SOLVE GAIN

Command SOLVE GAIN is used mainly at the observatory. With Walsh $\frac{\pi}{2}$ switching on the LO1 and demodulation in the backend, the sidebands which are superposed after the LO1 conversion can be separated. This allows the receiver sideband rejection, $g_{im}$ , to be measured by computing the ratio of the image sideband over the signal sideband amplitudes. This is done on the uncalibrated data by dividing each sideband by its $T_{sys}$ and correcting for the possible difference in atmospheric opacity between the sidebands.

This command also derives the mean baseband phases that are fed back to the system so that further online averages do not suffer decorrelation.