I'm looking for a way of measuring angular velocity in a spinning object at high speeds - up to 25,000 degrees per second. There's ridiculously expensive gyros in the market that can do this, but I'm hoping to find a way of using off the shelf, affordable sensors to do this.

2x thoughts come to mind

1 - using a magnetometer recording at a high frequency to record each time the sensor passes magnetic north. 2 - Using accelerometers to measure the centrifugal force of the spinning object.

I researched the concept of using 2x accelerometers which appears to for slower movements (less than 360dps).

I'm wondering if anyone else has been playing around with something similar or can point me in the right direction as to how we might be able to capture this data.

  • 1
    \$\begingroup\$ What is the spinning object in question? What are you ultimately trying to accomplish? \$\endgroup\$ – Nick Alexeev Jan 22 '18 at 1:15
  • \$\begingroup\$ So for clarity, you are talking about 3k - 10k RPM? \$\endgroup\$ – Harry Svensson Jan 22 '18 at 1:15
  • 2
    \$\begingroup\$ 25,000 degrees per second is about 4,100 RPM or about 68 rotations per second. Your title does not agree with that whether interpreted as rotations per second or radians per second. You're not planning on sticking this in a baseball or something like that are you? \$\endgroup\$ – Spehro Pefhany Jan 22 '18 at 1:29
  • \$\begingroup\$ Average angular speed is easy. What time-resolution do you require? Any motor tachometer, or stroboscope, can do the measurement if rotation is constant. \$\endgroup\$ – Whit3rd Jan 22 '18 at 1:53
  • \$\begingroup\$ \at speed you may not need a full magnetometer, a simple pickup loop should give an AC signal due to the ambient magnetic field. \$\endgroup\$ – Jasen Jan 22 '18 at 9:27

The big question you need to answer is: what update rate do you need and what resolution do you need? OK, that's two questions, but you get the point.

Resolution. Do you need to distinguish between 25,000 deg/sec and 24,999 deg/sec? How about between 25,000 and 24,900? More? Less?

Update Rate. Is a sample every millisecond OK? Do you need one every microsecond?

For modest accuracy and update rates, optical sensing will work just fine. 25,000 deg/sec is only 4,000 rpm, and automobile tachometers do that just fine. Of course, resolution isn't that great, but you can do better without much effort using optical sensing.

Take (for instance) an encoder wheel with 4 quadrants, alternating between white and black (for reflective sensors) or clear and opaque (for transmissive). If you use an LED/photodiode combination you ought to be able to get 1 usec time resolution without horrendous difficulty. Then one quadrant will take 3600 usec, and resolution will be about 7 degrees/second and update rate will be 278 updates/sec.

In principle you could use a finer pitch on the encoder and faster sensing electronics to get better resolution and/or better update rate. Or you can use a phase locked loop to do the same thing electronically, although there are possible error sources in that approach which need to be carefully considered.


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.