Installing SDR applications on a Raspberry Pi and Debian

This document will guide you through installing SDR applications on the platform built in the document: Installing drivers for the SDRplay RSP1A and HackRF One on Raspberry Pi and Debian and depends on that installation.

This document is intended to be followed in the order it is written. If you want to skip installing one of applications, there may be some missing prerequisites.

Install:

• Install QRadioLink. For GNU radio 3.10, see comments in the install section.
  Multimode SDR transceiver for GNU radio using the Internet for VOIP communication (radio over IP) with various analog, digital voice and digital video modes.
  • Prerequisites
    • mumble

      sudo apt install mumble mumble-server

    • Ensure that libusb-0.1-4 in not installed.

      # Look to see that "Installed" does not appear after libusb-0.1-4
      apt list -a libusb-0.1-4
      
      # If it is installed, remove it:
      sudo apt remove libftdi1 libftdi-dev libusb-0.1-4
      sudo apt autoremove

    • Install additional packages:
      Note: On Ubuntu 20.04 newer libjpeg-turbo8 libjpeg-turbo8-dev has to be installed instead of libjpeg62-turbo libjpeg62-turbo-dev
      Note: On newer systems such as Debian bookworm, use liblimesuite22.09-1 instead of liblimesuite20.10-1. You may also need cppzmq-dev
      • On Bullseye systems

        sudo apt-get install protobuf-compiler libopus-dev libspeexdsp1 libspeexdsp-dev libjpeg62-turbo libjpeg62-turbo-dev libconfig++9v5 libconfig++-dev libftdi1-2 libftdi1-dev libsndfile1-dev libsndfile1 libjpeg-dev libprotoc-dev liblimesuite-dev liblimesuite20.10-1

      • On Bookworm systems

        sudo apt-get install protobuf-compiler libopus-dev libspeexdsp1 libspeexdsp-dev libjpeg62-turbo libjpeg62-turbo-dev libconfig++9v5 libconfig++-dev libftdi1-2 libftdi1-dev libsndfile1-dev libsndfile1 libjpeg-dev libprotoc-dev liblimesuite-dev liblimesuite22.09-1

    • Make applications that need libftdi1 0.20 use the new packages

      sudo ln -s /usr/include/libftdi1/ftdi.h /usr/include/ftdi.h
      sudo ln -s /usr/lib/aarch64-linux-gnu/libftdi1.so.2.5.0 /usr/lib/aarch64-linux-gnu/libftdi.so.1
      sudo ln -s /usr/lib/aarch64-linux-gnu/libftdi1.so.2.5.0 /usr/lib/aarch64-linux-gnu/libftdi.so
      

  • Install

    cd ~/
    git clone https://codeberg.org/qradiolink/qradiolink
    cd qradiolink/
    # The default branch is the next branch, which is the latest.
    # To get the normal release, use the following command (without the "#")
    # git checkout master
    # Only if your GNU Radio version is 3.10, Switch to the gnuradio 3.10 branch with the following command
    git checkout gr_3.10
    sh ./build_debian.sh
    

  • To run

    ~/qradiolink/build/qradiolink

  • Click the Setup tab
  • Click the General Settings tab
    • To use a SoapySDR device
      • Select your receive device from the RX device args menu.
      • In RX antenna, enter RX
      • If you have a transmit device, select it from the TX device args menu.
      • In TX antenna, enter TX
      • Click Save.
    • To use the built-in gr-osmosdr device
      • For the SDRplay RSP1A
        • In RX device args, enter sdrplay=0,hwVer=255
        • In RX antenna, enter RX
        • Click Save.
      • For HackRF
        • In RX device args, enter hackrf=0
        • In RX antenna, enter RX
        • In TX device args, enter hackrf=0
        • In TX antenna, enter TX
        • Click Save.
      • For rtl-sdr
        • In RX device args, enter rtl=0
        • In RX antenna, enter RX
        • Click Save.
  • Click the Radio settings tab
    • For Output audio device, select alsa_output.platform-fef00700.hdmi.iec958-stereo or similar. It will vary between Operating Systems.
    • You may want to uncheck TX Limits if you are not in IARU region 1.
  • Click Save.
  • Click the Spectrum tab.
  • Click RX to start the SDR receiver.
  • Note: For the first time an SDR application is run after GNU Radio installation, you may have to wait 2-3 minutes for the terminal to print 3 lines of "gr::vmcircbuf_sysv_shm: shmat (3): Invalid argument". It will work correctly after that. There is a function in GNU Radio that does not work on the Raspberry Pi. GNU Radio will eventually fix itself.
• CubicSDR - build from source to get the latest version that supports the SDRplay RSP Duo.
  A very nice versatile SDR application.
  • Prerequisites
    • Install FFTW3 if not currently installed. Note libgtk2.0-dev is removed.
      Note: If the following apt-get install fails due to unmet dependencies blaming gir1.2-gtk-3.0 3.24.24-4+rpt2 or ...rpt3, remove it with the command below that matches the version specified in the error message:
      sudo apt remove gir1.2-gtk-3.0=3.24.24-4+rpt2
      sudo apt remove gir1.2-gtk-3.0=3.24.24-4+rpt3
      Repeat the install command.

