wish is granted!    :)

We modified the code to add a new checkbox in the Acquisition Options (v. 13.2.9) that now allows the user to only acquire the beam (and absorbed) current once per sample as shown here:


 
When this option is selected then the program only acquires a single beam (and absorbed) beam current measurement per standard or unknown sample and just re-load that first measurement for subsequent data points in that sample, as seen here:

Off-Peak Corrected or MAN On-Peak X-ray Counts (cps/1nA) (and Faraday/Absorbed Currents):
ELEM:     c ka   si ka   BEAM1   BEAM2   ABSD1   ABSD2
  194G    -.04   96.52 140.059 140.059 110.412 110.412
  195G    -.03   96.85 140.059 140.059 110.412 110.412
  196G    -.05   96.85 140.059 140.059 110.412 110.412

AVER:     -.04   96.74 140.059 140.059 110.412 110.412
SDEV:      .01     .19    .000    .000    .000    .000
1SIG:      .00     .26
SERR:      .00     .11
%RSD:   -16.08     .20
John J. Donovan, Pres. 
(541) 343-3400

"Not Absolutely Certain, Yet Reliable"

dawncruth

I like the Sort Stat and Data Grids by Geologic/Atomic Number in the Analyze window (see attached screenshot). Is there a way to turn that on all the time?

John Donovan

#590
Quote from: dawncruth on May 23, 2023, 12:01:01 PM
I like the Sort Stat and Data Grids by Geologic/Atomic Number in the Analyze window (see attached screenshot). Is there a way to turn that on all the time?

The only way you can have it "on all the time" is to simply enter (or change the order of) the elements in a sample. For example either when entering the analyzed elements manually from the Elements/Cations dialog or using the up/down arrows in the Combined Conditions dialog from the Acquire! window.

However, you can specify the default sort order by editing the Probewin.ini file as seen here:

[software]
GeologicalSortOrder=1      ; 0 = no sorting, 1 = traditional, 2 = low to high Z, 3 = high to low Z
John J. Donovan, Pres. 
(541) 343-3400

"Not Absolutely Certain, Yet Reliable"

JonF

Hi John,

I know there's been a lot of work in all the new deadtime corrections, and they look great, but would it be possible to decrease the available deadtimes to below 0.1us? I've tried changing the deadtime in both PfE and directly via the scalars.dat file, but I come up against an error both ways so I'm guessing there's a sanity check in the software.

0.0001us or something would be great  :)

Thanks

John Donovan

#592
Quote from: JonF on May 26, 2023, 02:45:13 AM
Hi John,

I know there's been a lot of work in all the new deadtime corrections, and they look great, but would it be possible to decrease the available deadtimes to below 0.1us? I've tried changing the deadtime in both PfE and directly via the scalars.dat file, but I come up against an error both ways so I'm guessing there's a sanity check in the software.

0.0001us or something would be great  :)

Thanks

"Sanity check" is a good way to describe the 0.1 us minimum dead time allowed in the SCALERS.DAT file...    :)

Seriously, why would you need this?  No WDS hardware I know of can go that low!  If you only want to turn off the dead time correction, just turn it off in the Analytical | Analysis Options dialog!

But actually there is a way... just use the Analytical | Update Dead Time Constants menu dialog to edit the values to any value between zero and 20 us.  It only updates values for the samples you select in the current run, but it's good enough for "what if" types of questions. You will get a warning to the log window if the value is zero though.

I'm just curious why you'd want to do this.... if you just want to try some modeling just use the TestEDS application and use the Output | Output Dead Time Calculations menu as described here:

https://smf.probesoftware.com/index.php?topic=1466.msg11025#msg11025
John J. Donovan, Pres. 
(541) 343-3400

"Not Absolutely Certain, Yet Reliable"

JonF

Quote from: John Donovan on May 26, 2023, 08:05:02 AM
But actually there is a way... just use the Analytical | Update Dead Time Constants menu dialog to edit the values to any value between zero and 20 us.  It only updates values for the samples you select in the current run, but it's good enough for "what if" types of questions. You will get a warning to the log window if the value is zero though.

Works like a charm, thanks!

Quote from: John Donovan on May 26, 2023, 08:05:02 AM
Seriously, why would you need this?  No WDS hardware I know of can go that low! 

The SXES can!

John Donovan

#594
Quote from: JonF on May 26, 2023, 10:35:50 AM
Quote from: John Donovan on May 26, 2023, 08:05:02 AM
Seriously, why would you need this?  No WDS hardware I know of can go that low! 

The SXES can!

Really!?   I didn't know that.  Too bad JEOL can't utilize the same pulse processing electronics for our WDS spectrometers!

I'm assuming that PFE is not reading a SXES spectrometer, correct? But if you are somehow able to read the SXES spectrometer in PFE I will edit the "sanity check" for reading the values in the SCALERS.DAT file.  How low can the dead time values get on the SXES?
John J. Donovan, Pres. 
(541) 343-3400

"Not Absolutely Certain, Yet Reliable"

JonF

Quote from: John Donovan on May 26, 2023, 10:50:20 AM
Really!?   I didn't know that.  Too bad JEOL can't utilize the same pulse processing electronics for our WDS spectrometers!

