Imaging Magnetism With A Hall Effect Camera

[Peter Jansen] is the creator of the Open Source Tricorder. He built a very small device meant to measure everything, much like the palm-sized science gadget in Star Trek. [Peter] has built an MRI machine that fits on a desktop, and a CT scanner made out of laser-

cut plywood. Needless to say, [Peter] is all about sensing and imaging.

[Peter] is currently working on a new version of his pocket sized science tricorder, and he figured visualizing magnetic fields would be cool. This led to what can only be described as a camera for magnetism instead of light. It’s a device that senses magnetic fields in two directions to produce an image. It’s cool, and oddly, electronically simple at the same time.

Visualizing magnetic fields sounds weird, but it’s actually something we’ve seen before. Last year, [Ted Yapo] built a magnetic imager from a single magnetometer placed on the head of a 3D printer. The idea of this device was to map magnetic field strength and direction by scanning over the build platform of the printer in three dimensions. Yes, it will create an image of field lines coming out of a magnet, but it’s a very slow process.

Instead of using just one magnetic sensor, [Peter] is building a two-dimensional array of magnetic sensors. Basically, it’s just a 12×12 grid of Hall effect sensors wired up to a bunch of analog multiplexers. It’s a complicated bit of routing, but building the device really isn’t hard; all the parts are easily hand-solderable.

While this isn’t technically a camera as [Peter] would need box or lens for that, it is a fantastic way to visualize magnetic fields. [Peter] can visualize magnets on his laptop screen, with red representing a North pole and green representing the South pole. Apparently, transformers and motors look really, really cool, and this is a perfect proof of concept for the next revision of [Peter]’s tricorder. You can check out a video of this ‘camera’ in action below.

47 thoughts on “Imaging Magnetism With A Hall Effect Camera

    1. Magnetic fields obey superposition, which means that you can’t image in the same way that you can with regular light.

      Effectively, at any single point you can separate light coming from two directions because light interferes by phase in addition to superposition. From a single point you can get different magnitudes (of light) in any direction.

      Magnetism adds in superposition, so from any individual point there is only 1 value of magnitude and direction, no matter which direction you look in.

      To make a light imager you only need a single point, which is a pinhole camera.

      To make a magnetic imager, you need to poll several points, and then you can only tell the magnitude and direction at each point. This is sufficient to identify things that are close to and smaller than your imager, but as you get further away the superposition property blurs everything together.

      A nice project though, and he’s taking the right approach by measuring multiple points. If he can couple it with some sort of local-offset position sensor (perhaps the accelerometer in the tricorder) he can get fractional pixel resolution images.

      1. Was thinking about this, how one still would be able to focus something. Could having multiple layers of sensors work. With every layer further away the image would blur more and more. And out this info how it blurs, render a somewhat deblurred image?

    2. as PWalsh says, it’s not really possible without a sensing proxy. In theory (not claiming its easy nor relatively feasible) one could use the Zeeman effect. I know the Zeeman effect has been used to calculate magnetic fields near celestial objects. If a molecule can be identified with a pronounced Zeeman effect, it should be possible to envelope it in a planar cavity, constructed from a reflective (diffuse or alternatively retroreflective) plane on one side and a transparent plane on the other, such that the Zeeman effective medium can be imaged through the transparent plane. The disadvantage would be a lot of upfront feasibility research, and a lot of trial and error, the advantage would be that one could leverage the cheap high pixel densities in digital cameras, a million magnetic field sensors on a PCB would turn out expensive.

  1. That is darn cool. I would love to see a longer range (inverse square law, I guess?) combined with a flir-one-like setup with a ‘real’ camera in the middle so you could overlay the image of the object you’re probing… Neat!

  2. This has been done before – years ago. A typical application is to qualify permanent magnets embedded in hand-held cast-acrylic production-level quantity “pucks” that are used for turning on/off functions in embedded medical devices through-the-skin non-evasively (e.g., Heart Pacers). I’m not saying the Author’s design is not worth looking at. But do recognize this has been done before.

