Multiple Sample Setups

[gmorgan@ou.edu]

John,

This is probably pretty basic, so maybe there are a few others out there that might have, or need, input as well.
I have a couple of basic questions about using multiple sample setups in an analytical session, in this case derived from a list of elements and standards contained in a single MDB. I created a database for analyzing two different mineral phases in a given set of samples, and it contains 15 analyzed elements, not all of which are to be analyzed in both phases. After acquiring standard intensities, from the initial sample and list of elements, I created two sample setups: one for amphibole, one for plagioclase, with slightly different sets of elements. Okay - easy so far. Now for two simple questions:

1. If I go to acquire standards again, will the standards be acquired as they were during the first acquisition (all 15 elements on all selected standards), or will they be acquired according to a chosen sample setup? I'm guessing all 15 as they were first acquired...?

2. When I digitize a sequence points for analysis, I believe the total time for analysis calculated by PFE corresponds to the method used to acquire all 15 elements as the standards were first acquired rather than on the basis of actual time required for the subsets of elements used in the two sample setups. Is this correct? I am guessing this because PFE is calculating a total time for the automated run (about 5.5 minutes per point) that is quite a bit longer than I calculate based upon the elements actually analyzed (about 3.5 minutes per point).

Thanks.

[Gseward]

George,

1. If I go to acquire standards again, will the standards be acquired as they were during the first acquisition (all 15 elements on all selected standards), or will they be acquired according to a chosen sample setup? I'm guessing all 15 as they were first acquired...?

The Standards will be acquired with a setup defined by the state of the buttons in the bottom right of the Automate! window. e.g. 'Use Last Unknown Sample' 'Use Digitized Conditions' 'Use Digitized Sample Setups' etc.

In your case it seems that you want to use the original 15 element general setup (edit:for the standards)? in which case, create a Setup for the standard data collection in the same way you did for the unknowns (edit: setups Amphibole and Plagioclase):
In Analyze! select an appropriate unknown with the setup that you want (i.e. the general setup you used to acquire the original standards), then click 'Add to Setup'. Now, select your standards and click the 'sample setups' button. Select the newly created 15 element general setup, and apply this to the standard positions.  Now when you run the data collection from Automate! using 'Use Digitized Sample Setups', Standards, Amphiboles and Plagioclase will be collected with the appropriate setup. Note that to collect Standards and unknowns in the same Automation you will need to use the 'All Samples' display option for the Position List so that you can select Standards AND Unknowns together.

2. When I digitize a sequence points for analysis, I believe the total time for analysis calculated by PFE corresponds to the method used to acquire all 15 elements as the standards were first acquired rather than on the basis of actual time required for the subsets of elements used in the two sample setups. Is this correct? I am guessing this because PFE is calculating a total time for the automated run (about 5.5 minutes per point) that is quite a bit longer than I calculate based upon the elements actually analyzed (about 3.5 minutes per point).

The time estimate is often longer than I would expect. I assume this is because there are many variables that must be estimate (spectro motion time, setting conditions, communications, etc. etc.). At this stage, you might consider running the experiment and then looking at the Time Error as displayed at the top of the Automate! window. See here:

http://probesoftware.com/smf/index.php?topic=42.msg2700#msg2700

This will give you a better idea of time estimates based on your instrument.

Gareth

[Probeman]

1. If I go to acquire standards again, will the standards be acquired as they were during the first acquisition (all 15 elements on all selected standards), or will they be acquired according to a chosen sample setup? I'm guessing all 15 as they were first acquired...?

Hi George,
What Gareth said is correct.  You can assign a different element setup to each standard. One other method I use sometimes is to first create a "mother of all" samples, which contains all the elements I want in all sample setups. I use this (MOA) sample to peak all spectrometers, test all PHAs and acquire standard intensities.

Then once the standard intensities are acquired, I make a new sample, name it something like "feldspar" for example, and delete everything that isn't "feldspar".  Same for the "amphibole" sample setup.

Then I use the "feldspar" and "amphibole" sample setups for unknowns (where appropriate) and use the MOA sample for re-standardizing.

2. When I digitize a sequence points for analysis, I believe the total time for analysis calculated by PFE corresponds to the method used to acquire all 15 elements as the standards were first acquired rather than on the basis of actual time required for the subsets of elements used in the two sample setups. Is this correct? I am guessing this because PFE is calculating a total time for the automated run (about 5.5 minutes per point) that is quite a bit longer than I calculate based upon the elements actually analyzed (about 3.5 minutes per point).

