What's Your IT Question?

Want to know more about this topic or about us? Contact us!

How to make a spectrum analyzer with Raspberry Pi

1. Hardware

To build this project you’ll need the following hardware:

  • Raspberry Pi, the model B+model B, or model A will work.
  • RTL-SDR software radio USB dongle
  • 3.5″ PiTFT display
    • Right now the software is optimized to run on the 480×320 3.5″ PiTFT display.  If you want to get your hands dirty in the code it should be possible to port it to smaller displays like the 320×240 2.8″ PiTFT by changing font sizes and other dimensions.

You might also consider some extra parts to extend the project:

Once you have the parts assembling the project is very easy.

Follow the 3.5″ PiTFT guide to learn how to install and use the PiTFT.  Make sure to start with the easy install PiTFT image so you can get up and running as quickly as possible.  Don’t go forward until you have the PiTFT working with the Raspberry Pi!

Finally plug in the RTL-SDR dongle to a USB port on the Raspberry Pi.

That’s it, your hardware is ready to go!  Continue on to learn how to install the software for this project.

2. Installation

Dependencies

Before you can install the software you’ll need to install a few depdendencies.  Make sure you’re running the Raspbian operating system and it has been configured to work with the PiTFT.  You will also need to make sure the Pi has access to the internet through a wired or wireless networking connection.

Connect to a terminal on the Pi and execute the following commands to install the dependencies:

 sudo aptget update
 sudo aptget install cmake buildessential pythonpip libusb1.00dev pythonnumpy git

Don’t worry if you see error messages that certain packages are already installed, you can ignore those and move on.

Next download, compile, and install the RTL-SDR library by executing:

cd ~
git clone git://git.osmocom.org/rtl-sdr.git
cd rtlsdr
mkdir build
cd build
cmake ../ DINSTALL_UDEV_RULES=ON DDETACH_KERNEL_DRIVER=ON
make
sudo make install
sudo ldconfig

Finally install the RTL-SDR Python wrapper by executing:

 sudo pip install pyrtlsdr

At this point all the dependencies should be installed and you’re ready to download and install the project software.

Installing the software for this project is as easy as downloading its Python source code.  At a terminal on the Pi run the following commands to download the source:

cd ~
git clone https://github.com/adafruit/FreqShow.git
cd FreqShow

After a few moments a splash screen should display on the PiTFT as the program loads:

raspberry_pi_DSC00701.jpg

If the program fails to load go back to the installation steps and carefully check all the dependencies are installed, then try again.

Once the program loads it will start by displaying an animated frequncy graph, and menu of options.  The graph displays the intensity of radio signals (measured in decibels) across a range of frequencies.  The taller a peak on the graph, the higher the intensity of the signal at that frequency.

Below is an over of the important parts of the graph:

raspberry_pi_DSC00687_annotated.jpg

To the left of the graph the minimum and maximum intensity values are displayed in decibels.  The top value is the maximum intensity, in this case 50 dB, which means a point at the very top of the graph has an intensity of 50 dB.  Similarly the bottom value is the minimum intensity, -3 dB, so points at the bottom of the graph have an intensity of -3 dB.  Points between the top and bottom of the graph have an intensity in between the minimum and maximum, so for example a point in the middle of the graph would have an intensity around 26-27 dB.

At the bottom of the graph the range of frequencies are displayed.  The minimum frequency is displayed at the bottom left, the center frequency in the middle, and the maximum frequency at the bottom right.  In the image above you can see a total range of about 2 MHz of frequencies centered at 90.3 MHz.

The frequency 90.3 MHz actually corresponds to an FM radio station (KEXP in Seattle, WA).  It’s interesting to see the spikes of intensity near the center frequency–the graph is visualizing what an FM radio plays as audio!

You can change the center frequency and other settings by pressing the CONFIG button in the upper left corner.  A list of settings such as the following should appear:

raspberry_pi_DSC00689.jpg

Press the CENTER FREQ button at the top to bring up a dialog to change the center frequency:

raspberry_pi_DSC00690.jpg

You can change the center frequency just like inputting numbers in a calculator.  Press the CLEARbutton to completely erase the current frequency value, enter new digits by pressing numbers (or the decimal point), and delete the last digit by pressing DELETE.

Once you’ve entered a new value you can press ACCEPT to accept it and change the center frequency.  If you don’t want to change the frequency press CANCEL to dismiss the dialog without changing the value.

Try changing the center frequency to that of an FM radio station in your area.  FM radio in the US is broadcast over 88 to 108 Mhz so pick a value within that range.  After accepting a new center frequency you should be returned to the settings list.  Press BACK in the lower left to return to the frequency graph.  Notice how the graph changes to show the intensity of the signal at the new center frequency.

Another useful way to visualize radio frequency data is with a spectrogram that shows the intensity of the signal over time.  In the main graph view press the SWITCH MODE button at the top of the screen.  You should see the graph change to start building a waterfall plot that scrolls up from the bottom to the top of the screen.

For example this is a waterfall plot that has entirely filled the screen:

raspberry_pi_DSC00686.jpg

The waterfall plot displays intensity over time.  Each row of the plot represents a point in time, and each pixel color in a row represents the intensity of the signal at that frequency.  Pure blue pixels are the minimum intensity, pure red pixels are the maximum intensities, and a gradient of blue-cyan-yellow-red represents in between intensities.

Notice in the picture above the highest intensity frequencies in the center are red, while the lowest intensity frequencies at the edges are blue.  Smaller peaks are cyan and yellow.

About once a second the graph will scroll up and a new measurement row will be added at the bottom.  Over time the waterfall plot will fill the screen, with the most recent measurements at the bottom and oldest at the top.  By looking at how the color of the watefall plot changes, you can see how the intensity of a radio signal changes over time.

See the image below for the full range of intensity colors:

raspberry_pi_gradient.png

To switch back to the instantaneous frequency graph press the SWITCH MODE button at the top of the screen again.  You can press SWITCH MODE to swap between instantaneous and waterfall plot modes.

Another useful feature of the graph is a full screen view.  Press anywhere on the graph image itself, like the middle of the screen.  You should see the graph maximize itself in a full screen view with no buttons or labels.  For example this is a full screen of the instantaneous mode:

raspberry_pi_DSC00688.jpg

The fullscreen view is great if you want to more clearly see the graph.  You can switch back to the normal view by pressing the middle of the graph again.

Now go back to the settings menu by pressing the CONFIG button.  Here’s a quick overview of each setting:

  • Center Frequency
    • This is the center frequency of the tuner and must be a value in megahertz.  The exact range of allowed values will depend on the radio tuner, but for the RTL-SDR in the store the range is about 24 Mhz to 1,850 Mhz.  Note that it can be illegal to tune certain frequencies so check the laws in your country!
  • Sample Rate
    • This controls how wide the range of frequencies are that will be displayed in the graph.  The tuner doesn’t support a large range of sample rates so your best bet is to stick with the default of 2.4 Mhz.
  • Gain
    • The internal gain of the tuner can be adjusted with this setting.  By increasing the gain you can remove noise and make a weak signal more easily visible.  The tuner has a limited range of gains from about 1 to 50 dB and will try to use a value as close to the configured gain as possible.  You can also choose AUTO as a gain option to have the tuner automatically adjust the gain for the best signal reception.  I recommend sticking with auto gain for simplicity.
  • Min and Max
    • These values at the bottom of the setting list control the minimum and maximum intensity values of the graph.  Like gain they can be set to AUTO, in which case they will adjust themselves based on the lowest or highest intensity values ever seen.  However if you want to restrict the graph to a certain range of values you can set explicit intensity values in decibels.  I’ve found a range of -3 dB to 50 dB is a good general range to set when using automatic gain.

Finally if you want to exit the tool, in the main graph view press the QUIT button in the upper right.  A confirmation dialog will show which you can accept to exit, or cancel to return back to the tool.

That’s all there is to using the Freq Show Raspberry Pi RTL-SDR scanner tool!  If you run into issues with the tool or wish to contribute to it, check out the tool’s home on GitHub.

Have fun exploring the world of radio waves around you!

raspberry_pi_DSC00705.jpg