The four-band tests (L, S, C, and X) done last week, tracking 3C147
from an elevation of 49 through to 19 degrees, has clarified some of the
characteristics of the switched power monitoring.
A) I've reported before about the unique triangular response of
antenna 18 -- PDif changes in time, with ~1 hour period, with a
triangular function. The amplitude is especially large in IF 'B'. No
such dependency is seen in PSum.
The new test shows that this occurs *at C-band only*.
B) Also reported earlier is a square-wave pattern in ea12, IF 'B'
only. The characteristics are: PSum and PDif both show a square-wave
amplitude dependence, with PSum varying by ~5%, and PDif by ~2%, on a
typically 15 second period. The period is not constant, but changes
steadily over the observation. The alternations are in anti-phase:
When PSum rises, PDif decreases.
The new test shows this occurs *at all bands*.
C) Also reported earlier was the square-wave pattern in ea25, IF 'B'
only. This has exactly the same characteristics as 12B, except that the
period is much longer -- at least a few minutes, with the period
shortening during the observation. Antenna 25 is at N02, and antenna 12
is at E08, suggesting a relationship with fringe rate?
D) Some antennas, at some bands, are showing PDif as noise but
normal PSum, suggesting that either the noise diode is not switching, or
that the demodulation is out of sync:
Antenna 9 at L-band
Antenna 12 at S-band
E) For nearly all antennas, the values of PDif is between 2 and 8
'counts'. The following list shows those which are significantly
different (not including the two listed in 'D' above):
Antenna 16 at X-band: much less than 1 'count'. This is
almost certainly related to a high system temperature for this antenna
at this band (hundreds of K).
Antenna 13 at L-band, in IFs 'B' and 'D': much less than 1
'count'. PSum is also low, indicating a wimpy Pcal. Tsys appears close
to normal.
Antenna 15 at L-band, in all four IFs. Again, much less
than 1 'count'. Tsys for this antenna is very high (over 100K).