Theoretical Consideration of Computing Errors of a Slow Type Electronic Analog Computer

When solving differential equations using analog computers, there are two causes of errors: one results from the integrators and the other results from the coefficient setting elements. The former is independent of the order of the differential equation if the integrating time constants are all equal, while the latter is dependent on the setup procedure. The amount of the error from the coefficient setting elements is comparatively small in the operating frequency range from ω≤rad/sec. Therefore, theoretical analysis considering errors due to the integrators only is quite sufficient. However, higher accuracy is required in ω = 10 to 100 rad/sec (for example in flight simulators) and consequently the error caused by the coefficient setting elements has become impossible to neglect. The authors have succeeded in formulating a generalized and practical equation developed from conventional formulas.