auto exposure time is too high - PDA warning

shtm100

<p>During an isocratic analysis using a PDA detector in 2D mode (215nm, Resolution = 3.6nm, Exposure Time = Auto), I recieved a warning that "The auto exposure time is too high". The corresponding chromatograms displayed increased area counts (6 to 25%) and/or a vertical baseline shift. Resetting the PDA detector through the Acquity Console and cycling the power on the detector did not correct the problem as the warnings continued. I converted the method to 3D mode (210 - 220nm, Resolution = 1.2nm, Exposure Time = Auto) and successfully ran my samples with no further warnings or errors. </p><p></p><p>Have others had issue with the auto exposure time (and how to reset it) or collecting 2D chromatograms with a PDA detector? Any thoughts or suggestions will be greatly appreciated. </p>

lizh

Hello,

We have discussed this at Waters and the warning message that was issued occurs where the A/D count of the brightest pixel exceeds a threshold that was felt to be safely under a level where noise problems had been known to occur, but there are no noise issues and frankly the chromatogram doesn't seem to match the symptom typically associated with this warning.

All in all we have decided to contact you directly as likely this may take some troubleshooting to get to the bottom of.

Expect a contact shortly!

We should plan to post the result of the investigation though!

Liz

BJ

Hi,

in our lab, we have multiple UPLC's all with this problem: sometimes you get the auto exposure time too high.

Each time they inject a sample, they give a bleep and then this warning is added in the logs.

Sometimes they get this warning, sometimes not.

Never found the root of the problem.

BJ

shtm100

When we run our PDAs in 2D mode, we've run into the problem described above - or a different problem where we lose small peaks and the baseline is perfectly flat. We don't have these problems in 3D mode, and now run accordingly.

lizh

Thanks for all these posts, so 3D is an excellent short term fix. I would like to look at all these cases indicidually and see if there are all the same issue and requiring the same fix.

May I ask if this has been discussed with your Field Service Engineers?

I will post further shortly.

Liz

lizh

All:

I think that if you are seeing the message, but not the shifts, and noise perfromance is fine, then Waters needs to consider removing the warning message as it is causing more concern than reassurance. WHat do you think?

Liz

shtm100

Hi Liz,

The "baseline" issue has been discussed locally. One concern was the detector wasn't zeroing properly and the suggestion was to cycle the detector once the system was equilibrated on mobile phase. This fix did not work. If I remember correctly, another possibility was the reference energy was being taken too close to the sample wavelength so we were seeing only the tops of very large peaks and losing the baseline and small peaks, as if there is a "baseline off-set" too high. In 2D, the suggestion was to collect a second wavelength away from the sample wavelength to avoid the reference being higher than the sample at zero. We have not evaluated the "second wavelength" fix. Internally, the problem had been overcome by 1) increasing bandwidth or 2) increasing sampling rate. This issue has been observed on our Acquity LCs and 2996 PDAs on our 2695 Alliance systems. There are no errors/warnings posted in the message center for this problem, but as before, this is not a problem with 3D collection. An abbreviated example is attached.

I will attempt to pull together examples of the baseline and auto exposure issues as quickly as I'm able. FYI, we're not yet on Empower 2 - I believe we run Empower Pro with service pack M - if that has any bearing.

Thank you for your attention,

Michael

shtm100

Our lab has 1) seen the warning, no shift and no apparant impact. 2) Seen the warning, no shift and an increase (5-10%) in peak area and 3) seen the warning and the shift with a 25% increase in peak area. I will seek examples and pass them along.