[evlatests] Super-RFI in P-band

[Rick Perley]

    Another test was run at noon today, and revealed (to me) a new and 
very strong source of RFI. 

    The spectra show an extremely strong signal, seen on all antennas, 
at 339.0 MHz.  This signal is orders of magnitude stronger than the 
noise (in the 250 kHz channel resolution), and was strong enough that 
the requantizer gains were turned down by about a factor of 100.  When 
this signal went away, there was no information left in subband #8. 
    Unfortunately, the power dumped in by this RFI was sufficiently 
strong that all the gains in subbands 1 through  7 are affected, as 
shown in the PSum values.  After the signal went away (it was on for the 
first ~ 30 minutes of the test, then went away and stayed off throughout 
the remaining 45 minutes), the system was stable. 

    The well-known repeater at 444 MHz was also active during this run.  
When on, it's even stronger than the 339 MHz signal.  Fortunately, it 
wasn't on when the gains were being set up, so the subband (#15) can 
still be used. 

    A remarkable curiosity was found in the spectra:  Subband #9 also 
had its requantizer gains turned way down -- but there was no strong RFI 
in that subband.  In looking at the switched power data, I find that the 
PSum values for subband #9 are *identical* to those in subband *8.  Yet 
the cross-power spectra are completely different, proving that the 
frequency ids are indeed correct.   It looks like the switched power 
information for subband #9 has been replaced with that belonging to 
subband #8!  But how can this be?