Error Message "cannot achieve constant peak widths"

[astroguille]

Hi people!

I work at a GC6890/MSD5973 (with a turbo pump and cuadrupole).
I execute the Autotune and then the system aborts, because of "cannot achieve constant peak widths".
Before that, I cleaned the ion source.
¿What can the problem be?

[lostcosmos]

Hi people!

I work at a GC6890/MSD5973 (with a turbo pump and cuadrupole).
I execute the Autotune and then the system aborts, because of "cannot achieve constant peak widths".
Before that, I cleaned the ion source.
¿What can the problem be?

Does standard tune work? If so you can try entering the the parameters from your last passing autotune in manual tune mode. Then save that tune as autotune.u and do an autotune.

[AICMM]

Astroguille,

Lostcosmos brings up an interesting point which I don't know if you have discussed. What does the tune look like when you do a manual tune? Specifically, abundance for 69, 219, 502, mass assignment and PW50's for each of these? Can you post the tune?

Best regards.

[dhan]

I had this happened for some time and my side board died. I spent $4000 to replace on my 5975.

Make sure you have either warranty or annual service which covers this expensive part.

Just my recent experience and not trying to scare you.

Also check vacuum level as compared to table provided in Agilent manual.

Dhan

[Marge]

I had this problem for the last month and only got it sorted recently. After several phonecalls to Agilent, the next step was to call out an engineer (not on contract so was really a last resort). It took him 5 minutes to fix the problem (which I could have done myself).

Basically, the tune file was corrupt, so the following needed to be done:
1.From the Windows desktop, open the Config programme
2.Go to Instrument 1
3.Keep clicking Next until the Finish button is present and then click Finish.

When autotuning the system, go to the Tune screen (in Instrument Control, select “View – Tune and Vacuum Controlâ€

[dspriya1986]

Dear Marge,

Do we need to uninstall the chemstation before doing this configuration step?

Looking forward your response.

Thank you!

[Steve Reimer]

No, you should not have to uninstall.
You should be able to load an old tune file an save it as autotune.u. Then you will have real values to start with. Old tune files are kept under msdchem\instrument 1\ ei tune or something like that( I'm not near an instrument). Copy the old file into the same directory as autotune.u and rename them both. Start the autotune again and watch the graphics flash by. If it can't find the peaks you have something else wrong. It it just takes a couple of tries but eventually tunes then your file was corrupt.
If you just had the source out for cleaning and the ion focus doesn't give you a maximum, but just looks like a noisy line when plotted, you may have switched the purple and orange wires when you put the source back in.
Good luck!

[lmh]

blue at the back.
And if you've managed to route these two wires so that they obstruct the cone where the column ends and there's a hole into the ion-source, as I stupidly did once, you will also get rubbish results on tuning, but that's another story.

[dspriya1986]

Thank you very much for your responses.

Blue and orange wires were connected as suggested. Replaced the old tune. Configured MSD. I baked out MSD for 6 hours.

(History : I cleaned ion source. Changed filaments, lens insulator and repeller insulator).

But still i am getting the same error message. I called Agilent and they said i need to change analyzer side board and it is too expensive.

Please suggest me other possible ways to solve this problem.

Thank you!!

[James_Ball]

If the side board is bad, not much else you can do. About $4000 but you can change it yourself just have to read up on how to adjust the quad driver screws (also known as DIPing the quads).

You may want to look closely at the side board and see if there is a swelled or broken resistor just to the rear of the box in the center of the board. I had one go on a 5971 once and it was a blown resistor there. If you have someone that could replace it with the proper size maybe you could save it.

Another thing to check inside the analyzer is the wires that go to the quads. Make sure they did not get pushed around when the analyzer was open. They are the bare wires that frame the outside of the quads. Of one of those is grounding on something it will give a similar problem.

Peak width problems are almost always associated with the quads or quad driver circuit which is located on the side board. Dirty quads will also cause the problem. Have you put it into manual tune and set it to profile and looked to see what the peaks look like?

Oh and another thing that can cause the problem, and everyone that has done it kicks themselves after figuring it out, is running out of the PFTBA tuning compound in the little vial on the front. Check that to make sure there is still liquid in it, plus if that solenoid valve goes bad you will get a similar error. To know if that is the problem make sure the analyzer pressure jumps higher as soon as the tune valve is opened.

