Saturated signal in band D
The anomaly was due to the fact that after a LOS calibration the nominal gain tables were not updated, so that the instrument measured with the high gain tables (used only for LOS calibration) in band D. The resulting band D spectra are all saturated in the period 7 - 18 Oct 2006. This can be seen in the figure below showing the ADC Max counts in all MIPAS channels as a function of time, where we can appreciate the start of the anomaly event.
The anomaly was detected during the weekly calibration. A warning was raised to the operational center and the corrected nominal gain was uplinked to the platform on 17 Oct.

All the MIPAS band D spectra recorded during these days cannot be used for the processing.