[evlatests] Abnormal reported Tsys values

[Rick Perley]

    The system temperature is calculated from the following (which 
presumes a linear response with zero offset):

    Tsys = Tcal * Poff/(Pon-Poff)

    where Pon and Poff are the powers, measured in 'counts', as detected 
in the synchronous demodulation on the station boards. 
    Note that the measured ratio is multiplied by Tcal, which is 
determined in the lab, and is tabulated in some database for the 
purposes of this calculation.  If the reported 'Tsys' is abnormal, there 
can be (at least) two origins:  either it really is abnormal, or the 
tabulated value of Tcal is incorrect. 

    The following antennas show abnormal Tsys values, taken from the 
test of last week.  In many cases, I can discriminate between the two 
postulated origins, as the known variation with elevation (due primarily 
to spillover and atmospheric emission) can be used to independently (but 
roughly) calibrate the temperature scale: 

    Antenna 7 at L-band.  Tsys ~ 70K, which I am sure is real. 
    Antenna 15 at L-band:  Tsys ~ 150 K, which is certainly real. 
    Antenna 24 at L-band:  Tsys ~ 5 K, which is certainly *not* real!  
This antenna is known to have an unusually (and deliberately!) high 
Pcal, so it seems clear that the problem is the Tcal listing is about a 
factor of 5 too low. 
    Antenna 28 at L-band:  Tsys ~ 20K, which seems too low to be 
believed ...
    Antenna 8 at S-band, in A/C only:  Tsys ~ 10K.  This must be a Tcal 
listing error.
    Antenna 9 at S-band, in AC only:  Tsys ~ 110 in A, and 66K in C -- 
clearly a Tcal error.
    Antenna 10 in S-band, in AC only:  Tsys ~ 50K.  Again, a Tcal error.
    Antenna 15 in S-band, in AC only:  Tsys ~ 90K, again a Tcal error.
    Antenna 24 at S-band, in AC only:  Tsys ~ 10K.  ditto
    Antenna 27 at S-band, in AC ony:  Tsys ~ 15K, ditto. 
    Antenna 16 at C-band in BD ony:  Tsys ~ 14K, clearly at Tcal error. 
    Antenna 2 at X-band in all IFs:  Tsys ~ 13K (AC) and ~20K (BD) -- 
probably Tcal error.
    Antenna 9 at  X-band in all IFs:  Tsys ~ 20K, which is likely a bit 
lower than real. 
    Antenna 16 at X-band in all IFs:  Tsys is very high:  180 in A, 650 
in C, 130 in B, and 220 in D.  These are correct, as the elevation 
effect is completely drowned out in the receiver noise. 
    Antenna 22 at X-band, all IFs:  Same situation as antenna 16, but 
not as severe:  Tsys ~ 70 in AC, and 130 in B/D.
   

    The rise in Tsys with decreasing elevation is easily seen at all 
four bands observed in this test.  For reference, I describe each, using 
antennas whose Tsys values appear correct:

    L-band:  Rise is nearly linear, adding about 6K between 49 and 19 
degrees elevation.

    S-band:  Rise is nearly linear down to ~25 degrees (adding about 4K) 
, then sharply rising below that:  5 more K by elevation 18!  (This is 
an expected effect, due to the location of the S-band feed). 

    C-band:  Linear rise down to ~30 degrees (adding about 1.5K), then 
increasing rate, adding about 3 more down to 18 degrees. 

    X-band:  Similar to C-band.