* “Phase” calibration

Load the default frequency configuration (16 MHz bw, freqs 597, 682, 853 and 938), and connect a tone generator to the inputs of the CoMos. Set up the CoMos to use filter 2 (e.g dbbcifa=1,agc,2). Set the tone to 784 MHz and ~-15 dBm. Check IF power levels are ok for all CoMos and then start the calibration with the DBBC command “calibration”. This will generate a file in C:\ with a name of yyyy-dddThhmmss-calib.txt which lists the measured power in the Cores assocated with each CoMo. You want to pick the values for each that correspond to minima as the tone should be outside of any of the selected frequency channels. These values should be added to the DBBC configuration file (the line above the CAT2 1024 entry).

* CoMo calibration

To get the optimal sensitivity out of the CoMos, you need to use the optimal power level. This can vary significantly between CoMos. This is best down using a phase cal signal signal, or another narrow band one in a input IF. Record a short scan of data in a 2-bit mode (e,g, in DIMino mode=ext,0xffffffff,2) and then use the Haystack utilies script bpcal to extract the amplitudes of the pcal signal. Repeat for consistency, and then repeat the measurements for different CoMo target levels. (settable via dbbcifa=1,agc,2,40000). You want to select the target power level that gives the highest measured amplitude for the phase cal. NB - since the measured power levels are per IF, you should average the amplitudes of those IFs which are all connected to the same CoMo.

* Autocorrelation/Bit statistics

For autocorrelation spectra and bit-distribution statistics, you can use either the Haystack vlbi2 and bstate scripts or the mark5access utilities m5spec and m5d. If the bit-statistics deviate significantly from 17/33/33/17, you may need to use the DBBC's mag_thr command to adjust the magnitude threshold. If there is a strong DC bias, this is usually indicative of a configuration problem.