AN-3008 RC Snubber Networks for Thyristor Power Control and ...

www.fairchildsemi.com
Application Note AN-3008
RC Snubber Networks for Thyristor Power Control and
Transient Suppression
Introduction
RC networks are used to control voltage transients that could
falsely turn-on a thyristor. These networks are called snubbers.
The simple snubber consists of a series resistor and capacitor
placed around the thyristor. These components along with
the load inductance form a series CRL circuit. Snubber
theory follows from the solution of the circuit’s differential
equation.
Many RC combinations are capable of providing acceptable
performance. However, improperly used snubbers can cause
unreliable circuit operation and damage to the semiconductor
device.
Both turn-on and turn-off protection may be necessary for
reliability. Sometimes the thyristor must function with a
range of load values. The type of thyristors used, circuit
configuration, and load characteristics are influential.
Snubber design involves compromises. They include cost,
voltage rate, peak voltage, and turn-on stress. Practical
solutions depend on device and circuit physics.
Static dV
dt
dV
What is Static ?
dt
dV
Static ------ is a measure of the ability of a thyristor to retain a
dt
blocking state under the influence of a voltage transient.
dV
( ) Device Physics
dt s
dV
Static ------ turn-on is a consequence of the Miller effect and
dt
regeneration (Figure 1). A change in voltage across the
junction capacitance induces current through it. This current
is proportional to the rate of voltage change ⎛dV
------ ⎞
⎝
. It triggers
dt ⎠
the device on when it becomes large enough to raise the sum
of the NPN and PNP transistor alphas to unity.
I1
ICN IJ
NPN
K
CJP
CJN
IBN
IK
IBP
IJ
A
IA
PNP
ICP
I2
TWO TRANSISTOR MODEL
OF
SCR
G
V
Figure 1.
dV
dt
t
dV
CJ
dt
IA =
1– (αN + αp)
CJ
CEFF =
1– (αN + αp)
dV
(
dt
) s
Model
Small transients are incapable of charging the self-capacitance
of the gate layer to its forward biased threshold voltage
(Figure 2). Capacitance voltage divider action between the
collector and gate-cathode junctions and built-in resistors
that shunt current away from the cathode emitter are responsible
for this effect.
CJ
G
A
PE
NB
PB
NE
K
C
INTEGRATED
STRUCTURE
dV
Conditions Influencing ( )
dt s
Transients occurring at line crossing or when there is no
initial voltage across the thyristor are worst case. The
collector junction capacitance is greatest then because the
depletion layer widens at higher voltage.
dV
STATIC (V/µs)
dt
180
160
140
120
100
80
60
40
20
20 100 200 300 400
MAC 228-10 TRIAC
TJ = 110°C
500 600 700 800
PEAK MAIN TERMINAL VOLTAGE (VOLTS)
Figure 2. Exponential
dV
(
dt
) s
versus Peak Voltage
REV. 4.01 6/24/02

AN-3008
APPLICATION NOTE
dV
Static ------ does not depend strongly on voltage for operation
dt
below the maximum voltage and temperature rating.
Avalanche multiplication will increase leakage current and
dV
reduce ------ capability if a transient is within roughly 50 volts
dt
of the actual device breakover voltage.
dV
A higher rated voltage device guarantees increased ------ at
dt
lower voltage. This is a consequence of the exponential
rating method where a 400 V device rated at 50 V/µs has a
dV
higher ------ to 200 V than a 200 V device with an identical
dt
rating. However, the same diffusion recipe usually applies
for all voltages. So actual capabilities of the product are not
much different.
Heat increases current gain and leakage, lowering ⎛dV
------ ⎞
⎝
,
dt ⎠s
the gate trigger voltage and noise immunity (Figure 3).
dV
STATIC (V/µs)
dt
180
160
140
120
100
80
60
40
Failure Mode
MAC 228-10
V PK = 800 V
20
20 100 200 300 400 500 600 700 800
T J, JUNCTION TEMPERATURE (°C)
dV
Figure 3. Exponential (
dt
) versus Temperature
s
dV
( )
dt s
dV
Occasional unwanted turn-on by a transient may be acceptable
in a heater circuit but isn’t in a fire prevention sprinkler
system or for the control of a large motor. Turn-on is destructive
when the follow-on current amplitude or rate is excessive.
If the thyristor shorts the power line or charged
capacitor, it will be damaged.
Static ------ turn-on is non-destructive when series impedance
dt
limits the surge. The thyristor turns off after a half-cycle of
dV
conduction. High ------ aids current spreading in the thyristor,
dt
dI
improving its ability to withstand ---- . Breakdown turn-on
dt
does not have this benefit and should be prevented.
Improving
dV
( )
dt s
dV
Static ------ can be improved by adding an external resistor
dt
from the gate to MT1 (Figure 4). The resistor provides a path
dV
for leakage and ------ induced currents that originate in the
dt
drive circuit or the thyristor itself.
dV
STATIC (V/µs)
dt
140
120
100
80
60
40
20
MAC 228-10
800V 110°C
R INTERNAL = 600Ω
0
10 100 1000 10000
GATE-MT 1, RESISTANCE (OHMS)
( ) s
dV
Figure 4. Exponential
dt
versus
Gate to MT 1 Resistance
Non-sensitive devices (Figure 5) have internal shorting
resistors dispersed throughout the chip’s cathode area. This
design feature improves noise immunity and high temperature
blocking stability at the expense of increased trigger and
holding current. External resistors are optional for non-sensitive
SCRs and TRIACs. They should be comparable in size
to the internal shorting resistance of the device (20 to 100
ohms) to provide maximum improvement. The internal resistance
of the thyristor should be measured with an ohmmeter
that does not forward bias a diode junction.
dV
STATIC (V/µs)
dt
2200
2000
1800
1600
1400
1200
1000
800
600
50 60 70 80 90 100 110 120 130
T J, JUNCTION TEMPERATURE (°C)
( ) s
dV
Figure 5. Exponential
dt
versus Junction Temperature
MAC 16-8
V PK = 600 V
2 REV. 4.01 6/24/02

