Active and passive implementations of self-sensing readout networks are stu
died in terms of system properties such as the poles and the zeros of their
transfer functions. The location of the zeros is shown to be dependent on
two non-dimensional parameters related to the balancing condition of the re
adout bridge and to the compensation of its losses, while it is not affecte
d by the active or passive implementation of the readout bridge. In the cas
e of ideal "loss compensation" a graphical procedure is employed to describ
e the migration of the zeros. Even if the self-sensing arrangement guarante
es the colocation of the sensing and actuating functions, the graphical pro
cedure shows the possibility of non-minimum phase zero coupls for balancing
conditions close to the so-called "electrical balancing". The effects of m
odel reduction techniques such as truncation or residualization is then stu
died, starting from a model of the system in terms of modal coordinates. Tr
uncation is shown to be better suited to determine the "electrical balancin
g" condition while residualization gives a better approximation of the syst
em zeros. The conditions of electrical balancing and loss compensation are
then related to a condition of minimum dependance of the self-sensing bridg
e output from the driving electrical input. This property can be practicall
y exploited to devise an adaptive readout bridge which can automatically re
ach the balancing conditions and the loss compensation or to identify the e
lectrical parameters of the piezoelectric transducer. Experimental tests pe
rformed on a beam and a plate structures provided with self sensing piezoel
ectric transducers are used to validate the analytical models. (C) 2000 Aca
demic Press.