      sudo apt-get install freeglut3-dev libgtk-3-dev libtinyxml2.6.2v5 libfec0 libfec-dev
      

    • liquid-dsp

      cd ~/
      git clone https://github.com/jgaeddert/liquid-dsp
      cd liquid-dsp
      ./bootstrap.sh
      CFLAGS="-march=native -O3" ./configure --enable-fftoverride
      make -j 4
      sudo make install
      sudo ldconfig
      

    • RtAudio

      sudo apt-get install libasound2-dev portaudio19-dev libtool libpulse-dev
      cd ~/
      Download http://www.music.mcgill.ca/~gary/rtaudio/release/rtaudio-5.2.0.tar.gz 
      tar -zxvf rtaudio-5.2.0.tar.gz
      cd rtaudio-5.2.0
      # ./configure --help # display the available options
      ./configure --with-jack --with-alsa --with-pulse
      mkdir build
      cd build
      cmake ../
      make -j 4
      sudo make install
      sudo ldconfig
      cd tests
      # Try one of these to see if it will play something. Press CTRL-C to exit.
      ./playsaw 1 48000
      ./playsaw 1 44100
      ./playsaw 2 44100 # 2 different sawtooth waveforms, one in each speaker
      
      # If you don't hear a sound from any of these, it may have selected the headphone output.
      # Set speakers as default
      pacmd set-default-sink 0
      

    • wxWidgets
      • CubicSDR wants to use a static installation of wxWidgets. CubicSDR has been updated to use wxWidgets 3.2.1. This will take quite a while.
        The installation path for this installation is included in the CubicSDR installation command.
      • Prerequisites:
        Remove all wxWidgets items using the Package Manager.
      • Install
        Note: replace '~/Develop/wxWidgets-staticlib' with your own path if you installed it in a different location. Use this path when building CubicSDR. Note that ./configure parameters are updated and they have a CXXFLAGS parameter which would conflict with the CXXFLAGS environment variable.

        cd ~/
        # Note: wxWidgets-3.2.6 works on Raspberry Pi OS Bookworm and most likely on other OSes.
        
        # If you wish to use git to install wxWidgets, use the commands below and skip the archive.
        # git clone --recurse-submodules https://github.com/wxWidgets/wxWidgets.git
        # git checkout v3.2.1
        
        # Download this archive to your home directory.
        https://github.com/wxWidgets/wxWidgets/releases/download/v3.2.1/wxWidgets-3.2.1.tar.bz2 
        tar -xvjf wxWidgets-3.2.1.tar.bz2
        # rename the install folder to wxWidgets
        cd wxWidgets/
        mkdir -p ~/Develop/wxWidgets-staticlib
        ./autogen.sh
        ./configure --prefix=`echo ~/Develop/wxWidgets-staticlib` --with-opengl --disable-glcanvasegl --disable-shared --enable-monolithic --with-libjpeg --with-libtiff --with-libpng --with-zlib --disable-sdltest --enable-unicode --enable-display --enable-propgrid --disable-webview --disable-webviewwebkit CXXFLAGS="-std=c++0x"
        make -j4
        make install
        sudo ldconfig
        

  • Build CubicSDR

    cd ~/
    git clone https://github.com/cjcliffe/CubicSDR.git
    cd CubicSDR
    mkdir build
    cd build
    
    # Choose one of the cmake statements below:
    
    # ALSA audio works on most systems.
    cmake ../ -DCMAKE_BUILD_TYPE=Release -DwxWidgets_CONFIG_EXECUTABLE=~/Develop/wxWidgets-staticlib/bin/wx-config -DUSE_HAMLIB=1 -DUSE_AUDIO_PULSE=0 -DUSE_AUDIO_OSS=0 -DUSE_AUDIO_ALSA=1
    