APPLICATION NOTE
AN-3008
Sensitive gate TRIACS run 100 to 1000 ohms. With an gate drive circuit needs to be able to charge the capacitor
dV
without excessive delay, but it does not need to supply continuous
current as it would for a resistor that increases ------
external resistor, their ------ capability remains inferior to
dt
dV
non-sensitive devices because lateral resistance within the
dt
dV
dV
Figure 7. Exponential ( versus
Device Physics
dt
) s
( )
dt c
gate layer reduces its benefit.
the same amount. However, the capacitor does not enhance
static thermal stability.
Sensitive gate SCRs (I GT < 200 µA) have no built-in resistor.
They should be used with an external resistor. The recommended
value of the resistor is 1000 ohms. Higher values
The maximum ⎛dV
------ ⎞
⎝
improvement occurs with a short.
dt ⎠s
reduce maximum operating temperature ⎛dV
Actual improvement stops before this because of spreading
------ ⎞
⎝
(Figure 6).
dt ⎠
resistance in the thyristor. An external capacitor of about 0.1
s
The capability of these parts varies by more than 100 to 1 µF allows the maximum enhancement at a higher value of
depending on gate-cathode termination.
R GK .
10 MEG
MCR22-2
One should keep the thyristor cool for the highest ⎛dV
------ ⎞
⎝
.
T A = 65°C
dt ⎠s
Also devices should be tested in the application circuit at the
A
10
highest possible temperature using thyristors with the lowest
1 MEG
V G
measured trigger current.
K
dV
TRIAC Commutating
dt
100K
dV
What is Commutating ?
dt
dV
The commutating ------ rating applies when a TRIAC has been
dt
conducting and attempts to turn-off with an inductive load.
10K
0.001 0.01 0.1 1 10 100 The current and voltage are out of phase (Figure 8). The
dV
TRIAC attempts to turn-off as the current drops below the
STATIC (V/µs)
dt
holding value. Now the line voltage is high and in the opposite
polarity to the direction of conduction. Successful turnoff
requires the voltage across the TRIAC to rise to the
dV
Figure 6. Exponential ( versus
dt
) s
Gate-Cathode Resistance
instantaneous line voltage at a rate slow enough to prevent
retriggering of the device.
130
R L
120
i 2
V LINE
MAC 228-10
G
V MT2-1
110
800V 110°C
1
dl
100
PHASE
dt c
90
ANGLE
φ
80
TIME
TIME
dV
70
i
V LINE
dt c
60
dV
0.001 0.01
0.1 1 Figure 8. TRIAC Inductive Load Turn-Off (
dt
)
GATE TO MT 1, CAPACITANCE (µF)
c
Gate to MT 1 Capacitance
A TRIAC functions like two SCRs connected in inverseparallel.
So, a transient of either polarity turns it on.
GATE-CATHODE RESISTANCE (OHMS)
dV
STATIC (V/µs)
dt
A gate-cathode capacitor (Figure 7) provides a shunt path for
transient currents in the same manner as the resistor. It also
filters noise currents from the drive circuit and enhances the
built-in gate-cathode capacitance voltage divider effect. The
VOLTAGE CURRENT
There is charge within the crystal’s volume because of prior
conduction (Figure 9). The charge at the boundaries of the
VMT2-1
REV. 4.01 6/24/02 3

AN-3008
APPLICATION NOTE
collector junction depletion layer responsible for ⎛dV
------ ⎞
⎝
is
dt ⎠s
also present. TRIACS have lower ⎛dV
------ ⎞ than ⎛dV
------ ⎞
⎝
because
dt ⎠c
⎝ dt ⎠s
of this additional charge.
P
N
N
N N N
REVERSE RECOVERY
CURRENT PATH
G
MT 2
+ –
MT 1
LATERAL VOLTAGE
DROP
Figure 9. TRIAC Structure and
Current Flow at Commutation
TOP
N N N
Previously
Conducting Side
STORED CHARGE
FROM POSITIVE
CONDUCTION
The volume charge storage within the TRIAC depends on
the peak current before turn-off and its rate of zero crossing
⎛dI
----⎞
⎝
. In the classic circuit, the load impedance and line
dt⎠c
frequency determine ⎛dI
----⎞
⎝
. The rate of crossing for sinusoidal
currents is given by the slope of the secant line between
dt⎠c
the 50% and 0% levels as:
⎛dI
----⎞
⎝dt⎠c
6fI TM
= --------------A/ms
1000
where f = line frequency and I TM = maximum on-state
current in the TRIAC.
Turn-off depends on both the Miller effect displacement
dV
current generated by ------ across the collector capacitance and
dt
the currents resulting from internal charge storage within the
volume of the device (Figure 10). If the reverse recovery
current resulting from both these components is high, the
lateral IR drop within the TRIAC base layer will forward
dV
bias the emitter and turn the TRIAC on. Commutating ------
dt
capability is lower when turning off from the positive direction
of current conduction because of device geometry. The
gate is on the top of the die and obstructs current flow.
Recombination takes place throughout the conduction period
and along the back side of the current wave as it declines to zero.
Turn-off capability depends on its shape. If the current amplitude
is small and its zero crossing ⎛dI
----⎞
⎝
is low, there is little
dt⎠
c
volume charge storage and turn-off becomes limited by ⎛dV
------ ⎞
⎝
.
dt ⎠s
At moderate current amplitudes, the volume charge begins
to influence turn-off, requiring a larger snubber. When the
current is large or has rapid zero crossing, ⎛dV
------ ⎞
⎝
has little
dt ⎠c
dI
influence. Commutating ---- and delay time to voltage
dt
reapplication determine whether turn-off will be successful
or not. (Figures 11,12)
VOLTAGE/CURRENT
0
Figure 10. TRIAC Current and Voltage at Commutation
NORMALIZED DELAY TIME
(td = W0 td)
VMT2-1
MAIN TERMINAL VOLTAGE (V)
0.5
0.2
0.1
0.05
0.03
0.02
VOLUME
STORAGE
CHARGE
V T
E
0
dI
dt
c
t d
IRRM
TIME
dV
dt
Figure 11. Snubber Delay Time
R L = 0
M = 1
I RRM = 0
V
DAMPING FACTOR
0.2
0.1
c
TIME
CHARGE DUE
TO dV/dt
0.001 0.002 0.005 0.01 0.02 0.05 0.1 0.2 0.3 0.5 1
V T
E
0.05
0.02
0.01
0.005
E
Figure 12. Delay Time To Normalized Voltage
4 REV. 4.01 6/24/02

