I want to create a new type of joystick, I have design idea I have been thinking.
I want to make use of a ball & socket joint. At first I thought to use entire neo-magnet as the ball, but that is way too complicated to mill & shape a magnet. So I will just glue a neo magnet to the bottom ball in steel ball mounted inside a socket joint.
So, I need a hall effect sensor to pick up the distance of this magnet mounted to the end of a ball in 2-axis. The distance is probably around 1 inch diameter.
Does anyone have experience or know about these hall sensors, can give advice?
Bet you're trying to make a contact-less joystick, problem I see with hall effect sensor is the field strength from the magnet varies with the inverse square root of distance (I think ), if the distance is doubled, field strength will be 1/4, also, the magnet would not be moving in a straight line, but swinging an arc. The math could get very hairy very quickly.
Optical is already contactless and immensely more reliable (no dissipation of resolution due to distance relationship), which is why it's used.
The ball and socket idea is aesthetic, really, as you gain nothing from 99%+ of the structure.
I can run over steel ball socket with a motorcycle and continue to use that. If i accidentally drop a plastic gimbals on the ground (which always lands stick-down for reasons unknown) -that's mostly guaranteed to break, with delicate sensor also taking a hit as well. Ball and sockets have least moving parts.
I'm going to ask the tech support people at Allergo, if a sensor exists - they will know about it. I am also told that magnets get picky and fussy, sensors lose calibration. I want a reliable sensor - I thought magnet would be a good idea, but I do not have much experience working with the sensors outside of using them for a temporary RPM gauge.
I agree with optical sensor, I thought about it for a while. Adopt the solution from an optical track-ball mouse. If I etch or print a pattern into underside of ball, use sensor to measure travel distance. I have to do this in such a way that a wear-pattern, dust/dirt can't destroy the pattern. The sensor will always be absolute calibrated 0,0 to whatever the stick's position is at the init/power - on for the micro-controller. The travel distance would be absolute, as well.
Paul: Thats a good observation you've made. I do finish projects for other people, though they are not as complicated as my own. I usually find myself revising completed tasks from unexpected issues. For my lawn-robot, for example -is complete and works. Now I begin new project to modify, adapt to remove lawnmower and add snow-blower - that must be complete before winter.
DocStein99:
At first I thought to use entire neo-magnet as the ball, but that is way too complicated to mill & shape a magnet.
Well you'd have failed, NdFeB magnets are a brittle sintered ceramic than cannot be machined, possibly could be
grounds (but clearing the highly magnetic debris would be difficult)
A 2D magnetic rotary encoder would be a nice thing to have indeed...
MarkT:
Well you'd have failed, NdFeB magnets are a brittle sintered ceramic than cannot be machined, possibly could be
grounds (but clearing the highly magnetic debris would be difficult)
A 2D magnetic rotary encoder would be a nice thing to have indeed...
A machine that grinds rocks into spheres might do it. My barber from many years ago was a rock hound and bought a machine to do just that and output spheres of various sizes. I never saw it, but he showed me quite a bunch of spherical pretty rocks!
MarkT:
Well you'd have failed, NdFeB magnets are a brittle sintered ceramic than cannot be machined, possibly could be
grounds (but clearing the highly magnetic debris would be difficult)
A 2D magnetic rotary encoder would be a nice thing to have indeed...
I would not machine a magnet of any type. I would machine the steel ball, insert a magnet of whatever shape needed. The ball joint is not bound to any axis, so it's actually possible to twist the ball as well - which would give it a 3rd axis for one more sensor to take a reading from.
Paul: They sell rock tumblers, is a motor with a cage that spins around in the abrasive fluid (and / or other rocks and sand) for hours, days, or whenever your patience runs out. Take out whatever you put in, and the surfaces will have glass finish. It's actually a fun machine since it's like a set & forget kind of thing if you have the willpower to not check it every 10 minutes.
), if the distance is doubled, field strength will be 1/4, also, the magnet would not be moving in a straight line, but swinging an arc. The math could get very hairy very quickly.