doble SFRA, User Manual, Page: 106 / 126

Sweep Frequency Response Analyzer (SFRA) User Guide
C-2 72A-2570-01 Rev. K 07/2011
There is a relationship between the geometric configuration of the internals of
a transformer and the distributed electrical elements inside the winding and
core assembly. These elements can be represented as an RLC network, and
such a network will have a frequency dependent transfer function. Changes to
the geometric configuration will affect the impedance of the RLC network,
and thus produce a different frequency response.
How SFRA Identifies Damage to Transformers
The primary objective of SFRA is to determine how the impedance of a test
specimen behaves over a specified range of frequencies. The impedance is a
distributive network of real and reactive electrical components. The components
are passive and can be modeled by resistors, inductors, and capacitors. The
reactive properties of a given test specimen depend on, and are sensitive to,
changes in frequency. The change in impedance versus frequency can be dramatic
in many cases. This behavior becomes apparent when we model impedance as a
function of frequency. The result is a transfer function representation of the RLC
network in the frequency domain.
Frequency response analysis is generally applied to a complex network of passive
elements. For practical purposes, we will consider only resistors, inductors, and
capacitors as passive circuit elements, and they are assumed to be ideal. These
three fundamental elements are the building blocks for various physical devices,
such as transformers, motors, generators, and other electrical apparatus.
It is important to understand the difference between the physical device and the
mathematical model we intend to use. When large and complex systems are
electrically analyzed, we are often faced with a poorly defined distributed
network. A distributed network contains an infinite number of infinitely small
RLC elements. For example, transmission lines are generally distributed in nature.
It is practical to model such distributed systems by lumping the basic RLC
components together, resulting in a lumped network. Lumping elements together
for a single frequency is a trivial task, but when system modeling requires
spanning a significant frequency interval, producing a suitable lumped model
becomes difficult.
When a transformer is subject to SFRA testing, the leads are configured to use
four terminals. These four terminals can be divided into two unique pairs—one
pair each for the input and output. These terminals can be modeled in a
two-terminal pair or a two- port network configuration (<Hypertext>Figure C-1).