APPLICATION NOTE
AN-3008
Conditions Influencing
dV
Commutating ------ depends on charge storage and recovery
dt
dV
dynamics in addition to the variables influencing static ------ .
dt
High temperatures increase minority carrier life-time and the
size of recovery currents, making turn-off more difficult.
Loads that slow the rate of current zero-crossing aid turn-off.
Those with harmonic content hinder turn-off.
Figure 13. Phase Controlling a Motor in a Bridge
Circuit Examples
Figure 13 shows a TRIAC controlling an inductive load in a
bridge. The inductive load has a time constant longer than
the line period. This causes the load current to remain constant
and the TRIAC current to switch rapidly as the line
voltage reverses. This application is notorious for causing
TRIAC turn-off difficulty because of high ⎛dI
----⎞
⎝
.
dt⎠
High currents lead to high junction temperatures and rates of
current crossings. Motors can have 5 to 6 times the normal
current amplitude at start-up. This increases both junction
temperature and the rate of current crossing leading to turnoff
problems.
The line frequency causes high rates of current crossing in
400 Hz applications. Resonant transformer circuits are doubly
periodic and have current harmonics at both the primary
and secondary resonance. Non-sinusoidal currents can lead
to turn-off difficulty even if the current amplitude is low
before zero-crossing.
Failure Mode
i
C
dV
( )
dt c
R S
L S
dl
dt
i
c DC MOTOR
– +
60 Hz R L
t
L
> 8.3 µs
R
dV
( )
dt c
⎛dV
------ ⎞
⎝
failure causes a loss of phase control. Temporary turnon
or total turn-off failure is possible. This can be destructive
dt ⎠c
if the TRIAC conducts asymmetrically causing a dc current
component and magnetic saturation. The winding resistance
limits the current. Failure results because of excessive surge
current and junction temperature.
c
dV
Improving ( )
dt c
The same steps that improve ⎛dV
------ ⎞ aid ⎛dV
------ ⎞
⎝
except when
dt ⎠s
⎝ dt ⎠c
stored charge dominates turn-off. Steps that reduce the
stored charge or soften the commutation are necessary then.
Larger TRIACS have better turn-off capability than smaller
ones with a given load. The current density is lower in the
larger device allowing recombination to claim a greater proportion
of the internal charge. Also junction temperatures are
lower.
TRIACS with high gate trigger current have greater turn-off
ability because of lower spreading resistance in the gate
layer, reduced Miller effect, or shorter lifetime.
The rate of current crossing can be adjusted by adding a
commutation softening inductor in series with the load.
Small high permeability “square loop” inductors saturate
causing no significant disturbance to the load current. The
inductor resets as the current crosses zero introducing a large
inductor into the snubber circuit at that time. This slows the
current crossing and delays the reapplication of blocking
voltage aiding turn-off.
The commutation inductor is a circuit element that introduces
time delay, as opposed to inductance, into the circuit.
dV
It will have little influence on observed ------ at the device.
dt
The following example illustrates the improvement resulting
from the addition of an inductor constructed by winding 33
turns of number 18 wire on a tape wound core (52000 -1A).
This core is very small having an outside diameter of 3/4
inch and a thickness of 1/8 inch. The delay time can be calculated
from:
( N A B 10 – 8 )
t s = -------------------------------- where:
E
t s = time delay to saturation in seconds.
B = saturating flux density in Gauss
A = effective core cross sectional area in cm 2
N = number of turns.
For the described inductor:
t s = (33 turns)(0.076 cm 2 )(28000 Gauss)(1 x 10 -8 )/(175 v) = 4.0 µs.
The saturation current of the inductor does not need to be
much larger than the TRIAC trigger current. Turn-off failure
will result before recovery currents become greater than this
value. This criterion allows sizing the inductor with the
following equation:
REV. 4.01 6/24/02 5

AN-3008
APPLICATION NOTE
I S
H S M L
= ------------------ where:
0.4 π N
H S = MMF to saturate = 0.5 Oersted
ML = mean magnetic path length = 4.99 cm.
(.5) (4.99)
I S = ----------------------- = 60 mA.
0.4 π 33
Snubber Physics
Undamped Natural Resonance
I
ω 0 = -----------Radians/second
LC
dV
Resonance determines ------ and boosts the peak capacitor
dt
voltage when the snubber resistor is small. C and L are
related to one another by ω 2 dV
0 . ------ scales linearly with ω
dt
0
when the damping factor is held constant. A ten to one
dV
reduction in ------ requires a 100 to 1 increase in either
dt
component.
Damping Factor
ρ =
The damping factor is proportional to the ratio of the circuit
loss and its surge impedance. It determines the trade off
dV
between ------ and peak voltage. Damping factors between
dt
0.01 and 1.0 are recommended.
The Snubber Resistor
dV
Damping and
dt
dV
When ρ < 0.5, the snubber resistor is small, and ------ depends
dt
dV
mostly on resonance. There is little improvement in ------ for
dt
damping factors less than 0.3, but peak voltage and snubber
discharge current increase. The voltage wave has a 1-COS(θ)
dV
shape with overshoot and ringing. Maximum ------ occurs at a
dt
time later than t = 0. There is a time delay before the
voltage rise, and the peak voltage almost doubles.
When ρ > 0.5, the voltage wave is nearly exponential in
dV
shape. The maximum instantaneous ------ occurs at t = 0.
dt
There is little time delay and moderate voltage overshoot.
R
---
2
dV
When ρ > 1.0, the snubber resistor is large and ------ depends
dt
mostly on its value. There is some overshoot even though the
circuit is overdamped.
C
---
L
High load inductance requires large snubber resistors and
small snubber capacitors. Low inductances imply small
resistors and large capacitors.
Damping and Transient Voltages
Figure 14 shows a series inductor and filter capacitor
connected across the ac main line. The peak to peak voltage
of a transient disturbance increases by nearly four times.
Also the duration of the disturbance spreads because of
ringing, increasing the chance of malfunction or damage to
the voltage sensitive circuit. Closing a switch causes this
behavior. The problem can be reduced by adding a damping
resistor in series with the capacitor.
+700
V (VOLTS)
0
-700
0
10µs
340 V
100µH 0.05
0.1
µF
0 10 20
TIME (µs)
Figure 14. Undamped LC Filter
Magnifies and Lengthens a Transient
dl
dt
Non-Inductive Resistor
The snubber resistor limits the capacitor discharge current
dI
dI
and reduces ---- stress. High ---- destroys the thyristor even
dt
dt
though the pulse duration is very short. The rate of current
rise is directly proportional to circuit voltage and inversely
proportional to series inductance. The snubber is often the
major offender because of its low inductance and close
proximity to the thyristor.
With no transient suppressor, breakdown of the thyristor sets
the maximum voltage on the capacitor. It is possible to
exceed the highest rated voltage in the device series because
high voltage devices are often used to supply low voltage
specifications.
The minimum value of the snubber resistor depends on the
type of thyristor, triggering quadrants, gate current amplitude,
voltage, repetitive or non-repetitive operation and
required life expectancy. There is no simple way to predict
the rate of current rise because it depends on turn-on speed
of the thyristor, circuit layout, type and size of snubber
capacitor, and inductance in the snubber resistor. The
equations in Appendix D describes the circuit. However,
V
VOLTAGE
SENSITIVE
CIRCUIT
6 REV. 4.01 6/24/02

