Page 4 - Application Guide Semiconductor Fuse Link
P. 4
Introduction
A fuse is a device for protecting an electrical system against the effects of overcurrents (excess
currents), by melting one or more fuse-elements, thus opening the circuit. Very fast-acting fuses are
widely used for the protection of diodes, thyristors, and other power semiconductors in a.c. and d.c.
power electronic applications, and provide excellent protection against the potentially damaging
effects of short-circuit currents.
Current-limiting fuses achieve this protection by limiting both the amount of energy produced during
an overcurrent and also the peak current which is allowed to fl ow.
The term ‘semiconductor fuse’ means a very fast-acting fuse specifi cally intended for the protection
of power semiconductors.
Fig.1 shows the construction of a typical semiconductor fuse. The fuse elements are usually made
of pure silver strips, with
regions of reduced cross- element
body
sectional area (often called
notches). There may be
several strips in parallel,
depending on the ampere
rating of the fuse. They
are enclosed within an blade
sand
insulating tube or ceramic
body, which is fi lled with
pure quartz sand. At each Fig.1 Construction of a typical semiconductor protection fuse
end there are terminals
with a variety of designs to
permit installation in fuse-
holders or connection to
busbars.
Fuses such as that shown in Fig.1, with the elements surrounded by sand, are called current-
limiting, or high breaking capacity fuses. Semiconductor fuses are sometimes referred to as
rectifi er or ultra-fast fuses. They are available with voltage ratings up to 12.5kV, and with rated
currents up to 10000A.
During normal circuit operation the fuse elements carry the required currents without melting. The
resistance is as low as possible to minimize the power loss. However, when a short-circuit or fault
occurs, the elements melt very quickly at the notches (regions of reduced cross-sectional area), and a
number of small electric arcs are produced in the notch zones, which causes the fault current to be
rapidly reduced to zero, and the arcs extinguish. The total time taken by a fuse to clear a fault is the
sum of the melting time (pre-arcing time) and the arcing time.
Special fuse designs are needed to provide protection of power electronic components. The p-n
semiconductor junction can be very easily damaged, and so a very fast-acting fuse is required. For
example, a typical thyristor may fail when subjected to a 10ms (one half-cycle in a 50Hz system)
pulse of only 10 times its nominal r.m.s. current rating. A low-voltage fuse designed for general
industrial applications may require 15 30 times its ampere rating to melt within 10ms, which is not
fast enough to provide protection, whereas a very fast-acting semiconductor fuse typically requires
only about 5-6 times its ampere rating, thus protecting the thyristor.
4