Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Static vs. dynamic acceleration with accelerometers

  1. Dec 20, 2011 #1
    In a project I'm working on at school I could only get static acceleration from the 3-axis accelerometer I have (MMA7361L). I used a simple set-up (4 pins: Tx, Rx, power and ground).

    As far as I understand, accelerometers are able to detect both types of acceleration but I don't know why I only get variations of data only when I tilt the sensor, not when I move itf from point A to point B. Can anybody point me in the right direction? what am I doing wrong? thanks a lot.
     
  2. jcsd
  3. Dec 20, 2011 #2

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    I guess it depends upon how much lateral acceleration you were giving it. OR, is there some 'stiction' in the accelerometer? By that I mean how sensitive is it to tilting? Does it respond to a tilt of just a few degrees?

    The principle of equivalence applies even at your School and the device should not be able to tell the difference. Have you checked that sensors for all three axes are working?
     
  4. Dec 20, 2011 #3
    Hi sophiecentaur, thank you for your reply. Well to be honest, I haven't tried moving the accelerometer too fast as the device (microcontroller, accelerometer and bunch of wires) is a bit delicate. However, I thought the accelerometer was able to get readings simply at "walking speed", this is what I need it for after all.

    I would say the sensor is quite sensitive, it responds to tilts of few degrees but you might be right when you suggest I should move it faster? I'll do that just to check, thanks.

    Oh, the three axes work fine.
     
  5. Dec 20, 2011 #4

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    OK, so each axis gives the same reading when it's vertical? i.e are all three accelerometers matched? When you tilt it on a slope, the vector sum (root of the sum of squares) of two should give you g, for instance. If that works then I think you can rely on the readings for acceleration. Remember, with 1g of acceleration (better than your average saloon car accelerating and about the max for normal brakes), a pendulum should tilt at 45 degrees. Perhaps you are being too gentle with it! Perhaps it just isn't sensitive enough for a good inertial navigation system.
     
  6. Dec 20, 2011 #5
    Throw it up in the air and catch it. The acceleration should be zero while it's flying. This tests both static and dynamic components (at zero that is.)
     
  7. Jan 3, 2012 #6
    Hey guys,

    Sorry for the short break, I'm on a trip and I don't have the device here with me but I'll continue with this discussion as soon as I can. Hopefully, you'll still be interested in helping me :)
     
  8. Jan 4, 2012 #7
    You mentioned Tx and Rx. So I assume your accelerometer does some preprocessing (amplification, filtering, ad-conversion, averaging ..) before it sends digital data to you. Do you know what exactly it does?
     
  9. Jan 4, 2012 #8

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    Good questions. An accelerometer with averaging / low pass filtering wouldn't be good for double integrating acceleration to give you distance because it would lose fast changes in speed.
     
  10. Jan 4, 2012 #9
    Tilting of course produces acceleration on x y z axes from Physics point of view.

    Did you accelerate the sensor when you were moving it laterally (in a straight line)?

    Are you sure what you have is MMA7361L? With the way you described it, I'm thinking maybe what you have is a gyro??
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook