# Micro Pressure Sensors for arduino?

I'll preface this with a little background. I'm a software engineer with little experience outside of a couple small arduino projects. I'm excited about the idea of combining both digital and physical components with frameworks/hardware such as arduino.

My question is: Are there any small/micro pressure sensors compatible with arduino? The idea would be to detect balance/distribution abnormalities in packaging.

Example: A grid comprised of 100x100 (the more the better) sensors used to recognize the weight distribution of a box or container (increased pressure on some sensors vs decreased pressure on others). Container dimensions are roughly 2' x 2' x 4' [lwh] and must remain upright. The height of the box is what makes the weight distribution important for stability.

Is this possible with arduino? If so, what hardware would I need? If not, is there something else I should be using?

Thanks all!

Somehow I feel that physics (especially mechanics) might hold a better answer than just throwing a matrix of sensors, which as @clabacchio says, is going to cost a nice little fortune.

Based on the material that the box (especially it's base, the bottom 2'x2' surface), you can choose various approaches as well.

What I mean is, things like -

• Checking tension in a grid of supporting wires that take the load of the box when placed on the grid
• Checking pressure on a grid of metal plates (i.e. you need not have an individual sensor covering each inch of the base)

Finally, "optics" could be a solution. There's a funky lot you can do with lasers (beam obstruction), or even with a good camera and some image processing. Of course, the last one (image processing) is not really a good option on Arduino.

Perhaps you could make use of Photoelasticity? That gets you the measurements with a fancy plate of plastic and a camera. Then it's all a software problem.

Pressure sensors exist, and also in MEMS technology, that means that the average dimension is 5x5x0.5 mm or less. But maybe (depending on the weight of the boxes) you may need strain gauges, and then the thing becomes complicated, as you need a bridge circuit and a very good differential amplification system.

Even using pressure sensors, each pressure sensor costs usually around 1$(roughly), and building a 100x100grid would mean 10k sensors (and some thousand$). Do you really need this resolution? If the package is solid, you may need only 4 sensors at the corners, or maybe a little more, but still less than 10k.

Another consideration: MEMS sensors are mainly for PCB mounting, so you'd have to build a structure with these sensors soldered, and wiring a good interface depending on which one the sensor uses.

• also 10k pins worth of analog to digital converters, plus associated circuitry to convey all their answers to one spot so they can be processed. – Jay Kominek Feb 1 '12 at 4:17
• @JayKominek yes, but I was also considering the use of MEMS sensors with built-in converter and - say - I2C interface – clabacchio Feb 1 '12 at 8:00
• sorry, i'm not trying to suggest anything was wrong here, just pointing out that having 10k sensors will be an electrical hassle in and of itself. (will bus capacitance become a problem when you're covering a 2'x2' area with a grid of wire?) – Jay Kominek Feb 2 '12 at 2:19
• @JayKominek probably, there are many reasons to avoid a 100x100 grid, but Justin seems not caring too much about this question – clabacchio Feb 2 '12 at 7:12