Hi, is it possible to parallel two hoverboard motors on the same motor output using the hall sensors from one motor for control ???
No, this would only work in a very specific and precisely aligned way which, IMO, would be extremely difficult to achieve - involving a shared and rigidly linked output shaft and aligning the rotors both relative to each other and simultaneously relative to their individual stators as viewed by the output forces.
The ODrive works, in principle, by varying the current through more than one of the 3 phases of the motor - a process called Field Oriented Control. To know which phases to apply current to, and how much current, it relies on feedback from one of a few options of sensors that tell it what position the motor shaft is in (and thereby the offset of the rotor/magnets relative to the stator teeth/phases).
So, imagine a situation where the first motor has the sensor, and the second motor does not, but the second motor is a bit “out of phase” relative to the first. ODrive will be applying current correctly to, say, phase A and B of the first motor, but for the second motor to behave the same way it may need current on phases C and A. The result may very well be the second motor attempting to turn in the opposite direction as the first.
In all but the perfect alignment of the two motors, you will have either an unworking setup, or a setup that “works” but the motors will cancel each others’ forces either partially or (if you happen to manage a precise but exactly out of phase offset between the two) entirely.
Edit: With that said… If I absolutely needed to do this, I would start by mating the motor mounts together and calibrating each motor individually using a high resolution encoder and then rotate the shaft to an accurate and repeatable angle and measure the setpoint of the encoder. Then I would key the shaft, machining the keys for each motor to be offset relative to each other by the difference between the previously measured setpoints. To achieve this, the two motors would need to remain mated to each other from start to finish - or at least able to realign the motor chassis back to the exact offset they were in during calibration. From there, you should be able to ditch the encoder and utilize just one motor’s hall sensors. Note that these motors and their output shaft would all be paired to each other forever, and replacing any of the 3 would require repeating the process.
Edit 2: It would probably be better to press the motor rotors onto the shared shaft rather than relying on the tolerance of the machined keyways. In this case you would need to machine a fixture that holds their rotors in the correct offsets while pressing them on the shaft.