UC Davis Students Build Coffee Can Radar Project Inspired By MIT


Blinking lights is a lot of fun, but if you’re getting an EE degree the cool stuff becomes a bit more involved. In this case, building your own radar is the thing to do. Here’s a coffee can radar setup being shown off by a group of UC Davis students. Regular readers will recognize the concept as one we looked at in December. The project was inspired by the MIT OpenCourseware project.

One of the cans is being used as a transmitter, the other as the collector. The neat thing about this rig is that the analysis is performed on a PC, with the sound card as the collection device. The video after the break shows off the hardware as well as the results it collected. About a minute and a half into the clip they show a real-time demonstration where a student walks in front of the apparatus while another takes a video of the plot results. As the subject moves away from the receiver the computer graph changes accordingly. The rest of the video covers some operational theory and pcb assembly.

[Thanks Gregory]

25 thoughts on “UC Davis Students Build Coffee Can Radar Project Inspired By MIT

  1. I know 100% originality isn’t a requirement to make it onto Hackaday, as many hacks are in fact based in other hacks, with different modifications and improvements…. But I was under the impression that these Senior Design Projects were supposed to incorporate something new and different. This seemed strikingly familiar to some ham radio experiments from one of my ARRL handbooks. I googled “coffee can radar”, and sure enough, this EXACT setup has been done many times over. Including the sound card based processing.

    Admittedly, I’ve never attended any EE program, so perhaps I’m mistaken in my understanding of these final projects. It just seems that after all that education your professor would expect more than you finding a completely documented project and assembling it. Unless the focus is on demonstrating a knowledge of assembly and function, not engineering your own solution to a problem.

    Still something I’d like to put together myself someday. I’m still trying to figure out what effective ranges could be squeezed out of something like this.

    1. I can’t speak for all EE programs, but at all of the undergrad programs I have heard of, as well as the one I am currently attending, the purpose of the design project is not to invent something new, but to prove that the student can take a design from an idea to a professional prototype. Perhaps in graduate and postgraduate work the research must be original, but the undergrad degree usually teaches only the basics of engineering, while the specialized information needed for a job is learned during an internship.

    2. I saw the project and thought the same thing. If you can get a degree from UC Davis rebuilding past projects, then where is the ingenuity? I understand if this was entry level where you are learning basics by building projects and figuring out how they work, but for degree-level stuff I would expect a bit more than following an instructable. Oh well, he’ll be making the big bucks one day.

    3. I think the real purpose of a senior project is to actually build something. Most of a EE degree consists of a lot of math and analysis. Actually picking a project and building it means learning a lot of real world interactions that isn’t covered in class. The steps to complete thas are not as straight forward as you might think, a lot is left out of documentation, and any mistakes made along the way have to be discovered and rectified by the students.

    4. I’m the instructor for the class. Thank you for the criticism. I have to clarify a misconception here. The purpose of a senior design class is not to invent something new, but rather to get the students’ hands dirty by forcing them to build a working system. There’s a lot to learn in that process.

      BTW, this year, we’ve made the radar in the form of an Arduino shield. We showcased it at Maker Faire this past weekend.

  2. Homodyne CW radars seem to be a popular topic in senior design projects…I did one as well, circa 1998. A few tips for anyone who wants to try it:

    Linearize the LO frequency sweep. The F/V characteristic of most VCOs is somewhat nonlinear, and applying a pure sawtooth voltage waveform will result in spectral smearing at the IF and a diminished ability to differentiate targets. Since the pulse repetition rate is in the audio frequency range, you may well be able to use an audio DAC to generate the sweep waveform.

    Reflected signal strength drops in amplitude by ~20dB/decade. Countering that by constructing an IF amp with a +20dB/dec gain slope will allow you to better control the dynamic range presented to the ADC.

    Gate the FFT so that the end/beginning of sweep discontinuities don’t pollute your data. This would be another good reason to use the computer’s sound card to generate the sweep waveform if possible.

