Capture To File

Capture to file is a simple utility that records USRP samples to a file for post processing.  The I/Q data pairs are stored as 32-bit IEEE float point values in little endian format.  This is version 1.0 of the utility.  The next version will automatically generate the output file and verify the input parameters.  For now, this should get you started.

  capture_to_file.py

The utility requires the following command line parameters:

file_name - Where file_name is the output filename and path.  (Like you couldn't figure that out).
IF_freq_MHz - USRP tuning frequency in MHz where a negative frequency number inverts the input spectrum.
decim_rate - Decimation rate down from 64MHz sampling clock rate.
duration_seconds - Length of recording in seconds where seconds is a valid floating pointer number

In the first example, we capture a narrow band, ±128KHz signal bandwidth.  In the second, we capture a wider ±2MHz bandwidth suitable for GPS signals.  In the third example, we capture a standard broadcast FM station with a bandwidth of ±256KHz.  In the final example, we capture a ±4MHz bandwidth suitable for NTSC video.
 

[mgray@socata ~/py]$ capture_to_file.py narrowband_256k_complex.dat 10.7 250 30 [mgray@socata ~/py]$ capture_to_file.py gps_4msps_complex.dat 10.7 16 0.5 [mgray@socata ~/py]$ capture_to_file.py wbfm_512k_complex.dat 10.7 125 15 [mgray@socata ~/py]$ capture_to_file.py ntsc_colorbars_8msps_complex.dat 10.7 8 0.5

AM

Demodulate good old fashioned AM (Amplitude Modulation) that can be received with a 3 cent diode and ear phone.  But you're using $2500 worth of computer hardware and lots of software.  Oh well, its all good.

  am_rcv.py

From the command line, run the program and specify the input file name and tuning frequency in KHz.  This application will run with only the GNU radio core and audio driver for your system.  No hardware is required. 

In this example, the first two commands were used to listen to a recorded RF file.  The first example tunes to 10.000MHz and is radio station WWV.  The second tunes to 9.975MHz and is a SW broadcast station.  In the last example, the USRP hardware was used instead of a file and we listened to 10.000MHz live.
 

[mgray@socata ~/py]$ am_rcv.py hf_10000khz_256_complex.dat 0 [mgray@socata ~/py]$ am_rcv.py hf_10000khz_256_complex.dat -25 [mgray@socata ~/py]$ am_rcv.py usrp 0

Upper/Lower SSB

Demodulate upper and lower single side band signals.  Code based on excellent set of samples from http://webpages.charter.net/cswiger

  ssb_rcv_file.py

From the command line, run the program and specify the tuning frequency, BFO (tune to get best audio quality), and side band selection.  This application will run with only the GNU radio core and audio driver for your system.  No hardware is required. 

For the OTA capture, the USRP IF was centered at 50.250MHz and the suppressed carrier signal was transmitted on 50.350MHz.  This 50Khz offset is the second parameter specified in the command line.  The BFO is used to tune for best or most natural sound.  SSB is very frequency sensitive and less than 50Hz offset between the transmitter and receiver will produce a "Donald Duck" sound.  I've not yet calibrated the sampling clock on the USRP hardware.  One note, the USB signal demod is not very clean.  More investigation is required. 
 

[mgray@socata ~/py]$ ssb_rcv_file.py ssb_lsb_256k_complex.dat 50000 1500 l [mgray@socata ~/py]$ ssb_rcv_file.py ssb_usb_256k_complex.dat 50000 1500 u

Narrow Band FM

The following code is useful for tuning and listening to the NBFM sample set above.  I've started to comment the code, but it will require more to make it useful.

  nfm_rcv_file.py

From the command line run the program and specify the sample file and desired tuning channel.  If you select a frequency without a signal, you'll hear the classic FM background hiss.  This application will run with only the GNU radio core and audio driver for your system.  No hardware is required.
 

[mgray@socata py]$ nfm_rcv_file.py nws_nbfm_4_stations_256k_complex.dat -50

The following table shows the tuning offset, its associated OTA frequency, and the transmitter location.  Because of terrain and the distance to the stations, signal strength varies.
 

Offset	OTA Frequency	Transmitter Location	Signal Level
-75KHz	162.400 MHz	Flagstaff	-97 dBm
-50KHz	162.425 MHz	Payson (Mt. Ord)	-85 dBm
+25KHz	162.500 MHz	Globe (Signal Peak)	-107 dBm
+75KHz	162.550 MHz	Phoenix	-111 dBm