Chromatography Forum

Hi,
I am running a 5890 which recently had an influx of water come in on the house air line. The filters were never drained and allowed water to bypass. So, I've rectified the issue and am ready to restart the gc. Is there any special precautions as I warm it back up? Any chance that my column (db wax) is still good? I am pretty sure water made it the full length as I got droplets in the detector (FID). Thanks!

Unless you have the GC set up completely wrong you are not putting air through the column - it is supplied only to the FID to support the combustion of the flame. So the column is probably OK, but you may have damaged the flow control for the air, so it is probably a good idea to measure the air flow to make sure that the controller is working. You should let the FID sit at its operating temperature overnight with all gasses flowing before you switch on the electrometer.

Peter

Another suggestion:

Find the person who shut off your chiller (which acts as a drier for your house air) without killing the house air first and throttle them.

Regardless, I would suggest putting some water traps just before your GC. In the past I've had good luck with Balston products.

Ok, so I baked out the column and detector overnight and fired it up this morning and a new problem arose. It doesn't seem to be sending a signal to chemstation. It downloads the method fine and tracks the oven temp but won't display a signal. If I check the sig 1 output, I get a nice steady reading (around 25) that reacts if I mess with the detector (ie: extinguish and relight hydrogen) and when the oven temp rises I get a rise on the detector as well. It just won't send the data to the computer. As far as I can tell, all the settings are correct...the detector is on and FID lit. Attenuation and range are both set to 0. Signal 1 is assigned to detector A, no det. b installed. Before I discovered the water issue, the detector was maxed out at 830000. Is it possible to have part of the comm board die? Thoughts? Thanks!

Also, we don't have a chiller as part of our air system. We simply have a 5 micron air/water trap right after the compressor and a second one right before the GC, both installed by me. Is this something we should look into? At the moment, the only other instrument running on air is an icp-oes. It has multiple traps before going into the instrument.

The detector maxed out at 830000 was probably due to water shorting the detector. It might be worthwhile to take the detector apart for a good cleaning.

Yeah, your detector signal indicates shorting out, time to take stuff apart and clean, or at least try hotter FID temperature overnight to see if that dries it out.

And you're complaining of WATER in your house air ??? At our old lab, after about 25 years the air compressor went down, and to "save money" over our objections they purchased an oil-lubed replacement, swearing that the maintenance and safeguards would not let any oil into the house air.

Guess what ? That's right, the whole building air supply became contaminated with oil; so we used glass-sided traps with cotton in them to absorb the oil, and to give us a visual determination of effectiveness. We were lucky, too, as our GC air purifier was mounted up high on the wall, and gravity helped oil from getting into that.

We also had a 5890 GC shipped to us from a warehouse at manufacturing that was saturated with oil, someone had hooked it up to an oil-lubed air supply there. There was liquid oil all through the FID and its controls, and the insulation surrounding the column oven was also saturated, we cut away as much as we could, replaced with fiberglass tape. We could take the GC up to 190 C without smell or smoking after that, so we dedicated it to a single lower-temperature analysis.

If you are getting a signal at the front panel then the detector is working. Maybe not working perfectly, but at least working, so don't mess with it any more at this stage.

It is possible for any part of any board to die at any time. HP GCs used to have a test signal generator built in - if your has that try it and see what you get on the Chemstation.

Peter

Ok, so I can't get the instrument to start the test plot with chemstation connected. With the GC running in stand-alone I can. However, when I click to show the second signal (usually off as there is no detector b), I get the baseline trace. I was not (am not) getting any trace at all on signal 1. In addition, I went to check on it this morning and the detector output was maxed again with water sitting in the bottom of the detector. I am wondering if we got a bad tank of either helium or hydrogen. The traps on the air line are dry so I don't think it is coming through there. Is there any way to easily test my tanks? I think I have several problems...=(

At what temperature are you running your FID? If it's not significantly above 100C, you'll get condensation from the hydrogen combustion. Try setting it above 300C.

As for your hydrogen and helium supply- are they ultrahigh purity (UHP)? What traps do you have on those two gases?

FID temp is 250. We use high purity helium, but regular? grade hydrogen. No traps were used on this instrument, but I am working on correcting this.

FID temp is 250. We use high purity helium, but regular? grade hydrogen. No traps were used on this instrument, but I am working on correcting this.

You cannot get liquid water in a detector at 250C - even if there were water drops coming in with the air or hydrogen they would evaporate completely at that temp. Check the detector temperature with an independent thermometer or thermocouple or whatever.

Peter

Where is the detector temp taken? I have a non-contact thermometer and the best I can get is approx 75 degrees. I took off the chimney to get a reading on the inside ( water is still present) and that is the hottest I can get. Am I missing something? Is there an easy way to swap out temp probes? I have a couple of spare fid's I can grab parts from.

Oh, and I was mistaken on the quality of H2, it is 5.0 grade, so high purity.

So I am still chasing the temp/water issue. However, I did some work to see why I was not getting a signal from channel 1. I replaced the comm board not realizing that only signal 2 goes through that board. Anyway, I assigned the detector to signal 2 and am now getting a signal in chemstation. Assuming signal 1 goes through the main board, I am starting to suspect that it is my problem or at least one of them. Hopefully, someone will chime in as to if I am making the correct assumptions before I do the swap. I am wondering if the incorrect detector temp could also be linked to a bad board.

I am wondering if the incorrect detector temp could also be linked to a bad board.

Not your question, but I've got spare 5890 boards and some other available used and reasonable.

On the 5890 front panel, one must set up to turn detector on, which channel, etc., don't overlook that.