Strange noise with FID 5890

[AZBiodiesel]

Hi all, i don't have time right now to post pictures of the problem but here is whats happening. I am running a 5890 with FID on column injection with 15m capillary column plus guard column. I am getting an FID signal that fluctuates and looks like electrical noise. Normally the signal on the machine is stable around 4 all the time, but yesterday it started fluctuating steadily from 4 to 30 up and down up and down cycling very rapidly. I ran a sample on it and the chromatogram looks as it normally should, all the peaks are where they should be, but it looked like it was inverted since there was so much noise spikes covering up and over the peaks. So we left it alone for the night, since it was late, and in the morning it had dissapeared, we ran two samples on it and it started doing it again. So i took apart the FID and cleaned it, checked for leaks on the column, Checked the air and hydrogen bottles, and started the machine up again, and the noise was gone, the chromatogams looked good nice normal peaks. However it has come back and started cycling again however now the signal cycles only from 4 to 7 or from 4 to 5. What could be causing this? Im figuring its an electrical thing but we already replaced the wires from the recorder a couple months ago, im hoping its not something with the board. Thanks

[aldehyde]

When you cleaned the FID how did the insulators look? Were they discolored, cracked etc? If you turn the FID off does the signal go down to 0.0 or is there some residual current (a few pA)? Also, while looking at the signal try tapping/lightly banging around the FID. I have seen improperly grounded FIDs that will have noise spikes from vibrations in the instrument. I have also seen carbon threads grow between the collector and the electrometer, causing noise problems (though they usually aren't rapid cycling noise). Does the noise increase when the oven, inlet, or detector are heating/cooling?

Is the instrument on its own circuit or does it share with other equipment? Have you installed anything in the room recently?

If you have a voltmeter check the line voltage and make sure you are getting the 120 V required.

Dumb question but have you restarted the instrument?

Probably unrelated but when is the last time the jet was changed?

[AZBiodiesel]

aldehyde, when i cleaned the FID the white insulators were not discolored or cracked, there did not seem to be a lot of soot or buildup or particulates on any of the pieces. There is no residual current when the flame is out. The jet is almost new i put in a new one a month back. It is on its own circuit and I have turned the machine completely off and back on.
I recently pulled out the board and blew some of the dust off and pushed it back in and I have not seen the noise since. Hopefully it won't come back I have my fingers crossed.

Although I thought maybe when it was cooling it was the fan blowing on the column and maybe it was a loose loop of column moving back and forth due to the force of the fan, so I rewound the loop and made sure it wasnt moving around. But i don't see how that could create electrical interference. Anyway it has seemed to stop. Thanks for the help.

[aldehyde]

Could be that the board was working its way loose from the slot on the mainboard via thermal expansion. Hopefully it doesn't come back but keep us updated if it does .

If some of the column were touching the oven wall when it was hot and the fan was blowing on it I guess you could potentially see some cycling noise but that would be pretty unlikely.

[AZBiodiesel]

Ok its back! The noise has come back, I am thinking it is due to moisture contamination from one of the air tanks, since we are broke we decided to skip putting a moisture trap in line from the air and hydrogen tanks. The problem started shortly after we put in a new air tank. Since the last post I took the FID out to check it again for particulates and about 3mL of water poured out of the collector housing after i unscrewed it from the machine and took it apart on the bench. As you can see the column is fine the peaks appear but they are covered in noise.
Here are some of the chromatograms, i hope i am doing it right:

[Ron]

I suspect you may have a bad signal cable. I have seen something similar although not as bad when a heater cable was too close to a detector signal cable. There was RF interference when the heater cycled on and off, and it was worse at higher temperature. I determined the signal cable was an issue by wrapping it in aluminum foil to shield it and the noise went away. I replaced the signal cable and rerouted it to try to avoid the problem in the future.

[Pheroman]

AZB-

What temperature is your detector at? I would be surprise if it is possible to get that much water in your detector if you run it at normal (250-300C) FID temps.
Are you sure your detector heater is working? Does your 5890 power off while the detector is still lit? If so, condensation from the flame will build up, depositing moisture and eventually rust.

[AZBiodiesel]

Phero- the detector is at 380degC at all times. However I think the FID went out with the gas on over the weekend and moisture from the tank accumulated in the FID, however it is a lot of water for just contamination from a tank (i think). I always leave the oven at 50 deg C and the Detector at 380. I don't think there was ever a situation with the FID on and the detector not hot.
The only other thing i can think of to explain the water is if the other lab person turned it off, or maybe he checked the flows with our bubble meter and water got into it from the tube to the bubble meter. But he says that didn't happen.
The reason why i don't think it is the signal cable is because we had the GC serviced 3 months ago and I think the detector cables were replaced then, (these are the wires with silver insulation coating right?). Thanks

[AICMM]

AZBiodiesel,

I would be very surprised to have that much water from a detector kept at 380C. Even though I have seen water accumulate in 5890 FID's it is usually when they are run quite a bit cooler. Therefore, my suspicion would be that the wrong heated zone is hot (I have seen this before.) Does your instrument have two detectors? If so, someone may have swapped the heaters and you are heating B while you think you are heating A. If you put you hand over your detector with the flame off, at 380, it should be much to warm to touch.

Best regards.