Specs for a magnet to trigger a magnetic reed switch?

[Yooniv]

Hello, for a project I'm working on i need a magnet that is attached to a spinning wheel to trigger a magnetic reed switch that's about a cm away every time that the wheel makes a rotation.
I don't know which parts to buy because if the magnet is too weak or the switch too hard to trigger it won't work, and the magnet has to be very small to fit on the wheel. I don't understand the specs of various magnets and reed switches being sold online, so i would appreciate it if someone who has a decent understanding of the topic could tell me what kind of magnet and reed switch i would need.
Thanks!

 

[berkeman]

Hello, for a project I'm working on i need a magnet that is attached to a spinning wheel to trigger a magnetic reed switch that's about a cm away every time that the wheel makes a rotation.
I don't know which parts to buy because if the magnet is too weak or the switch too hard to trigger it won't work, and the magnet has to be very small to fit on the wheel. I don't understand the specs of various magnets and reed switches being sold online, so i would appreciate it if someone who has a decent understanding of the topic could tell me what kind of magnet and reed switch i would need.
Thanks!

Welcome to the PF.

a 1cm separation is pretty far for triggering a reed switch with a small magnet. Is there any chance you can decrease that distance somehow? Look at typical bicycle speedometer setups -- the separation between the magnet on the spoke and the pickup on the downtube is only a couple mm. Is that the size magnet you want to use? What's the biggest magnet you could accommodate?

A better approach, especially if you need the 1cm separation, is to use a Hall Effect sensor to detect the changing magnetic field as the magnet sweeps by. Hall effect sensors require power though -- what sorts of power supply voltages do you have available? What happens when the reed switch or Hall device triggers? Are you just counting revolutions?

BTW, could you just use a light sensor and retro-reflector instead of a reed switch and magnet?

 

[Yooniv]

Welcome to the PF.

a 1cm separation is pretty far for triggering a reed switch with a small magnet. Is there any chance you can decrease that distance somehow? Look at typical bicycle speedometer setups -- the separation between the magnet on the spoke and the pickup on the downtube is only a couple mm. Is that the size magnet you want to use? What's the biggest magnet you could accommodate?

A better approach, especially if you need the 1cm separation, is to use a Hall Effect sensor to detect the changing magnetic field as the magnet sweeps by. Hall effect sensors require power though -- what sorts of power supply voltages do you have available? What happens when the reed switch or Hall device triggers? Are you just counting revolutions?

BTW, could you just use a light sensor and retro-reflector instead of a reed switch and magnet?

Thanks for the warm welcome! And yes, I do think i could decrease that distance. It would be hard to build, but i guess i could get it so the magnet and the switch are almost touching. I'm trying to fit the magnet inside the rim of the wheel, which i measured out to be exactly 3mm deep, but it could stick out maybe a couple more mm without hitting anything (maybe that would even be good to get it closer to the switch). And it would have to be flat, like a disc or something so i could glue it on, and the flat side couldn't be more than a cm on each side.

So i guess it would have to be about as small as 3mm x 10mm x 10mm or 300 cubic mm(235 cubic mm if it's a disc), so maybe we could calculate a magnetic force or something with that? Sorry i don't know anything about magnetism XD. I looked into your Hall effect sensor suggestion and it sounds like a great idea! I actually looked into reed switches more and now i understand that it would take magnetic field parallel to the switch, which would be harder to set up than a perpendicular magnetic field which works with hall effect sensors. But do you know if that size magnet at a close distance would be good enough to trigger a hall effect sensor? And I have a 5 volt power supply, so will that be enough? And would the magnet work with a reed switch? The advantage i think of a reed switch i think is that it would be easier to wire, but not by much so I think even if that will work i think I'll probably go with the hall sensor idea that you suggested :)

Unless of course, I use the light sensor ;). I'm not totally sure I know how that would work though. So I would have the light sensor facing where the reflector is going to be when the wheel comes around, right? But how would the reflector light up? Would I have a light above the sensor that would shine on the reflector whenever it came around? If that is how to do it, then yes, I think I could probably set that up, but I think I would still prefer the hall effect switch method though because It would be simpler (two parts instead of three if I understood the light sensor right) and also from what I've seen of light sensors online most of them would be pretty awkward to hook up to my project.

Oh and also yes, I'm just counting revolutions. End goal is linear distance, so I would just have the hall effect sensor send a signal to my microcontroller every time it turns on. I might have it send a signal multiple times a revolution (multiple magnets) , to reduce error from what position the wheel starts in (like we might get an off by almost one revolution error if a signal is only sent once per revolution)
Thank you!

 

[berkeman]

Ah, so yes, you should do something slightly different to cut down on errors. I'd recommend going with the optical reflector idea, and putting stripes all around the rim of your wheel. That way you will have much better resolution on the rotation of the wheel, and your uC will be able to know with much better granularity what the change in wheel position is.

Use a combined emitter/detector unit similar to this (google optical emitter detector reflective):

http://robotgestation.com/images/OpticalEncoderReflective.jpg

 

[Yooniv]

Ah, so yes, you should do something slightly different to cut down on errors. I'd recommend going with the optical reflector idea, and putting stripes all around the rim of your wheel. That way you will have much better resolution on the rotation of the wheel, and your uC will be able to know with much better granularity what the change in wheel position is.

Use a combined emitter/detector unit similar to this (google optical emitter detector reflective):

http://robotgestation.com/images/OpticalEncoderReflective.jpg

Ok, that sounds like a great idea. Thank you!