APPLICATION NOTE
AN-3008
the values required for the model are not easily obtained
except by testing. Therefore, reliability should be verified in
the actual application circuit.
Table 1 shows suggested minimum resistor values estimated
(Appendix A) by testing a 20 piece sample from the four
different TRIAC die sizes.
Table 1. Minimum Non-inductive Snubber Resistor for
Four Quadrant Triggering.
Peak V C
TRIAC Type Volts
Non-Sensitive 200
Gate
300
(I GT > 10mA) 400
8 to 40 A (RMS) 600
800
R S Ohms
3.3
6.8
11
39
51
dl
dt
A/µS
170
250
308
400
400
dl
Reducing
dt
dI
TRIAC ---- can be improved by avoiding quadrant 4
dt
triggering. Most optocoupler circuits operate the TRIAC in
quadrants 1 and 3. Integrated circuit drivers use quadrants 2
and 3. Zero crossing trigger devices are helpful because they
prohibit triggering when the voltage is high.
Driving the gate with a high amplitude fast rise pulse
dI
increases ---- capability. The gate ratings section defines the
dt
maximum allowed current.
dI
Inductance in series with the snubber capacitor reduces ---- .
dt
It should not be more than five percent of the load inductance
dV
to prevent degradation of the snubber’s ------ suppression
dt
capability. Wirewound snubber resistors sometimes serve
this purpose. Alternatively, a separate inductor can be added
in series with the snubber capacitor. It can be small because
it does not need to carry the load current. For example, 18
turns of AWG No. 20 wire on a T50-3 (1/2 inch) powdered
iron core creates a non-saturating 6.0 µH inductor.
A 10 ohm, 0.33 µF snubber charged to 650 volts resulted in a
dI
1000 A/µs ---- . Replacement of the non-inductive snubber
dt
resistor with a 20 watt wirewound unit lowered the rate of
rise to a non-destructive 170 A/µs at 800 V. The inductor
gave an 80 A/µs rise at 800 V with the non-inductive resistor.
The Snubber Capacitor
A damping factor of 0.3 minimizes the size of the snubber
dV
capacitor for a given value of ------ . This reduces the cost and
dt
physical dimensions of the capacitor. However, it raises
voltage causing a counter balancing cost increase.
Snubber operation relies on the charging of the snubber
capacitor. Turn-off snubbers need a minimum conduction
angle long enough to discharge the capacitor. It should be at
least several time constants (R S C S ).
Stored Energy
Inductive Switching Transients
1 2
E -- L I
2 O Watt-seconds or Joules
I O = current in Amperes flowing in the inductor at t = 0.
=
R
V PK
C
Resonant charging cannot boost the supply voltage at turnoff
by more than 2. If there is an initial current flowing in the
load inductance at turn-off, much higher voltages are possible.
Energy storage is negligible when a TRIAC turns off
because of its low holding or recovery current.
The presence of an additional switch such as a relay, thermostat
or breaker allows the interruption of load current and the
generation of high spike voltages at switch opening. The
energy in the inductance transfers into the circuit capacitance
and determines the peak voltage (Figure 15).
dV
dt
L
I
= V PK = I
C
L
C
(a.) Protected Circuit
I
OPTIONAL
FAST
SLOW
(b.) Unprotected Circuit
Figure 15. Interrupting Inductive Load Current
Capacitor Discharge
⎛ 1
The energy stored in the snubber capacitor E C = --CV 2 ⎞
⎝ 2 ⎠
transfers to the snubber resistor and thyristor every time it
turns on. The power loss is proportional to frequency
(P AV = 120 E C @ 60 H Z ).
Current Diversion
The current flowing in the load inductor cannot change
instantly. This current diverts through the snubber resistor
dV
causing a spike of theoretically infinite ------ with magnitude
dt
equal to (I RRM R) or (I H R).
Load Phase Angle
Highly inductive loads cause increased voltage and ⎛dV
------ ⎞
⎝
at
dt ⎠c
turn-off. However, they help to protect the thyristor from
transients and ⎛dV
------ ⎞
⎝
. The load serves as the snubber
dt ⎠
s
REV. 4.01 6/24/02 7

AN-3008
APPLICATION NOTE
inductor and limits the rate of inrush current if the device
dV
does turn on. Resistance in the load lowers ------ and V
dt PK
(Figure 16).
/ (E W0)
dV
dt
dV
dt
NORMALIZED
1.4 2.2
E
dV 2.1
dt
1.2
2
1.9
1
1.8
M = 1
M = 0.75
1.7
V PK
0.8
0.6
0.4
0.2
M = R S / (R L + R S)
0
0 0.2 0.4 0.6 0.8 1
DAMPING FACTOR
R
M = RESISTIVE DIVISION RATIO = S
R L + R S
I RRM = 0
Figure 16. 0 to 63%
M = 0.5
M = 0.25
M = 0
dV
dt
Characteristics Voltage Waves
NORMALIZED PEAK VOLTAGE
VPK/E
Damping factor and reverse recovery current determine the
shape of the voltage wave. It is not exponential when the
snubber damping factor is less than 0.5 (Figure 17) or when
significant recovery currents are present.
VMT2-1 (VOLTS)
500
400
ρ = 0 ρ = 0.1
300
0.1
200 0.3
1
ρ = 0.3 ρ = 1
100
0
0
0
0.7 1.4 2.1 2.8 3.5 4.2 4.9 5.6 6.3 7
TIME (µs)
dV
0-63% = 100 V/µs, E = 250 V,
dt s
R L = 0, I RRM = 0
1.6
1.5
1.4
1.3
1.2
1.1
1
0.9
NORMALIZED PEAK VOLTAGE AND dV
dt
2.8
2.6
2.4
2.2
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
Figure 18. Trade-Off Between V PK and
dV
A variety of wave parameters (Figure 18) describe ------ .
dt
Some are easy to solve for and assist understanding. These
dV
dV
include the initial ------ , the maximum instantaneous ------ , and
dt
dt
dV
the average ------ to the peak reapplied voltage. The 0 to 63%
dt
⎛dV
------ ⎞ and 10 to 63% ⎛dV
------ ⎞
⎝
definitions on device data sheets
dt ⎠s
⎝ dt ⎠c
are easy to measure but difficult to compute.
Non-Ideal Behaviors
Core Losses
10-63%
dV
dt
0
0 0.2 0.4 0.6 0.8 1.0 2.2 1.4 1.6 1.8 2.0
DAMPING FACTOR ( ρ )
(R L = 0, M = 1, I RRM = 0)
NORMALIZED
dV
=
dt
dV
dt
dV
dt
E
dV/dt
E ω 0
MAX
o
0-63%
dV
dt
V PK
10-63%
NORMALIZED V PK
= VPK
E
dV
dt
The magnetic core materials in typical 60 Hz loads introduce
losses at the snubber natural frequency. They appear as a
resistance in series with the load inductance and winding dc
dV
resistance (Figure 19). This causes actual ------ to be less than
dt
the theoretical value.
Figure 17. Voltage Waves for Different Damping Factors
8 REV. 4.01 6/24/02

