Recently I encountered a misunderstanding with applying Newton's second law to rotational dynamics. Imagine a yoyo of radius $R$ on a plane (with non-zero friction) being pulled towards the right.
Then imagine that the force applied at some inner radius $b$ (red circle) will be precisely equal to that of the magnitude of the static friction $f$ on the yoyo at the contact point with the ground.
$f=Mg\mu$
$F=f$
Then, if we sum the torques on the yoyo about the contact point with the ground we get:
$\tau=-(R-b)F$
Clearly, there is some nonzero torque which will cause some sort of angular acceleration, and thus, a translational acceleration along the plane. Summing the torques about the circles center of mass yeilds the same result; a non-zero torque being applied onto the yoyo causing rotation.
But when we apply Newton's second law to the system, my understanding begins to fall apart. $F=ma$ applied on the yoyo's center of mass yields
$F_{net}=ma=F-f=0$
It sums to zero, since as we remember $F=f$. This suggests no translational acceleration whatsoever, in conflict with the result we derived using rotational dynamics. (Remember: all forces applied to a rigid body can be treated as being applied to its center of mass).
How is this conflict between $F\propto a$ and $\tau\propto\alpha$ then explained?
TLDR: I know I am making a mistake using $F=ma$ somewhere, but I'm not sure where. Could somebody point it out with an intuitive explanation and a resolution of the two equations that adds up?