On 29jun07, Dana hooked up 1 channel of his new
board to look at the bandpass. The setup was:
The plots show the
bandpasses for the 3 IF center frequencies (.pdf):
- The noise source in the IF/LO was used as the signal. Only polA
was hooked up.
- The pdev input was set to 1 sigma=50 counts on the rms voltage
- Data was sent thru the mixer Low band (using the 250Mhz low pass
filter) into the box103 polA lowband inputs.
- The bandpass was integrated for 30 seconds using 8192 channels
and 170 Mhz bandwidth.
- The bandpass was centered at 3 IF center freqeuncies: 160, 175,
and 200 Mhz.
- The mixer uses a 250 Mhz lowpass filter for the lowband that was
tested. It should be down about 1.2 db at 250 Mhz.
- The I,Q baseband signals have a 75 Mhz anti aliasing filter.
- The first page plots the spectra vs the bandwidth. The 3 colors
are the different IF center frequencies. The dashed blue line is 75 Mhz
from the center of the band where the anti aliasing filter starts to
fall off (at least on the high side).
- Black: IF cfr 160Mhz. The bandshape is different than the other
- Red: IF cfr = 175 Mhz. This is the design center frequency for
- Green: IF cfr=200 Mhz. This was taken to see how fast the IF
filter is falling off.
- 2nd page: The spectra are plotted vs their IF frequency. The
dashed lines are the edges of the 170 Mhz sampled bands.
- The 175 Mhz band is down 10db at an IF of 250 Mhz. It was only
supposed to be down by 1 - 2 db.
- The 200 Mhz band is down 2 db at 250 Mhz. This is the expected
value for the 250 Mhz lowpass filter.
- The 160 Mhz bandshape is different than the other two. I would
have expected it to be more similar.
- The pdev testing of the
boxes also showed a bump of around 60 Mhz about the center of the
band for all the boxes for polA low band. These measurements used a
different mixer setup and the center of the band was at 190 Mhz.