Author Topic: LaB6 and High voltages  (Read 187 times)


  • Emeritus
  • *****
  • Posts: 1875
  • Never sleeps...
    • John Donovan
LaB6 and High voltages
« on: April 01, 2019, 10:05:32 am »
An interesting discussion on the SX50 listserver regarding Cameca LaB6 filaments and high operating voltages started by Philipp Poeml, which I have reproduced here:

Hi all,

another question: I have an SX100R, which basically has the gun of a SX100. We are equipped with a LaB6.

Normally we do not go higher than 25 kV. But now I would like to do that.

Did anyone use 30 kV or even 35 kV yet?

What could I kill in the worst case?


Hi Philipp,

I have some notes here from a CAMECA user meeting, that >30keV there is a possibility for the beam to arc to the side. There were rumors of people who had run at 35 or 40keV without incident.

I suppose that the stray current should just go to the ground, if everything is actually well grounded. But, probably good to confirm that with measurements.

Jared [Singer]

I think you could certainly blow the tip and I guess an arc-over wouldn't be great for the HT tank. [Mike Matthews]

  Though I don’t remember why, I know Dan Pekle from Cameca said that when running at voltages above 25kv, you are suppose to switch the upper column tube (and holds an aperture) for a shorter one.

Joe Boesenberg
Brown University


We had a bad experience (very bad) the first time we tried to go to high voltage, beyond 15 to 20 kV (I think we tired to go to 50 kV).  This was many many years ago and involved an SX-50, so I’m not sure how  relevant our experience might be to newer instruments.  However, the experience was bad enough I felt compelled to mention it.  The person operating our instrument in those days - Ray Guillemette - has now retired.  However, if you are interested in the details let me know and I can provide his email contact information.  I am in infrequent contact with Ray and I imagine he would be willing to share what he remembers.

Will [Lamb]


I remember one of Cameca's service engineers saying that when running at >30 kv it is necessary to change a jumper on the power supply and also that the column should be very clean and perfectly aligned.

Let us know if you take it up to the 50 kv design limit!  It would be interesting to hear how it performs.



Hello Philipp,

In the notes I took during installation of our SX-Five (July 2017), I wrote that 50 kV is the limit of the HV tank (17 L) and 30 kV is the limit of the LaB6. Our installation engineer did a beam alignment with W-filament up to 30 kV. After this we installed the LaB6 and did a few tests at 25 kV, but besides a bit of imaging we did not conduct any analyses at this condition. Maybe this info will be of use for your SX-100.


Will [Nachlas]

Back ~25 years ago, I learned that Bruce Robinson (CSIRO Perth SX #125) and colleagues in Australia ran their SX50 at 50 kV (that’s why its called the 50), although they did some modifications I think, including realigned the ion pump to be off-axis. Bruce edited back in the late 1980s (1986-1990) 4 issues of an  SXusers (hardcopy: remember mimeograph?—young whippersnappers can google it!) newsletter.

(I have the newsletters and skimmed them and see no great detail of their modifications to the sX50.

They didn’t say if LaB6 or W…my assumption was W.

They showed that you can get lower/quicker detection limits at 50 kV for metals.

John [Fournelle]


B.W. Robinson, N.G. Ware and D.G.W. Smith wrote a chapter entitled:

“Modern electron-microprobe trace-element analysis in mineralogy”

This was published as Chapter 6 in Modern Approaches to Ore and Environmental Mineralogy, MAC Short Course 27 (1998).

From the section “Precautions for EPMA Trace-element analysis”:

“If there is a risk of instrumental damage when running at the manufacturer's maximum kV, then it will be prudent to reduce the accelerating voltage a little. It is for that reason that much of our work at CSIRO has been done at 40 kV rather than the 50-kV maximum warranted for our instrument.”


Andrew [Locock]

Yep, for beam stable specimens (e.g. garnets) we ran at our SiroTrace software at 40kV and 450nA.

Our SX50 (#125) used a W filament.

This has now gone to the great electron beam lab in the sky, along with a 2nd hand SX50 we purchased

from Uni of Hawaii. Our lab no longer has an electron microprobe.


Greg [Hitchen]

Finally Philipp Poeml thanked everyone and posted some archived replies on this subject which I quote below:

Hi everyone,

thanks for all the replies and advice.

I will let you know how far I get.

Xavier also found some old conversation about it, see below.

Right now I am fighting with the beam regulator aperture... Does anyone have the instruction set to adjust the potentiometers in the picoammeter box?


Let me raise a related topic: "high" accelerating voltage.  Rather than  get 249 messages to the list (that's the number of subscribers today),  let me specify ONLY if you run at 30 or higher kV -- please specify the  model of your probe and any problems/issues at these "high" voltages you  may have encountered.

