Hi,
I'm running into a bit of a mind-roadblock upon 'converting' our JEOL analysis routines to run using PfE, while also still getting my head around the logic of PfE and using sample setups.
1. If a user wants to analyze several different phases (e.g., feldspar, pyroxene, glass) in one session (e.g., over several days) is this done from within just one mdb file or several? If several files, then I can't run things sequentially overnight as I would do using JEOL software, using serial analysis?
2. Also, let's say it is all in one file, how can I analyze different phases that utilize different analytical conditions, but are also using some of the same standards? To elaborate further, I have a glass recipe/setup, using combined conditions (8nA for Na, K, Si, other elements 10nA) which is working well. I also created a recipe for feldspars and pyroxenes in one mdb file, everything measured at 12nA, starting with a general setup for wavescan, peaking, PHA and calibration including all standards and elements I need for both phases, then split it off into 2x new sample setups where I adjust elements, counting times and standard assignments. Then assign to my unknown samples which one it should use. Also works well. This is as far as the youtube tutorials and other posts on here have gotten me thus far.
What about if I wanted to add glass setup in as well? Or other minerals, that use other conditions again (e.g. I'd measure olivine with 20nA current)? We've got a student coming on in the next few weeks where I know there will be olivine, feldspar, pyroxene, amphiboles and glass, and it'll be 10-20 thin sections, with not that many points per each. I can fit 3 into the probe at the same time, so would need to finish everyting on those 3, then swap for new ones, rinse and repeat. With JEOL I'd have all the calibrations and a quant file for each phase sitting in serial analysis, then able to jump between files to set points, finally assign calibrations and run everything overnight. Swap for new samples the next day to keep going. Just needs good record keeping of quant files, but otherwise straightforward. Without any glitches or system crashes in between of course ;)
I think two main issues I'm trying to understand if I was to integrate all into one mdb file using sample setups?
1. There are standards that are used for both glass and the px/fsp routine (e.g., TiO2, Cr2O3), but the current is different. Only one sample name in the list of standards. How would that work with different conditions and calibration? I do need to recalibrate 'in between', maybe every 1-2 days depending on weather conditions and can't just calibrate once and rely on those intensities for a whole week.
2. I don't want the instrument to change the current up and down while going through the unknowns during analysis with assigned different conditions. There are crystal flips as well as current differences between setups and I do not want to change current/flip at every point. Best would be all pyroxenes first, then feldspars, and so on. Just making the change once, e.g. the crystal flip. Definitely not keen on back and forth all the time.
I'm sure this is routine for most people and I just haven't quite figured out how to do it yet. Any insights appreciated, even if it's just for one of my questions :)
Cheers,
Kerstin
Hi Kerstin,
Yes I'd do this in one mdb, with sample setups
Calibrate all elements
Then break it into different unknown samples - which store as sample setups, and in automate select which sample setup to use
For different beam currents during an analysis - use combined conditions - using PCC files to store column conditions may be quicker - and if there is a beam shift can correct the beam shift.
Ben
I think Ben is correct. You can probably do everything you want to do using sample setups in a single MDB file, by creating a "mother of all samples" sample. Then tuning up all the elements, then breaking that sample into your mineral specific samples with the count times, calculation options, etc. that you want for each specific mineral. Then save those mineral specific sample setup from the Analyze! window using the Add To Sample Setups button. Then assign those sample setups to the positions in the Automate! window using the Sample Setups button.
This is explained starting on page 200 of the User's Reference Manual PDF available from the PFE Help menu (but who reads manuals these days!).
(https://smf.probesoftware.com/gallery/1_15_10_25_9_26_02.png)
Either way, it is worth mentioning a few more helpful items, for example, what exactly is the difference between sample setups and file setups in the Acquire! and Automate! windows?
Samples setups contains any and all parameters that are specific to samples: elements, count times, background positions and/or types, interference assignments, calculation options, etc. They can be saved and utilized as described by Ben and I above.
File setups are a specific sample setup from another MDB file. They contain everything that a sample setup contains, but they also load global parameters that are specific to that MDB file into the current MDB file.
For example, the options specified in the Special Options dialog in the Acquire! window for TDI (volatile) acquisitions. For example, one could have two different MDB files, one using TDI (volatile) acquisitions and another MDB not using TDI acquisitions. By switching between those two MDB files using the File Setups assignments from the Automate! window one can decide whether or not to utilize TDI acquisitions for each automated position sample.
