Help creating a recessed ceiling mount

jonatan

Hi! I'm working on a presence detection sensor that I want to mount on the ceiling. I want this sensor to also be able to rotate a bit so I can adjust the detection zone. On this onshape file you will find one of my ideas so far, using magnets (keep in mind this is just a quick one to show what I'm thinking, sizes and alignment of parts might not be correct). I'm wondering if there's a better way to do this so I want to ask all of you!

Some of my needs are:
1. Should be able to 3D print
2. Should be able to snap in and out (the sensor part) so I can work on the sensor, and change components (that's why I'm using a magnet)
3. Should have some friction so it doesn't move easily

Quick video showing how it would look and move on the ceiling

martin_kopplow

I see how you want to snap the sensor housing into the ceiling recress. The axial placement of the magnets, though, isn't going to help you much: Neither with precise positioning, nor with the required friction. Try orienting the magnet radially. Use one magnet and one iron core per side (to eliminate sweet spot), or even think the term 'axis' much much bigger and make a cylindrical face on the upper side of the sensor housing that slides in the recess part. Large diameters: That's how you get apprpriate friction. Put some magnets and a bent iron strip in these surfaces. Then, everything else will just fall into place.

robert_scott_jr_

Hello Jonaton. Magnets would be an easy-to-do, uncomplicated method of a 'snap' connection. Perhaps coupled with the swivel method below. - Scotty

https://cad.onshape.com/documents/4077757f5c556437f0666985/w/b3d3812a977e2c6ced4e78a9/e/9cc01f5c45243e4aae897f41

nick_papageorge073

Is the sensor portion off-the-shelf, and you roughly modeled it in CAD? Or are both the sensor half and the "holder" half your own design? With two pivot ears coming off the side of the sensor portion, I'd probably put a ramp and rely on the holder portion to flex to snap in the sensor. There could be a lead-in and lead-out ramp so you can install/remove. If the flexing of the holder is not enough, you can add a flexible portion to the printed model by adding a few cuts to turn it into a snap finger.

nick_papageorge073

Here is a different idea that uses a magnet in the flat orientation.

martin_kopplow

So, we have a sketching competionion on? :0)  If we already have a magnet, why not use it to make the joint, the friction and the snap-in? At the same time, the gap between sensor and recess could be minimized:

... and because we are in Onshape, you could place a rotate mate in the virtual axis of the cylindical slide and see what angle you get with what gap and so on. This is also much easier to print than snap-joints, I suppose.

robert_scott_jr_

Excellent idea, Martin. Sensor retention with adjustment all in one! - Scotty

jonatan

martin_kopplow said:

I see how you want to snap the sensor housing into the ceiling recress. The axial placement of the magnets, though, isn't going to help you much: Neither with precise positioning, nor with the required friction. Try orienting the magnet radially. Use one magnet and one iron core per side (to eliminate sweet spot), or even think the term 'axis' much much bigger and make a cylindrical face on the upper side of the sensor housing that slides in the recess part. Large diameters: That's how you get apprpriate friction. Put some magnets and a bent iron strip in these surfaces. Then, everything else will just fall into place.

True! I was planning to add a bit of friction in the 3D printed case itself but I think I'm overthinking it too much

nick_papageorge073 said:

Is the sensor portion off-the-shelf, and you roughly modeled it in CAD? Or are both the sensor half and the "holder" half your own design? With two pivot ears coming off the side of the sensor portion, I'd probably put a ramp and rely on the holder portion to flex to snap in the sensor. There could be a lead-in and lead-out ramp so you can install/remove. If the flexing of the holder is not enough, you can add a flexible portion to the printed model by adding a few cuts to turn it into a snap finger.

It's just my own design, I will have to design both the box for the sensor and the holder that attaches to the ceiling. Currently, I'm only doing tests with the sensor itself in a rounded box

martin_kopplow said:

So, we have a sketching competionion on? :0)  If we already have a magnet, why not use it to make the joint, the friction and the snap-in? At the same time, the gap between sensor and recess could be minimized:

... and because we are in Onshape, you could place a rotate mate in the virtual axis of the cylindical slide and see what angle you get with what gap and so on. This is also much easier to print than snap-joints, I suppose.

Uhhh I like this idea, seems quite simple. I guess instead of adding a steel strip I could just add another magnet inside the rounded part of the sensor (add a half circle on top of it as the steel strip) and add some texture to the plastic to generate some friction?

jonatan

So more of less like this sensor https://sonoff.tech/product/gateway-and-sensors/snzb-06p/

martin_kopplow

jonatan said:Uhhh I like this idea, seems quite simple. I guess instead of adding a steel strip I could just add another magnet inside the rounded part of the sensor (add a half circle on top of it as the steel strip) and add some texture to the plastic to generate some friction?

That would probably be possible, though two magnets will always tend to center upon each other, limiting the possible angle of adjustment. It depends on how great an angle you need. A steel strip could be derived form an old tomato paste tin can, an empty paint bucket, or the like. You could also put in one or two rows of countersunk screws, just to give the magnet someting to cling to at all angles.

When FDM-Printing, you will get visible layers anyway, you could call that an intentional feature and use these as kind of a 'ratchet' surface for your friction. Have fun! ;0)

jonatan

That’s true, might not work with two magnets.

i think I will try to do something that’s 3D printed + magnets or something else only, so I can easily reproduce it