---

Intro:

• The control software runs on the cadi cpu (under linux).
  
• The users are root and cadi
  
• A vnc session is available for login.
• warning (the vnc session currently belongs to root since it is started in the rc.local without su )
  
• the ionosonde is normally run every 15 minutes via a root crontab file.
• programs on cadi cpu:
  
• /usr/sbin/cadi  configfile.dat
• runs the ionsonde, gathers the data, and writes to output file (xx.MDL)
• this is what the crontab runs every 15 minutes
• data is written to /var/cadidata
• See the data file format:
  
• /usr/sbin/cadidisplay -f filename  
  
• will read the .MDL file and plot the ionogram.
• there are some scripts under /home/cadi that have been used to copy data from the cadi cpu to other computers on aonet,.
  
• Data locations;
• the .MDL files are written to /var/cadidata/xxx.MDL
• eventually it gets moved to /home/cadi/data/yyyy
• historically data has been copied to
  
• /share/aserv00/xserve0/raid/cadi/data/yyyy
• 2003 - 2013 is on disc
• Offline software
• idl routines (under cadi) that let you read in the .mdl files and generate an ionogram.
  

180130: initial cadi tests

    The cadi was run 30jan18 through 01Feb18 (together with a world day).
Processing:

• The data from the cadi host was copied to /share/aserv00/xserve0/raid/cadi/data/phil/2018
• the monnamtonum.sc script was run to change the month names to numbers (so the file order is now date order).
• the idl routine: /share/megs/phil/x101/cadi/ldcadifiles.pro was run to load all of the data files into
  
• hdrAll[n]: 1 header per datafile
• frqAll[300,n]: freq channel info for the 300 freq channels of each file
• had freq, avg noise power (averaged over heights), gainflag, thresholding flag, numbers of heights/dopper channels recorded, a pointer in eAll[m] for start of this freq info.
• eAll[m] : this holds the echo data recorded by the cadi program
• Only height,doppler channels above a specified threshold are recorded.
• this contains:
• ifreq - first index into frqall[300,n] for this entry
• ih      - height index for this entry.
• for our config, heights go 90km, to 510 km in 3km steps
• so ih =0 is 90km ih=1 is 93km...
• dopBin- doppler bin for this entry.
• our cadi processing takes 8 ipps (25 ms each) at a single frequency
• removes the 13 length barker code on the received data,
  
• then does an 8 point fft on each height
• the dopbin is 0..7 , the doppler freq bin for this entry.
• iq[2]: iq voltage for this entry. units are a/d counts
• this is probably after the 8 point fft.

Plots:
    The plots show the recorded voltages and powers  for the 3 days. (.ps) (.pdf):

• Page 1: echo voltages:
• Top: i/q voltages for recorded echo power (for the 3 days)
• Black is the I digitizer , red is the Q digitizer
• These are 8 bit digitizers (-128 to +127)
• The echo data is only recorded if it is above a preset threshold.
• The I digitizer has a 9 level DC offset. The Q digitizer has a 1 level DC offset
• Bottom: histograms of the i,q data
• black is digitizer I, red is digitizer Q.
• Page 2: noise and echo power.
• Black: the average noise power for an ionogram vs hour from 30jan18
• I integrated over all heights and frequencies.
• Red: the power from the echo returns that were recorded.
• Page 3: over plot average noise power (over hghts) vs freq for all data
• an offset has been added to each spectra for display
• there is an rfi spike visible at 2.3 Mhz.

Dynamic spectra of the average noise power vs time (.gif)

• For each frequency channel,  the noise power is averaged over all heights (giving a 300 channel spectra for each ionogram).
  
• I limited the x axis to 1 to 10 MHz.
• the tick marks in the middle are not correctly spaced since the freq spacing is not linear (someday I'll fix the image plotting routine that asks for the min, max value!). The constant vertical lines is actually at 2.36 MHz.
  
• The white strips horizontal stripes (at 10-15 hrs) and 34-37 hours are when we got good ionograms.

Compare fof2 from ionogram and clp data
    Data taken 20180120, 11:45 ast is shown for the ionosonde, and for the clp data taken with the echotek recorder

• ionogram from cadi (.jpg)
• fof2 at  5.95
  
• plasma line profile from 430tx and echotek recorder (.jpg)
• peak at 436.07 MHz
  

movies of the ionograms for the 3 days (.avi) (.mov)

• I made ionograms from each of the 285 .mdl files
• data was input with cadi_inpfile
• the ionograms was created with cadi_doimg
• jpegs for each ionogram were written with mkkpeg
• the movie was then made with mkmov.sc
• all in x101/cadi/
• Not many of the ionograms are usable.
• the better ionograms correspond to when the average noise power was lower (see above).

Summary

• the cadi was run for about 3 days.
• fof2 freq for 11:45 30feb18 had cadI; 5.95, clpData:6.07. the 430 data was about 75Khz higher.
  
• there is a 9 a/d level offset in the I digitizer
• most ionograms were not usable
• there were 2 periods (of about 4 hours) when we had good ionograms.
• there is probably something not working correctly

processing: x101/180206/caditest.pro

Documentation

• cadi users manual (.pdf)
• The data format is not complete. The description has unaligned data. The linux data as been aligned for the data type.
  
• cadi software Manuel (.pdf)
  

<- page up
home_~phil