Hi Ben & John,
thank you, that's what I've done my setup for feldspar and pyroxene (all elements calibrated first and then break into separate setups). But what about the different currents across the different phases - the beam current has to be the same for calibration? So I can't calibrate all elements with one current and then split it off into something that would use a different current on unknowns? (e.g., glass at 8/10 using combined conditions, fsp/pyroxene at 12, olivine/spinel at 20nA). And all needing a calibration on the Cr2O3 standard.
Do I do different calibrations, within different setups (e.g. glass calibration, save as setup), then apply that specific one to unknowns? Different calibration for px-fsp etc., save as setup and apply? How do I know it is using the correct standard calibration, and won't it be overwritten if there is just one entry in the 'standard' list of samples? If I add another standard it adds it as a second 'set' but the name itself is 'locked' to what's in the standard database?
This also doesn't solve the issue of not wanting to change the current/flipping crystals during analysis of unknown samples around all the time if they are listed in sequence (entered by user as they go), but then have assigned different sample setups (and conditions) for analysis. Can I specify somewhere to acquire 'all samples with the assigned setup A' first, all samples with assigned setup B second, and so on? There will be largely individual unknown sample entries, not one unknown sample with many lines, as needing to keep track of naming of points to track location/sample/etc, if that makes sense. I think it could work if I could individualize/edit the names of lines part of an unknown sample and have all the points as 'lines' within one sample. Then I could have some sort of sequence of unknown samples if all pyroxenes are contained in one unknown, all olivine in another, and so on.
Cheers,
Kerstin
Quote from: KerstinGruender on October 15, 2025, 03:11:28 PMthank you, that's what I've done my setup for feldspar and pyroxene (all elements calibrated first and then break into separate setups). But what about the different currents across the different phases - the beam current has to be the same for calibration? So I can't calibrate all elements with one current and then split it off into something that would use a different current on unknowns? (e.g., glass at 8/10 using combined conditions, fsp/pyroxene at 12, olivine/spinel at 20nA). And all needing a calibration on the Cr2O3 standard.
Sure you can.
All measurements are normalized to cps per nA so you can use different beam currents for different samples and even different beam currents for different elements within a sample, using the combined conditions option that Ben mentioned:
https://smf.probesoftware.com/index.php?topic=5.0
Quote from: KerstinGruender on October 15, 2025, 03:11:28 PMDo I do different calibrations, within different setups (e.g. glass calibration, save as setup), then apply that specific one to unknowns? Different calibration for px-fsp etc., save as setup and apply? How do I know it is using the correct standard calibration, and won't it be overwritten if there is just one entry in the 'standard' list of samples? If I add another standard it adds it as a second 'set' but the name itself is 'locked' to what's in the standard database?
No your previous standardizations won't be overwritten by subsequent standard acquisitions.
Unlike the Cameca and JEOL software, Probe for EPMA time tags each acquisition. Standard intensities are utilized based on the time of the unknown acquisition. This is how the standard intensity drift correction works (if a standard is acquired both before and after an unknown point (or pixel). And the standard intensity drift correction is automatically applied:
https://smf.probesoftware.com/index.php?topic=168.0
Quote from: KerstinGruender on October 15, 2025, 03:11:28 PMThis also doesn't solve the issue of not wanting to change the current/flipping crystals during analysis of unknown samples around all the time if they are listed in sequence (entered by user as they go), but then have assigned different sample setups (and conditions) for analysis. Can I specify somewhere to acquire 'all samples with the assigned setup A' first, all samples with assigned setup B second, and so on? There will be largely individual unknown sample entries, not one unknown sample with many lines, as needing to keep track of naming of points to track location/sample/etc, if that makes sense. I think it could work if I could individualize/edit the names of lines part of an unknown sample and have all the points as 'lines' within one sample. Then I could have some sort of sequence of unknown samples if all pyroxenes are contained in one unknown, all olivine in another, and so on.
Crystal flipping is an issue that you need to decide for yourself. Sometimes it doesn't have much effect (e.g., trace elements), sometimes it can (major elements), depending on the reproducibility of the intensities during crystal flipping:
https://smf.probesoftware.com/index.php?topic=210.0
https://probesoftware.com/download/Presentation%20of%20John%20Donovan_China_EPMA_2023.mp4
From the Automate! window you can create the feldspar/olivine/whatever positions samples (based on a particular sample setup), and append new digitized points to them as you digitize points, and then run the position samples in any order you see fit:
https://www.youtube.com/watch?v=7qhu-QvdkcI&t=1s
Quote from: John Donovan on October 15, 2025, 04:55:32 PMAll measurements are normalized to cps per nA so you can use different beam currents for different samples and even different beam currents for different elements within a sample
Ahhh, so that is where my thinking was going wrong! So it does work with just one calibration and I will only do the standards once. And also fine to realibrate in between. It will be a 'godmother of all samples', but I can see that working now. Thank you!
