5973 Cannot achieve constant peak widths - analyzer side board issue?

I recently have experienced an issue on not one, but TWO of our Agilent 5973 detectors that we have connected to GCs. The detector seems completely dead, if I run a sample I get a complete flatline. I can't do an autotune because it says "Tune action stopped: Cannot achieve constant peak widths". If I do a manual tune it looks like a lawn, maybe I see a little higher peak corresponding to the PFTBA standard ion peaks but overall it just looks like noise. Extremely few peaks, only 6 reported with an abundance of 962. This issue happened after I vented, turned off the GCMS, cut the column (we had a small leak), turned the GCMS back on, pumped down, and then tried to run a tune. The instrument was working perfectly fine, we just had a slight leak which was causing our peaks to look small (N2 abundance was 20%). Here are some of the things I've tried:

1) I tried re-cutting the column and making sure that it was only sticking out ~1-2mM into the analyzer.

2) Perhaps a corrupt tune file? I changed atune.u to atune.old and loaded a fresh copy of atune.u. No luck. I tried all the other tune files on our system, no luck.

3) We changed the filaments and cleaned the ion source very recently. Regardless, we tried another filament change and no luck. We haven't changed the EM Horn, could this be it? If so, how do we test for this?

4) I tried resetting the MSD by pressing the "menu" button until I saw the "controller" item. Then I pressed item and clicked reset controller. The MSD reset but no luck.

5) The PFTBA does look low but I was able to run a perfectly fine tune just minutes before I turned the system off to cut the column. I hear the solenoid click everytime I do a manual tune "profile" or "scan".

6) Vacuum & He flow look fine. I also did a few power cycles of everything. I checked and all cables are connected just fine.

I'm really at a loss and the leading theory is that the analyzer side board is broken and needs to be replaced. Is there a way I can verify this is the case? Furthermore, what are the chances that this breaks on two of our instruments just a few months apart? Am I doing something to damage the side board?

Below are some screenshots that help show the issue.

Thank you all so much

  • What's the vacuum level?  Column flow ~1.2 ml/min?  

    Turn the Source Temperature and Quad Temperature both to 40°C. Save that as COLD_Tune.U.   That will start it cooling off if you need to vent it.   Then, before venting it to check out inside, go into Manual Tune.  Change the AMU Gain to 400 and AMU Offset to 50.  Change Entrance Lens offset to 12 and the Entrance Lens to 28. Click on the Profile Button bottom left. Do you see anything remotely resembling a really wide peak?  If not, turn the EMV up, up, up... a hundred or so volts at a time. Do you see anything remotely resembling a peak?  With AMU Gain and Offset at 0, the 'peak' may be wider than the window and that's OK for now. Even with no peaks, you want to see if turning the EM up changes the signal. If it does, great! If it doesn't there is another test to do.

    If the signal goes up, and up, and up...the EMV and signal processing are probably working fine.  Whatever EMV you got some large response, keep that level, save the tune file, again as COLD_Tune.U.  Most of that won't actually ever be used to tune. Having a tune that gets everything ready to vent is handy.

    Try Filament 2.  Better? Same? Worse?   

    Then vent it.  When it's ready, open the analyzer door, take a picture of the end of the transferline, and share it here.  The column can be barely sticking out the end of the metal transferline, pulled back flush, or even a tiny bit back inside -- I tested one at 50mm pulled back inside and only then started to have some odd chromatographic peak shapes, so a few millimeters probably won't cause any issues with most applications.  If the column is sticking out too far it blocks the electron beam and you lose essentially all the signal.  This happens quite a lot, especially since the very last thing you did was clip the column.   How much did you cut off?   Did you use the proper long grph/vespel ferrule with the narrow end into the nut?   A giant leak can masquerade as a blown filament, too.

    Verify that all of the wires are connected properly. If the entrance lens and ion focus wires are swapped the symptom can look like this. In fact the best thing would be to remove the source, take it apart and check it out, too. I mean you have the system vented and the source is right there, so checking it now is smart.  I've seen column and ferrule chunks inside the ionization chamber.

    Use a voltmeter and measure the filaments. They should both read very low Ohms.  Move the one you had in F2 to F1 and visa versa - just because it feel good to move them. The system needs them both installed. Depending on where one of them is broken it might or might not be able to run with one good and one bad. 

    Reassemble, reinstall, pump it down, wait a full hour for good vacuum - it takes way longer than most folks think for good vacuum, and then try it again. In MSD Chemstation you would do Reset to Defaults -- in tune, save it as atune.u, and then try it again. 

    Make sure that all of the cables are connected between the electronics module and the side board before turning on the power to the MSD.  Do not plug or unplug those cables in with the power on --as that will damage the side board.

    and let us know.........

  • Hi Paul,

    Thank you SO much for your suggestions! Here's what I found:

    1) I totally accidentally lied about the filament! It seems it was another instrument we replaced the filaments on. When I took apart the ion source the filaments looked terrible - black and deformed. I will definitely change these in the morning. I wouldn't be surprised if this fixes the issue although I am surprised that this issue appeared only exactly the first time after I had opened the analyzer. The ion source also looks very dirty so I will clean it tomorrow.

    2) When I do the manual tune settings you suggest I do get a peak at EMV of 1200 for about a few seconds. Then it goes back to looking like grass. As I increase the EMV the peaks still look like grass and have a very low abundance. If I let the instrument sit for several hours then try again I can see the initial peak again for a few seconds but then it goes away. If I switch to the second filament I also only see this peak for a little bit and then get grass. Could this be the result of the filaments dying? Or perhaps I actually have run out of PFTBA.

    The peak I see for a few seconds:

    What the peak looks like shortly after (doesn't change even if I pump the EMV):

    I had cut about 6 inches off the column and I did replace the ferrule/ms interface nut. The column is flush with the metal transferline - that's great to know that it can be quite close to flush and still be ok.

    I will change filaments, refill PFTBA, and clean the ion source and let you know if this solves any of the issues. I still am surprised if this is the issue because the symptoms only began after I vented the analyzer to cut the column and then pumped back down and started the instrument again.

    Thank you very much again!

  • Great.  Make sure to let the system pump down sufficiently before trying anything --- any air in there when the filaments turn on will cause premature damage.  Don't even take a peek, a quick look.  Wait an hour or more before testing.

    and let us know.

  • Hi Paul,

    We fixed the machine!! Thank you so much for your time and advice. If it wasn't for your manual tune tips I would have just written this off as a sideboard issue. The problem was that we ran out of PFTBA... Embarrassing since I said I didn't think this was the issue but alas it was. 

    I think the issue was also hard to pinpoint because I didn't know that you shouldn't try to run a tune within an hour of pumping down. I pumped down and immediately tried to tune. I didn't get anything (because we were out of PFTBA) and as a result I kept running tunes and the filaments blew out and the ion source was extremely dirty.

    After replacing filaments, cleaning ion source, and replacing PFTBA we are up and running again and the instrument is in the best shape I can remember. Thank you so much again, I was quite desperate!

  • My mentor a very long time ago told me, "it's not a machine, it's an instrument!  Machines do work, instruments make measurements."

    Glad it's working.  Again a very long time ago learning troubleshooting the instructor removed the PFTBA vial, dumped it out, and reinstalled it. The system still tuned five or six times before it wouldn't.  It doesn't need much, but it needs it for sure!

    Please watch these source cleaning videos when you get a chance as they are packed with great information: Eliminate the Fear of Mass Spec

Was this helpful?