0
$\begingroup$

I was wondering how I would justify a sampling rate of 125 Hz for accelerometer and gyroscope data from wearable sensors. This is a rate used in a lot of biomechanics literature, but I can't seem to find a justification.

I have been looking up the Nyquist rate, but I'm not sure how to find the bandwidth of the IMU signal.

Any help in this justification would be great

$\endgroup$
  • $\begingroup$ It's probably "high enough" and "commonly used". Don't expect a justification. $\endgroup$ – Marcus Müller Sep 2 at 15:19
  • $\begingroup$ It depends on your mode of operation. If it is bluetooth, 125 Hz is a reasonable sampling rate. However, if it is wired, it can go all the way up to 1000 Hz. $\endgroup$ – Maxtron Sep 2 at 16:08
  • $\begingroup$ It is a wireless bluetooth sensor, I am just wondering because I might have to justify it for an exam $\endgroup$ – Eric Sep 2 at 16:20
  • $\begingroup$ @Eric It should be easy to compute right. You have six channels (three-axis accelerometer and three-axis gyroscope). For each channel, you have the header, data, and check sum information. Compute the number of bits per second and multiply with six. Then, show that the total rate is less than bluetooth max capacity. I hope this helps. $\endgroup$ – Maxtron Sep 2 at 17:09
  • $\begingroup$ @Maxtron seeing that bluetooth can easily do Mb/s of data, there's no direct causation between wireless or not and sampling rate, I'd say. The amount of data that even a 6-axis 8bit device would do, 48 kB/s=284 kb/s, is really tiny for modern wireless links. $\endgroup$ – Marcus Müller Sep 2 at 17:32
1
$\begingroup$

I'm not sure I can state anything better than Marcus's or Maxtron's comments (it's that way because it's always been that way / because the hardware is limited to that rate), but another way to think about it is:

  • The Nyquist rate is 62.5 (=125/2) Hz.
  • Can your acceleration or rotational angles (the things measured by the accelerometer and gyroscope) change at a frequency faster than that?

If your instrumented system cannot, then this sampling rate is OK. If your system can move faster, then the sampling will miss something.

$\endgroup$
1
$\begingroup$

One random thought, the 125-Hz sampling frequency allows the use of notch filters to cancel 50-Hz and 60-Hz periodic powerline noise ? It would be possible to cancel this periodic noise in hardware but it would cost more, it would be harder to integrate in wearables (large capacitor values), etc. ...

$\endgroup$

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.