Yes, the crystal flipping is an issue I'm aware of. It's been working ok on our instrument for what we've used it. Only doing it for two particular routines, while avoiding flipping most of the time where possible.
Quote from: John Donovan on October 15, 2025, 04:55:32 PMFrom the Automate! window you can create the feldspar/olivine/whatever positions samples (based on a particular sample setup), and append new digitized points to them as you digitize points, and then run the position samples in any order you see fit
Any order you see fit, meaning I can set or change the order of acquisition? Don't think I've come across a how-to for that. Youtube also only goes as far as digitizing a series of unknown samples, unless I've missed it?
Quote from: KerstinGruender on October 15, 2025, 10:54:42 PMQuote from: John Donovan on October 15, 2025, 04:55:32 PMFrom the Automate! window you can create the feldspar/olivine/whatever positions samples (based on a particular sample setup), and append new digitized points to them as you digitize points, and then run the position samples in any order you see fit
Any order you see fit, meaning I can set or change the order of acquisition? Don't think I've come across a how-to for that. Youtube also only goes as far as digitizing a series of unknown samples, unless I've missed it?
There are several videos on using the Automate! window, so be sure to watch these additional videos. Actually you should probably watch all of them! ;D
You can change the order of acquisition for unknown and wavescan samples using the spin button here:
(https://smf.probesoftware.com/gallery/1_16_10_25_9_24_35.png)
Remember, to see a brief description of every control in Probe for EPMA, just hover the mouse over the control and the pop up help will show...
If you want to get fancy you can export the position samples to a POS text file, then edit them using a text editor such as NotePad, then import them back into the Automate! window.
Remember, to add points to an existing position sample, be sure to double-click the position sample in the Automate! list to see the currently digitized points for that position sample, then use the Digitize! window to add (or delete) more points to that position sample.
Thanks, John :) We did get fancy and exported our position list, reordered it, even inserted extra secondary standards this way and then ran the different mineral phases in sequence. Nice!
One more question in that topic though, regarding multiple setups. I set mineral and glass points in the same mdb file, with one large MOA calibration setup, then created setups for them, including two that would be for the glasses (majors and volatiles). Then applied a mineral setups to the mineral points and majors to the glasses, then a second setup to glass points where I also wanted to acquire volatiles. However, I couldn't actually start the run as it first prompted me to tick 'run using multiple setups' - which I did - but then it wouldn't let me start it either because the mineral points only had one setup? Is there something I missed upon setting this up and assigning sample setups? Or can I not combine a run where some points have only one setup and others several?
I ended up running the minerals first one day and glass the next, but we now needs extra time because there weren't that many points for each phase and we had wanted to run it all over night in sequence.
Cheers,
Kerstin
Quote from: KerstinGruender on November 25, 2025, 05:23:10 PMOne more question in that topic though, regarding multiple setups. I set mineral and glass points in the same mdb file, with one large MOA calibration setup, then created setups for them, including two that would be for the glasses (majors and volatiles). Then applied a mineral setups to the mineral points and majors to the glasses, then a second setup to glass points where I also wanted to acquire volatiles. However, I couldn't actually start the run as it first prompted me to tick 'run using multiple setups' - which I did - but then it wouldn't let me start it either because the mineral points only had one setup? Is there something I missed upon setting this up and assigning sample setups? Or can I not combine a run where some points have only one setup and others several?
I ended up running the minerals first one day and glass the next, but we now needs extra time because there weren't that many points for each phase and we had wanted to run it all over night in sequence.
I think I know what the problem is but let me first illustrate an example of using different numbers of sample setups for different position samples in Probe for EPMA from the Automate! window.
