Abstract
The results of an analysis and comparison of mantissa only, sign bit, true amplitude, automatic volume control, and diversity stack signal trace summations are presented. The results show that under conditions frequently encountered for low-level sources, all these summation methods preserve the fidelity of input signals to within a time variant scale factor. The value of the signal-to-noise (S/N) average power ratio of the true amplitude summer is shown to be much more sensitive to intertrace noise variance differences than the other summing techniques. Excessive signal distortion may occur for mantissa only or sign bit summers when instantaneous input signal magnitude is large in comparison with the input noise standard deviations of the summed data. There will be no significant differences in the performance of mantissa only and sign bit summers under conditions frequently encountered for low-level sources. The effects of signal summation methods are not directly dependent upon any instantaneous gain ranging characteristics of the amplifier used to create the input data.