Probe Software Users Forum
Hardware => JEOL => Topic started by: Bill_Mushock on October 15, 2019, 01:41:09 PM
-
So here is the problem. When we digitize an image for the purpose of doing a line profile analysis, the OL fine focus resets to some default value which in our case is 132. This leaves our beam in a slightly defocused/larger state. You may not notice this if you are taking low mag images (<3-4kx). You will also not notice a change in the JEOL EOS Monitor window until you turn the OL Focus on the console one click which updates the monitor. The real issue for us is that we are now digitizing a line traverse with a slightly larger probe diameter. This behavior has been observed on at least one other 8900.
Can anyone else out there confirm this problem on your 8900. Have you come up with a work around? If your doing line traverses in tight spots, you definitely want to check this out.
-
Hi Bill,
Not having a JEOL instrument, my question would be: on the JEOL 8900 how does the objective lens focus relate to the beam size command?
For example, on our Cameca instrument we set the beam size to zero and then adjust the fine focus to get the sharpest image. Once that is done, that focus is "remembered" as the focus for a zero size beam and fully focused image.
After that you can perform your quant analyses at different beam sizes (e.g., 5 um or 10 um), but when you go to acquire another image (whether for documentation or digitizing points or traverses) in Probe for EPMA, the size gets set to zero automatically which should (if the instrument is stable), give one the best image focus because the instrument remembers what the exact OL value is for a fully focused beam.
So again, if you have the beam size set to 0, and you adjust the focus, does the instrument not remember what the focus is if the beam size is changed?
-
Hi,
In my testing in the JEOL software on my 8900 imaging works similar to your CAMECA instruments. I find a feature of interest, focus using the focusing knob and the objective lens values are remembered. Imaging and all other activities remember the OL values in the JEOL software even changing between beam sizes the OL values don't change unless one actively changes them using the software or the focus knob.
I'm not sure why when PfE sets the beam size to 0 when imaging that there is a change. It might be one of those weird order-of-operation problems that 8900s seem to have when sending it settings. Maybe sending the set to 0 command should be the first thing that is sent?
-dave
-
Hi,
In my testing in the JEOL software on my 8900 imaging works similar to your CAMECA instruments. I find a feature of interest, focus using the focusing knob and the objective lens values are remembered. Imaging and all other activities remember the OL values in the JEOL software even changing between beam sizes the OL values don't change unless one actively changes them using the software or the focus knob.
I'm not sure why when PfE sets the beam size to 0 when imaging that there is a change. It might be one of those weird order-of-operation problems that 8900s seem to have when sending it settings. Maybe sending the set to 0 command should be the first thing that is sent?
-dave
Hi Dave,
You're saying there is a change in the objective lens (from what one would consider to be focused), when the beam size is set to zero on the JEOL 8900 (I wonder if this is something also seen on the JEOL 8200/8500?).
Does this happen even when the beam size is already zero or does it happen only when the default beam size is non-zero? And what OL value is it changing to and why?
Another question: on the Cameca setting the beam from zero to non-zero changes the objective lens setting by a fixed amount. When change the beam back to zero size doesn't the OL return to the previous value?
Maybe someone could create a table showing what is going on with the interaction between the beam size and OL? I really don't understand exactly what the relationship is on the 8900.
Another thing to try is to check the "Skip Beam Scan Modes" checkbox in the Imaging window, and then no column conditions are sent, prior to the image acquisition.
-
Hi John,
When using the JEOL software there is no change in the objective lens when the beam size is set to zero or any other beam size. The behavior that Bill and I are seeing occurs when using PfE to image. I can only speculate as to what is going on between PfE and the instrument. Based on observations it may be when PfE tells the 8900 to set the beam to zero before beginning the image acquisition.
The behavior I see on my machine is that even when the probe diameter is zero that the OL values change when beginning an image in PfE.
When changing the beam diameter on the 8900 the objective lens values DO NOT change. They remain at the values set when focusing under a beam diameter set to zero.
-dave
-
Hi Dave,
I really appreciate the additional information, but I'm still having difficulty understanding what it going on, so sorry if I sound a little dense.
The thing I'm having trouble with is that the beam size command that is sent, when changing the beam size in the analytical conditions dialog, is *exactly* the same command that is sent from just prior to acquiring an image. In both cases it is the driver J8K_SetSize function.
Can we do a sanity check and check the "Skip Set Beam Scan Modes" checkbox in the Imaging window and see if the OL still changes prior to acquiring an image? I expect the OL might not change and then we can do some further testing to track down the actual culprit
For example, one can temporarily turn off the set beam size function in the probewin.ini file by setting the BeamSizePresent flag equal to zero. Maybe it's got nothing to do with setting the beam size!
Also it would be helpful to know if there is any consistency in how much and in which direction the OL is changing when acquiring an image... a little data might be nice. :)
-
When using the JEOL software there is no change in the objective lens when the beam size is set to zero or any other beam size. The behavior that Bill and I are seeing occurs when using PfE to image. I can only speculate as to what is going on between PfE and the instrument. Based on observations it may be when PfE tells the 8900 to set the beam to zero before beginning the image acquisition.
