Agilent 7890B with Empower 2

[bor9455]

Hello all. I'm hoping that I can get someone who knows how to get my 7890B to play nice with Empower. I'm an analyst with a pharma company and we just recently purchased and qualified some new 7890s, one A and one B. The 7890A is working like a champion, and I was the first to run on the 7890B today, or so I thought.

We have recently had a Waters rep out and he installed the ICOPS for the 7890B, and was able to get the two systems talking, which is a nightmare. I can run a monitor of the test plot, or back det output, etc and have no problem at all. The second that I tell Empower that there is a column in this machine, it starts throwing Instrument failures and I can't even monitor the system, nor can I perform a test injection. I tried everything, went to the GC and wrote down all the parameters in about every menu I could find, and then went to Empower and made sure that everything matched up. I even tried setting column 1 and column 2 to the same thing and did the same in Empower, and to no avail. My Empower install is super great in that I can't see what the hang up between the two talking is. I was able to set up 3 different comm tests and get the test plot, atmospheric pressure and the back det output without a column, which is an FID without a flame, or just a flat line of zero. I don't know what could be the hangup here. The column I have was on the back inlet and back detector, but I'm planning to use a headspace, which we run manually without Empower controlling. So I did set the injector to none on the main screen of Empower. I'm thoroughly confused as to why I continued to get system failures for hours on end. Even my Empower administrator who knows very little about the GCs didn't have any idea.

I know I described a lot there, but if someone has any ideas about how to get these two systems talking, I would appreciate it.

[DR]

nothing above and no more detail than "Instrument Failure"?

[bor9455]

nothing above and no more detail than "Instrument Failure"?

When I hit "setup" with my comm tests it says "System Idle" and then if I hit "monitor" I will get whatever I set in channels 1-3 for the output. I set up a basic instrument method with the out put as test plot, atmospheric pressure and back det output, and got all three channels when I did a single injection and when I hit "monitor". The second I went back into the instrument method and added in Column 1 on the back inlet and back detector it didn't work. I would hit setup and get a system failure. I couldn't monitor, and nor could I perform a single injection. I had assumed that it was an issue between the setup on the instrument and the Instrument Method in Empower, which has happened before with our 6890s and usually it is something stupid like Flame Enable not being checked or something silly, but this one has thrown a number of us for a loop. The GC is giving Empower data so they are talking, but why is it once I add in a column, which kicks open the back inlet and back detector tabs do I suddenly not get it to work.

[DR]

Is there a "ready" cable involved? If not, this could be the problem. Empower may be willing to monitor regardless of current conditions, but won't work if the GC is waiting for a "ready" signal that is not forthcoming.

[dlbenach]

Make sure the column configuration in your method matches the column configuration entered in the front panel of the GC. Check both Column 1 and Column 2 on the GC to make sure there are no conflicts.

We have had problems when the column configuration in the GC has not matched the column configuration in Empower. For instance, if the last person to use the GC configured from the GC control panel that Column 2 is connected to the front inlet and back detector. Then the next analyst downloaded a method via Empower with Column 1 configured for the front inlet and front detector. You would think that the Empower settings would overide the GC settings but they don't. It "confuses" the GC. I'm not sure we ever got an instrument failure but it does cause the column flow to do strange things.

Did Empower message center give you any info?

[bor9455]

Make sure the column configuration in your method matches the column configuration entered in the front panel of the GC. Check both Column 1 and Column 2 on the GC to make sure there are no conflicts.

We have had problems when the column configuration in the GC has not matched the column configuration in Empower. For instance, if the last person to use the GC configured from the GC control panel that Column 2 is connected to the front inlet and back detector. Then the next analyst downloaded a method via Empower with Column 1 configured for the front inlet and front detector. You would think that the Empower settings would overide the GC settings but they don't. It "confuses" the GC. I'm not sure we ever got an instrument failure but it does cause the column flow to do strange things.

Did Empower message center give you any info?

I will have to take a look at this on Monday when I get back into the lab. Our Empower message center is just as useful as a stopped up toilet. I talked to our administrator and he didn't have any idea why I would be getting system failures, especially because we set up Column 1 and Column 2 to be the exact same in both Empower and on the front panel. What is really odd is that we have a new 7890A that plays very nice with Empower, and the new headspace is a rockstar when it comes to true <467>.

[mrussell]

Direct control of agilent GCs via empower can be problematic. Pretty much all of the problems we have had have been due to the autosampler. We have a bunch of 7890s. Our older 6890s are controlled with sat/in boxes,which are terrible,as it digitizes at rather low resolution and you cant get the full dynamic range of the instrument. The direct control gets high quality data,but as I mentioned,the autosamples are sqirrley.

First,we have to disable "Inject Ahead". That simply crashes the instrument.It used to take the entire LAC/E and all four instruments down. Now at least you can (usually) just restart the run and it goes. But sometimes we still lose the entire LAC/E. Additionally,we cant get the sample depth setting to work.

It sounds like you want to run the new instrument in a different configuration than the old. I would suggest first though,that you get it running in the identical configuration to the other one. If you didnt get a autosampler and the other one has one,then take its autosampler. At that point,you can be sure the problem is the difference in configuration.

Waters tells us that some of the problems we have like inject ahead failing and the very long delay (3 or 4 minutes) between injections after the instrument comes ready are agilent problems and even chem stations have problems like that. Thats not true. We have a chem station on a 7890 there is no significant delay and the inject ahead feature works perfectly.

One thing Ive noticed is that when controlled by Empower,the injector is set for "Large Volume Injection". It should not be. Thats telling me that they dont have the instrument configured right. They claim that agilent provides them with everything they need to make it work. My guess is,Agilent provide them with the basics,if even that. My guess is that they figured out how to read and write the configuration of the instrument and then have the driver send that information. I suspect they may be telling the truth when they say chemstations have that problem. But probably only THEIR chemstation. So they copy the bad config and it does the same thing.

Im working on getting IT to let me isolate one of the LAC/E boxes and a GC and putting a protocol analyzer on it so I can watch whats happening and see if can figure out the difference in how the LAC/E configures the instrument vs how a chemstation does it.