[evlatests] Internally-generated RFI!

[Rick Perley]

    A recent observer has reported a remarkable comb of RFI in her data, 
taken Sunday afternoon. Lynn Matthews reports a comb, with spacing 39.7 
kHz, is on many baselines, but always on antennas 6 and 22 (located at 
the end of the west arm).  This immediately raises the suspicion of some 
radiator near the barn, LWA, or technical services area, were it not for 
an extra, curious fact -- the RFI comb can be made to disappear by 
changing the frequency by 1 MHz -- either up or down. 

Being a disbelieving sort of guy, I used a 30 minute segment of software 
time to repeat a portion of her file. 

    Everything reported by is present, precise!  Specifically:

    BW - 781 kHz
    Correlator mode = 2AC
    NChan = 256

    With the center frequency set to 1421.19067 MHz, a remarkable comb 
appears all across the spectrum on baseline 6-22, in both polarizations. 
    With the center frequency set 1 MHz higher, or lower, and on the 
same sources, the comb disappears!!!
    The comb spacing is 40 kHz, nearly exactly. 

    The comb is not visible on spacing involving antennas at the end of 
the north or east arms.  Thus, it must originate from one of our 
antennas (either VLA or EVLA) near the end of the west arm. 

    Suggestions?  We need to find out what is going on here.