OK, I just saw you say that changing the beam size in the JEOL software does not change the OL value on the 8900!
So if you set the beam size to zero or non-zero in the PFE analytical conditions dialog, does the OL value also not change?
-
Dave,
Could you could try this test for me please.
1.With JEOL software set the spot size to 0
2. Set your Ol coarse one click lower or higher and refocus with the OL Fine only. (You should be able to do this for several coarse settings)
3. Select a 5um spot size
4. Set the spot size back to 0
5. Is the image focused? If so this problem is unique to my instrument.
-
Let me explain the last post. I did some tests on the 8900 this morning and found if I focus with the beam set to 0 on the JEOL software, then change the spot size to something other than 0 and then change back to 0, the OL fine resets to another value. So, its apparent the JEOL software is not storing my settings.
I have no idea when the OL values are set for a 0 spot size. And this week the value for OL fine is 195 vs 135 last week. So obviously I did something to change it.
-
Hi Bill,
Well I'm relieved that you're a little confused because I certainly am! :) Why don't you and Dave do some additional testing and get back to us when you've got more information.
As you might have seen above, I suggested some other tests that could also be performed by temporarily turning off some of the functions that are usually called when operating Probe for EPMA.
There's also two very low level test apps that you can obtain from us if you like: TestType and TestImage. They allow one to test each instrument function call one at a time. Nice for troubleshooting the instrument. They are in the TestProbe.msi installer. Just let me know by email if you want them.
https://probesoftware.com/smf/index.php?topic=566.msg3207#msg3207
I'm sure we'll get to the bottom of this eventually!
john
-
OK, this is what I discovered: When using the JEOL software, if you open the EOS monitor, focus the beam and then open the Probe Diameter window and select
another diameter besides 0, the software will store the OL fine 0-beam size value. As far as I know so far, you would then need to close the EOS monitor window, then reopen, refocus and switch to another beam size to record the new 0-beam value. BTW, it only stores the OL fine. This could present a problem for the PfE software if one
were to change the OL coarse and refocus with the OL fine without going in to the JEOL EOS moitor and storing a new OL fine value.
-
Hi Bill,
OK, that makes some sense I guess.
So are you OK with the Imaging window setting the beam size to zero prior to starting an image acquisition?
john
-
Actually better off not setting the beam to zero before imaging. I assume most users will focus before acquiring an image anyway. Telling the beam to go to zero spot size calls up the values stored in the JEOL software which may de-focus the beam if not correct. Subsequent digitized spot and line analysis will also be with a potentially de-focused beam.
A possible solution to this would be to define the zero beam within the PFE software. At least for the 8900. I suggest a routine in the Analytical Conditions window where once the beam is focused you could press a button to store OL zero beam size coarse and fine values. Alternative beam sizes would use this value as a reference.
-
Actually better off not setting the beam to zero before imaging. I assume most users will focus before acquiring an image anyway. Telling the beam to go to zero spot size calls up the values stored in the JEOL software which may de-focus the beam if not correct. Subsequent digitized spot and line analysis will also be with a potentially de-focused beam.
Hi Bill,
Thanks for the explanation.
The original reason we have been sending a beam size zero command prior to image acquisition is that the "default" beam size might be 5 or 10 microns. If we don't send the zero beam size prior to imaging the resulting image will be out of focus.
We then thought about maybe sending a zero size beam after each sample acquisition, and then the beam will already be focused prior to the image acquisition. But if someone manually removes the faraday cup, the beam will be fully focused and then might cause sample damage. Though it is true that if someone is manually or automatically acquiring data, the program will set the proper beam size before unblanking the beam and sample acquisition.
What do you all think about this? We would like to hear some feedback before making such a change to the default operation of the software.
A possible solution to this would be to define the zero beam within the PFE software. At least for the 8900. I suggest a routine in the Analytical Conditions window where once the beam is focused you could press a button to store OL zero beam size coarse and fine values. Alternative beam sizes would use this value as a reference.
So you're suggesting that PFE store these objective lens values internally and use them when resetting the beam size to zero?
By the way it is an interesting thing, but on the JEOL 8900, one can read both the objective lens coarse and fine values, but on the JEOL 8200/8500 one can *only* read the objective coarse value through the communication interface.
Hope you don't mind but here I quoted, from your email to me, your explanation of getting the JEOL instrument to "remember" the current beam size zero focus settings from the UNIX box:
But in the process I discovered the OL fine/ Zero beam value somehow is tied to opening the EOS monitor and then the Probe diameter window. Once a non-zero size is selected in the PD window, it appears the OL fine/Zero beam value is reset to the current value. This makes sense I suppose but if PfE is instructing the JEOL to go to Zero beam size, one needs to be aware of this.
