synchonizing speeds of two dc motors.

[WHONOES]

Having thought about this a bit more, it is not necessary to use either a Tacho or an optical encoder. If the motors are driven using using PWM, it is possible to sample the back emf of the motors during the PWM off period and use this to gauge any difference in motor speed. It would then be possible to adjust one of the PWM pulse widths to adjust the speed of the errant motor. All this could be achieved, apart from the PWM, in the analogue domain.