I'm assuming that PFE is not reading a SXES spectrometer, correct? But if you are somehow able to read the SXES spectrometer in PFE I will edit the "sanity check" for reading the values in the SCALERS.DAT file.  How low can the dead time values get on the SXES?

The SXES is using a CCD rather than proportional counters, so they can work incredibly quickly. The "deadtime" they experience is probably more akin to that of EDS, with the deadtime being proportional to the frequency of the channel reads/resets. I haven't actually run in to a deadtime issue yet, and I know Nick Wilson from CSIRO gave a talk back in M&M 21 that showed a straight line when plotting peak count rate vs beam current up in to the hundreds of nanoamps (as opposed to the curves that we see with GFPCs), and I've seen similar with mine plot cps/na vs cps.

One big advantage of the SXES over GFPCs is the ability to simultaneously acquire backgrounds and integrate peaks for the light elements (I'm currently doing boron and nitrogen), but this leads to large count rate values (100ks of counts) and hence I'm running in to issues with the deadtime correction, not just in PfE but CalcImage too. It'd be great if you could alter that sanity check for the SCALERS.DAT file for both PfE and CalcImage!

John Donovan

#596
Quote from: JonF on May 29, 2023, 02:20:33 AM
The SXES is using a CCD rather than proportional counters, so they can work incredibly quickly. The "deadtime" they experience is probably more akin to that of EDS, with the deadtime being proportional to the frequency of the channel reads/resets. I haven't actually run in to a deadtime issue yet, and I know Nick Wilson from CSIRO gave a talk back in M&M 21 that showed a straight line when plotting peak count rate vs beam current up in to the hundreds of nanoamps (as opposed to the curves that we see with GFPCs), and I've seen similar with mine plot cps/na vs cps.

Nice to hear that the SXES spectrometer can handle such high count rates and still have a linear response. However, the non-linear response of our GFPC electronics (for count rates over 50 kcps) can be dealt with using the new logarithmic dead time expression that we just published a couple of weeks ago in M&M:

https://smf.probesoftware.com/index.php?topic=1466.msg11858#msg11858

This non-linear expression (and the constant k-ratio procedure outlined in the paper) using Probe for EPMA, is outlined in a pdf available from the Help menu in PFE.

Quote from: JonF on May 29, 2023, 02:20:33 AM
One big advantage of the SXES over GFPCs is the ability to simultaneously acquire backgrounds and integrate peaks for the light elements (I'm currently doing boron and nitrogen), but this leads to large count rate values (100ks of counts) and hence I'm running in to issues with the deadtime correction, not just in PfE but CalcImage too. It'd be great if you could alter that sanity check for the SCALERS.DAT file for both PfE and CalcImage!

I'm still not following: why do you need to read in dead time values less than 0.1 us from the SCALERS.DAT file (the sanity check code is the same for both PFE and CalcImage)?  Are you able to read data from the SXES into PFE somehow?
John J. Donovan, Pres. 
(541) 343-3400

"Not Absolutely Certain, Yet Reliable"

JonF

#597
Yeah, I've got the SXES data going in to PfE and CalcImage.

It'd be great if I could put a dead time value as low as 0.0001us or something ridiculously small in to scalers.dat. My worry is that the very large CPS/nA values I get from integrating the SXES peaks is getting inappropriately dead time corrected.

John Donovan

#598
Quote from: JonF on May 29, 2023, 12:05:33 PM
Yeah, I've got the SXES data going into PfE and CalcImage.

It'd be great if I could put a dead time value as low as 0.0001us or something ridiculously small in the scalers.dat. My worry is that the very large CPS/nA values I get from integrating the SXES peaks is getting inappropriately dead time corrected.

Very, very interesting.  Let me see what we can do.  So is the SXES spectrometer sp5 or something in the SCALERS.DAT file?

In the meantime you can certainly just use the Analytical | Update Dead Time Constants menu dialog to select samples and update the dead time values.

Of course dividing the SXES intensities (standard and maps) by a factor of 10x or 100x works too since there's no dead time to speak of.

You should post some results to the CalcImage board. I'd be very interested to see this data.
John J. Donovan, Pres. 
(541) 343-3400

"Not Absolutely Certain, Yet Reliable"

JonF

SP3 at the minute, I hadn't thought of trying to add SP5 to our instrument. I might have a think about that!
Our SXES-LR on the 8530F takes up the SP5 position, hence our instrument only having 4 "traditional" WDS.

For the Analytical | Update Dead Time Constants - is there an equivalent way of doing this for CalcImage?
At the minute I'm doing exactly as you suggest - dividing the measured intensities across the board by an order of magnitude or two, which wouldn't make a difference in the grand scheme of things, but it would be nice if I didn't have to!

I'll see what I can do about creating some demo data for quant mapping with the combined WDS and SXES: the work I've been doing has all been on client data. I work quite a lot with light elements, and the SXES is a game changer for this.