Beam regulation

[AndrewLocock]

I am inquiring about beam regulation (stability) on a Cameca SX100 using PfE software when using Combined Conditions.

The project involves analyzing steel for 5 minor/trace elements and 2 major elements. The Combined Conditions involve switching from 200 nA – 1 micron beam – 80 s on peak, to 20 nA – 1 micron beam – 30 s on peak.
The PfE Analysis Calculation Options are:
- Beam Drift Correction ON
- Automatic Drift Correction ON
PfE is setting the conditions for each point.
The run involves hundreds of points, so a lot of work for the beam regulator.
 
As the run progresses, I am seeing an increasing number of high totals (105 to 115 wt%), which I presently attribute to mis-reading of the current.
 
The instrument was serviced last week (cleaned the lower graphite tube, cleaned Faraday cup, installed new beam regulator).
Thus, the instrument is (was) in good shape.

Is there something that I should be doing differently in the PfE software, or should I expect that the hardware simply require more servicing?

I appreciate your insights.

Thanks,
Andrew
 

[Probeman]

I am inquiring about beam regulation (stability) on a Cameca SX100 using PfE software when using Combined Conditions.

The project involves analyzing steel for 5 minor/trace elements and 2 major elements. The Combined Conditions involve switching from 200 nA – 1 micron beam – 80 s on peak, to 20 nA – 1 micron beam – 30 s on peak.
The PfE Analysis Calculation Options are:
- Beam Drift Correction ON
- Automatic Drift Correction ON
PfE is setting the conditions for each point.
The run involves hundreds of points, so a lot of work for the beam regulator.
 
As the run progresses, I am seeing an increasing number of high totals (105 to 115 wt%), which I presently attribute to mis-reading of the current.
 
The instrument was serviced last week (cleaned the lower graphite tube, cleaned Faraday cup, installed new beam regulator).
Thus, the instrument is (was) in good shape.

Is there something that I should be doing differently in the PfE software, or should I expect that the hardware simply require more servicing?

Hi Andrew,
Excellent questions. First let me start by saying that using combined conditions itself should not cause any beam drift issues. Yes, it does work the beam current regulator a lot more than a single condition does, but if the beam current regulator is clean and working, it should just cycle between the conditions.  You mentioned using 200 nA and 20 nA, I assume 20 nA for the majors (run first correct?), then 200 nA for the traces?

We do a lot of combined condition acquisitions on melt inclusions (usually 10 or 20 nA on the majors with the TDI correction turned on and 50 to 100 nA for S, Cl, P, F, etc.) and we see no beam drift problems on our Sx100. That said, it would be helpful if you could post some data since PFE acquires a beam current both before and after each analysis.  For combined condition samples it looks like this:

Faraday or Beam Currents:
ELEM:    na ka   si ka    k ka   al ka   mg ka   ca ka    s ka   fe ka   cl ka   ti ka    p ka    f ka
   71G  10.010  10.010  10.010  10.010  10.010  10.010  50.020  10.010  50.020  50.020  50.020  50.020
   72B  10.002  10.002  10.002  10.002  10.002  10.002  49.973  10.002  49.973  49.973  49.973  49.973
   73B   9.985   9.985   9.985   9.985   9.985   9.985  49.994   9.985  49.994  49.994  49.994  49.994
   74B  10.007  10.007  10.007  10.007  10.007  10.007  49.992  10.007  49.992  49.992  49.992  49.992
   75B   9.994   9.994   9.994   9.994   9.994   9.994  50.023   9.994  50.023  50.023  50.023  50.023

AVER:   10.010  10.010  10.010  10.010  10.010  10.010  50.020  10.010  50.020  50.020  50.020  50.020
SDEV:     .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000

Second Faraday or Beam Currents:
ELEM:    na ka   si ka    k ka   al ka   mg ka   ca ka    s ka   fe ka   cl ka   ti ka    p ka    f ka
   71G   9.998   9.998   9.998   9.998   9.998   9.998  49.970   9.998  49.970  49.970  49.970  49.970
   72B   9.994   9.994   9.994   9.994   9.994   9.994  50.028   9.994  50.028  50.028  50.028  50.028
   73B  10.003  10.003  10.003  10.003  10.003  10.003  49.997  10.003  49.997  49.997  49.997  49.997
   74B  10.000  10.000  10.000  10.000  10.000  10.000  49.984  10.000  49.984  49.984  49.984  49.984
   75B  10.015  10.015  10.015  10.015  10.015  10.015  50.029  10.015  50.029  50.029  50.029  50.029

AVER:    9.998   9.998   9.998   9.998   9.998   9.998  49.970   9.998  49.970  49.970  49.970  49.970
SDEV:     .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000

In this example, the majors were acquired at 10 nA and the traces at 50 nA. Note that PFE utilizes the average of the two beam currents for the beam drift correction (if both are present) for performing the beam drift correction. However, if the nominal beam is set to zero, there will be no drift correction applied to the x-ray intensities.  This is the same as unchecking the Use Beam Drift Correction in the Analytical | Analysis Option menu dialog.  The software should warn you if this is the case.

Perhaps you could post an example of the beam currents you are seeing as I did above.  For example: is the beam current drift you are seeing occurring within a single analysis point?  Or is the drift occurring over time, say overnight?  The latter brings up temperature control concerns in the lab.

Now, next is the standard drift correction you mentioned, which works in conjunction with the beam drift correction, and which is applied to each analysis point. If the standard drift correction is turned on in the Analytical | Analysis Options menu dialog, the software will automatically correct *each element* unknown intensity for any changes over time based on the most recent standardization acquired previously and subsequently to the data point being analyzed. That is assuming you have more than one standardization in a run. This means that you should acquire a standard set at the beginning of your run and also at the end of your run.  The Automate! window provides an option for this is the Automation Actions group of controls as seen here:



You can also set the software to automatically acquire a standard set every n hours as seen here:



Here is an example from my lab with a number of standardization automatically acquired during a long run:

http://probesoftware.com/smf/index.php?topic=332.msg1741#msg1741

The manual discusses the standard drift correction here:

http://probesoftware.com/download/PROBEWIN.pdf#page=334

My first take on your beam drift issue is that your instrument should not be experiencing drift of 5 to 15% in the beam current over time, but how much does temperature vary in your lab over a 24 hours period?  I know I'm bragging here but here is some monitoring of our lab temperature (see attached below).
john

[AndrewLocock]

Hi John,
Thanks very much for the discussion and guidance.

As it turns out, the Faraday cup on our Cameca SX100 was behaving erratically (flaky)! We were having occasional extremely-sudden changes in current from 20 nA down to 16 nA. Thus, in a series of analytical points, 3 to 7% of them were giving odd results, with no apparent periodicity over time.

We finally replaced the Faraday cup this morning (August 5), and now have excellent beam stability (RSD 0.007% for 1000 current measurements with the replacement Faraday cup).

With regard to the issue of temperature, we monitor our lab temperature hourly, and over the previous 1.2 years have a mean of 23.5 degrees Celsius, with a standard deviation of 0.6 degrees (74.4 Fahrenheit, +/-1). Although temperature fluctuations were not the cause of our problems, I do appreciate the suggestion.

I look forward to continuing to use Combined Conditions with the outstanding PfE software.

Thanks again,

Andrew

[John Donovan]

Hi John,
Thanks very much for the discussion and guidance.

As it turns out, the Faraday cup on our Cameca SX100 was behaving erratically (flaky)! We were having occasional extremely-sudden changes in current from 20 nA down to 16 nA. Thus, in a series of analytical points, 3 to 7% of them were giving odd results, with no apparent periodicity over time.

We finally replaced the Faraday cup this morning (August 5), and now have excellent beam stability (RSD 0.007% for 1000 current measurements with the replacement Faraday cup).

With regard to the issue of temperature, we monitor our lab temperature hourly, and over the previous 1.2 years have a mean of 23.5 degrees Celsius, with a standard deviation of 0.6 degrees (74.4 Fahrenheit, +/-1). Although temperature fluctuations were not the cause of our problems, I do appreciate the suggestion.

I look forward to continuing to use Combined Conditions with the outstanding PfE software.

Thanks again,

Andrew

Hi Andrew,
No worries.

I sometimes think that the most difficult aspect of troubleshooting problems on the probe is differentiating hardware issues from software issues. These two systems are now so integrated in modern instruments that it is sometimes quite non-trivial to separate them!

But I'm relieved to hear that you tracked it down!

And thank-you for the compliment!
john

[Probeman]

Hi John,
Thanks very much for the discussion and guidance.

As it turns out, the Faraday cup on our Cameca SX100 was behaving erratically (flaky)! We were having occasional extremely-sudden changes in current from 20 nA down to 16 nA. Thus, in a series of analytical points, 3 to 7% of them were giving odd results, with no apparent periodicity over time.

We finally replaced the Faraday cup this morning (August 5), and now have excellent beam stability (RSD 0.007% for 1000 current measurements with the replacement Faraday cup).

With regard to the issue of temperature, we monitor our lab temperature hourly, and over the previous 1.2 years have a mean of 23.5 degrees Celsius, with a standard deviation of 0.6 degrees (74.4 Fahrenheit, +/-1). Although temperature fluctuations were not the cause of our problems, I do appreciate the suggestion.

I look forward to continuing to use Combined Conditions with the outstanding PfE software.

Thanks again,

Andrew

Hi Andrew,
No worries.

I sometimes think that the most difficult aspect of troubleshooting problems on the probe is differentiating hardware issues from software issues. These two systems are now so integrated in modern instruments that it is sometimes quite non-trivial to separate them!

But I'm relieved to hear that you tracked it down!

And thank-you for the compliment!
john

I'm wondering if we need to clean or replace our faraday cup as well as we are having beam stability issues very similar to what you describe.

Also if we set a beam current, say 20 nA, and watch the picoammeter we see the beam current continue to rise past 20 to 25 or even 30 nA and more.  Then if we turn off the regulation it reverts close to the 20 nA.

Did you actually replace with a new faraday cup, or just clean it?

[Probeman]

Ok, more data on this beam current instability issue (and fixed some typos below!).

