Hi Joe
If cleaning the source and a new filament doesn't do it, then the problem is likely the source magnets; you may need to replace them.  Sensitivity drop over time is a typical response of the source magnets loosing their magnetism due to the heat near the filament.  I have changed only 3 sets over 25 years of working with the IRMS; 2 sets on VG machines (one magnet cracked from the spot weld) and one set on Thermo.

The other possibility is that one or both of the magnets poles are in the wrong direction after removing and reinserting them (e.g. source cleaning).  Make sure the south pole is facing of the filament and the north pole is facing the trap.

Regards
Gilles  

Gilles St-Jean
Chercheur / Research Scientist
Université d'Ottawa / University of Ottawa
Sciences de la Terre / Earth Sciences
140 Louis Pasteur, Ottawa, Ontario, Canada K1N 6N5
Tel: 1-613-562-5800 xt 6830 (Bureau / Office)
                                   xt 6839 (Bureau / Office Lab)
                                   xt 6836 (IRMS lab)
Téléc. / Fax: 1-613-562-5192
Courriel / E-mail: [log in to unmask]
Toile / Web: www.isotope.uottawa.ca 

You can please some of the people all of the time, and all of the people some of the time, but you can not please all of the people all of the time". 
Mod. from Abraham Lincoln (attributed); 16th US president (1809 - 1865)

 

> -----Message d'origine-----
> De : Stable Isotope Geochemistry 
> [mailto:[log in to unmask]] De la part de Joe Lambert
> Envoyé : 09 janvier 2009 10:51
> À : [log in to unmask]
> Objet : [ISOGEOCHEM] Delta Plus Beam Sensitivity Decay
> 
> Hi all,
> 
>     I am writing to see if anyone else has experienced 
> similar behavior with their machine.  In a nut shell, our CO2 
> beam from REF1 (27 PSI) on our GasBench has dropped from 
> ~5800 mV (mass 44) to below 2000 mV over the span of about 2 
> months.  We have a Delta Plus that is about 7 years old and 
> our Gasbench is less than 2 years old.  We're typically 
> running very clean sample gases (CO2 and H2, never S2) and 
> probably have less sample throughput than most labs. The bulk 
> of the work lately has been carbonate powders where our 
> sample heater block is at 50 oC and we are doing manual acid 
> injection.
> 
>     Problems first appeared the end of November, 2008 when 
> both reference and sample peaks during a 25-hour GasBench run 
> dropped dramatically (but
> gradually) from sample 1 to 96. We stopped measuring samples 
> and started to diagnose the problem. What we discovered was 
> that if we did a peak center followed by a scan of the 3 beam 
> intensities through time, was that the beam intensities 
> dropped (see linked PDF below).  The rate of "beam decay" 
> varied but could be ~200 mV per hour.  Our 5800 mV beam was 
> down to 5400 mV after just a couple hours; however, the next 
> morning the beam would be back up to 5800 mV (pump time 
> helped?). We introduced CO2 from a separate source (Conflo 
> III) and from the dual inlet system and saw similar results 
> (PDF).  The Conflo and GasBench share a common CO2 ref tank 
> and He carrier tank. Never did we see the CO2 beam increase 
> during a scan through time. We've tried using a different He 
> tank but the results are the same and the CO2 tank is ~1-year 
> old and does not appear to be at the end of it's life.
> 
> http://www.geo.ua.edu:16080/asil/Documents/sensitivityissues.pdf
> 
>     We have put in a new filament/cathode and tried multiple 
> tunings for the source settings, none of which has had an 
> effect.  Our water background looks reasonable (PDF). A mass 
> spectrum (background) scan with the SGE/needle valve closed 
> resembles a scan we did before the problems started.  Several 
> months ago (>6) we did have some minor problems with the 
> sample needle possibly sucking in some acid but I'm not sure 
> if this could be a problem that is just now showing up in the 
> source.  It should also be noted that it has been about 2 
> months since any real samples have been measured so I would 
> think the system should have purged itself by now...but maybe 
> acid is stickier than we realize.
> 
>      We also experienced times where the beam would decay 
> through time (one day to the next), but in most cases we 
> could adjust source setting to reacquire the full beam.  
> However, over a long time period our maximum possible beam 
> has gone from ~5800 mV at the end of November, to ~3000 mV 
> before Christmas break, to <2000 mV today.  I do not know if 
> we have two problems superimposed on each other or if one 
> problem is causing the other.
> 
>     Possible causes of the problem in my mind could be 1) 
> contaminated CO2 or He gases/capillaries, 2) contaminated 
> source, or 3) electronics issues. 
> Since the behavior is similar between GasBench, Conflo, and 
> dual inlet gas sources I would somewhat lean toward a 
> contaminated source; however, the recovery time with pumping 
> suggests something in the gases.  As mentioned, our machine 
> is 7 years old but we have never cleaned the source because 
> it looks almost as clean as day 1 and we've never experienced 
> any "dirty source" type problems (until now?).  Would a dirty 
> source result in such a quick decay in the beam?  I had 
> noticed a few comments on ISOGEOCHEM regarding similar 
> problems caused by electrical issues, but I am not sure if 
> they symptoms were the same as ours.  Our machine has 
> experienced many harsh uncontrolled shutdowns because of no 
> back up power supply so wear and tear on the electronics is 
> not out of the question.
> 
>    Sorry for the long message but I wanted to cover as many 
> things as possible to hopefully rule some things out.  I'm 
> sure I forgot to clarify some things though.  
> 
> Thanks in advance,
> 
> Joe Lambert
> Alabama Stable Isotope Laboratory
>