[evlatests] P band

[Barry Clark]

It looks to me that we are not turning down the requantizer gain enough
for P band.  Our standard setup has a gain of 30 for the high bands, 
where we have much of a GHz of bandwidth into the digitizer.  At P band, 
effectively all of the power from the digitizer comes out of the 
deformatter, whereas at the high bands, the FIR filter in the deformatter 
removes about 1000 MHz/ 64 MHz of the power.  Therefore, the gain of the 
requantizer should be turned down by a factor of 4, to about 8.  It is 
currently set to 15.

4 Band and P band are in the same digitizer band.  We are attempting 
to set them to approximately the same spectral power density (or maybe
a little higher for 4 Band).  So the power coming from the FIR filter
in the two cases, P and 4 bands, is roughtly proportional to the 
respective bandwidths, 40 MHz for P and 1.7 MHz for 4.  However, the
same argument I just used above also applies to the analog filters in
the T5 module - they will turn the gain up by about the same factor, 
so the requantizer gain can remain the same at 4 band as it is at P
band.  It is currently set to 5.

The A & C IFs are leveled according to the detector readings.  This should
produce the same RMS on the link as we see at the high bands.  The same
should also govern the B & D setups, which are set by hand, at P band, 
to set to the same RMS.

I shall take some time tomorrow and see if turning down the Pband requantizer
gain to 8 or 10 helps in practice (too late to try 4 band - it's gone).