CHEBYTOP is actually a set of subroutines for solving variable thrust and constant (constrained) thrust problems. It requires a calling program which organizes inputs and outputs. The program dates from the late 1960's and early '70's and was written by the then Boeing Co. under contract to NASA Ames. CHEBYTOP uses a series of approximations to the optimal control problem and reduces that problem to a series of classical calculus optimizations. It may actually be the first collocation based trajectory optimization code. CHEBYTOP basically presumes that the constant thrust trajectory will "look like" the variable thrust trajectory of comparable mass and power. (The variable thrust trajectory can be found exactly in most cases.) The non-structured, compact FORTRAN of the program is little changed from the original Boeing work which culminated in 1973 with the release of CHEBYTOP III. Forrester Johnson of Boeing is the author of the solution technique and the program.
The Analysis and Integration Group has two variants of CHEBYTOP. One is the original Boeing version, the other is a JPL version which Carl Sauer created and combines features of CHEBYTOP II and CHEBYTOP III. We tend to think that the JPL version is a little more accurate and stable than the original. CHEBYTOP is usually within 5% of the propellant estimates that VARITOP generates for nearly planar trajectories and very dubious for highly inclined trajectories or vehicles with thrust to weight ratios greater than .001.