HI Melanie--

I take it you checked the leads already, so no shorts there.  And you've 
probably cleaned the source a few times--and checked/cleaned the 
spacers.  You may want to double check the lenses past the slit--make 
sure they have the slit "centered" and are not twisted after tightening 
them down.  Also, have you checked the LED's on all the control boards?  
One of them may have failed.  Sometimes it can even be a "dull" LED.

High water in the mass spec will through things off, especially when 
levels are changing a lot.  And yeah, you're instrument will get to the 
point where you'll need to condition it a bit before starting runs.

Unfortunately chasing linearity issues is a problem.  Autotune can 
actually hurt linearity.  You can use it, but then try "defocusing" the 
beam a bit.  Also do background scans with the without CO2 going into 
the source--make sure the baseline hasn't gone off a bit.  And a hv scan 
on the CO2 peak will show if you have a good peak shape too.

take care,


On 7/24/2019 3:37 AM, Melanie Gimpel wrote:
> Hi all
> I apologize for asking about the same topic again, but we are really at a loss now.
> After a long time we put an older MAT 253 back into operation and do not get any linearity and actually I do not like
> the stability either.
> Source is clean (cleaned by Thermo), the filament is quite new, vacuum is good, background values are ok, water is
> around 1000 mV.
> At the moment it’s connected only to a ConFlow III and we measure CO2. We have disconnected the EA and TC / EA in
> order to rule out a faulty influence.
> At the beginning we had a lot of water in the system, especially in the needle valve. If we turned on the needle valve
> heater, we had 50V.
> The stability is ok after about 30 to 60 minutes (Stdev 10 pulses around 0.05), but in the first std_on_off runs of
> each day there is an enormous drift (min value to max value up to 4 permille over 10 minutes). I've never seen anything
> like that before.
> If we felt that the MAT was now quite stable and we wanted to focus on linearity, we found that it does not matter
> which extraction is set. We do not get any linearity.
> If anyone still has any idea, please let me know!
> Regards
> Melanie
> /// Wissenschaft im Dienst des Menschen ///