Multibody analysis: position constrain equations of the slidercrank example

Hi everyone.

I have a doubt:

On the example at "kinematic exports help", the position constrain equations are not the same as the one made by hand. In fact, they should be as:

l1·cos(theta)+l2·sin(beta)-s = 0

l1·sin(theta)-l2·cos(beta) = 0

I know that a bit difference might be due to de sector of the angle beta. However, what I cannot undestand is why the term "s" appears multiplying to "cos(theta)" in the example.

Can anybody please help me?

Regards.

Thanks for make learning easier!