FID signal maxed out

[skunked_once]

Today I helped someone troubleshoot a maxed out FID signal which occurred right after changing the air tank. At first I suspected a bad tank of air but he changed back to the previous tank which was clean and the problem did not go away. The next step was to instruct him to cool, disassemble, and clean the detector. This resulted in a normal signal so the problem was in the detector. He told me afterward that when he took the detector apart, there was a clear liquid inside the detector body. This was the "AHA" moment. I have seen condensation in the detector if the flame is lit before the detector is up to operating temperature. The moisture must have shorted out the detector and produced the maxed out signal. Lesson learned, always be sure the detector is hot before turning on the flame. Our first GC lesson of the new year!

[tpmcginnis]

I've noticed the condensation issue with FID's a couple of times like you describe. I recently saw something like this, but with a twist.

We've had the air tanks completely run out due to negligence on our part. The detector has stayed hot, but the flame has gone out. After sitting in this condition for awhile (let's say over the weekend) the signal is maxed out at 830000 (HP FID). Cool the unit down, start to take apart the detector.....and it's filled with water!

With the detector at 300C the whole time it didn't seem likely that water would be there, but it was. And, the detector parts were corroded as well (the collector body, the chimney, etc.). Wondering if the gas contained water contamination in the tank that carried over as it ran out to completely empty?

[skunked_once]

I'm assuming that the hydrogen flow was still going after the air ran out. Could it be possible that the hydrogen combined with oxygen from the atmosphere to form water? If you have the proper traps hooked up there shouldn't be any moisture from the air tank. It's still hard to imagine that water would stay in the hot detector but it did somehow. Someone should start a thread on the strange things that have been experienced in the field of chromatography. I'll bet service reps have a few stories to tell

[Ron]

This water issue wasn't that uncommon with the early HP 5890, I never did get a good answer from anyone at HP as to how you can condense water inside an FID sitting at 300 degrees.

[AICMM]

A particular pet peeve of mine concerning the Agilent/HP series of detectors. The answer to the water issue is quite simple really. Agilent has a heater, a relatively small heater, 2 1/2 inches away from the top of the chimney. Furthermore, the heat from the heater only gets transferred through the little neck on the detector body which reduces the overall transmission. Your detector heater might be at 250C but you can probably touch the top of the igniter tower where the water comes out. For people running halogenated solvents, I recommend a secondary heater sitting on the top of the detector body; prevents condensation and prevents acidic pitting.

Best regards.

[gcguy]

I have seen this problem as well but only with 5890 and only those with EPC control on the carrier. If the gas supply drops below a certain level then the system assumes there is a leak and shuts down all the heaters. If there is no EPC on the detector then there is often enough pressure left in the line to keep the flame lit after the detector has cooled down. Hence the water condensate. This is particularly bad if using helium as the carrier, then the flame will stay lit after the carrier gas has run out. Still a problem if using hydrogen.
One solution is to reduce the pressure set point on the gas inlet to the instrument. We use hydrogen as carrier and normally set this at 50psi using one of the epc units regardless of the column flow/pressure set on the column epc unit. If this is reduced to 10-15psi above the max pressure for the column epc unit
, if the carrier runs out then the flame goes out before the detector cools down. This only works when using hydrogen as the carrier.

This has turned into a bit of a ramble, hope it makes sense.

GCguy