    If the RF/microwave work makes you nervous, check out Gunn transceivers, also known as Gunnplexers. These are integrated assemblies that include the oscillator and mixer mounted in a piece of waveguide. All of the RF is encapsulated, all that’s exposed is the low-frequency IF and VCO control voltage. Making a waveguide horn antenna is straightforward, there are lots of patterns and designs online.

  3. Using a can(-shaped) antenna seems like a amateurish and poor approach, I though we had many designs available because over the years we learned plenty on the theory and practical elements of antennas.
    But hey maybe it’s because they followed the book mentioned by ‘dukeofmuffins’ in the other comment, and they did not do any effort to put that in a more clever/modern setup

    Oh and MIT students need to use a hacked toaster-oven for reflow? I thought they were rolling in money and pro-equipment.

    1. Whats amateurishand poor about using cantennas, they are actually pretty high gain and directional, which is what you want. And they are cheap to make. If you went with a parabolic dish, you would need two (as this design uses serperate tx/rx antennas) and it would be a pain to align everything. While cool for playing in the lab, pretty useless for anything else. you aren’t going to have enough resolution or power to do much with this thing besides play.

      And a gunnplexer horn with VCO? gunn diodes are CW and mechanically tuned cavilty devices, i.e. you have to adjust the size of the cavity in the horn to change the frequency. They require a bias voltage that is DC, you may get away with pulsing it, but I’ve never seen a VCO gunn diode…? At least not outside some abstract papers and you aren’t going to find one at hobby prices. A K-band gunndiode horn like you’d find in an automatic door opener is not a VCO device, nor is a police radar horn, else kustom signals or decatur electronics would have made radar detectors obsolete by now!

      1. You’re right, Gunn diode oscillators are fundamentally fixed-frequency and cavity tuned, however the frequency can be pulled within a few percent and they’re often paired with a varactor diode to create a VCO. Examples can be found here:
        Yes, those prices will cause you to spit out your drink. They can often be found more cheaply on the surplus market, example here:
        Datasheet here:
        (note the date, 4/77!!)

        The design that I did in college used a 24 GHz Gunn transceiver with a horn antenna, and it worked pretty well. We did characterize the tuning voltage characteristic and synthesized a voltage waveform to linearize the sweep, and that helped a lot.

    2. Feedhorns used with parabolic reflector antennas are nothing more than a can with an RF probe. They are somewhat directional, providing a 30 degree wide beam (more or less, depending on feedhorn geometry). It’s good enough for this project, as the goal was to develop a radar system, not a radar antenna.

    3. Hmm, I said MIT students but it’s inspired by the MIT thing, they are UC davis students though, so I guess that’s why they have to make do with a toaster oven they hack themselves.

      And there is nothing wrong with learning by trying some projects incidentally, I do think I came a cross a bit too derisive maybe. In fact once they have this working it would be a nice starter to see what else works as antenna and compare the results, since you can easily replace the cans.

  4. If I duct tape the 3 Hot Wheels Radar Guns I’ve bought (at 2nd hand stores for less than $5 each) to a board, and hook their battery leads to an external power supply, will I get a BSEE from UC Davis too?

    1. Well, if you cook your soup in the cantenna first, they may at least give you credits for thermodynamics, I’ll ask the guy I know that got his PhD from UC Davis and report back.

      In all seriousness, most of the young engineers coming out of school that I know these days are woefully unprepared for real world electrical engineering. No practical knowledge at all, and yes kids, sometimes you do have to actually know the polarity of something before you hook it up to a power supply… And you have to know what polarity means in regards to electrical current, and if you come at me with conventional theory flow i will slap the teeth out of your face…

      1. Better yet is those that don’t know the difference between a safety/earth ground and a common ground. Guy at work (self-proclaimed nerd of course) announces the other day that he discovered some “idiots” error in connecting one side of a car accessory to the frame of his car, rather than a direct line back to the battery. Facepalm wasn’t even enough, I had to break out the facedesk. It’s why I describe myself a tinkerer rather than an EE student, or even a nerd. It gives me wiggle room when I make silly errors myself.

          1. I recall it being a tail light, I’m thinking a brake light? I was under the impression that the majority of electrical connections in a car involved a hot connection to the battery/switch and the other side to the frame.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.