I've been working on a HP 5890 GC with and FID detector. I have little idea what maintenance has been done on it in the past five or more years. After replacing several pieces of the instrument, I was finally getting it to work decently until all of a sudden the signal shot up to 830000 and wouldn't come down even after I fiddled with the spring interconnect assembly. Finally, was advised by Agilent to buy a new detector assembly and interconnect assembly. Upon installation, the FID worked well for a couple of days, but the signal has again gone up to 830,000 with nothing running and I am at a complete loss. I assume there is an electronic problem, but have no clue how to proceed. Any suggestions?
By anon on Wednesday, July 28, 2004 - 06:44 am:
I think there is some moisture collecting in the spring ...clean the detector assembly.
By GC user on Wednesday, July 28, 2004 - 09:24 am:
I have seen moisture do this, especially in older 5890s. There is something shorting out in the detector, which is the cause of the 830000. Agilent should refund the cost of the new assembly, as this did not solve the problem.
By Anonymous on Wednesday, July 28, 2004 - 10:05 am:
Is the signal 830000 even with the flame off? If so, somethings shorting electrically. If it's only with the flame on, maybe tons of signal is coming through the column (lower the column temperature to 25C to see if signal goes down). We just received a 5890 from QC which literally had 25 ml of oil in the FID and the lines going to it, and it readily pegged out at 830000 until we repalced the detector plumbing and cleaned everything out.
By ced on Thursday, July 29, 2004 - 01:56 am:
Did you try another electrometer.
By WK on Thursday, July 29, 2004 - 06:58 am:
Do you know how 25mL of oil got into the FID and gas lines from QC? Sounds jolly dangerous to me.
Did Agilent advise on changing the assembly?
In which case they might refund - but I think its useful to have a spare anyway.
(I had to replace a PE FID assembly once too - never could really pinpoint the exact part within it causing the noise).
By Anonymous on Thursday, July 29, 2004 - 07:13 am:
Apparently QC hooked up the air supply to an oil-lubricated compressor, and didn't use any oil separator or filters (great money-saving idea !!!). However, equally dumb, our R&D lab's compressor died five years ago and an oil-lubed compressor with an oil separator was the replacement (to save money) but no one in the labs was consulted, or even knew the compressor was dying or was being replaced. Anyway, Maintenance was too understaffed or forgot to drain the oil out of the separator, and now our entire R&D center's air lines are contaminated with oil. We had to buy our own oil-less compressor, after the oil had killed our nitrogen generator, dry air supply, etc.; another great cost savings, and EPA and cost prevent cleaning out those lines. I'd bet replacing all those and getting a large oil-free compressor would cost several hundred $K.
By WK on Thursday, July 29, 2004 - 07:35 am:
I think a lot of GC operators will be on their way to their maintenance depts as we speak (or type!).
Thanks for sharing that tale of woe.
By elizabeth18 on Thursday, July 29, 2004 - 09:09 am:
So, I consulted Agilent again, and after some fiddling with the machine we have pretty much determined that the FID electronic board is bad. I asked the engineer if it was possible to fix the existing board or get a refurbished one, and he said no, that I'd have to buy a new one. So, I though I'd pass the same question to all of you GC people....do you know of a way to get a board fixed, or should I just spend the $650 and get a new one.
By GC user on Thursday, July 29, 2004 - 09:57 am:
An electrometer is a very delicate piece of equipment, and the chances of it working well after repair are not great, plus at instrument repair labor rates it wouldn't be cheap. Replacement is the simplest solution.
By Consumer Products Guy on Thursday, July 29, 2004 - 10:27 am:
I'd substitute a known working board in first to make sure that's the problem; I assume you mean the circuit board accessible after taking off the right cover (viewed from the front) which has the "rod" connecting to the FID.
By elizabeth18 on Thursday, July 29, 2004 - 10:58 am:
Considering that the Agilent guy said electronic board damage is a very rare occurence, is there any other component that would cause this board to short out that we should think about before we stick a new one in the machine?
By GC user on Saturday, July 31, 2004 - 09:35 am:
There is actually a test procedure to verify the proper functioning of the electrometer. I can't remember the details (almost 10 years ago), but I performed the test, verified the electrometer was bad, then replaced it and the problem went away. Push Agilent to either test the board or tell you how to do it.
By elizabeth18 on Thursday, August 5, 2004 - 01:58 pm:
Just to let ya'll know, I switched out the electrometer board and that fixed the problem. Thanks for all your input.
By Consumer Products Guy on Thursday, August 5, 2004 - 04:05 pm:
Thanks for posting your fix, that helps all of us learn.
By ced on Friday, August 6, 2004 - 12:53 am:
Thanks for the feedback