[evlatests] First SDM1 dataset

[Rick Perley]

    Michael had me look at the first data from  SDM1.  Some good news:

    a) The amplitudes and phases are smoothly continuous in time.  No 
jumps or drops (but see below).

    b) The eight subbands (128 MHz each, spanning 1024 MHz of L-band) 
are perfectly continguous. 

    c) All frequencies are correct (based on known interferors).  The 
half-channel error is gone.

    Then, the not-so-good news:

    a) The header information is correct only for subband #1.  The 
even-numbered subbands have headers indicating a zero frequency, and a 
bandwidth equal to the channel width.  The channel separation value is 
correct.  The odd-numbered subbands are generally incorrect:  subband 3 
contains subband 2's information, subband 5 has subband 3's, and subband 
7 has subband 4's information. 

    b) A single all-zero record is found (in 8 minutes duration).