AN9768 Datasheet, PDF (2/8 Pages) Littelfuse – Transient Suppression Devices and Principles

Application Note 9768
Crowbar-type devices involve a switching action, either the
breakdown of a gas between electrodes or the turn-on of a
thyristor, for example. After switching on, they offer a very
low impedance path which diverts the transient away from
the parallel-connected load.
These types of crowbar devices can have two limitations.
One is delay time, which could leave the load unprotected
during the initial transient rise. The second is that a power
current from the voltage source will follow the surge
discharge (called “follow-current” or “power-follow”). In AC
circuits, this power-follow current may not be cleared at a
natural current zero unless the device is designed to do so;
in DC circuits the clearing is even more uncertain. In some
cases, additional means must be provided to “open” the
crowbar.
Filters
The frequency components of a transient are several orders
of magnitude above the power frequency of an AC circuit
and, of course, a DC circuit. Therefore, an obvious solution is
to install a low-pass filter between the source of transients
and the sensitive load.
The simplest form of filter is a capacitor placed across the
line. The impedance of the capacitor forms a voltage divider
with the source impedance, resulting in attenuation of the
transient at high frequencies. This simple approach may
have undesirable side effects, such as a) unwanted
resonances with inductive components located elsewhere in
the circuit leading to high peak voltages; b) high inrush
currents during switching, or, c) excessive reactive load on
the power system voltage. These undesirable effects can be
reduced by adding a series resistor hence, the very popular
use of RC snubbers and suppression networks. However,
the price of the added resistance is less effective clamping.
Beyond the simple RC network, conventional filters
comprising inductances and capacitors are widely used for
interference protection. As a bonus, they also offer an
effective transient protection, provided that the filter's front-
end components can withstand the high voltage associated
with the transient.
There is a fundamental limitation in the use of capacitors
and filters for transient protection when the source of
transients in unknown. The capacitor response is indeed
nonlinear with frequency, but it is still a linear function
of current.
To design a protection scheme against random transients,
it is often necessary to make an assumption about the
characteristics of the impinging transient. If an error in the
source impedance or in the open-circuit voltage is made in
that assumption, the consequences for a linear suppressor
and a nonlinear suppressor are dramatically different as
demonstrated by the following comparison.
A Simplified Comparison Between
Protection with Linear and Nonlinear
Suppressor Devices
Assume an open-circuit voltage of 3000V (see Figure 2):
1. If the source impedance is Z_S = 50Ω
With a suppressor impedance of Z_V = 8Ω
The expected current is:
1 = ₅--³--₀-⁰----+⁰---⁰--₈- = 51.7A and V_R = 8 × 51.7 = 414V
The maximum voltage appearing across the terminals of a
typical nonlinear V130LA20A varistor at 51.7A is 330V.
Note that:
Z_S × I = 50 × 51.7 = 2586V
Z_V × I = 8 × 51.7 = 414V
= 3000V
2. If the source impedance is only 5Ω (a 10:1 error in the as-
sumption), the voltage across the same linear 8Ω sup-
pressor is:
V_R = 3000 ₅-----+⁸-----₈-- = 1850V
However, the nonlinear varistor has a much lower
impedance; again, by iteration from the characteristic curve,
try 400V at 500A, which is correct for the V130LA20A; to
prove the correctness of our “educated guess” we
calculate I,
3000-400V
I=
5
= 520A
Z_S x I = 5 x 520 = 2600V
V_C =
400V
= 3000V
which justifies the “educated guess” of 500A in the circuit.
Summary
TABLE 1. 3000V “OPEN-CIRCUIT” TRANSIENT VOLTAGE
ASSUMED SOURCE IMPEDANCE
50Ω
5Ω
PROTECTIVE DEVICE PROTECTIVE LEVEL ACHIEVED
Linear 8Ω
414V
1850V
Nonlinear Varistor
330V
400V
Similar calculations can be made, with similar conclusions,
for an assumed error in open-circuit voltage at a fixed source
impedance. In that case, the linear device is even more
sensitive to an error in the assumption. The calculations are
left for the interested reader to work out.
The example calculated in the simplified comparison
between protection with linear and nonlinear suppression
devices shows that a source impedance change from an
assumed 50Ω to 5Ω can produce a change of about 414V to
1850V for the protective voltage of a typical linear
suppressor. With a typical nonlinear suppressor, the
10-103