Baseline increase with 5890II (after FID runs out of air)

[hakanolcay]

Hello,

I have had this problem twice so far. It happens every time when the FID runs out of air.

As the air pressure in the source line decreases first my signal drops with the flame still on.

First, I turned off H2 and air flows, and changed the air cylinder and fired it up again. My signal which used to be 10 was now 400. And it wasn't stable at all no matter how long I waited for. It was fluctuating around 400.

I tried shutting the GC down completely, checking all the flows, re-starting the GC, and even re-conditioning the column. --didn't work..

I ran the GC at this baseline level for almost two weeks until the GC ran out ouf air again. (I wish I had changed the tank one day before.) Again the flame was on. I turned off the flows, and fired it up after replacing the tank. This time, the baseline was 4000!

Previously, I had been shutting the GC down to change the gas tanks and re-starting, but the baseline had always stayed at 10.

Here is the other information:

GC: HP 5890II
Column: AT-Q
Column flow: 4 cc/min
Aux flow: 26 cc/min
H2 flow: 30 cc/min
Air flow: 400 cc/min
Total flow: 4 cc/min. (i.e. splitless) Although the split vent is open I don't really see any gas coming out because I am sending everything to the column. I am using an on-line valve for sampling and trying not to lose anything through the vents. Septum purge valve is closed during analyses.
Oven temp: 40 to 180 at 15 C/min
FID: 200 C
Inj: 200 C
GC equipped with Electronic Pressure Control system

With ramping, my signal is obviously increasing somewhat. But the thing is every time the GC runs out of air it starts with an increased baseline and this signal never goes back. If anyone of you has already come up with such a problem or has any idea can you help me solve this?

Thanks,
Hakan

[Peter Apps]

One possibility is that the GC is autozeroing while the air is low and the FID signal is below its normal baseline. Then when the air is back on again and the baseline is at its normal level the signal readout is correspondingly high.

Peter

[gcguy]

Try giving it a thump. A bit unscientific I know, but sometimes the jets on 5890s get a little bit of dirt stuck on it and giving the top of the detector a gentle tap usually sorts it out.
If this fails you may need to strip down the detector and clean everything.

GCguy

[hakanolcay]

I took out the detector and it was full of dirt, kind of coke.. I cleaned the collector using hexane and methanol, and passed compressed air through the jet. Now it shows the exact same signal as it used to do!!

Thanks guys for your replies!

Hakan