I don't know what a "40" is. Mine are N42s, and 0.0625 cubic inch in volume.
The 3/4" is the distance between the magnets, they're about 3/8" from the blade.
-
Ok, I see the basic problem. The OEM magnets in both my RCBS (Ohaus) 5-0-5 and my Lyman (Ohaus) M5 are a type of shallow horseshoe magnet. They all have notches in them, and there is a S pole on one side of that notch, and a N pole on the other side. The magnets are positioned with their notches
facing each other, and their magnetic moments are
parallel to each other. Magnet A's S pole is directly opposite magnet B's N Pole, and Magnet A's N pole opposite B's S pole. Hence the pair form a circuit loop, with the force lines and the magnets themselves tracing a squared circle - not unlike an electrical circuit. The copper blade is exposed to two main force lines "flowing" in both directions near the ends of the magnets, with a sort of neutral area between the middles of the magnets where the flow lines are mostly parallel to the copper blade:
The effect of this arrangement is to confine the blade inside the wide neutral area, but within that area to not exert much influence on the blade.
When I tried my N42 block magnets, the blocks' magnetic moments (vectors) were aligned
coaxially. The magnets are positioned end-to-end magnetically, across from each other, with a single N pole directly across from the other magnet's S pole, with the blade between. The other S and N poles are on the
back sides of the magnets, facing away from the copper blade. The force lines are flowing in a single direction, and largely perpendicular the the blade. Also rather than a neutral area where the blade resides, the flow lines are closest together (strongest) near the center of the blade:
In this arrangement, with powerful neodymium magnets, they try to center the blade in the strongest force area, directly between the two magnets.
As proof of that, consider what happened when I mounted a single N42 magnet with one pole pointing at the copper blade. The magnet attracted the blade, centering the blade on the magnet's magnetic moment vector. I could move the scale's leveling leg adjustment up and down through the
full range of adjustment (more than an inch) and the beam pointer stayed centered on the magnet (and pointing to zero on the scale's indicator) from one end of the adjustment to the other!
-