Wave in the basline of the GC-FID

[shadu]

Hi,

I am using a Agilent 2890 serie GC-MS equiped with a FID detector and coupled with a purge-and-trap (tekmar 3100).

We are developping methane analysis on this GC and installed the FID for that purpose. But form the beginning, we did get a periodic wave in the baseline. Periodicity = 6 min and amplitude of 2000 (response). The small conpagnie that installed the FID did change the FID place in the electronic boardd (front to back) then changed the FID with major result.

We did use nitrogen/hydrogen/air from a generator and bottle without change. We did cap the FID and try helium and hydrogen as carrier gas without effect.

Any idea?

Sebastien

[chromatographer1]

Since you capped off the FID, and I assume you had clean supply lines from your gases to the GC and you did not contaminate the EPC controllers or were they contaminated at the start, the only logical source is electronics.

That sadly is the most expensive link in the chain.

best wishes,

Rod

[GasMan]

Generally, if you have a SLOW variation in a signal, it is not internal electronics on modern instruments. It could be external pick up from a piece of equipment nearby that is switching on and off at the period you mention. If you turn the flame off, do you still get the problem. If the answer is yes, you probably have an electronic problem. If the answer is no, you probably have a pneumatic problem. A common problem can be an external gas regulator which is sticking and it will will periodically 'burp' and cause a change in pressures or flows.

Gasman

[chromatographer1]

Gasman has a good point.

Put the FID out and see if the signal still waves. That is a good diagnostic step. No fire, no wave, then it is a contamination problem, not a voltage or power supply problem. If it persists then look into the building's electrical supply and possible interferences, then look inside the GC itself.

If not electrical or electronic, you look for a possible grease contamination problem from the air or hydrogen EPC controllers and trouble shooting is the required first step.

If the EPC is contaminated it may be necessary to replace, not clean them.

Good hunting.

Rod

[shadu]

Thanks you for your response.

I just try the turn off flame tips and i have no signal at all so it is not a electronic/electric background problem. The pattern is perfectly perriodic (at all 6 minutes) and it will be very strange to have a random problem like regulator sticking making this perfect pattern over and over. I did put a image of the baseline on the FTP of our compagnie. Take a look.

ftp://admceaeq:MtwVCXro@ftp.mddep.gouv. ... seline.JPG

At first we were thinking about a variation in the current of the detector cause by the heat cycle of the FID but the compagnie who did the installation of the FID tell us they made this trouble shooting first. And all the valve and regulator IN the GC were installed with the FID so they are brand new.

Ty for your help

Sebastien

[dblux_]

... The pattern is perfectly perriodic (at all 6 minutes) ...

If your FID gets air from the compressor check the interval of compressor switching on (and compare with 6 min).

[shadu]

Already done dblux. It didn't match at all as the central compressor have no cycle in it. We also have a new GC-MS (3 months) with FID and during the installation the FID was tested and no wave affect were seem in the 5 injections. After that the FID was not used (running none stop on MS).

However i will try to switch the air form the compressor to the backup system (bottle) later this week.... Air are used on others instruments that run almost 24h/24h so it is a little difficult.

Thanks

Sebastien

[GasMan]

Sebastian,

I would look at the hydrogen supply as the source of your problem. If you look at the characteristics of the FID, you will see that the ratio of hydrogen to carrier gas is important. If you vary the hydrogen flow and keep the carrier flow constant, you will find that by starting with a low hydrogen flow and increasing it, your signal will increase to a maximum and then decrease. On the other hand, if you vary the air flow, you will find that the signal will increase and then remain constant over a fairly large flow. So changing the air flow from 350 ml/min to 450 ml/min will not have the same effect as changing the hydrogen flow from 25ml/min to 30 ml/min.

Gasman

[Peter Apps]

Hi,

We did use nitrogen/hydrogen/air from a generator and bottle without change. We did cap the FID and try helium and hydrogen as carrier gas without effect.

Any idea?

Sebastien

Hi Sebastien

I would have put this down to an artifact of the gas generators - most likely the nitrogen, which will be operating two scrubbers on "pressure swing", so can you confirm that you have run the FID with bottled gas for all three gasses and seen no improvement in the baseline ?

Although it is very unlikely that a mechanical regulator would generate a regular cycle, it is entirely possible with an electronic regulator. You need to actually measure the volume flow of each FID gas at the FID with the column connection capped to see if there is any change over a cycle of 6 min - use an external flow meter, not the onboard meters of the FID.

Peter

[GasMan]

The suggestion that a mechanical regulator causing the problem was not made on guess work but from experience, on more than one occasion. If dirt gets into the regulator, it is possible for the needle to stick in the orifice of the regulator. The pressure will change inside the regulator until the spring can overcome the 'sticktion' and force the needle back to its correct position. As you are supplying the same flow, the pressure decay will always be the same, so the time when the spring reacts will always be the same. Hence the reproducible time.

The same is true for electronic regulators, 95% of the time it will be the electro-mechanical valve inside the unit that is causing this type of problem.

Gasman

[AICMM]

shadu,

I am not familiar with 2890 so please tell me, do you have EPC or manual pneumatics on the FID? Also, what supply pressure are you giving the instrument and how many other instruments are hooked to this set of gas lines?

Best regards,

AICMM

[ant78se]

If nothing helps ... move your mobile phone away from the GC. Or PC and other wireless devices.