One first starts (as you did) by creating the several sample setups they want to automate from the Acquire! window. These samples setups could be different sets of elements (minerals, glasses, etc.) with different backgrounds, interference assignments, etc.:
https://smf.probesoftware.com/index.php?topic=1559.msg13587#msg13587
Or for automated acquisition of multiple keV samples for thin film characterization:
https://smf.probesoftware.com/index.php?topic=111.0
Then once all your various sample setups are properly configured, save them from the Analyze! window as shown here:
https://smf.probesoftware.com/index.php?topic=531.msg12600#msg12600
Finally you assign your multiple sample setups (one or more) to your digitized position samples from the Automate! window as shown here:
(https://smf.probesoftware.com/gallery/1_26_11_25_1_34_15.png)
Note that some position samples only have one sample setup, others have two or even three. Each position sample can have different numbers of (one or more) sample setups.
In the above example, I acquired different numbers of setups using different keV conditions, 10, 15 and 20 keVs and assigned them to the position samples as named. The results after automation are here, as seen in the Analyze! window:
(https://smf.probesoftware.com/gallery/1_26_11_25_1_10_19.png)
Note that some samples were acquired with two sample setups, others with more or less numbers of sample setups.
Now here's where you might have gone wrong. If you have different sample setups per position sample, you'll want to run them using the default option with the Run Multiple Sample Setups option UNchecked:
(https://smf.probesoftware.com/gallery/1_26_11_25_1_09_46.png)
This subtle point is also discussed here:
https://smf.probesoftware.com/index.php?topic=531.msg12600#msg12600
That is, only check the Run Multiple Sample Setups One at a Time checkbox if you want to run the first sample setup for all samples, then the second sample setup for all samples, etc. If you check the Run MUltiple Setups One at a Time checkbox, all selected samples have to have the SAME number of setups assigned! This feature is especially useful for running the Constant K-Ratio dead time calibration discussed here:
https://smf.probesoftware.com/index.php?topic=1466.msg11102#msg11102
where one wants to run all samples at say 10 nA, then all samples at 20 nA, then all samples at 50 nA, etc., etc. Hoping this helps and let me know if you have further questions.
Thanks John! In hindsight I think I may also have assigned sample setups (where it's just one) with the 'sample setups' button and then multiple setups with the 'multiple setups' button where it's several - which would likely also result in it complaining. I'll double check this.
I do think I'd want the Run Multiple Sample Setups One at a Time checkbox ticked for it to not switch beam current conditions at every position sample (with multiple setups, 8nA and 60nA current) in order to run them all at one current first, then move back to the same positions and run the second setup? It would skip the position samples where there is no second sample setup assigned? Sounds like this will pop up an error though if it absolutely expects the same amount of setups for every position. I'll try this out next week!
Quote from: KerstinGruender on November 26, 2025, 08:23:16 PMThanks John! In hindsight I think I may also have assigned sample setups (where it's just one) with the 'sample setups' button and then multiple setups with the 'multiple setups' button where it's several - which would likely also result in it complaining. I'll double check this.
That's not the way to think about the automation in PFE.
Basically you use either the current unknown as the sample basis or the digitized conditions, sample setups, file setups or multiple setups as shown here:
(https://smf.probesoftware.com/gallery/1_27_11_25_7_22_36.png)
Remember, one can have a single sample setup assigned when using the Multiple Sample Setups mode, or more than one, so this mode should work for you I think.
What both of the sample setup and multiple sample setup methods will do is check that you have at least one sample setup assigned to all selected position samples, when you start the automation.
Quote from: KerstinGruender on November 26, 2025, 08:23:16 PMI do think I'd want the Run Multiple Sample Setups One at a Time checkbox ticked for it to not switch beam current conditions at every position sample (with multiple setups, 8nA and 60nA current) in order to run them all at one current first, then move back to the same positions and run the second setup? It would skip the position samples where there is no second sample setup assigned? Sounds like this will pop up an error though if it absolutely expects the same amount of setups for every position. I'll try this out next week!
No. When the Run Multiple Sample Setups One at a Time checkbox is checked it basically runs the sample setups in a loop. So the number of sample setups assigned to each position sample needs to be the same.
But these sample setups can be in any order. For example, the first position sample could have a 10 keV and 15 keV setups assigned, and the next position sample could have a 12 and 20 keV setups assigned. If the Run Multiple Setups One at a Time checkbox is checked, it will run these position samples like this:
First position sample, at 10 keV
Second position sample, at 12 keV
First position sample, at 15 keV
Second position sample, at 20keV
By the way, there should not be any problem running samples at different beam currents using automation. The modern JEOL instruments handles this fine. In fact, you might want to consider using the Combined Conditions samples and run all the elements at different beam currents within the same sample as described here:
https://smf.probesoftware.com/index.php?topic=5.0