How-to measure rpm from inside a ball

I’m looking for a way to measuring rpm from within a ball.
The ball will be rotating anywhere from 200 - 700 rpm and the accuracy needs to be fairly high, preferably < +/- 10. Is there any sensor that can do this that’s small enough to be placed inside an object the size of a football?

Will the ball be in freefall or rolling on a flat surface or what?

You can get modules with electronic gyros and accelerometers, take readings a few times per millisecond and still have time to process (avoiding floats, AVR has no FPU) the data.

Inside of a spin there's an outward force.

In freefall you can't detect gravity from inside while rolling on a flat you can.

It'll be rolling on a flat surface (bowling lane).

Two other data points that would be really interesting is axis tilt and rotation. In the swing the ball will (at least in the last few ms) traveling towards the "target" down lane, using that path as a relative center could I somehow determine axis rotation and axis tilt while the ball is rotating?

Could I also calculate it speed?

zyberse:
It’ll be rolling on a flat surface (bowling lane).

That suggests to me that you could get all the data you need from a camera looking down on the bowling lane and image processing software on a PC.

…R

That means a permanent installation that I would like to avoid and also, rpm, axis tilt & rotation would most likely be very difficult to see as bowling balls come with lots of different patterns and colors.

Measuring the path can and have already been done with lidar, works great but is expensive and requires a permanent installation.

I would like to find a simple, low cost solution and I thought it could be done with a gyro and accelerometer?

“I would like to find a simple, low cost solution and I thought it could be done with a gyro and accelerometer?”

Good, fast, cheap, chose any two. The more you study your project, the more complex you will find it to be. You might go with the external video option and use a done as the video platform if you can’t use a parallel lane with a camera(s) stationed there.

zyberse: I would like to find a simple, low cost solution and I thought it could be done with a gyro and accelerometer?

I was not aware you could open a bowling ball to insert something in it. Even if you can, won't it upset the balance and behaviour of the ball?

Maybe use a drone to carry the camera ?

...R

I thought it could be done with a gyro and accelerometer?

Possibly RPM, with the gyro, but none of the other items on your list.

Hi, You need to realise that what ever you use in the ball to measure rotation, the rotation will not be consistently about ONE axis. In fact the bowling ball does not roll in the sense of a wheel, it actually TUMBLES then assumes a basic rolling axis. The position of the axis in the ball is not predictable.

What data do you want from the rolling ball?

Tom... :) PS, The motion is even more complex if you have an experienced bowler launching it to make strikes. (spin).

Hi, You could investigate how the satellite and space engineers get motion data from satellites.

Tom... :)

Accelerometers placed off-center could tell a lot, the direction down is part of every reading rolling on the floor. They get centripetal, gravity and every wobble and you only gotta separate them.

How do you plan to recharge the device? To get data back out?

A black ball with 12 equally-spaced white dots could tell a lot in a picture. Every dot and 2 adjacent actually are points of an equilateral triangle, 12 vertexes and 20 faces.

jremington: Possibly RPM, with the gyro, but none of the other items on your list.

Use three gyros. The math to calculate everything is way beyond my capability, but the relative motions of three gyros would give you all you need.

TomGeorge: Hi, You need to realise that what ever you use in the ball to measure rotation, the rotation will not be consistently about ONE axis. In fact the bowling ball does not roll in the sense of a wheel, it actually TUMBLES then assumes a basic rolling axis. The position of the axis in the ball is not predictable.

What data do you want from the rolling ball?

Tom... :) PS, The motion is even more complex if you have an experienced bowler launching it to make strikes. (spin).

I was going to mention the same action based on my highschool boiling class, so many years ago. ALSO, anything you do to change the complex density of the ball will effect the spinning/rolling axes.

Paul

The biggest problem with installing any sensors inside a ball will be debugging the project. I reckon it would be necessary to make a wire-frame ball for the testing process so you could visually identify the orientation of the sensors to determine if the data makes sense. Also you will need ready access to modify the code.

Use a camera.

...R