For instance, if you change the Ol coarse and refocus using the OL fine, without going back in to the JEOL software and resetting it, Zero beam size will not be zero. Or for that matter if you were not exactly focused when the OL fine were stored, you could have issues with spot size and image focus using PfE.
I don't quite understand the process you're describing here, but I guess my question is, why can't 8900 users simply be aware that they need to get the instrument to remember what the OL values are before they start trying to acquire images in PFE? Would this be so onerous? Is this what JEOL 8200/8500 users also have to do to get the instrument to remember the beam size zero focus?
The other possibility as described above, is to always have the PFE software reset the current beam size to zero whenever it finishes a sample acquisition. Then the user would be seeing a focused beam any time they manually unblanked the beam. Then they can focus manually easily. But again, I'd like to hear from users of all instruments before we commit to this.
Or maybe we could just add a new flag in the probewin.ini file for this behavior of forcing the beam back to zero after each sample acquisition? I'd like to hear from users of all instruments before we make any such change.
-
Here is my input on this. I am observing behavior within Digitize Image on the 8200:
We focus and stigmate manually on the 8200. Then open Digitize Image and acquire the image, then use Save As to save the image with its .acq file, then finally we open that image in Picturesnap and digitize all further points, linescans, etc. within Picturesnap.
There is no observed change in beam diameter or focus when the image is acquired in Digitize Image. That is, the probe current and beam diameter definitions for the current sample are not enforced and that is the correct mode; for example, the current sample might be 25nA and 10 micron diameter but you want to collect the image at whatever probe current and focused beam is used.
I'm not in favor of sending beam control commands after the user has carefully set the image focus, only to see that overridden by values that might be stale.
I have confirmed that on the 8200, the probe diameter is a separate lens function and does not change the objective lens dac value. This is confirmed using the EOS monitor; you can click on different beam diameters and set back to zero beam with no change in objective dac value. There may be some difference on the 8900 but the EOS monitor is the same as far as I know.
Possibly the issue is elsewhere.
Paul
-
I'm not in favor of sending beam control commands after the user has carefully set the image focus, only to see that overridden by values that might be stale.
Hi Paul,
Good to know the 8200 doesn't have this issue.
I agree with your statement that we should not be sending commands that change the focus prior to image acquisition. But currently, in PFE, the Digitize Image and Imaging windows *do* send a beam size zero command just prior to the image acquisition in PFE. This is because if one is using a default beam size of 5 or 10 um, then the image will be acquired out of focus if the beam size zero command is not sent.
So you, I and Bill Mushock (and just spoke with Dave Adams) all agree that we should change this behavior so that we *do not* send a beam size zero command just prior to image acquisition. That is the default state of the beam size should always be zero, unless a sample acquisition is being performed.
So we propose the following:
1. When a sample acquisition is finished, reset the beam size back to zero, even the if the default beam size is not zero.
2. Prior to the image acquisition, do not set the beam size to zero. That is, always assume that the current beam conditions are correct just prior to image acquisition.
3. However, for image acquisition *during automation*, we will have to set the beam size to zero prior to image acquisition, since the beam will already be at the default beam size which might be non zero.
How does this sound to everyone?
-
That sounds like the best plan to me.
-
That sounds like the best plan to me.
There's always a complication!
Something I just thought of: what if the user just set the beam size to non-zero from the Analytical Conditions dialog? We'd want it to be reset to zero when acquiring an image. Or if they re-set the beam size to non-zero from another application, such as the JEOL software?
I have to think more about this...
Edit by John: OK, the beam size does automatically get reset to zero after an acquisition *if* the Default Focus checkbox is checked in the Analytical Conditions dialog!
I think I know what to do now. Thanks for everyone's help on this!
-
OK, we've made a small change in the imaging code that we think should avoid the issue that Bill Mushock is reporting on his JEOL 8900.
But it should first be pointed out that the "Default Focus" feature, seen here:
(https://probesoftware.com/smf/gallery/1_24_10_19_10_19_41.png)
is key to this situation. Basically, when this box is checked (in the Analytical Conditions dialog in PFE), whenever an acquisition or automation completes, the software re-sets the beam size to zero. The idea being that one can then simply unblank the beam and image with a fully focused beam. This is the default mode in PFE as many of you already know, but it essentially has no effect if the default beam size is already zero (fully focused).
However, if the user prefers to leave the beam size non-zero (for whatever reason), simply uncheck this box and the beam will always stay at the specified (non-zero) beam size.
Now, when it comes to image acquisition, the latest v. 12.7.8 PFE will first check if the beam size is already zero, and if it is already zero, it will not change the beam size. We think this will take care of the OL changes that Bill is seeing on his JEOL 8900 when the beam size is reset to zero.
But if for some reason the beam size is left at a non-zero value, or reset by other software to be non-zero, the imaging code will reset the beam size to zero, ensuring that the image acquired is fully focused.
Please let us know how this works for you all.
Edit by John: In summary, I suggest everyone use the Default Focus feature so the beam is generally at beam size zero. Then when one starts an image acquisition, the beam is already at size zero, so the beam size command is not sent.