I don't find the current implementation of the mems controlled cursor very intuitive.
Remember that the mems is 3D, not 2D only.
For intuitive operation, you want to imitate something physical, that you can hold in your hand and move around, but then add smarts, based on what's available,
The build environment's not working for me, so I'm not offering to make the changes, bu these are my suggestions:
1) Use the Z axis as a stickiness control (i.e. decelerator). Lift up, the cursor becomes mired, wherever it was heading, diminish it's velocity proportional to the amound of extra G force momentarily experience, as the operator lifts the stmcircle vertically, and past a minor at-rest threshold.
2) Using additional angle to move the cursor further, is a disturbing experience, as it takes the LCD out of optimum viewing angle under normal human usage. Better, to have the X-Y axis control be primarily for null zone and direction, Use >1.0 G measurements to directly control velocity, but through a to the power of relationship, i.e. lurch the circle abruptly downward while angling it slightly in the desired cursor motion direction, and the speed will be very fast, drop it in slow motion, and the cursor will creep.
With respect to X-Y axis controlling direction, I'm thinking to use enough bits that the cursor can move with realistic angular granularity, so we are now looking at X-Y for angle, and Z for speed/lock.
Keep in mind, that a good solution can form the basis for more than merely the toy functionality presented on the STM32 primer, it can be translated across to RC model flight control, game joysticks, and various other similiar applications.
Graham
