Improvements to binary floating-point digital differential analysers
G. Hannington · Electronics Letters · 1980
A description is given of a novel method of providing a residue in a binary floating-point digital differential analyser. The error bounds for a sine/cosine generator utilising floating-point integrators employing Euler integration both with and without a residue are given and compared. Also described is a method for avoiding the extra delay incurred in using an Euler integrator module as a coefficient multiplier.