Page 373 - Power Electronics Handbook

Electrical machine control 363
14.3 Electrical machine control
Electrical machine control marks one of the most important applications
for power semiconductors, especially thyristors and triacs. Due to their
small size these power devices have given practical reality to systems which
previously were only an experimenter’s dream (Dewan et al., 1984; Pillai,
1995; Shepherd et al., 1996).
Control electronics bridges the gap between two widely differing
sciences. On the one hand, the power semiconductor, with its associated
control circuitry, may be regarded as a tool of the electronic engineer,
whereas the machine, which differs only slightly from conventional
designs, is the instrument of the electrical engineer. It is not possible to
treat power semiconductors and machines separately if they are to work
together as a system. The characteristics of the machine affect the
performance of the power semiconductor and vice versa, but the rapid
growth of electronic control systems has resulted in instances where
electronic engineers have developed systems whilst having no knowledge
of machine fundamentals, or electrical engineers have blundered into
electronic control schemes whilst being unfamiliar with transistor circuitry.
A control engineer is a unique animal who must have a good working
knowledge of both electronics and electrical machines. The previous
chapters have dealt with power electronic circuitry at some length, and the
present section will consider the principles of electrical machines and how
they can be controlled by power electronic devices.
Although power semiconductors are in widespread use for machine
control, sometimes because they represent the only practicable method for
obtaining a control system, often they are in direct competition with
traditional controllers and must then prove themselves to be either
technically or economically superior. The advantages which power
semiconductor drives have over conventional methods, such as
Ward-Leonard systems, are:
(i) Quicker response. This is specially advantageous when fast accelera-
tion or high-speed accuracy is required, as in reversing-mill tables.
(ii) High operating efficiency due to the small voltage drop across a
power semiconductor, even when it is handling large amounts of
power.
(iii) Less weight and lower installation cost. A power semiconductor
control cubicle occupies a small floor area, is light and does not
require any special floor surface or mounting. This compares very
favourably with the aligning and mounting requirements for
motor-generator systems.
(iv) Ease of maintenance. A power semiconductor system can be
designed to be easily maintained by incorporating diagnostic circuits
during the design stage and providing plug-in card replacements for
electronics. Servicing of the controller is, of course, not required due
to the absence of any moving parts.

The disadvantages of power semiconductor systems are:
(i) The low overload capability, which is caused primarily by the low
time constant of the power semiconductor devices. This means that in