    # Pulse audio works better on systems that use pipewire such as Raspberry Pi OS Bookworm and Ubuntu.
    cmake ../ -DCMAKE_BUILD_TYPE=Release -DwxWidgets_CONFIG_EXECUTABLE=~/Develop/wxWidgets-staticlib/bin/wx-config -DUSE_HAMLIB=1 -DUSE_AUDIO_PULSE=1 -DUSE_AUDIO_OSS=0 -DUSE_AUDIO_ALSA=0
    
    make -j 4
    # You can now run CubicSDR from the build folder.
    x64/CubicSDR
    
    # Install CubicSDR (and launcher in Applications menu)
    sudo make install
    sudo ldconfig
    

    • There will be an icon for CubicSDR in the Applications Menu / Other. It will appear after a restart.
• gqrx. Works with GNU radio 3.7 or higher.
  Can operate as an AM/FM/SSB receiver with audio output or as an FFT-only instrument. There are also various hooks for interacting with external applications using network sockets. One of the few that has a noise blanker and allows custom bandpass adjustment.
  • The version in the packager is fairly recent. If you install that, skip to the "To start gqrx..." section below.
  • Prerequisites

    # portaudio should have been installed for RtAudio.
    sudo apt install portaudio19-dev
    

  • Install

    export CXXFLAGS='-O2 -march=native -mtune=native'
    export CFLAGS='-O2 -march=native -mtune=native'
    
    cd ~/
    git clone https://github.com/gqrx-sdr/gqrx.git gqrx
    cd gqrx
    mkdir build
    cd build
    cmake ..
    make -j 4
    sudo make install
    sudo ldconfig
    

  • To start gqrx, type gqrx at the command prompt and press enter.
  • The Configure I/O devices dialogue box will appear.
  • Select your SDR device from the Devices menu.
  • Select an audio device
  • The defaults for the other items should be OK.
  • Click OK.
  • Click the Start DSP processing button on the top left of the gqrx window.
  • Note: When using AM, be sure to uncheck DCR in the "^{. . .}" button to the far right of Mode if there is no audio.
• SparkSDR, works with SDRplay.
  An sdr application that can run multiple radios and virtual receivers simultaneously with built in digital modes PSK31, WSPR, JT9, JT65, FT8, FT4, FST4, FST4W.
  • Prerequisites (normally installed in pevious steps)
    • The SDRplay API version 3.07 or greater. It is installed on the drivers, etc. page.
    • portaudio, installed previously.
    • WSJT-X from the Installing utility software on a Raspberry Pi page.
  • Install SparkSDR
    • Go to https://www.sparksdr.com/downloads
      Download the Linux installer to ~/Downloads. Example for arm64 (rpi 64 bit): SparkSDR.2.0.33.linux-arm64.deb
      There is a Preview Release available that will support more of your SDRs. Download SparkSDR.2.0.986.linux-arm64.deb and use it in the commands below.

      cd ~/Downloads
      sudo dpkg -i SparkSDR.2.0.33.linux-arm64.deb
      

    • Run SparkSDR from the Sound & Video group on the start menu.
    • If you hear audio dropouts, try Pulse. Click the "^{. . .}" virtual receiver settings menu button. Under Audio Output, select Pulse.
  • Help can be found here: https://www.sparksdr.com/manual and the SparkSDR Google Group.
• SdrGlut and iqSDR
  A simple software defined radio player. It is small, fast and efficient. Very low CPU load. Has Search and Scan Features. Supports SDRplay RSPduo running in Dual Tuner mode.
  • Prerequisites
    • RtAudio (installed earlier in this document)
    • liquid-dsp (installed earlier in this document)
    • They need:

      sudo apt-get install libopenal-dev libalut0 libalut-dev freeglut3-dev libsndfile1 libsndfile1-dev libwxgtk3.2-dev libglew-dev libusb-dev

    • Setup for the SdrGlut V1.24+ Voice Command Module
      • For Bullseye and older systems

        pip install SpeechRecognition
        sudo apt install python3-pyaudio

      • For Bookworm and later systems
        With recent python versions, pip install will fail. We will install it from source.
        
        Download the source archive from https://pypi.org/project/SpeechRecognition/#files
        Extract it to your home directory and change to the directory created from the archive.

        cd ~/speechrecognition-3.11.0
        sudo python3 setup.py install
        
        #Test installation:
        python3
        import speech_recognition
        # There should be no error
        ^D to exit.
        

