Wavelength inaccuracy

Using Sciex Analyst software, we get incorrect wavelengths for peak maxima of caffeine (e.g. 238 nm instead of 273 nm). When we used chemstation, the maxima were fine.

From this I assume that the detector is functioning correctly, but there is either a problem with the software or the connectors.  As I can't find a technician with knowledge of both Agilent hardware and Sciex software we have not been able to resolve this for years now.

Any ideas would be gratefully accepted.