CAD flood (charged aerosol detector)

[lynzjm]

Hi all,

we've had a fault with our Agilent HPLC which has randomly stopped after injection 9 (of 24) in a run. There was no method switch, method looks good and also no obvious fault although we seem to have the system randomly going "red" (we use chemstation software) and we believe it could be a circuit overload with other items on the bench.

The result of this however has meant mobile phase (0.05M ammonium acetate) has been runninmg through the detector for approx 12 hours before it was discovered and the gas had run out (nitrogen). The CAD has flooded and is reading -499.996 (when it should be reading about 2...). We've drained the detector and are leaving gas to run through it but any suggestions would be gratefully received otherwise its going to be a very expensive engineer call....

Cheers.

[lynzjm]

"error 200" now showing as the issue if anyone can shed some light, google seems very unhelpful on the subject!

[lmh]

I can't tell you how to rescue your CAD, but I can tell you how to avoid this happening again, if that helps!

It is very probable that Chemstation did produce some error, particularly if everything is going red. When an error happens, each module in an 1100 or 1200 hplc system will start to run an error method. The error method is stored in the module, and not on the control PC. To prevent the pump from continuing to run when an error happens, what you need to do is:
(1) Using the method and instrument control view, set up a method where the pump is safe (i.e. flow = 0.000).
(2) Click on the pump to get the menu, and go to "control..."
(3) On the pump control page, choose "take current method"

This will send the current pump method (0.000 flow!) to the pump and store it there semi-permanently as the error method. This flow rate will then be selected in the future whenever an error happens anywhere in the system.

Incidentally, you may still be able to trace what went wrong. You can use "View" "Logbook" to look at Chemstation's log, but don't look at "current logbook" even if the fault has only just happened! This will show you an instrument log with little information. You need the sequence logbook, which is a file called something.log in the same location as the sequence file. Use "open logbook" and browse around until you find it. It is generally very useful, and will often explain what brought the sequence to a halt.

[jpwoods]

If running nitrogen gas through the detector for a day or two doesn't bring it back then the only real option is to have the manufacturer fix it (unless you have someone that is trained to do that maintenance). I had a CAD that I sent back to routine maintenance and they said it was the worst case of flooding they had ever seen. I knew I had never left the liquid running when there was no gas flow so I was perplexed. I eventually discovered that the HPLC was leaking a small amount of buffer even though the pump was turned off and after a couple of weeks of sitting there with that small leak the CAD had become flooded even though the pump was not on.