APPLICATION NOTE
AN-3008
L
C
R
Core remanence and saturation cause surge currents. They
depend on trigger angle, line impedance, core characteristics,
and direction of the residual magnetization. For example,
a 2.8 kVA 120 V 1:1 transformer with a 1.0 ampere load
produced 160 ampere currents at start-up. Soft starting the
circuit at a small conduction angle reduces this current.
L DEPENDS ON CURRENT AMPLITUDE, CORE
SATURATION
R INCLUDES CORE LOSS, WINDING R. INCREASES
WITH FREQUENCY
C WINDING CAPACITANCE. DEPENDS ON
INSULATION, WIRE SIZE, GEOMETRY
Complex Loads
Figure 19. Inductor Model
Many real-world inductances are non-linear. Their core
materials are not gapped causing inductance to vary with
current amplitude. Small signal measurements poorly
characterize them. For modeling purposes, it is best to
measure them in the actual application.
Complex load circuits should be checked for transient voltages
and currents at turn-on and turn-off. With a capacitive
load, turn-on at peak input voltage causes the maximum
surge current. Motor starting current runs 4 to 6 times the
steady state value. Generator action can boost voltages above
the line value. Incandescent lamps have cold start currents 10
to 20 times the steady state value. Transformers generate
voltage spikes when they are energized. Power factor correction
circuits and switching devices create complex loads. In most
cases, the simple CRL model allows an approximate snubber
design. However, there is no substitute for testing and measuring
the worst case load conditions.
Surge Currents in Inductive Circuits
Inductive loads with long L/R time constants cause asymmetric
multi-cycle surges at start up (Figure 20). Triggering
at zero voltage crossing is the worst case condition. The
surge can be eliminated by triggering at the zero current
crossing angle.
i (AMPERES)
90
0
240
VAC
40
20 MHY
i
0.1
Ω
ZERO VOLTAGE TRIGGERING, I RMS = 30 A
80 120 160 200
TIME (MILLISECONDS)
Figure 20. Start-Up Surge For Inductive Circuit
Transformer cores are usually not gapped and saturate
easily. A small asymmetry in the conduction angle causes
magnetic saturation and multi-cycle current surges.
Steps to achieve reliable operation include:
1. Supply sufficient trigger current amplitude. TRIACS
have different trigger currents depending on their quadrant
of operation. Marginal gate current or optocoupler
LED current causes halfwave operation.
2. Supply sufficient gate current duration to achieve
latching. Inductive loads slow down the main terminal
current rise. The gate current must remain above the
specified I GT until the main terminal current exceeds the
latching value. Both a resistive bleeder around the load
and the snubber discharge current help latching.
3. Use a snubber to prevent TRIAC ⎛dV
------ ⎞
⎝
failure.
dt ⎠c
4. Minimize designed-in trigger asymmetry. Triggering
must be correct every half-cycle including the first.
Use a storage scope to investigate circuit behavior
during the first few cycles of turn-on. Alternatively,
get the gate circuit up and running before energizing
the load.
5. Derive the trigger synchronization from the line instead
of the TRIAC main terminal voltage. This avoids regenerative
interaction between the core hysteresis and the
triggering angle preventing trigger runaway, halfwave
operation, and core saturation.
6. Avoid high surge currents at start-up. Use a current
probe to determine surge amplitude. Use a soft start
circuit to reduce inrush current.
Distributed Winding Capacitance
There are small capacitances between the turns and layers of
a coil. Lumped together, they model as a single shunt capacitance.
The load inductor behaves like a capacitor at frequencies
above its self-resonance. It becomes ineffective in
dV
controlling ------ and V PK when a fast transient such as that
dt
resulting from the closing of a switch occurs. This problem
can be solved by adding a small snubber across the line.
Self-Capacitance
dV
A thyristor has self-capacitance which limits ------ when the
dt
load inductance is large. Large load inductances, high power
factors, and low voltages may allow snubberless
operation.
REV. 4.01 6/24/02 9

AN-3008
Snubber Examples
Without Inductance
Power TRIAC Example
Figure 2l shows a transient voltage applied to a TRIAC
controlling a resistive load. Theoretically there will be an
instantaneous step of voltage across the TRIAC. The only
elements slowing this rate are the inductance of the wiring
and the self-capacitance of the thyristor. There is an exponential
capacitor charging component added along with a
decaying component because of the IR drop in the snubber
resistor. The non-inductive snubber circuit is useful when the
load resistance is much larger than the snubber resistor.
R L
10V/µs
R in 1
6 180 2.4k
V CC
2 MOC R 1 R 2
3020 0.1µF C1
3021 4
0.63 (170)
φ CNTL
DESIGN
dV
dt
APPLICATION NOTE
T2322D
1V/µs
(0.63) (170)
= = 0.45V/µs
(2400) (0.1µF)
TIME
240µs
dV
dt (V/µs)
Power TRIAC
Optocoupler
0.99 0.35
1A, 60Hz
L = 318 MHY
170V
E
e
R S
C S
Figure 22. Single Snubber For Sensitive Gate TRIAC and
Phase Controllable Optocoupler (ρ = 0.67)
e
E
t = 0
V step = E
R S
R S + R L
TIME
τ = (R L + R S) C S
e(t = o+) = E
R S
R S + R L
e – t/τ + (1 – e – t/τ )
RESISTOR CAPACITOR
COMPONENT COMPONENT
The optocoupler conducts current only long enough to trigger
the power device. When it turns on, the voltage between
MT2 and the gate drops below the forward threshold voltage
of the opto-TRIAC causing turn-off. The optocoupler sees
⎛dV
------ ⎞
⎝
when the power TRIAC turns off later in the conduction
cycle at zero current crossing. Therefore, it is not neces-
dt ⎠s
sary to design for the lower optocoupler ⎛dV
------ ⎞
⎝
rating. In this
dt ⎠c
example, a single snubber designed for the optocoupler
protects both devices.
Figure 21. Non-inductive Snubber Circuit
Opto-TRIAC Examples
Single Snubber, Time Constant Design
Figure 22 illustrates the use of the RC time constant design
method. The optocoupler sees only the voltage across the
snubber capacitor. The resistor R 1 supplies the trigger
current of the power TRIAC. A worst case design procedure
assumes that the voltage across the power TRIAC changes
instantly. The capacitor voltage rises to 63% of the maximum
in one time constant. Then:
0.63 E dV
R 1 C S = τ = --------------- where ⎛------
⎞ is the rated static dV ------
⎛dV
⎝
------ ⎞ dt ⎠s
dt
⎝ dt ⎠s
for the optocoupler.
V CC
1
2
3
MOC3031
4
5
6
100
1N4001
51
MCR265-4
MCR265-4
100
1N4001
(50 V/µs SNUBBER, ρ = 1.0)
0.022
µF
Figure 23. Anti-Parallel SCR Driver
1MHY
430
120V
400 Hz
Octocouplers with SCRs
dV
Anti-parallel SCR circuits result in the same ------ across the
dt
optocoupler and SCR (Figure 23). Phase controllable optocouplers
require the SCRs to be snubbed to their lower ------
dV
dt
rating. Anti-parallel SCR circuits are free from the charge
storage behaviors that reduce the turn-off capability of
TRIACs. Each SCR conducts for a half-cycle and has the
next half cycle of the ac line in which to recover. The turn-off
dV
------ of the conducting SCR becomes a static forward blocking
------ for the other device. Use the SCR data sheet ⎛dV
------ ⎞
dt
dV
dt
⎝ dt ⎠s
rating in the snubber design.
10 REV. 4.01 6/24/02

