For belt scale's error estimation we shall reject constraints (idlers), which will be replaced by their corresponding responses as shown in fig.7. There was added a coordinate system with origin in a point
.
The rejected constraints have been substituted by force's responses, that is force
,
and
. Thus, if you recall,
-- it is a response of a spring-bias deflected idler and magnitude of this deflection obeys the law:
![]()

Figure 6: System of force are existing in the mech system with properly counted assumptions. Here is no any response reaction of idler H.

Figure 7: Modernized mech system where constraints(idlers) were thrown off being repalced by its reactions.
Obviously, that
.
Further there shall be used "the principle of a solidification" which widely known in theoretical mechanics. As the belt scale system is in an equilibrium we can replace two branches of a flexible string with pivot-joint rods/beams as shown in fig.8. As the system is symmetric we can consider a condition of an equilibrium for its halfs concerning a line of a symmetry
. In a word it can be split in this place. In fig.9 the new conventional scheme of mech system is presented. There I have tried to depict force keeping a belt scale factor for the further course of operations could be clear.

Figure 8: Modernized mech system where string sections have been replaced by rods having use ``principle of solidification''.

Figure 9: In regard of system's symmetry we can consider a half of one. Increased tension force
is shown dotted.
There are force
and
(fig.9) as projections to coordinates' axes of varied (increased) string tension force
. Increased tension
shown dotted. The resultant force of a tension is always directed along a string or in our assumption along a rod. Such force were called as sliding in mechanics and we have "the right" to displace its along a rod as the state of a system's condition will remain the same. Well, we shall take advantage of this "right" and we shall move component of tension force to a point
as it shown on fig.10. I am sorry! There are so much force and the point
such a tiny, but I think you've seen, what has been conceived.

Figure 10: Belt scale mech system where sliding force have been moved at point

Position: