Build a DIY Doppler Radar for Short-Range Tracking: Tips and Techniques

[moonman239]

I'd like to know how to build a short-rang (maybe 30-100 ft) Doppler radar. I will use the radar to pinpoint any object it sees on a map. This radar will be used with a PC or iPhone, so the unit does not make any calculations whatsoever.

Also, I'd like the unit to be as small as possible.

 

[Bill Simpson]

It generally seems like a good idea to start any project with a small simplified prototype to see if you can get anything at all to work.

Can you, for example, program your sound card to simultaneously output a good quality sine wave which linearly increases in frequency and then linearly decreases in frequency WHILE you are using the sound to capture incoming sound? To try this idea you need it to be doing both at the same time and have very good quality, no jittering around with the input or output sampes.

IF you can get it to do that then aim the speaker off towards an open part of the room and position the microphone beside and perhaps somewhat behind the speaker. You are trying to get the microphone to, mostly, pick up sound reflected back from the room and not be overwhelmed by the volume coming directly from the speaker.

Now think about the frequencies that will be present in the recorded sound. There will be one very large spike which is coming directly from the speaker, but you are trying to minimize that with the configuration. It is the other, potentially much smaller amplitude, frequencies in the sound. Those represent the "old" frequency from the speaker as the sound traveled across the room, hit an object, bounce off and traveled back across the room to your microphone. The difference in frequency between the "current" speaker frequency and the "old" frequency is your doppler signal that you are interested in.

Can you then write a bit of software that will identify the main peak directly from the speaker and the other peaks from echoes, determine the difference between those frequencies and display this?

The prototype can be as small and simple as this. Then you start looking at all the reasons why it isn't working as well as you would like and start to remedy those.

You might include one tiny bit of hardware outside the PC that could make your task much easier. There are inexpensive little integrated circuits called "four quadrant multipliers." This takes in two, positive or negative, voltages, multiplies them and gives you that out as a positive or negative voltage. If one input were your audio output and the other was your suitably amplified microphone input then the multiplier would be acting as a "mixer" and with a little filtering of the resulting frequencies you would have preprocessed doppler radar you might be able to input via your sound card. That would get rid of the need to identify peaks, measure differences in frequencies, etc. To understand that you might want to read up on "mixers", not what the usual audiophile calls a mixer, and find and understand the mathematics behind mixing.