Many existing audio amplifiers employ various error-correction techniques somewhat different from ordinary feedback. Out of these techniques two generalized error-correction structures might be singled out, each consisting of main and auxiliary channels. The subsidiary amplifier operates with an error signal caused by the departure of the main path from its desired value. Transfer functions, sensitivities of overall gain to changes, and stability criteria are determined separately for each error-correction structure. It is shown that, except for genuine feedforward, other topologies have no advantages over ordinary feedback as far as sensitivity to changes in the main amplifier are concerned. Feedforward frequency compensation is considered as a specific example of an error-correction structure with active error feedback. The aspects of the feedforward compensation by means of parallel summing are discussed in detail.
This paper costs $33 for non-members and is free for AES members and E-Library subscribers.