[evlatests] P band power levels

[Vivek Dhawan]

with plots!

Prompted by Frank I took a look at P band power levels. In summary,
the astronomy data are really quite nice, but some power levels might
be be worth cleaning up.

Thing 1:

T301 TOTAL_POWER_AVG is as close to the 4/P front end as I can get.
This is after the splitter to VLITE, which should see about the same.
The levels at that monitor point are roughly -50dBm +_5dB, on all
antennas, 'R' and 'L'. So that is pretty good I think.

However, some data appear to be jumpier than others, e.g. ea01 R,L
are more variable than ea02 R,L. ea01, 4L, 5,8,18,20,28 are more
variable than the rest.  I think this is not interference, which
is seen on all antennas, including the 'calm' ones, but the jumps
seem to have a different character. It may be flaky TP detectors,
and not real.

One day's worth of monitor data are plotted here for 4 antennas.
Note the bursts of RFI at around 9 and 11 UT. The Y-axis is 40dB
range for each antenna, centered at -55dBm.

Thing 2:

T302 TOTAL_POWER_AVG is monitored at the input to the LSC converter.
Looking at a day of this is also instructive.  As the antennas get
setup to various bands during the day (all bands were used this day).
The power level into LSC is normally around -55dBm at this monitor
point. (ea02 is always at -40dBm regardless of which band is in use.)

Notably, when 4/P is in use, the power into LSC goes up by 10 to 25dB,
on all antennas, (except ea02 as above). So it looks to me like we are
losing 10-20dB of headroom on the LSC converter when we use 4/P. Is
this true, and is it a design feature or an accident?

Also, there was bad/no data from that monitor for ea18R,L and ea21L.

Again one day's data are in plot 2, yesterday. Same axes. Note the
jump of +10 to +20dB at UT 22:30, near the end,  when I used P band
for 40 min.  (The other jumps are from P band at the start, plus
people working on antennas - ignore except to get the nominal level.)

Finally, I note that by the time the P band data get into the T304s,
they are all nicely levelled, and the fringe amplitudes (signal-to-
noise) on a point source in an interference-free band at 328 MHz, are
all within 10% of each other across the array.


-------------- next part --------------
A non-text attachment was scrubbed...
Name: t301p.57961.png
Type: image/png
Size: 74003 bytes
Desc: not available
URL: <http://listmgr.nrao.edu/pipermail/evlatests/attachments/20170728/5f72b640/attachment-0002.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: t302p.57961.png
Type: image/png
Size: 67264 bytes
Desc: not available
URL: <http://listmgr.nrao.edu/pipermail/evlatests/attachments/20170728/5f72b640/attachment-0003.png>