Thinking about this possibly being a mass calibration problem, can you
1) name the MS instrument you are using,
2) are you the sole user, or is this a walk-up, open access system (OA)?
3) if it is OA, does the first user of the day run a mass calibration?
4) if it is OA, does every user know that you cannot retune or tweak lens voltages etc once the system has been calibrated, without having to again calibrate?
When I was analyzing samples by ESI-MS under GLP/GMP protocols, first job of the day was to tune with CsI for peak shape and sensitivity; scan m/z 125-1250, in say 3 sec and acquire a calibration file of say 5-10 scans; then repeat that acquisition and save as the calibration check file; run samples; at the end of the day, run another calibration check file to ensure that the calibration had not drifted over the work day.
If a sample came in that required a different mass range, say m/z 500-2500, I would save it for the end of the day, do every other low mass sample as described above, and then re-tune, take a new Cali file (m/z 50-2500, 3 sec), repeat this to give the Cali check file; then analyze the sample and follow with another Cali check file.