APPLICATION NOTE
AN-3008
A SCR used inside a rectifier bridge to control an ac load
will not have a half cycle in which to recover. The available
time decreases with increasing line voltage. This makes the
circuit less attractive. Inductive transients can be suppressed
by a snubber at the input to the bridge or across the SCR.
However, the time limitation still applies.
The current through the snubber resistor is:
t
V ⎛ – -
τ⎞ i ----- ⎜1–
e ⎟ ,
R τ ⎝ ⎠
and the voltage across the TRAIC is:
=
dV
( ) c
dV
Opto
dt
Zero-crossing optocouplers can be used to switch inductive
loads at currents less than 100 mA (Figure 24). However a
power TRIAC along with the optocoupler should be used for
higher load currents.
LOAD CURRENT (mA RMS)
80
70
60
50
40
30
20
10
C S = 0.01
C S = 0.001
NO SNUBBER
0
20 30 40 50 60 70 80 90 100
T A, AMBIENT TEMPERATURE (°C)
(R S = 100Ω, V RMS = 220V, POWER FACTOR = 0.5)
Figure 24. MOC3062 Inductive Load Current versus T A
A phase controllable optocoupler is recommended with a
power device. When the load current is small, a MAC97
TRIAC is suitable.
Unusual circuit conditions sometimes lead to unwanted
operation of an optocoupler in ⎛dV
------ ⎞
⎝
mode. Very large
dt ⎠c
currents in the power device cause increased voltages
between MT2 and the gate that hold the optocoupler on.
Use of a larger TRIAC or other measures that limit inrush
current solve this problem.
Very short conduction times leave residual charge in the
optocoupler. A minimum conduction angle allows recovery
before voltage reapplication.
The Snubber with Inductance
Consider an overdamped snubber using a large capacitor
whose voltage changes insignificantly during the time under
consideration. The circuit reduces to an equivalent L/R series
charging circuit.
e = i R S
The voltage wave across the TRIAC has an exponential rise
with maximum rate at t = 0. Taking its derivative gives its
value as:
⎛------
⎞
⎝ dt ⎠0
VR
= ---------- S
L
Highly overdamped snubber circuits are not practical
designs. The example illustrates several properties:
1. The initial voltage appears completely across the circuit
inductance. Thus, it determines the rate of change of
dV
current through the snubber resistor and the initial ------ .
dt
This result does not change when there is resistance in
the load and holds true for all damping factors.
2. The snubber works because the inductor controls the
rate of current change through the resistor and the rate
of capacitor charging. Snubber design cannot ignore the
inductance. This approach suggests that the snubber
capacitance is not important but that is only true for
this hypothetical condition. The snubber resistor shunts
the thyristor causing unacceptable leakage when the
capacitor is not present. If the power loss is tolerable,
dV
------ can be controlled without the capacitor. An example
dt
is the soft-start circuit used to limit inrush current in
switching power supplies (Figure 25).
E
E
AC LINE SNUBBER
L
AC LINE SNUBBER
L
Snubber with no C
RECTIFIER
BRIDGE
dV
dt
R S
= ERS
φ L
R S
Figure 25. Surge Current Limiting
For a Switching Power Supply
G
G
RECTIFIER
BRIDGE
C1
C1
REV. 4.01 6/24/02 11

AN-3008
APPLICATION NOTE
TRIAC Design Procedure
dV
(
dt
) c
dV
1. Refer to Figure 18 and select a particular damping factor
(ρ)giving a suitable trade-off between V PK and ------ .
dt
dV
Determine the normalized ------ corresponding to the
dt
chosen damping factor.
The volage E depends on the load phase angle:
E 2 V RMS Sin (φ) where φ = tan – 1 X =
⎛------
L ⎞ where
⎝R L
⎠
φ = measured phase angle between line V and load I
R L = measured dc resistance of the load.
dV
( Safe Area Curve
dt
) c
dV
Figure 26 shows a MAC16 TRIAC turn-off safe operating
area curve. Turn-off occurs without problem under the curve.
The region is bounded by static ------ at low values of ⎛dI
----⎞
dt
⎝dt⎠c
and delay time at high currents. Reduction of the peak current
permits operation at higher line frequency. This TRIAC
operated at f = 400 Hz, T J = 125 °C, and I TM = 6.0 amperes
using a 30 ohm and 0.068 µF snubber. Low damping factors
extend operation to higher ⎛dI
----⎞
⎝
, but capacitor sizes
dt⎠c
increase. The addition of a small, saturable commutation
inductor extends the allowed current rate by introducing
recovery delay time.
Then
V RMS 2 2
Z -------------- R
I L + X L X L Z 2 2
= = – R L and
RMS
X
L = --------------------- L
2 π f Line
If only the load current is known, assume a pure
inductance. This gives a conservative design. Then:
V RMS
L = --------------------------------------- where E = 2 V
2 π f LINE I RMS
RMS
For Example:
120
E = 2 120 = 170 V; L = ----------------------------------- = 39.8 mH
( 8 A) ( 377 rps)
(V/µs)
100
10
c
dV
dt
1
0.1
10
dl
dt
-ITM = 15A
c = 6 f ITM x 10-3 A/ms
WITH COMMUTATION L
14 18 22 26 30 34 38 42 46 50
dV
AMPERES/MILLISECOND
dt c
MAC 16-8, COMMUTATIONAL L = 33 TURNS #18,
52000-1A TAPE WOUND CORE 3/4 INCH OD
Read from the graph at ρ = 0.6, V PK = (1.25) 170 = 213V.
dV
Use 400V TRIAC. Read ------ = 1.0
dt ( ρ = 0.6)
2. Apply the resonance criterion:
ω ⎛ 0 spec dV ------ ⎞ dV
= ⁄ ⎛------ E⎞
⎝ dt ⎠ ⎝ dt ( P)
⎠
510 × 6 V/S
ω 0 = ---------------------------- =
( 1) ( 170 V)
29.4 × 10 3 rps
1
C = ------------- = 0.029 µF
2
ω 0 L
3. Apply the damping criterion:
R S 2ρ L 39.8 × 10 – 3
= --- = 2( 0.6)
C ------------------------------ 0.029 × 10 – 6 = 1400 ohms
Static
dl
( ) c
( ) c
dV
Figure 26.
dt
versus
dt
T J = 125 °C
dV
dt
Design
There is usually some inductance in the ac main and power
wiring. The inductance may be more than 100 µH if there is
a transformer in the circuit or nearly zero when a shunt
power factor correction capacitor is present. Usually the line
inductance is roughly several µH. The minimum inductance
must be known or defined by adding a series inductor to
insure reliable operation (Figure 27).
One hundred µH is a suggested value for starting the design.
Plug the assumed inductance into the equation for C. Larger
values of inductance result in higher snubber resistance and
dI
reduced ---- . For example:
dt
12 REV. 4.01 6/24/02

