It's not a good idea to have a neodymium magnet next to an SBC, is it?

[DiTBho]

for a really crazy project I am thinking about drilling a couple of 8mm holes for each neodymium magnet which will be glued inside its hole.

I don't know if it's a problem or not, but these hole will be drilled on a poly-carbonate frame, which is the same frame where I am thinking about installing an single computer board.

I have no technical information on the magnetic field generated by each magnet, I cannot find a technical data sheet, speaking of the mechanical plane, each magnet will probably be 18-24mm near the SBC.

The purpose of the two magnets is to create a mounting/unmounting mechanism to quickly mount and remove the SBC-unit from the cabinet, so this solution would be elegant and practical.

But ... I am not sure it's a good idea for the SBC 

[Ian.M]

It depends on the SBC.   The main issue is the static magnetic field will push inductors closer to their saturation point, or even in very close proximity fully saturate them which can interfere with or even destroy switching regulators.  Vibration of the SBC PCB relative to the magnets could also be problematic due to induced currents.

[DiTBho]

it's an embedded Linux board, similar to the RPI, but PowerPC-based.
It eats 0.9A @ 12V

[Ian.M]

So odds are its got various buck converters on-board.   You'd need to determine their locations and avoid positioning the magnets in close proximity to their inductors.  Its not an insurmountable problem as many phone cases have magnetic latches, and apart from interference with compass sensors, they typically dont cause any issues with the phone itself.

[T3sl4co1l]

Why do they have to be so big?  That's a ton of magnetic energy!

Use a recessed hole in a piece of mild steel.  You'll get the same holding force in a fraction of the size.

The steel acts as a pole piece, carrying the back-side field around to the front (concentrated around the rim and face of the hole), not just doubling the effect of the field (instead of one pole face, it's two) but increasing it significantly as well (instead of flux density diluted into free space, it's concentrated in the pole pieces).  Only a steel plate is needed for the opposite face, or you can use the same geometry with opposite polarity for somewhat more holding.

Needless to say, the steel focuses the field, so it's not going out all over the place, but confined mostly to the front face.  This also means the grab distance is shorter, but that should be fine for a latch.

Ideal geometry would be something like, say a 2mm dia 1mm thick magnet, in a 3-4mm I.D. hole (centered -- maybe use a stepped flat-bottomed hole to get it centered more easily), with minimum O.D. 6-8mm say, and minimum thickness behind hole / of catch plate, say 0.8mm.  So, it can be as small as the original proposal.

If you can't mill/turn flat-bottomed holes, then perhaps these can be found off the shelf as latches/catches?

As for snapping in place, there are zillions of less aesthetic options.  Metal and plastic tabs locking into holes; edge catches/latches; slides; clamps; etc.

The

[DiTBho]

Why do they have to be so big?  That's a ton of magnetic energy!
Use a recessed hole in a piece of mild steel.  You'll get the same holding force in a fraction of the size.

It's a complex mechanical design with a lot of parts. Some are not on the polycarbonate-frame, so I won't mention here.

There is a big and heavy poly-carbonate frame, and two boards on it, plus other stuff, for something like 1.2Kg. Then there is a metal box that I cannot drill too much.

The poly-carbonate frame goes inside the metal box, there is a poly-carbonate bed mounted to one of the wall of the metal box, and there are two guide bars that keep the poly-carbonate-frame aligned to the correct horizontal and vertical position, but it needs something to keep the poly-carbonate frame attached to the bed, there there is a free degree of freedom and the frame can move.

The two big magnets have demonstrated to do the job, at least mechanically.

The alternative is to drill, the metal box, the poly-carbonate-bed and the poly-carbonate frame to use a screw with a nut, with the nut on the  poly-carbonate frame.

[DiTBho]

Use a recessed hole in a piece of mild steel.  You'll get the same holding force in a fraction of the size.

anyway, I will reconsider something like this, thanks 

[Kleinstein]

It really makes sense to guide the magnetic field, so that little of the magentic field is at the inside. This can save quite a lot on the magnets. The less the magnetic field has to travel through air, the stronger the field. So a reasonable construction, could reduce the needed size of the magents to maybe 1/10 and the field to the circuit to maybe 1/100. The simplest form would be some iron / steel plate on the backside of the magents and than have the magents oriented in opposing direction.

[m3vuv]

Whats an SBC?.

[T3sl4co1l]

Single board computer.

[eugene]

Anecdotal story: I once built a bit of hardware that mounted to a snow plow with a pair of magnets. Since snow plows bounce around a lot, I used two neodymium magnets, one north-up the other south-up, bridged by 0.125" thick piece of mild steel. As predicted, it stuck much more than twice as strongly as a single magnet. Surprisingly, the steel was not enough to contain the magnetic field as the microcontroller would not work. I added a second, similar piece of steel spaced apart from the first another 0.125" to catch the stray field and it worked again.

I never have figured out exactly why the thing wouldn't work with the strong field. The circuit had no switching supply so no inductors other than a small ferrite. I imagine the electrons were all trying to go in circles instead of where they were supposed to go.