My headspace GC-MS method includes a 20 minute incubation and has a cycle time of 11.5 minutes. For weeks, we were getting consistent injection times and the overlap was working very well. But something has changed. Now, the time between injections is variable. Digging deeper into the audit trails (mslogbk), the incubation time is anywhere between 20 minutes and 48 minutes. It will typically perform the first 2 injections correctly, then times will start to vary (with no pattern but never less than 20 minutes). This is even the case when we set a very long cycle time to force the instrument not to do overlap. What could be causing this?
I am using a 7890 GC with a HES source and coupled to a PAL RSI 120 autosampler. Software is the ECM compliant version of MassHunter. Software version and build have not changed, PAL sampler software and firmware versions are unchanged. "Script" for autosampler was last changed in 2021. The exact same version of the method has been used, it has not been updated at all between good sequences and bad sequences and the "increment incubation time" setting is not enabled. There are no error messages in the logbook and it is not waiting for GC ready state during incubation. Preventative maintenance was performed by Agilent between good and bad sequences and some parts were replaced in the HES source.
My next thing to try is to re-build the method from scratch and wiggle all the cables to ensure the connections are secure. If this doesn't work, I have no more ideas so any suggestions are welcome. Thanks!
