Intro

    SiriusXm digital radio has satellites that transmit  2320 to 2345 MHz (more info). On 11nov22 voltage sampling was tested while sampling 5MHz about 2322 MHz. This included the siriusXm band.

---

Setup

• Telescope sitting at az=137.8, el=40.5 deg
• 1 mock spectrometer was used
• clock: 160MHz
• bw divisor 32 ( for the digital filter to generate a 5MHz output
• 8bit data storted (i,q)
• rms was set to 30 counts (should have been set to 20)
  
• rf rfcfr 2322 MHz.
• 10 seconds of data was taken: 1 byte*2iq*2pol*5mhz=20MBytes/sec

---

Looking at the data

    The plots show the complex voltage data and the computed spectral density (.ps)  (.pdf)

• Page 1: rms and mean values for each rec
• 64k samples were processed per rec (.0131 seconds of data)
• The mock rms was set to 30 cnts=1sigma before starting the scan
• should have used 20 cnts/sigma for 8bit sampling
• Top 2 frames: voltage Rms vs time for polA and polB
• black=idig, red=q dig
• the rms was computed every 64k samples.
• Pola hyas rms's around 49 counts
• Initially the rms was set to 30counts, but it probably did not use enough data to see the stronger rfi.
  
• polB has rms's around 30 counts
• There was very little rfi in polb
• Bottom 2 frames: mean voltage value
• Page 2: histograms of the voltages.
• voltage histograms were made using the 10 seconds of data.
• The y axis has been normalized to the average number of counts  per a/d bin.
  
• top: polA i,q digitizer
• bottom: polB i,q digitizer
• polA
  
• is a lot wider than polB because of the rfi.
• It has extra counts in the most negative channel. This is probably from the dc offsets in the digitizers,.
• Page 3: spectral density.
• The spectral density was computed for each 64k samples and then averages over the 10 seconds.
• .013 secs of data gives a resolution of about 32 hz.
• The dc value (averaged over the 10 secs of data) was removed from each record before computing the spectra,.
• Black is polA,red is polB
  
• Top: full spectra
• The X axis is the rf sky frequency.
  
• there are two spikes close to Dc and .5 MHz below Dc
• middle:blowup around DC
• The x axis is now relative to the a/d DC.
  
• The spike is about 1Khz above DC
• It looks like the voltage was ac coupled (missing power around DC). I thought the mocks were not ac coupled,.
• Bottom:
• blowup .5 MHz below DC
• The spike is .5 MHz below DC, (not below the 1khz spike).
  
• This must be an instrumental birdie.
• Page 4: Blowup of the rf freq scale.
• The xm signal is about 200 Khz wide (in this channel).
• There is very little leakage into polB.
• Our rcvr is circular. They must be transmitting circular.
  

 processing: x101/221118/voltsamp.pro

----------------------------------------------------

• click on link above
• right click on a file
• save link
• each datafile is about 400MBytes