Changing PE Autosystem from packed column to capillary

[ratwing]

I'm installing a megabore capillary (to look for glycols in glycerin) in to an old PE Autosystem GC previously fitted with a packed column but there doesn't seem to be any facility for detector make-up gas - the instrument has only one detector and the manual makes no mention of make-up gas when installing capillary columns. The first few injections gave not only tiny peaks but they were negative too. (None of the wiring has been changed and the polarity is greyed out in the Totalchrom software).
None of us have much experience in this area so any suggestions would be gratefully received Thanks guys

[CE Instruments]

Detector type ?
Flow rate you now use on the Megabore column. If used like your packed you may not need make up gas. (Some GCs use Capillary designed detectors so no make up needed for Capillary)

If you put the packed column back do you get proper peaks ?

[ratwing]

Its FID. The flow rate is 5ml/min but with the temperature program (100 to 220 degrees) it drops to about 3ml/min by the end. I've tried increasing the flowrate at the start of the run so that it ends up at 5ml/min at 220 degrees but apart from different retention times.
The column is 0.53mm x 30m, injector 220, detector 250, oven 100 to 220 linear program over 16 minutes, 10:1 split ratio. I won't get time today but might try a higher flowrate - worth a go do you think?

Unfortunately the packed column got broken so it can't be tried...

[AICMM]

Ratwing,

The fact that you broke your last column means it was 1/4" glass which means you should be able to easily build yourself a makeup adapter. Take a 1/16" T; on the branch run a gas line outside the GC oven to a needle valve capable of 0-80 mL/min, on the straight run you run your megabore all the way through the T and on the other part of the T you run a 1/16 piece of tubing with an id big enough for the column to slide all the way up. Install the 1/16" tubing into the FID and you are good to go. In this way, you can run the column at what ever is optimal and run the FID flow with added make-up.

Not beautiful but it definately works. I guess my other question is why you are doing a split, you didn't do that before with the packed column did you?

Best regards.