APPLICATION NOTE
AN-3008
100 µH
20A
LS1
10 0.33 µF
< 50 V/µs
MAC 218-6
8A LOAD
R
68Ω
0.033 µF
L
120V
60Hz
340
V
12Ω
HEATER
dV
dt
s
= 100 V/µs
dV
= 5 V/µs
dt c
Figure 27. Snubbing for a Resistive Load
Given E = 240 2 = 340V
Pick ρ = 0.3
Then from Figure 18, V PK = 1.42 (340) = 483 V.
Thus, it will be necessary to use a 600 V device. Using the
previously stated formulas for ω 0 , C and R we find:
5 0× 10 6 V/S
ω 0 = ---------------------------------- = 201450 rps
( 0.73) ( 340 V)
1
C = --------------------------------------------------------
( 201450) 2 ( 100 × 10 – 6 = 0.2464µF
)
100 × 10 – 6
R = 2( 0.3)
--------------------------------- 0.2464 × 10 – 6 = 12ohms
Variable Loads
The snubber should be designed for the smallest load
dV
inductance because ------ will then be highest because of its
dt
dependance on ω 0 . This requires a higher voltage device for
operation with the largest inductance because of the corresponding
low damping factor.
dV
Figure 28 describes ------ for an 8.0 ampere load at various
dt
power factors. The minimum inductance is a component
dV
added to prevent static ------ firing with a resistive load.
dt
Figure 28. Snubbing For a Variable Load
Examples of Snubber Designs
Table 2 describes snubber RC values for ⎛dV
------ ⎞
⎝
. Figures 31
dt ⎠s
and 32 show possible R and C values for a 5.0 V/µs ⎛dV
------ ⎞
⎝ dt ⎠c
assuming a pure inductive load.
dV
Table 2. Static Designs
dt
(E = 340 V, V peak = 500 V, ρ = 0.3)
L
µH
dV
R L V step V PK
ρ
dt
Ω MHY V V V/µs
0.75 15 0.1 170 191 86
0.03 0 39.8 170 325 4.0
0.04 10.6 28.1 120 225 3.3
0.06 13.5 17.3 74 136 2.6
C
µF
5.0V/µs 50V/µs 100V/µs
R
Ohm
C
µF
R
Ohm
C
µF
R
Ohm
47 0.15 10
100 0.33 10 0.1 20
220 0.15 22 0.03 47
3
500 0.06 51 0.01 110
8 5
100 3.0 11 0.03 100
0 3
REV. 4.01 6/24/02 13

AN-3008
Transient and Noise Suppression
Transients arise internally from normal circuit operation or
externally from the environment. The latter is particularly
frustrating because the transient characteristics are undefined.
A statistical description applies. Greater or smaller
stresses are possible. Long duration high voltage transients
are much less probable than those of lower amplitude and
higher frequency. Environments with infrequent lightning
and load switching see transient voltages below 3.0 kV.
APPLICATION NOTE
The natural frequencies and impedances of indoor ac wiring
result in damped oscillatory surges with typical frequencies
ranging from 30 kHz to 1.5 MHz. Surge amplitude depends
on both the wiring and the source of surge energy. Disturbances
tend to die out at locations far away from the source.
Spark-over (6.0 kV in indoor ac wiring) sets the maximum
voltage when transient suppressors are not present. Transients
closer to service entrance or in heavy wiring have
higher amplitudes, longer durations, and more damping
because of the lower inductance at those locations.
10K
1000
RS(OHMS)
100
0.6A RMS
2.5A
5A
10A
20A
40A
80A
The simple CRL snubber is a low pass filter attenuating
frequencies above its natural resonance. A steady state sinusoidal
input voltage results in a sine wave output at the same
frequency. With no snubber resistor, the rate of roll off
approaches 12 dB per octave. The corner frequency is at
the snubber’s natural resonance. If the damping factor is
low, the response peaks at this frequency. The snubber
resistor degrades filter characteristics introducing an up-turn
at ω = 1/(RC). The roll-off approaches 6.0 dB/octave at
frequencies above this. Inductance in the snubber resistor
further reduces the roll-off rate.
WITH 5µHY
Figure 32 describes the frequency response of the circuit
in Figure 27. Figure 31 gives the theoretical response to a
3.0 kV 100 kHz ring-wave. The snubber reduces peak voltage
across the thyristor. However, the fast rise input causes a
10
dV
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 high ------ step when series inductance is added to the snubber
dt
DAMPING FACTOR
resistor. Limiting the input voltage with a transient suppressor
dV
Figure 30. Snubber Capacitor For (
dt
) = 5.0 V/µs
Figure 32. Snubber Frequency Response
c
PURE INDUCTIVE LOAD, V = 120 V RMS,
reduces the step.
I RRM = 0
400
dV
WITHOUT 5 µHY
Figure 29. Snubber Resistor For (
dt
) = 5.0 V/µs
c
WITH 5 5 µHY AND
450V MOV
AT AC INPUT
1
0
WITH 5 µHY
80A RMS
-400
0 1 2 3 4 5 6
40A
TIME (µs)
0.1
20A
Figure 31. Theoretical Response of Figure 33
Circuit to 3.0 kV IEEE 587 Ring Wave (R SC = 27.5 Ω)
10A
5A
+10
0.01
2.5A
0
-10
0.6A
100µH
0.001
V
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
in V out
-30
10
DAMPING FACTOR
12 0.33µF WITHOUT 5µHY
-20
5µH
-40
PURE INDUCTIVE LOAD, V = 120 V RMS,
10K
100K
I RRM = 0
FREQUENCY (Hz)
1M
CS(µF)
14 REV. 4.01 6/24/02
V MT2-1 (VOLTS)
VOLTAGE GAIN (dB)
V
(
out
V in
)

