Make Your Own TSA “Naked” Scanner

Have you ever wanted to ability to see through objects? Perhaps you have been looking for something special for your own personal TSA role playing adventures? Well, [Jeri Ellsworth] has your back. She has managed to cobble together her own millimeter centimeter wave scanner using a hacked set of Feed Horns (like from a satellite dish) to create the image. By reversing the power transistor on one of the Feed Horns, one of the horns is made into a transmitter, while one of the other horns stays as a receiver. This data is then fed into a FPGA by way of an A2D converter, where an image is assembled when the scanner is moved over a surface. X and Y axis tracking is handled by an optical mouse also controlled by the FPGA, and the whole setup is output to a monitor.

Right now there is no text write up, or any specific details as the hack will vary by whatever Feed Horn is available. However, the video does a great job of explaining some of the electrical concepts, as well as some very useful schematics. Be sure to watch the whole video after the break, and don’t blame us for any health complications, whether the radiation is ionizing or not.

[youtube =

82 thoughts on “Make Your Own TSA “Naked” Scanner

  1. i must say that Jeri Ellsworth is the most intelligent hacker i’ve seen on HaD, with the most informative videos. i will also say that i don’t know shit about transistors, what they really do.. i’ve probably put about 3 of them in a circuit before, but i was following instructions.

  2. Very impressive hack!

    What’s interesting is the RF stage is designed to process Doppler shift, but the scanner moves laterally across the surface. Where does the Doppler component come from?

    Are you getting getting a baseline reading where no target is present or is there no frequency present at all?

    Ideally if there is no target reflection then the only signal coming from the mixer would be slight L.O. leakage which the amp/adc could not resolve. I’m wondering if the material being penetrated slows the RF enough to cause a shift…

  3. I believe the technical name for one of those MM wave scanners the TSA uses is “pornotron” as in “Please either step into the pornotron or you’ll be subjected to a standard TSA gate rape.”

  4. “im not an alien, i have more than three fingers”

    I think female hackers are pretty alien in this carbon based society. too bad we don’t thrive in arsenic rich environments.


    What is the method you use for displaying the signal? us script kiddies are curious.

  5. The only person I’ve seen to touch Jeri was the Brit who built a ruling engine that trumped all the major players back 40+ years ago. Instead of trying to make a perfect screw, he used an interferometer to regulate an imperfect screw. Because his errors were random, there were no Rowland’s ghosts. A few years earlier Scientific American had a series on ruling engines which made much of amateurs failing and destroying their lives in the process.

    Way to go Jeri!

  6. There is a certain change in social and behavioral interaction patterns that can occur when hacker types start to gain a large audience, be it paying clients or the blogosphere.

    One discovers that notoriety of speech and opinion tends to increase your audience, especially if your attitudes, opinions or abilities are “interesting” to a large segment of that audience… or even just interesting to those individuals in charge of discretionary spending.

    That means that the more “out there” elements JE adds to her videos, the larger her audience becomes. I know a number of people who have found interesting niche consulting gigs based on striking a sympathetic note with patrons of the arts – even when those arts happen to be AI, mathematics or biology.


  7. Hmmmm…I have a set of feed horns in the junk pile from my Wifi dish project a while back. Not that I could come anywhere near this hack, but hey, a guy can dream.

    And since everyone seems to tiptoe around it on Jeri posts, I’ll say it:

    A smart, creative, articulate woman who also happens to be quite attractive: Proof that there is a God.

  8. @Jeri – Just curious, why did you do the video generation and everything in an FPGA, rather than just feeding it into a computer?

    I’m always surprised to find another person that prefers writing RTL to just programming a CPU.

  9. Jeri, when you build your starship will you please let me visit it before leaving Earth? Thanks!

    BTW It would also be interesting to use those modified feed horns to communicate. Basic frequency modulation of the reversed transistor shouldn’t be that hard to obtain unless one doesn’t aim at very precise deviation and/or frequency stability. I believe sending voice and low to medium data rate should be possible.

    1. Radio amateurs have already been doing that for 20-30 years using re-purposed satellite equipment. I am astounded that people are so amazed by this “hack”…

  10. Did you try to pick the RF LO to feed the transmitter after the Wilkinson power divider instead directly from the buffer LO output so that the receiver downconverter LO fed will be less disturbed?

    May be that with a proper LO feed to the TX the distance range (or S/N ratio) could be incresased.
    (The printed filters on the PCB could be removed (soldering iron+razor blade) and a microstrip line (super)glued all along the way)

    I think that circular polarization helps in this case. The bounced signal will reverse the polarization while if it were linear the losses probably will be much higher.

