Clearly, if the perturbing link had not been added immediately
after the initial link, or indeed if the initial link had been
originally added as a new tree branch, the results would be
the same. As a consequence, a perturbing scalar multiple of a
link can be added using the standard formula of eqn. 1 and
the extra common factor is then removed by division.
It has been shown that the common factors that arise
out of adding perturbing links can be identified, and therefore
removed. The removal of the common factor is trivial if the
perturbing elements are resistive, inductive, or capacitive
2-terminal or 4-terminal elements. Hence the perturbation
technique using Kron's link-at-a-time approach will be highly
effective, and it is feasible to attempt
s
plane optimisation by
this technique.
R. K. JARROTT
17th May 1974
60 Buckley Crescent
Fairview Park
South Australia 5126
References
1 DOWNS, T.: 'Symbolic evaluation of transmittances from the noda
admittance matrix',
Electron. Lett.,
1969, 5, pp. 379-380
2 BARHAM, R. A.: 'A network analysis method with reduced storage
requirements'. Proceedings of 13th national IREE electronics con-
vention, Melbourne, Australia, 1971, pp. 22-23
3 BRANIN, F. H., JUN.: 'The relation between Kron's method and the
classical methods of network analysis'. IRE Wescon Convention
Record, 1959, Pt. 2, pp. 3-28
NEW ACTIVE-GYRATOR CIRCUIT
Indexing terms: Active networks, Gyrators
Ideal passive gyrators can be made using two operational
amplifiers, and it has been proved that they cannot be made
with only one amplifier. However, this latter proof led to the
discovery that an ideal active gyrator can be made with only
one amplifier. A circuit for doing this is presented here, and
its behaviour when used in an inductance-simulating mode
is analysed.
The gyrator is now well established as an element in the design
of inductorless networks, being used, together with a capacitor,
primarily in an inductance-simulating mode, either directly
1
*
2
or indirectly.
3
The practical nonreciprocal components that
are presently available for making a gyrator are either transis-
tors or, more conveniently, complete transistor amplifiers of
which, for this purpose, the operational amplifier is the most
commonly used form.
Over the past few years, several
ideal* gyrator circuits have been described in the literature,
all being variants of the original Riordan circuit
4
and using
two such amplifiers in conjunction with a few resistors. In
contrast to this, no one has described an ideal gyrator circuit
using only one amplifier, although a few nonideal 1-amplifier
circuits have been published.
5