  • Install

    cd ~/Desktop
    git clone --depth=1 https://github.com/righthalfplane/SdrGlut.git
    
    # Install SdrGlut
    cd SdrGlut
    make -f makefileUbuntu -j 4
    
    # Install iqSDR
    cd iqSDR
    make -f makefileUbuntu -j 4
    
    Run the programs:
    ~/Desktop/SdrGlut/sdrglut.x
    or
    ~/Desktop/SdrGlut/iqSDR/iqSDR.x
    

  • It is not uncommon for audio to be distorted. I you are having trouble, try this fix, which will work for most installations:

    sudo mousepad /etc/openal/alsoft.conf
    

  • Find the line: #periods = 3  Remove the "#" and change it to periods = 4
  • Reboot and run SdrGlut. The audio should be good.
• OpenWebRX
  This procedure assumes that you have gone through the basic system installation. It also depends on items installed on the utilities page. This procedure is based on this manual installation.
  • Prerequisites
    • On Bullseye and Bookworm systems

      sudo apt-get install libfftw3-single3 libfftw3-double3 python3-setuptools python3-paho-mqtt netcat-openbsd libsndfile1 liblapack3 libreadline8 gfortran libgfortran5 libgomp1 libasound2 ca-certificates libopus0
      sudo apt-get install sox libtool libfftw3-dev libsamplerate-dev gnuplot gnuplot-x11 libudev-dev libprotobuf-dev libicu-dev

    • On Ubuntu 24.04 systems

      sudo apt-get install libfftw3-single3 libfftw3-double3 python3-setuptools python3-paho-mqtt netcat-openbsd libsndfile1 liblapack3 libreadline8 gfortran libgfortran5 libgomp1 libasound2t64 ca-certificates libopus0
      sudo apt-get install sox libtool libfftw3-dev libsamplerate-dev gnuplot gnuplot-x11 libudev-dev libprotobuf-dev libicu-dev

    • Install csdr

      cd ~/
      git clone https://github.com/jketterl/csdr.git
      cd csdr
      mkdir build
      cd build
      cmake ..
      make -j $(nproc)
      sudo make install
      cd ../..
      sudo ldconfig
      

    • Install pycsdr

      cd ~/
      git clone https://github.com/jketterl/pycsdr.git
      cd pycsdr
      sudo python3 setup.py install install_headers
      cd ..
      

    • Install JS8Call

      cd ~/
      git clone https://[email protected]/widefido/js8call.git
      cd js8call
      mkdir build
      cd build
      cmake ..
      make -j $(nproc)
      sudo make install
      cd ../..
      sudo ldconfig
      

    • Install js8py library

      cd ~/
      git clone https://github.com/jketterl/js8py.git
      cd js8py
      sudo python3 setup.py install
      cd ..
      sudo ldconfig
      

    • Install owrx_connector

      cd ~/
      git clone https://github.com/jketterl/owrx_connector.git
      cd owrx_connector
      mkdir build
      cd build
      cmake ..
      make -j $(nproc)
      sudo make install
      cd ../..
      sudo ldconfig
      

    • Install optional digital voice
      Digital voice requires an AMBE server (hardware or software).
      A software MBE server is installed in these procedures.

      sudo apt-get install libprotobuf-dev protobuf-compiler libudev-dev
      
      cd ~/
      git clone https://github.com/jketterl/codecserver.git
      cd codecserver
      mkdir build
      cd build
      cmake ..
      make -j $(nproc)
      sudo make install
      cd ../..
      
      # For Raspberry Pi OS, Bullseye only: move the service file to the correct location (other systems should work without this)
      sudo mv /usr/local/lib/aarch64-linux-gnu/systemd/system/codecserver.service /lib/systemd/system
      
      sudo ldconfig
      
      # Create the codecserver user and set permissions:
      sudo adduser --system --group --no-create-home --home /nonexistent --quiet codecserver
      sudo usermod -aG dialout codecserver
      # The username "codecserver" is used for the systemd service unit.
      
      # Copy the configuration file
      sudo mkdir -p /usr/local/etc/codecserver
      sudo cp ~/codecserver/conf/codecserver.conf /usr/local/etc/codecserver
      
      # Add configuration for mbelib:
      sudo mousepad /usr/local/etc/codecserver/codecserver.conf
      
      # Add these 3 lines to the end of codecserver.conf:
      [device:mbelib]
      driver=mbelib
      unvoiced_quality=3
      
      # Save and close the file.
      
      sudo systemctl daemon-reload
      sudo systemctl restart codecserver
      

    • Install mbelib

      # If MBELIB is already installed, please skip this step.
      cd ~/
      git clone https://github.com/szechyjs/mbelib.git
      cd mbelib
      mkdir build; cd build
      cmake ..
      make -j $(nproc)
      sudo make install
      cd ~/
      sudo ldconfig
      