  3. I saw something similar at a science exhibition recently – it was a demo of detecting cracks in ferrous parts, by placing a magnet one one side and looking for anomalies in the resulting field.
    Not sure if it was a commercial device or research prototype but it was a similar array of hall or GMR sensors, with fancier processing of the display

      1. Thanks. What you are doing is awesome as well. Everyone here in the lab likes reading about projects of your caliber. Wish we could help but all of us have signed our souls away and replaced them with NDAs so all we can do is throw enigmatic hints and non-descriptive rants.

        1. Can one just purchase those bare spintronics MR heads in harddrives (or mounted in e..g a SOT-23 case) from somewhere (like Digikey) in low volumes? Or do you need the larger units that you guys used in your paper, with what (I’m guessing) is a low noise amplifier inside the black case? The idea of moving from uT to nT would open up a lot of measuring possibilities

  4. This is one of the coolest things I’ve seen in a long time.
    I would love to see this combined with RF sensors in order to visualize the interaction between the electrical and magnetic fields being radiated from an antenna.

    1. When the field initially leaves the antenna, the magnetic field lags 90 deg behind the electric field. As the field reaches the far-field region this difference increases to 180.

      Why; I haven’t a clue beyond my intuition, which blames everything on entropy. Your proposal is intriguing.

  5. Could an array like this be baked into a single slice of silicon (IC)? Would love to see resolution and “pixel” (magxel?) density increased!

    Ah and call me old-fashioned but I always get shivers seeing rare-earth magnets waved close to a computing device. Yeah I know, no spinning magnetic platters anymores…

    Disclaimer: I’m a software guy… ;)

  6. You guys think you could implement some sort of comment quality system? I keep coming down here for insight and instead find pedantry, Hacker-dick measuring contests, and outright trolling.

  7. I wonder if a magnetic lens like the ones in an electron microscope could be used to help shape the field for greater resolution. It might require some calibration to eliminate the interference from the lens so that the only fields sensed would be from the object source.

    1. As someone else mentioned, light radiates, magnetic fields don’t. Electrons don’t radiate but they travel. I don’t think you can sense a magnetic field unless you’re actually in the field itself, or send something like an electron through it.

      Still maybe there’s a possibility, send a load of electrons from a point to a flat detection plate. Or perhaps better, from a point, to a point, but via an intermediate point in the field you’re measuring. I guess that would get complicated really quickly though, trying to steer an electron beam, then having an unknown magnetic field steer it as well, and detect where.

      1. (Very) theoretically you could sense deformations of the Earths magnetic field and thus view whatever you’re trying to view without being in it’s magnetic filed…

      2. Magnetic fields can have quite the range though, but drown out obviously in the ever weakening signal vs other signals, but if you can somehow shield them effectively to only detect from a very narrow direction you could I guess image from a direction and get some modest range improvement.
        How do you do that though? Per pixel shielding, I have no idea myself if and how that could be done.

  8. A less expensive detection method would be to use a magnetic field viewer film, available on ebay. Next to that Ferromagnetic fluid in a flat transparent case is best. For a computer to recognize the specific shape of that field would require a camera and some vision software. Expanding on your current approach, if you used a color capable dot matrix display to represent field strengths, and having the dots in the same overlaid geometry as the hall sensors, then you could create a simple visual compact field detector. I could easily see it succeding as a test device in industries that have machinery using magnetic fields as sensors and detectors, and using it to detect weak fields would be key. Good luck with your project!

    1. But don’t the particles in ferromagnetic fluid itself become magnetic and thus distort the result? They also stick to each other rather than just reacting to the external field right? And because of that are also slowing the response time and create ‘blur’ when an object moves across it. I guess you might be able to create a static lattice of small transparent bubbles with the fluid in it though, to alleviate the problem a bit.

  9. This is very cool, and I hope others with the talent to build this stuff create higher density arrays, ones with faster framerates, and push the sensitivity of the sensors (maybe these could lead to seeing the radiation patterns of antennas). In the short term, I hope we get videos of motors, transformers, speakers, and other interesting stuff.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s