0
\$\begingroup\$

In hobby-grade MEMS IMUs and MARGS such as MPU-6050, you can change certain registers on the chip and modify the sensor's range. For example 2 g or 8 g for an accelerometer. Of course, because the ADC resolution doesn't change, you get lower maximum accuracy for higher ranges.

What exactly changes when you modify the range registers? Obviously the mechanical and physical structure of the sensor can't change, but I don't know what enables it to measure higher or lower values.

\$\endgroup\$
1
  • \$\begingroup\$ Hopefully they aren't just boosting the gain of the internal DAC. Preferably they would increase the drive amplitude so that the sensor is more sensitive but saturates more easily. \$\endgroup\$ – DKNguyen May 21 '20 at 19:56
0
\$\begingroup\$

There's very different ways of building a MEMS accelerometer, but the MPU-6050 uses a proof mass method:

Imagine a plate capacitor, with one plate held into place using springs (~):

    ~~~~~~|~~~~~~~
          |   |
- --------|   |------ +
          |   |
    ~~~~~~|~~~~~~~

as you can imagine, an acceleration in left or right direction shifts the spring-hung plate and thus changes the capacity of that capacitor.

You can have one of such mechanism per axis, and build a 3-axis accelerometer.

The adjustability in sensitivity can be achieved through different ways, for example:

  1. Adjusting the sensing range of the capacitance detector. Capacitance is usually measured by observing the frequency of an oscillator built using the capacitor. If one is to change e.g. the reactive component in that oscillator, then that changes the frequency, and higher frequencies are better to measure low capacitances, but make it hard to measure large ones accurately. You can even make the the springs part of the oscillating system: applying a voltage between plates of a capacitor very much applies an attracting or repelling force!
  2. actually adjust the springs: in some MEMS accelerometers, the mass is actually not standing still, but in a mechanical oscillation, caused by the springs being excited by Piezo crystal components, excited with an external frequency. As you might remember, "ideal" spring forces are proportional to elongation, and thus you can extract spring "pre-tension" through external acceleration from the oscillatory properties.
  3. simply have a different system for each sensitivity range, if that's easier than adjusting the physics of your one system.

So, be careful about

Obviously the mechanical and physical structure of the sensor can't change

since this is a micromechanical system.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.