Notifications
Clear all
Project Corner
12
Posts
3
Users
0
Likes
947
Views
Topic starter
2021-08-28 11:36 pm
Hello! First post. Excited to be here.
I'm designing / printing a ball that acts like an aircraft ADI, but in 3 dimensions instead of two. The mechanical design is done and I'm working on control programming.
Each axis has its own stepper motor. I can obviously count steps on the stepper to know where I am, relative to where I started, but I'm having a hard time figuring out how to make sure it goes to the zero position when I power up. Or rather, how do I sense the zero position?
I think a hall effect sensor and magnet for each axis would work, right? What specific hall effect sensors would you recommend? The ones I've found want me to buy a few thousand.
Is there another method? I considered having two wires touch when they are in the right position, but that seems hard and unreliable.
Is there a vision based solution? The exterior of the ball will have markings on it, could I use a camera to see the position?
2021-08-29 12:28 am
Welcome to the forum.
How do you rotate the ball with the steppers ? is it free to turn more than 360 degrees in any direction ?
I'm wondering if you might be able to use a shiny dot on the sphere and an IR proximity sensor.
Experience is what you get when you don't get what you want.
Topic starter
2021-08-29 1:16 am
Yes, all 3 axes can rotate continuously, not limited to one revolution or anything.
I'd need three dots for IR prox, which is fine. What sensor would you use?
2021-08-29 1:47 am
I'm having trouble visualizing the driving mechanism. You'd need to be able to rotate the sphere along only one axis at a time find a magnet or IR reflector.
How do you have 3 axes of rotation ? I can see one for east/west and another for north/south ... what's the third ?
Experience is what you get when you don't get what you want.
Topic starter
2021-08-29 3:02 am
Yeah I can drive one axis at a time. Think of a 3 axis gyro, that's how it's built, it's just mostly surrounded by a ball.
Which do you think would be more accurate? Magnet on each axis or ir reflector? Ball is 6" diameter.
The axes are roll/pitch/yaw. So a location on the sphere (that's two of the axes - pitch and yaw) and a specific roll at that location (that's the third).
2021-08-30 2:18 am
If the ball is a light colour, then the magnet is probably a better choice.
Experience is what you get when you don't get what you want.
Topic starter
2021-08-30 2:24 am
@will excellent, thanks. recommendation on a good hall sensor? I haven't been able to find ones that I can buy just a few of.
2021-08-30 2:31 am
They're about $14 CDN for 20 transistor pcs on Amazon ...
Or you can get 5 Hall effect modules for about the same amount from Amazon as well.
Experience is what you get when you don't get what you want.
Topic starter
2021-08-30 2:46 am
boom! thanks for the help. will report back.
2021-08-30 7:38 am
Generally a flag interrupting an optical sensor is the way most machines implement zeroing a stepper drive system. They are repeatable and precise.
2021-08-30 12:53 pm
Is a typical and randomly chosen example. We kinda used to buy them by the handful and never really bothered with stringent "equivalence" research. They pretty much have to work, provided the photo transistor is the correct polarity and you get the pinout correct.
You're challenge (assuming I have the right picture in my mind of what you are hoping to build) will be getting the mechanical side of things where you need them to be. That said you will find the "Common or Garden" optical sensor reliable, robust, free from interference / noise and not requiring finicky calibration.
Kind Regards
M