Pdev time domain mode
28aug07
Time domain dumping of data was added to the pdev
spectrometer in aug07. To test this, data was taken with the if/lo
noise source going through the temporary mixers. The setup was:
- 170 MHz sampling clock
- low pass filter with a hanning window decimating by 170 (giving 1
MHz bw).
- upshift (hr_shift) by 13. The rms at the digitizers was about 35.
The rms of the output samples was output the same.
- dump 8 bit data, 128k*4 +8 bytes per dump (iq lowband, iq
hiband). dump 1000 of the 128K datasets.
- The data rate was 170e6/170*4*2 = 8Mb/sec.
- transform the time data using 128K transform. Average 1000 of the
spectra.
- The resolution is :170e6/170/128K=7.6 hz.
The plots show spectra from the time domain data.
(.pdf):
- Page1: Full 1 MHz span:
- Black is polA, red is polB.
- The top plot is the low 170MHz, the bottom plot is the upper
170 MHz.
- The drop off around DC is caused by the amps after the mixers.
The regular mixers will not have this. The upper band has a birdie at
+/- 78.5 KHz.
- Page 2: +/- 1 KHz about Dc
- The bottom plot shows lots of 60 Hz. The top plot is relatively
clean.
There are two separate mixer chassis (that are not
identical) being used: one for the lower band and one for the
upper band. Both mixer chassis are being power by an external dc power
supply. The upper band (lots of dc) mixer has an extra negative
voltage being provided to it (without the corresponding common).
processing: x101/070828/pdev.pro
<-
page
up
home_~phil