This drawing reveals how complicated the design is. Notice how the escape wheel and the impulse pallet appear to roll together, instead of sliding across one another, and therefore why the escape wheels and the pallets need not be lubricated.

There are several differences between my drawing and the original specifications. I made these changes in order to enhance the actions of the escapement during the simulation. The gap between the escape wheels was not drawn in proportion to the original, but the only effects this had were to change the impulse pallet's circle radius and the diameter of the upper locking pallet. The convenience of using the gap of half an inch, in allowing me to use the "snap to grid" function in the computer's software, outweighed the compulsion to draw it as closely to the original as possible. Furthermore, I chose an angle of 60º for the rotation of the roller jewel during engagement with the pallet fork. The original design called for 24º, but increasing the angle also increases the depth of the roller jewel inside the fork: by exaggerating the actions of the escapement, the actions become more easily visible during the simulation. Other side effects of increasing this angle from 24º to 60º are that the pallet fork becomes longer and the roller jewel's circle radius becomes smaller by half. The depth increases by a factor of three.

This example demonstrates how creating your own drawings would allow you to change the variables and observe the consequences. The idea is not necessarily to create the drawing as closely to the original specifications as possible, but rather to make changes and to experiment. You may be able to improve a design or even invent an all-original design.