I ask because I would like to be able to run my SX51 at 35 kV, as I have  an LIF 220 xtal and can acquire Ka x-rays of Sr, Y, Zr, Nb and Mo,  useful for various projects where L lines of these elements have  interferences or other issues. (There's a story behind this and a  publication in Surface and Interface Analysis. 2005 37:1012-16). (Also,  having a Xe gas detector would be optimal.)

With the anode spacers set properly, and with the  jumper on the HV  board set at the 50 kV position rather than 30 kV, the highest I can run  "normally" is 29 kV; at 30+ kV, on benitoite, the beam pulses at a rate  of a beat a second (it's alive!) which I take as a signal "don't mess  with me" and so I back down to 29.

Does _anyone_ with an electron probe (any flavor) run at 35 (or higher)  kV?




I believe that Todd had run at higher than 30 kV on our SX50 (#293), but  we have not done so in recent years, since he retired in 1999. He asked  Cameca to make some special modifications for this when we received the  instrument in 1989, one of which included adding an extra ground jumper  to the gun.  We also have what is billed to be a special anode for high
V, which also has a jumper strap. I think Stefan remembers the  modifications, and will probably see this message and can comment.

I just walked into the back room where there's a 1983 Camebax, ran it up to 45 kV, regulated the beam to 20 nA with success, then dropped the beam current back to 10 nA and got a decent SE image at several thousand magnification.  I would have gone the additional 5 kV to the top of the dial at 50 kV, but that involved tweaking the gun voltage fine adjustment pot and I could already imagine the 24-year-old insulating oil in the HV tank swirling around at high speed. That's not exactly an exhaustive test, but it did work (and I remember doing this years ago, too).

The SX100 gun is rated to 50 kV, but we haven't tried it out there. Haven't had the need.




To qualify to answer, we have had to run at high KV..... Well, we qualify, not only have we run at high KV but we have destroyed one of the old fashioned epoxy filled HT tanks.  We were trying to punch through the tin layer on tinplate to excite the Fe Ka line (the SnLA:FeKa is a non-linear fn of tin thickness).   In this case, the tin coating on the plate was thicker (~1um) and the operator quite sensibly ran the KV up to 50 to get max penetration. I was on holidays at the time and came back to a dead EPMA.

HV flashovers in the gun are an issue at the higher kV and once you have had one or more, it seems to be that they are easier to come by.  Bruce Robinson looked extensively at 50kV work and found that by disconnecting several of the ion pump elements it was possible to get a stable supply. I'm not sure whether this had anything to do with residual ions in the
gun or fields.

Evidently it is known that gun vacuum degrades prior to a flashover. So, the perceived wisdom is to run the HV up in small steps and make sure the gun vacuum does not degrade. Our service engineer has rigged up a DVM to the tapping points on the vacuum logic board that monitor the voltage from the various vacuum sensors.  For high voltage work, the ion pump sensor can be watched to monitor the gun vacuum.  More significant figures are needed than are provided in our steam driven (PDP model SX50) vacuum dialogue for this to be done with the vacuum system dialogue display.

Wishing you many high energy electrons without the smell of charred electronics.



Yep, we here regularly run at 35kV for trace element analyses with a 1987 SX50 (with an upgraded version III HV oil tank).
Bruce Robinson has recently retired from our lab and did a fair bit of work at higher kV (40-45, not sure if he actually went to 50 though, he'll probably let me know!).

We haven't gone above 35kV for ages as is seems that we would relatively regularly risk a flashover, which would normally
result in having to pull boards out and replace chips.

 From memory the flashovers were related to a "periodic" pressure increase due, I think to the poor pumping of Ar (and ion pumps "burping"?). Hopefully Bruce will chime in with some more details.


« Last Edit: April 01, 2019, 11:23:47 am by Probeman »
The only stupid question is the one not asked!

Karsten Goemann

  • Global Moderator
  • Professor
  • *****
  • Posts: 177
Re: LaB6 and High voltages
« Reply #1 on: April 03, 2019, 04:23:05 pm »
Hi Philipp,

On the SX100 here at UTAS (SX846, installed 2003) we originally couldn't get to 30kV when using LaB6.
Above something like 27kV we'd have arcing.

This is one of the main reasons why in 2007 the system was modified to have the new side-mounted ion pump design that the newer microprobes have. Originally we had the ion pump with the 4 elements around the Wehnelt. The modified system was tested successfully up to 30kV for LaB6 (which was the specification) but not higher. We actually never used it at voltages like that. Most of the time we wouldn't go beyond 20kV, maybe 25kV occasionally for specific trace element methods.