structured singular value, state-space, active filter, performance
This paper considers an application of the structured singular
value, mu, to the robust performance of active filter circuits that are subject
to real parameter component uncertainty. The focus is on the different Linear
Fractional Transformations that are required for the calculation of safe upper
bounds on filter performance without using conventional Monte-Carlo based
frequency gridding techniques. State-space ¹ results are applied so that worst
case maximum and minimum filter gain are separately determined using a
non-probabilistic test on a static matrix. The methodology is illustrated by a
typical analysis on an active Butterworth low pass filter of order 4.