[dspriya1986]

Thank you!

I have not vent the system yet. I will check the analyzer wires later. But I checked the PFTBA level and it is 1/4th of the vial. Is that sufficient? Also could you please guide me with the manual tuning?

Thank you so much for your time.

[James_Ball]

Thank you!

I have not vent the system yet. I will check the analyzer wires later. But I checked the PFTBA level and it is 1/4th of the vial. Is that sufficient? Also could you please guide me with the manual tuning?

Thank you so much for your time.

That should be plenty of PFTBA.

When you are in the Instrument Control window, where you see the buttons to edit methods and such, at the top there is a menu item that says VIEW. That should have a sub menu of Tune and Vaccuum Control, choose that and you will be in the tune window.

It should load the tune file being used by the currently loaded method. At the top you will see a menu called Parameters and in that will be the sub menu Manual Tune, select it.

When that window opens use the button labeled Profile, to see what the mass peaks look like. If you can get a screen shot of a profile scan and post it it may help us to better identify your problem.

[dspriya1986]

Thank you for your time.

I could not attach the picture here. Hence i attached the link for its location in dropbox.

https://www.dropbox.com/sc/a93o31yyaj9r7ue/rqFqiTHhlT

Please have a look and help us to find a problem.

[James_Ball]

Thank you for your time.

I could not attach the picture here. Hence i attached the link for its location in dropbox.

https://www.dropbox.com/sc/a93o31yyaj9r7ue/rqFqiTHhlT

Please have a look and help us to find a problem.

That scan actually looks very good on the peaks and abundance. I do see one major problem, turbo speed and HiVacuum. Turbo speed should be at 100% and the HiVacuum should be in the 10^-5 torr range not 10^-4 torr, that is too high. With such high pressure in the system the autotune is probably having trouble keeping things in the proper ranges. The turbo speed being less than 100% is either a sigh of a very large leak, super high column flow rate or a failing turbo pump or roughing pump.

What column are you using and what is the column flow setting?
Has the instrument gone through pumpdown?

Where you found the manual tune window there is also a Vacuum menu, what does it give if you select Vacuum Status?

I am having trouble figuring out how the source and quad temperature is up without the turbo speed being so low, usually that would trigger a vent mode and shutdown after a little while.

Where you are there scanning in manual tune, use the MoreParameters menu and use the Tune Parameters menu and change the 219 and 502 mass boxes to 28 and 18 and do a scan and see if your air and water peaks are higher than the 69 base mass of the PFTBA.

The only reason I can think that the HiVacuum would be so high if everything else is working corrrectly is if you are using Hydrogen as carrier since it doesn't pump as efficiently with a turbo pump.

[manman]

I had a similar problem with my old 5972A. Autotune could hardly find m/z 69, 212, 502 and when that was possible, the peaks were too wide, with unstable width and with too much noise on top of them (the peak of each m/z was split at several points and it did not look like a single, gaussian peak). I firstly suspected my old electron multipler. Though, I noticed that when I was changing the EM voltage, the intensity of detected ions (if you can call ions all these noise peaks) was changing accordingly. I then suspected the cables connecting to the topboard of the MS analyzer. I disconnected-connected, pushed, pulled but nothing happened. The message "cannot achieve constant peak widths" continued to pop-up and the autotune could not complete. The deletion of "Instrument 1" and re-addition of new "Instrument 1" using MSConfig, did not help a lot. I finally started playing with the options in the Diagnostics/Vacuum Control. Mass Spec Status was "OK" and the same was for "Mass Spec I/O Test" implying that there was not an obvious problem in my electronics. I finally clicked "Initialize MS Interface" in the Diagnostics panel and after a few seconds the model of the MS was shown on the screen. After that, the Autotune was able to find m/z 69, 212 and 502, the noise gradually went away and the Autotune procedure completed normally, as it used to do in the past. I don't know if that was just a coincidence or if this "Initialization" solved the problem. Having spent two days with this issue, I wanted to mention this trick and hope it will save time from other facing a similar problem.