    This video remembered me your magnetic field “magic wand” scanner from the old happy days. Thanks Jeri.

  11. Can’t watch the video, it has freaking ads on it, I don’t watch crappy ads, to hell with the new google who refuses to advertise on torrentsites and changes searchresults for the mediamafia and then puts ads on youtubes.

  12. heh, this is really cool.

    certainly should be useful for us HV’ers for examining flybacks, triplers and suchlike (helpfully plastic casing) to find out where to drill to series connect them..

    also useful if you need to detect hidden transmitters etc in wood or plastic ornaments..

  13. Fluidic – everything happens down one feedhorn. The other two are unused.

    S – I’m a microwave novice, so I’m sure there are many ways to improve this. I could see reflections from this 5-6ft away in open air.

    me – I’m in a down time right now. With a little luck I’ll continue my contract after the holidays, although I still devote a lot of hack time during work. I’m a nerd with no life.

  14. After staring for over an hour, I still don’t see any brass knuckles in that picture ;)

    Wonderful hack. What about using the center feedhorn as the transmitter, and the two outer ones as receivers? I realize my feeble attempt at “stereoscopic vision,” for this will be met with, “won’t someone think of the low resolution,” but wouldn’t you interpolate some based on the increase in information available (see dimensionality)? Okay, I guess I’ll try it (I happen to have an unused feedhorn here).

  15. Jeri, you are awesome. And it is getting rare to have people (esp. young and female) who knows there analog stuff these days.

    I like how you using a mouse as tracking and a FPGA as the ‘brain’.

    BTW, I have one of those razor scanner too.. ;)

    Good job, well done.

  16. @Jeri, Excellent build.. I’m considering a couple of ideas that may replace the mouse but I’m gonna have to think about it for a while before I know if they’re feasible.. Regardless, this was very well put together.

  17. I like the way an optical mouse tracks the movement. Are the optics modified to give an infinite focus or is there an extra lens or two in there somewhere? I think if there is a problem with it drifting, it’s possible to read the camera directly over PS2 and do a better tracking algorithm yourself.

    Nice knuckle duster too :-) The first thing I did after CNCing my mill was to carve one of those from a bit of scrap steel plate.

  18. A few years ago when I was at school I did some work experience with a company called QinetiQ in my home town Malvern. I got to spend one of the days in a lab were they where developing mm wave imaging, from what I can remember they had just one sensor (or whatever it is) and the waves were reflected off tinfoil inside a large foam rotating disk (over a meter in diameter) so that an image could be made up one pixel at a time. I think I the researchers name was Rupert. This must be going back to 2001. I returned in 2003 and got to see the machine again, this time it was so much more refined, and now its being used for real, awesome. I have really fond memory working with QinetiQ.

  19. nes – Currently it needs to be placed on a flat surface. I scanned my leg with a pair of pliers laying on them. I used a piece of thin rigid plastic for the surface. I could tell what was going on, but it wasn’t a good demo for the video.

    I had considered several options before choosing the mouse. My first plan was to make an x/y table, but I didn’t think I had strong enough steppers. I almost built this with potentiometers on arms to register the location.

  20. For majolsurf,
    The Doppler component actually comes from the lateral sweep. It helps to think of the scanner as being a stationary 2-dimensional object.

    Doppler shift measures velocity, either toward or away from the transmitter/scanner.

    When you sweep the thing back and forth, that is indistinguishable from the object being scanned moving back and forth.

    What the doppler shift measures in this case is the velocity of movement found in the contours of the object being scanned.

    Imagine it like if you’re holding the scanner stationary, and the object is 12 inches away, then you start to sweep back and forth, and curves and contours in the object being scanned are at different (slightly) distances than 12 inches.

    This will register as a doppler change in velocity.

    1. “Radiation hormesis is the hypothesis that low doses of ionizing radiation…”

      This is microwaves – not ionising, therefore the link is not relevant (non-thermal effects may still be an issue, but not ionising effects)

  21. Airport ones are also backscatter, but have better amp bus and use wider physical “tunnels” so they don’t need tracking. Also the DAC data is scaled to higher resolution for clarity.

    The other type that does 3D aren’t used anywhere except maybe NASA and government forensics.

    This can be made better even cheaper.

  22. cf – I missed your question last time I stopped in. I used an FPGA, because I could prototype this in a much shorter time. I’m a chip designer by day, so working with PC’s and all the overhead of OS’s and non-realtime annoys me. This could be done with just the instrumentation amp and a sound card audio input.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

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