Why does the potentiometer become more sensitive after calibration?
Before calibration, part of the potentiometer's rotation range is spent on positions the engine can never reach. After the upper limit is taught, the GS740-3P redistributes the whole range of the potentiometer between idle and the taught top speed as 0-100 percent. The dead zone thus disappears and the same physical rotation angle corresponds to smaller speed steps.
The issue is where the resolution is spent.
Before the upper limit is taught, the potentiometer's range and the travel actually used by the throttle lever do not overlap exactly; the part after the lever rests on its stop remains in practice as a dead zone. The result: nothing happens at one end of the potentiometer, the useful region stays narrow and fine adjustment becomes difficult.
When the upper limit is taught this mapping is redone: the zero position of the potentiometer corresponds to idle, its full position to the taught top speed, and everything in between is distributed as 0-100 percent. The same physical rotation now sweeps a narrower speed band, that is, the same angular movement produces a smaller speed change.
On the position side the component that determines resolution is the AS5600 magnetic encoder: 12-bit, that is 4096 steps per full turn. Since it is contactless there is no wiping contact to wear out; the position measurement is more stable than in potentiometer-based solutions.
Note: this sensitivity is throttle position sensitivity. It does not keep the engine speed constant under load.
Full article: Electronic speed and throttle control on a diesel engine