Here's a little science for the uninitiated: these two images show floating ball compasses, a design using a fluid to provide very low friction triaxial freedom to (presumably) point to magnetic north.
Spot the difference?
One has the arrow aligning with the north symbol on the dial, while the other has it aligning with the south symbol.
Which is correct?
On a conventional compass that is viewed from above, the arrow indicates magnetic north, so when you stand facing north and you look down on the ball, the arrow should be pointing away from you. So far that would suggest that either design may be correct.
However, the dial is designed to be read from the side. This means that raising the compass to eye level, you should be seeing the N, not the S. This makes the first image the correct design.
But wait, there's more!
I ordered a ball magnet to incorporate into some nostalgic custom automotive instrumentation, an excuse to fire up the CAD and 3D printer.
I got one matching the second image. The arrow pointed north just fine but so did the N, meaning I would see an S when driving north, so I hit the 'return' button, got a refund and could keep the item (gee, thanks).
Next, I ordered one matching the first image, but when it arrived its arrow pointed south, so I'm stuck with precisely the same outcome. Yep, hit 'return', get the refund, gee thanks again.
I can't get a break, can I? How can millions of these ball magnets be manufactured completely useless? Do you have one of these in the house? Are you wondering whether yours points north or south?
So as an alternative, I designed my own compass to rotate on a spindle in a fixed horizontal plane, but friction is proving a major challenge.
The answer?
I'm about to do some surgery on one of these ball compasses. If I can manage to separate the clear plastic hemispheres neatly, I may be able to get to the magnet in the float and reinsert it 180 degrees out, then refill and rejoin the hemispheres. With proper quality control in manufacturing, this really shouldn't be necessary.
Wish me luck!
_____________
OK, so I managed to pry the hemispheres apart and the fluid appears to be an extremely light machine oil. The float consists of a black plastic platform housing a rectangular magnet at its centre, all trapped in an air bubble by a glued-on painted aluminium dome.
After prying the ball apart, trying to pry the dome from the platform, I noticed the craft knife's blade had punched through instead, so I used the blade to slice away the dome from the platform, leaving a wafer-thin remnant of aluminium still glued in its track. This was easily pulled off.
The edge of the dome was somewhat deformed by the knife strokes but was reformed by pressing it onto the surface of an automotive globe of a slightly larger diameter.
The dome has been reattached at 180 degrees using high temperature silicone sealant and is setting while I consider how to refill the ball before rejoining it.