[evlatests] Using S-band is Tough!

[Rick Perley]

    I've finally gotten around to trying to properly calibrate the 
S-band data from the Hercules A 'demo' run, from C-configuration -- 
taken last December. 

    It ain't easy ...

    The major problem is RFI, the majority of which is from 
geostationary satellites.  Because Herc A is near zero declination, it 
essentially travels right along the equatorial belt, passing by these 
satellites in turn.  Six of the 16 subbands are essentially unusable due 
to this -- subbands 2 through 5, and 13 through 16.   The first three 
are taken out by XM radio, plus the emitter at 2190 MHz (whose origin I 
forget -- but it most clearly is a geostationary satellite), the latter 
three are taken out by TV downlinks. 

    The emission is so strong in subbands 2 through 5 that it causes the 
accumulators to overflow, which then completely trashes the entire 
subband.  It would be useful, when we can find the time, to utilize the 
flag which is aware of accumulator overflow, so these ruined spectra 
never make it to the user. 

    An interesting question is whether we should be changing the 
requantizer gains to accommodate the extra power in these subbands.  I 
think the answer is 'no', but we should probably discuss the issue at 
some point. 

    On the brighter side, the calibrator observations on 3C286 and 3C48 
were not seriously affected by any of these emissions -- the antenna's 
forward gain is sufficient to suppress these transmission.  So observing 
in the north will generally work.  The specific frequencies allocated to 
the emission in subbands 2 through 4 may be unuseable, but the adjacent 
channels will be fine.  What we don't know is how close to the satellite 
we can point, and still get useable data.