Our engineer pulled out the Faraday cup this morning and wasn't able to open it up, but he was able to scrap around inside it and dumped out a bunch of carbon dust. However, after we replaced it and pumped down, the beam current was exactly as unstable as before.

Interestingly we noticed something else though.  If we set the beam current to regulate at say 30 nA, and watch the picoammeter, the instrument starts at 30 nA, but then slowly descends downward until it stops around 22 nA. Then if we turn off the beam current regulation, the beam current goes right back to 30 nA.

Likewise, but differently, if we set the beam current to 50 nA, it starts at 50 nA but then slowly climbs upwards until it stops around 94 nA. Again, turning off the beam regulation snap the beam current right back to 50n nA.

So now I've checked the Do Not Send Conditions During Acquisition checkbox in PFE's Acquisition Options dialog, and am running samples *without* the beam current regulator turned on. It's actually quite stable with the beam regulation turned off for some reason!

So our latest hypothesis is that the beam current regulation circuit has gone bonkers and now our engineer is looking at the IC chips on it.

[Philipp Poeml]

Hi John,

what was the outcome of this? We see similar issues, even if not so severe, but bothering enough that sometime we used "Do Not Send Conditions During Acquisition" as well.

When we set 20 nA often we arrive at 22 nA. If we set 5 nA we arrive at 5.6 nA. 100 nA could end up as 96 or 107 nA. A friend at Cameca knows the issue, but did not provide a solution yet. The person said it is worse at the beginning of a range (0.5-5, 5-50, 50-500) of the picoammeter. And it is correct. 5 nA is really bad, but 40 nA does work.

Did your engineer find a solution?

[Probeman]

Hi John,

what was the outcome of this? We see similar issues, even if not so severe, but bothering enough that sometime we used "Do Not Send Conditions During Acquisition" as well.

When we set 20 nA often we arrive at 22 nA. If we set 5 nA we arrive at 5.6 nA. 100 nA could end up as 96 or 107 nA. A friend at Cameca knows the issue, but did not provide a solution yet. The person said it is worse at the beginning of a range (0.5-5, 5-50, 50-500) of the picoammeter. And it is correct. 5 nA is really bad, but 40 nA does work.

Did your engineer find a solution?

Hi Philipp,
Yeah, we had a hard time believing it was the beam current regulator because we had just changed it out a month before, but somehow it was contaminated with an aluminum fluoride material which was very difficult to remove, but when that was re-installed the beam current was totally stable again.  We have no idea where that contaminant came from.
john

[Philipp Poeml]

Ok, so for you it was the beam regulator aperture. I don't think this is the case for us, we will see. Thanks anyway for the info.

[Probeman]

We have been having some difficulty regulating the beam at 50 nA, but our amazing instrument engineer (Steve Wiemholt), rebuilt our faraday cup and I have to say the stability is now amazing.

Here are the standard intensities over an 18 hour period:

Drift array background intensities (cps/1nA) for standards:
ELMXRY:     c ka   al ka   mo la   cr ka   fe ka   mn ka   ni ka   cu ka   si ka    o ka
MOTCRY:  1   PC2 2  LTAP 4   PET 4   PET 5   LIF 3  LLIF 5   LIF 3  LLIF 2  LTAP 1   PC2
INTEGR:        0       0       0       0       0       0       0       0       0       0
STDASS:      506     513     542     524     526     525     528     529     514     913
STDVIR:        0       0       0       0       0       0       0       0       0       0
           13.07    5.50     .34    1.07     .24     .74     .31    1.26   12.10   29.30
           12.88    5.40     .32    1.07     .23     .70     .30    1.26   12.40   27.14
           13.01    5.47     .33    1.07     .27     .74     .31    1.29   12.38   26.69

Drift array standard intensities (cps/1nA) (background corrected):
ELMXRY:     c ka   al ka   mo la   cr ka   fe ka   mn ka   ni ka   cu ka   si ka    o ka
MOTCRY:  1   PC2 2  LTAP 4   PET 4   PET 5   LIF 3  LLIF 5   LIF 3  LLIF 2  LTAP 1   PC2
STDASS:      506     513     542     524     526     525     528     529     514     913
STDVIR:        0       0       0       0       0       0       0       0       0       0
         1017.92 2637.60   46.41  130.17   38.86  140.87   18.46   26.34 2624.76  335.23
         1001.47 2630.10   46.51  129.20   39.00  140.86   18.30   25.85 2623.39  335.77
         1024.92 2658.47   46.46  129.73   38.77  141.03   18.19   25.98 2633.08  335.84

Drift array interference standard intensities (cps/1nA):

1st assigned interference elements
ELMXRY:     c ka   al ka   mo la   cr ka   fe ka   mn ka   ni ka   cu ka   si ka    o ka
INTFELM:     cu                              mn      cr                                 
INTFSTD:     529                             525     524                               
           39.97                             .02     .35                               
           42.53                             .02     .38                               
           42.55                             .02     .37                               

For the primary standard intensities, we see Si Ka that's around 0.3% drift and for O Ka, that's around 0.1% drift. That is *with* a background correction, which slightly worsens precision, but I can live with that!