    • Install codecserver-mbelib-module

      cd ~/
      git clone https://github.com/fventuri/codecserver-mbelib-module.git
      cd codecserver-mbelib-module
      mkdir build
      cd build
      cmake ..
      make -j $(nproc)
      sudo make install
      cd ..
      sudo ldconfig
      
      sudo systemctl restart codecserver
      

    • Install digiham

      cd ~/
      git clone https://github.com/jketterl/digiham.git
      cd digiham
      mkdir build
      cd build
      cmake ..
      make -j $(nproc)
      sudo make install
      cd ../..
      sudo ldconfig
      

    • Install pydigiham

      cd ~/
      git clone https://github.com/jketterl/pydigiham.git
      cd pydigiham
      sudo python3 setup.py install
      cd ..

    • Install codec2 from source:
      Check to see if the codec2 package is installed. Removing it will uninstall many other required packages..

      # If codec2/freedv is already installed, please skip this step.
      # Do not install this if the codec2 packages are installed.
      # If you install the freedv package, it will also install libhamlib4.
      cd ~/
      git clone https://github.com/drowe67/codec2.git
      cd codec2
      mkdir build
      cd build
      cmake ..
      make -j $(nproc)
      sudo make install
      
      # Manually install freedv_rx since it's not part of the default install package
      sudo install -m 0755 src/freedv_rx /usr/local/bin
      sudo install -m 0755 src/freedv_tx /usr/local/bin
      cd ../..
      sudo ldconfig
      

    • Install m17-cxx-demod

      sudo apt-get install libboost-program-options-dev
      cd ~/
      git clone https://github.com/mobilinkd/m17-cxx-demod.git
      cd m17-cxx-demod
      mkdir build
      cd build
      cmake ..
      make -j $(nproc)
      sudo make install
      cd ../..
      sudo ldconfig
      

    • DRM/Dream has not yet been tested with OpenWebRX. We may have to build a special version.
    • For Packet / APRS, Direwolf must be installed as described on Installing utility software on a Raspberry Pi.
    • Install the aprs-symbols to be able to display APRS positions on the map (you can change the destination, if you update the config accordingly)

      sudo git clone https://github.com/hessu/aprs-symbols /usr/share/aprs-symbols

    • wsjt-x must be installed as described on Installing utility software on a Raspberry Pi.
    • Data storage

      sudo mkdir /var/lib/openwebrx
      # If necessary, change the user from pi to your actual user name.
      sudo chown pi. /var/lib/openwebrx
      
      # Warning, This step will overwrite any existing user database.
      sudo sh -c "echo [] > /var/lib/openwebrx/users.json"
      # If necessary, change the user from pi to your actual user name.
      sudo chown pi. /var/lib/openwebrx/users.json
      sudo chmod 0600 /var/lib/openwebrx/users.json
      

  • Install OpenWebRX

    cd ~/
    git clone https://github.com/jketterl/openwebrx.git
    cd openwebrx
    
    # Create an initial user for the web configuration interface. Replace [username] with the name you will be using.
    ./openwebrx.py admin adduser [username]
    # Enter and confirm the password
    

  • Run OpenWebRX.

    # Start OpenWebRX
    sudo systemctl start codecserver
    cd ~/openwebrx
    ./openwebrx.py
    

  • Open the GUI in any web browser.

    http://localhost:8073
    
    # Using the IP address of your Raspberry Pi, you can open it on another computer on your local network.
    
    ip address
    
    # IP addresses for all of your interfaces will be displayed. Substitute that for localhost in the URL as in the example below.
    http://192.168.1.12:8073
    
    # If you can set up your router, you can access OpenWebRX from the Internet.

  • You should log in as the administrator user that you set up and configure it for your system. See the documentation.
    • To add an SDR device, for example, HackRF
      • In Settings, click SDR devices and profiles.
      • Click Add new device...
      • In Device name enter a name for your HackRF.
      • In Device Type, select HackRF.
      • Click Apply and save.
      • In Additional optional settings, select options such as Device Gain.
      • Click Apply and save.
    • To add it to the devices menu, create a profile.
      • Click the new profile tab.
      • Create a profile for a frequency range that you want to receive. Do not set the Center frequency to the same as the frequency you want to receive.
      • In Sample rate, select MS/s and enter something from 2 to 4 MHz. It may work with higher rate.
      • Click Apply and save.
  • If you wish to Install OpenWebRX as a systemd service, see the documentation.