The Automate! window time estimation percent is great for checking the entire automation time estimate accuracy which on my Sx100 is usually within 5% or so.  But you can also check the single point time estimation accuracy.  Basically when the counting on the spectrometer(s) finishes, you'd like to see the yellow elapsed time bar level with the element gauges as seen here:



In this particular example, the Cr Lo peak measurements still has a couple of seconds "on the clock", but all the other elements have finished.

[Gseward]

Hi George,
What Gareth said is correct.  You can assign a different element setup to each standard. One other method I use sometimes is to first create a "mother of all" samples, which contains all the elements I want in all sample setups. I use this (MOA) sample to peak all spectrometers, test all PHAs and acquire standard intensities.

Then once the standard intensities are acquired, I make a new sample, name it something like "feldspar" for example, and delete everything that isn't "feldspar".  Same for the "amphibole" sample setup.

Then I use the "feldspar" and "amphibole" sample setups for unknowns (where appropriate) and use the MOA sample for re-standardizing.

Based on George's original post, I think this is what he was already doing, minus the 'MOA' Setup, which he hadn't created. I was trying to suggest how to create the MOA setup for the standards, and use it. I guess this wasn't obvious. I'll edit the OP.

[John Donovan]

Hi George,
What Gareth said is correct.  You can assign a different element setup to each standard. One other method I use sometimes is to first create a "mother of all" samples, which contains all the elements I want in all sample setups. I use this (MOA) sample to peak all spectrometers, test all PHAs and acquire standard intensities.

Then once the standard intensities are acquired, I make a new sample, name it something like "feldspar" for example, and delete everything that isn't "feldspar".  Same for the "amphibole" sample setup.

Then I use the "feldspar" and "amphibole" sample setups for unknowns (where appropriate) and use the MOA sample for re-standardizing.

Based on George's original post, I think this is what he was already doing, minus the 'MOA' Setup, which he hadn't created. I was trying to suggest how to create the MOA setup for the standards, and use it. I guess this wasn't obvious. I'll edit the OP.

At first I wasn't quite sure what George was asking, but after reading your response it made much more sense.  I was just agreeing with you...
john

[gmorgan@ou.edu]

John,

Even stranger... The first session went fine, but when I ran a second set of samples, even though the feldspar samples in the digitized list clearly had the feldspar sample setup declared/attached, they were analyzed with the amphibole sample setup. I just went back and checked, and if I select a one of the feldspars in the digitized list and click on Sample Setup it clearly shows "Feldspar Template" (and I double checked this prior to starting the analysis) but it was analyzed with the amphibole template (a different set of elements).
Am I missing something, or is not the sample setup attached to a digitized point via "Sample Setups"?
Is this a common occurrence? Can I not trust the program to do as instructed?

[John Donovan]

Even stranger... The first session went fine, but when I ran a second set of samples, even though the feldspar samples in the digitized list clearly had the feldspar sample setup declared/attached, they were analyzed with the amphibole sample setup. I just went back and checked, and if I select a one of the feldspars in the digitized list and click on Sample Setup it clearly shows "Feldspar Template" (and I double checked this prior to starting the analysis) but it was analyzed with the amphibole template (a different set of elements).
Am I missing something, or is not the sample setup attached to a digitized point via "Sample Setups"?
Is this a common occurrence? Can I not trust the program to do as instructed?

Hi George,
Pretty sure you are missing something as many of us have been using this feature almost everyday for years!  You might want to triple check your sample setup assignments!     

This could help: note the Run | List Selected Sample Positions menu in Automate which you can access using a right click on the position sample list as seen here:



This will list all or selected position samples and any sample setup assignments if assigned.

For greater clarity in these discussions I would suggest saying "acquired" when discussing acquisition, and saying "analyzed" when discussing the conversion of intensities to concentrations. Because these are separate steps in the PFE software it makes things easier to understand.

Another thought- did you accidentally select the Use Last Unknown Sample as the automation basis for your 2nd acquisitions, instead of Use Digitized Sample Setups?

