There really is not a "normal" value as it is going to depend on your method conditions. We do guarantee 1000 hours of lamp use. We also provide the ability for you to enter the date that you changed the lamp and then you can monitor the energy and see how its use changes over time relative to your requirements. As in practice lamps can be used for longer than the recommended lamp hours.
The problem is your Signal to noise requirements will vary. If you run assays vs. impurity profiles clearly you can squeeze more lamp life Here is the short answer. It would be good to see what other users recommend. But if you have a check method that you use all the time I would run that and check the signal to noise and see how it varies as the lamp ages. Then with this first lamp you will know when the S/N gets too bad for your use. Note when that occurs relative to lamp hours. For now go to the Console and the Diagnostics and enter the date when the system was installed as the change date of your lamp and then set a time, say 1000 hours as a lamp life warning limit to remind you to look at S/N performance.
Best practice says to use a signal to noise value limit, using a given method, mobile phase, flow rate and injection peaks as a far simpler and more meaningful test.I would also add tracking the values for auto-exposure for these given set of samples where issues have arisen is very valuable. As the lamp ages the exposure time will lengthen as the lamp energy decreases. You will then know what that value will be on the day your S/N value is too low. Also for a given method and set of samples the auto-exposure determined should not change so it is a great diagnostic. If it changes then something has chnaged.
The use of the read energy diagnostic is not recommended as pass/fail criteria of the detector. This is because the detector has the ability to set the exposure automatically for optimal detection and because each assay is different in it's requirements for acceptable signal.
For example it would be possible to replace a lamp that has abundant energy for most assays prematurely.
It would also be possible just using the read energy diagnostic to misdiagnose a flow cell problem as if it were a lamp issue.
The only way that the read energy diagnostic would be useful would be to record the values from when the detector is new at wavelength 230nm (at this wavelength the deuterium lamp has the most energy) and track the energy value at a 5 msec exposure until the signal becomes unacceptable for your typical separations. At this point the lamp could be changed, but it would also be critical to compare the energy of the new lamp from the flow cell to a shunt flow cell to avoid confusing cell problems to optical degradation (optics bench versus flow cell mechanics).
Though the diagnostic exists it was created to ensure that there was sufficient signal, although counts are reported the diagnostic was not mean to be quantitative, but rather qualitative.
Hope this helps,
Also more comment from a more technical source that I thought would be of interest.
The UPLC PDA is specified to wavelength accuracy of +/- 1nm. Upon each power up (confidence test) we engage the filterwheel and move it to the Erbium filter position where we verify that the 256.7, 378.9 and 486. are all within 1nm. If all are within 1nm then the detector is ready for use. Note: this is a verification of the detector wavelength accuracy. This can be confirmed by injections of known standards such as those in our SQT product and samples for wavelength accuracy. A user can also perform a verification of calibration accuracy by using the Console, as follows:
(1) From the Maintain menu, click Verify calibration to check and verify against the last calibration.
(2) In the Calibration Verification window click Start, the erbium filter is put into position
(3) A progress bar is displayed while the erbium and deuterium lines are located.
NOTE: After one and a half minutes, the results will display to confirm pass or fail. Maximum deviation needs to be less than or equal to 1 nm.
There is no absolute number value for energy as the optical design contributes to the value.The optics design contains these components as part of the light transmission path: a D2, M1, lamp housing window, filterwheel assembly, the Flow cell, M2, Aperture, grating, PDA. All these contribute to the energy value. The Waters UPLC PDA design is such that upon each power up we perform several internal confidence tests. We offer autoexposure that utilizes a combination electronics and lamp optimization software that make the best use of the energy present to optimize instrument performance. This design also compensates for the lamp energy changes that occur across the deuterium spectrum as well as lamp aging. So it is difficult to just give a specific number value.
Thank you for your best detailed explanation !
Now i know ,for lamp energy,tracking the value of S/N and auto-exposure time for a given method is more meaningful