Linearity problems

[paddypower]

Hi All
I ran the Agilent linearity standard on my 6890 GCFID ,Splitless injection
there are 6 hydrocarbons in the mix C17 C18 C19 C20 C22 C24 with a massive conc difference between C17 and C17 in the order of 100,000
Normally this works really well even if the peaks at the higher conc are poor shape(column overloaded)
this time however the C24 peak which should be 10 times higher area than the C22 peak is in fact only 6 times which from c17 to C22 is perfectly linear
I have changed all parts liner ,septum cut column both ends changed liner collector used new standards used 3 differentcolumn and got another analyst to run system with same result
Its all very consistent just seems the linear range has been reached
has anyone seen this before and could suggest a remedy
thanks
Martin

[tkubowicz]

Hello

Have you got the same liner type? Glass wool position (center/bottom or no glass wool) is very important as it provides additional surface area for complete volatilization of the sample to minimize thermal discrimination.

Regards

Tomasz Kubowicz

[paddypower]

I used many different types including split ,splitless and headspace liner(just to provea point) and all gave same ratio although there were variation on thre areas and peak shape
I was thinking that it maybe my electrometer that is going as the peak areas for all peaks is much lower than i would expect