[gmorgan@ou.edu]

John,

As I've moved on and no longer have that MDB loaded and have deleted the samples in the position list, I'm not sure can only guess is that perhaps "Use Last Sample Setup" may have been enabled rather than "Use Digitized File Setups". I'm pretty sure I didn't change that after the first session, but does it reset when an acquisition automation finishes and/or when the position list is cleared to starts over?
Dang, one more little flag to have to remember. Why doesn't the software see multiple setups in the list and automatically select, or ask the user to select, the conditions flag? Or does it?....

[Probeman]

As I've moved on and no longer have that MDB loaded and have deleted the samples in the position list, I'm not sure can only guess is that perhaps "Use Last Sample Setup" may have been enabled rather than "Use Digitized File Setups". I'm pretty sure I didn't change that after the first session, but does it reset when an acquisition automation finishes and/or when the position list is cleared to starts over?
Dang, one more little flag to have to remember. Why doesn't the software see multiple setups in the list and automatically select, or ask the user to select, the conditions flag? Or does it?....

Hi George,
Well I guess we'll never know for sure, but since the software *does* remember your last automation basis selection, even if you close and re-open the run (test this for yourself please, but it does!), I suspect you accidentally clicked the Use Last Unknown Sample option some time before you launched your 2nd automated acquisition...  the software tries to help you, but it does allow the user to change their mind- and apparently you did- whether you meant to or not!

[John Donovan]

I just wanted to make a comment about the "Run Multiple Sample Setups One at a Time" feature in the Probe for EPMA Automate! window under Multiple Setups.

To utilize multiple sample setups per position sample, you first open the Automate! window and digitize the positions that you want to run multiple sample setups on.  Note that sample setups can be added to your sample setup list using the Add To Setups button in the Analyze! windows as described here:



Then, from the Automate! window, you select the position sample(s) that you want to run multiple samples on:



Then you click the Multiple Setups button and select the sample setups that you want to run on those position samples:



Now, there are two ways that you can run these multiple samples setups on your positions samples. The default is to run all the samples setups on each position sample.  Then it proceed to the next position sample...

"Run Multiple Setups One at a Time" is NOT checked:
position sample 1 (5 points)  ---->  sample setup 1
position sample 1 (5 points)  ---->  sample setup 2
position sample 1 (5 points)  ---->  sample setup 3

position sample 2 (5 points)  ---->  sample setup 1
position sample 2 (5 points)  ---->  sample setup 2
position sample 2 (5 points)  ---->  sample setup 3
etc.

This is the mode one would usually utilize when running say multiple accelerating voltages on thin film samples so thin film composition and thickness can be calculated using the method of Pouchou in the STRATAGem or BadgerFilm applications:

https://probesoftware.com/smf/index.php?topic=111.0

But by checking the "Run Multiple Setups One at a Time" checkbox, Probe for EPMA will instead run all the selected position samples in the Automate! window using the first selected sample setup, and then run all the position samples using the next sample setup...

"Run Multiple Setups One at a Time" is checked:
position sample 1 (5 points)  ---->  sample setup 1
position sample 2 (5 points)  ---->  sample setup 1

position sample 1 (5 points)  ---->  sample setup 2
position sample 2 (5 points)  ---->  sample setup 2

position sample 1 (5 points)  ---->  sample setup 3
position sample 2 (5 points)  ---->  sample setup 3
etc.

This acquisition mode is preferred when attempting to utilize the "Constant K-Ratio" method developed by John Fournelle and myself for the calibration of ones dead time, picoammeter linearity and determining ones spectrometer effective take-off angles:

https://probesoftware.com/smf/index.php?topic=1466.msg11196#msg11196

The reason is that the primary and secondary standards will be acquired at the same beam current so any non-linearity in the picoammeter will be nulled out.  Though one must also be careful to always run the primary standard before the secondary standard, and make sure that the standard intensity drift correction is turned off!  These steps ensures that the primary standard is utilized for the secondary standard measured at the same beam current...

I hope that makes sense to everyone. Please let me know if you have any questions.

[John Donovan]

We made some minor changes to the latest version of Probe for EPMA v13.8.3, but the only thing worth mentioning is that we added an "Interactive Help": button to the Multiple Setups dialog which is accessed from the Automate! window as seen here:



which now connects the user to the post above.  I think it's worth having because even I sometimes have to think for a second about which option I want to utilize with regards to the "Run Multiple Samples Setups One at a Time" checkbox!