lbw 1230-1470 filter (filter number 3)
The lbw cal measurements
have shown a
jump in calDefl/Tsys for polB when the 1230 to 1470 Mhz filter is installed.
To test this the hcorcal was fired using the 1 second toggle for about
2400 seconds with the 1230-1470 filter in (1 second cal on , 1 sec cal
off). This was followed by 1200 seconds of calon,off with the 1230-1470
Mhz filter out. The telescope was sitting at az=280, za=10 for these measurements.
The lst was around a hours.
The data reduction consisted of:
compute Tsys=TcalOff/(TcalOn-Tcaloff) * CalvalueInKelvins.
use corblauto to fit a linear plus second order harmonics to each bandpass
throwing out outliers.
Use the constant term of the Fit for Tsys for each of the 4 25 Mhz bandpass.
These were centered at 1350,1375, 1400, 1425 Mhz.
The plots show Tsys
versus time (.ps) (.pdf):
A Tsys measurement was made every 2 seconds (1 sec calon,
1 sec caloff). The positive going spikes are sources drifting through
the beam. The blue 1425 Mhz Tsys is about 7% less then than the 1350,1375,
and 1400 Mhz Tsys. This may be a problem in the calvalues in use
(if so, the error in in both polA and polB). The transition from filter
to no filter at 14.14 hours is smooth for polA. For polB the 1425 Mhz (blue)
tsys jumps down. This is not the 1230-1470 filter since it would affect
all 4 of the bands.
Top: Tsys polA versus time. The 25 Mhz bands are plotted in different
colors: (black 1350, red 1375, green 1400, and blue 1425 Mhz).
The vertical dashed line is where the 1230 to 1470 Mhz filter was switched
bottom: Tsys polB versus time.
If the problem is the filter bank relay not closing completely a better
test would be to take 20 seconds of filter in, 20 seconds of filter off
, and then repeat this for a few tens of cycles.