GC problems with an FID

[soak]

Hi folks,
We're are currently having a very weird problem on one of our GC's. We are detecting N20 on an ECD and CO2 and CH4 on an FID with methaniser. The CO2 is behaving nicely for all standard samples, but the CH4 changes dramatically depending on user! We have one person using it (filling evacuated vials with a standard gas) who gets very good reproducibilty and accuracy, while someone else doing exactly the same procedure gets very different results. Both get the same CO2 data but the CH4 is erractic for the second person. Both are analysed on the same detector so I would expect the CO2 to be weird as well if it was purely a user error. Any suggestions on how to eliminate this?

Thanks in advance

S

[Johnny Rod]

How are you preparing and injecting the samples? I don't quite follow

[soak]

Standards are from a tedlar bag filled from a cylinder. A sample is taken from with a syringe into a evacuated glass vial and then sampled by autosampler needle on the GC. I have watched each person fill evacuated vials and nothing glaringly obviuos is amiss.

[Johnny Rod]

is it a mixed gas standard or are you making the mixture in the vial? It's a funny way to handle gases. Do you break the vacuum inthe vial with another gas or is it still under vacuum when the autosampler goes in?