The size of the cal deflection is not a function of rotation angle.
It was hypothesized that the
size of the cal deflection might depend on the rotation angle of the alfa
receiver (2004 Dec20 Holistic
estimates of Alfa cal values... Heiles, Action item 3,pg 22).The idea
was that some of the cal was going out a horn, bouncing around, and then
coming back in a different (or the same) horn.
To test this, data was taken on 01jun05 using the
wapps in pulsar mode.
The setup was:
The data acquisition consisted of moving from rotation
angle -90 to +90 in 100 seconds, then +90 back to -90. This was repeated
to give 4 passes through the -90 to +90 range.
The azimuth was at 270 degrees and the dome was at 8.834 degrees.
A 100 Mhz bandwidth was taken with 256 channels centered at 1420 Mhz. The
spectra were sampled at a 1 millisecond rate.
The 25 hz cal was used to drive the cal diode continuously. The cal was
on for 20 milliseconds and then off for 20 milliseconds. It was synchronized
with the dumps via the 1 second tick.
The data analysis was:
The plots show the normalized
caldeflection versus rotation angle (.ps) (.pdf):
Compute the total power by averaging the spectra for each millisecond.
Split the data into calons and cal offs. For each 20 milliseconds of a
state ,compute the median value throwing out the first and last millisecond
of each state (to allow for cal transitions).
For each pass compute the calDeflection (calOn-calOff). Normalize this
to the median caldeflection for a pass (of 100 seconds).
Fig 1: The fractional change in cal versus rotation angle is plotted.
Black is polA, red is polB. The 4 passes have been averaged together. The
positive and negative spikes are the faa radar at 1350 Mhz. Each frame
is a separate pixel
Fig 2: The data from Fig 1 was smoothed by 25 (using a robust mean that
threw out outliers). This gave a sample time of once a second (100 points
for the 180 degrees of rotation). There is no structure down to .3% of
The cal deflection was measured as a function of feed
rotation angle. The cal deflection is not a function of rotation angle
down to .3% of the cal value.