APPLICATION NOTE
AN-3008
dV
The noise induced into a circuit is proportional to ------ when
dt
dI
coupling is by stray capacitance, and ---- when the coupling
dt
is by mutual inductance. Best suppression requires the use of
a voltage limiting device along with a rate limiting CRL
snubber. The thyristor is best protected by preventing turn-on
dV
from ------ or breakover. The circuit should be designed for
dt
what can happen instead of what normally occurs.
In Figure 30, a MOV connected across the line protects
many parallel circuit branches and their loads. The MOV
dI
defines the maximum input voltage and ---- through the load.
dt
dV
With the snubber it sets the maximum ------ and peak voltage
dt
across the thyristor. The MOV must be large because there is
little surge limiting impedance to prevent its burn-out.
V MAX
In Figure 32, there is a separate suppressor across each
thyristor. The load impedance limits the surge energy delivered
from the line. This allows the use of a smaller device
but omits load protection. This arrangement protects each
thyristor when its load is a possible transient source.
Figure 34. Limiting Thyristor Voltage
It is desirable to place the suppression device directly
across the source of transient energy to prevent the induction
of energy into other circuits. However, there is no protection
for energy injected between the load and its controlling
thyristor. Placing the suppressor directly across each
thyristor positively limits maximum voltage and snubber
dI
discharge ---- .
dt
Figure 33. Limiting Line Voltage
REV. 4.01 6/24/02 15

AN-3008
APPLICATION NOTE
Examples of Snubber Applications
REV
In Figure 35, TRIACs switch a 3 phase motor on and off and
reverse its rotation. Each TRIAC pair functions as a SPDT
switch. The turn-on of one TRIAC applies the differential
voltage between line phases across the blocking device without
the benefit of the motor impedance to constrain the rate
dV
of voltage rise. The inductors are added to prevent static ------
dt
firing and a line-to-line short.
115
91 0.1
91
R S
T2323D
FWD
0.1
C S
46 V/µs
MAX
L S
3.75
330V
500µH 5.6
MOTOR
1/70 HP
0.26A
Figure 36 shows a split phase capacitor-run motor with
reversing accomplished by switching the capacitor in series
with one or the other winding. The forward and reverse
TRIACs function as a SPDT switch. Reversing the motor
applies the voltage on the capacitor abruptly across the
blocking thyristor. Again, the inductor L is added to prevent
⎛dV
------ ⎞
⎝
firing of the blocking TRIAC. If turn-on occurs,
dt ⎠s
the forward and reverse TRIACs short the capacitors (C S )
resulting in damage to them. It is wise to add the resister R S
to limit the discharge current.
Figure 36. Split Phase Reversing Motor
Figure 37 shows a “tap changer.” This circuit allows the
operation of switching power supplies from a 120 or 240
vac line. When the TRIAC is on, the circuit functions as a
conventional voltage doubler with diodes D 1 and D 2 conducting
on alternate half-cycles. In this mode of operation,
dI
inrush current and ---- are hazards to TRIAC reliability.
dt
Series impedance is necessary to prevent damage to the
TRIAC.
SNUBBER
φ1
100µH
300
4
MOC
3081
FWD
2 1
6
G
91
22Ω
2W
WIREWOUND
0.15
µF
SNUBBER
300
4
MOC
3081
REV
2 1
6
G
SNUBBER
91
SNUBBER
ALL MOV’s ARE 275
V RMS
ALL TRIACS ARE
MAC218-10
1/3 HP
208V
3 PHASE
91
φ2
100µH
300
4
MOC
3081
FWD
2 1
6
G
SNUBBER
91
6
MOC
3081
4
43
G
1
2
SNUBBER
2 1
φ3
300
4
MOC
3081
REV
6
G
91
N
Figure 35. 3 Phase Reversing Motor
16 REV. 4.01 6/24/02

APPLICATION NOTE
The TRIAC is off when the circuit is not doubling. In this
state, the TRIAC sees the difference between the line voltage
and the voltage at the intersection of Cl and C2. Transients
on the line cause ⎛dV
------ ⎞
⎝
firing of the TRIAC. High inrush
dt ⎠s
dI
current, ---- and overvoltage damage to the filter capacitor are
dt
possibilities. Prevention requres the addition of a RC snubber
across the TRIAC and an inductor in series with the line.
Figure 37. Tap Changer For Dual
Voltage Switching Power Supply
Thyristor Types
Sensitive gate thyristors are easy to turn-on because of their
low trigger current requirements. However, they have less
dV
------ capability than similar non-sensitive devices. A nonsensitive
thyristor should be used for high ------ .
dt
dV
dt
TRIAC commutating
static
dV
------
dt
SUBBER INDUCTOR
120 VAC
OR
240 VAC
ratings.
D
D 1
2
+
C 1
–
D 3
D 4
240V
0
+
C
120V
2
–
dV
------
dt
R S C S
G
R L
ratings are 5 to 20 times less than
Appendix A
Testing Snubber Discharge
dI
dt
AN-3008
The equations in Appendix D do not consider the thyristor’s
turn-on time or on-state resistance, thus, they predict high
dI
values of ---- .
dt
Figure 38 shows the circuit used to test snubber discharge
dI
---- . A MBS4991 supplies the trigger pulse while the
dt
quadrants of operation are switch selectable. The snubber
was mounted as close to the TRIAC under test as possible to
reduce inductance, and the current transformer remained in
the circuit to allow results to be compared with the measured
dI
---- value.
dt
What should the peak capacitor voltage be? A conservative
approach is to test at maximum rated V DRM , or the clamp
voltage of the MOV.
What is the largest capacitor that can be used without limiting
resistance? Figure 39 is a photo showing the the current
pulse resulting from a 0.001 µF capacitor charged to 800 V.
dI
The 1200 A/µs ---- destroyed the TRIAC.
dt
Is it possible for MOV self-capacitance to damage the
TRIAC? A large 40 Joule, 2200 A peak current rated MOV
was tested. The MOV measured 440 pF and had an 878 volt
breakover voltage. Its peak discharge current (12 A) was half
that of a 470 pF capacitor. This condition was safe.
dV
Phase controllable optocouplers have lower ------ ratings than
dt
zero crossing optocouplers and power TRIACS. These
should be used when a dc voltage component is present, or to
prevent turn-on delay.
dV
Zero crossing optocouplers have more ------ capability than
dt
power thyristors; and they should be used in place of phase
controllable devices in static switching applications.
REV. 4.01 6/24/02 17

AN-3008
APPLICATION NOTE
120 VAC 60Hz
SWEEP FOR
DESIRED V Ci
VTC
5-50
2.5kV
2000
500W
5000
200W
X100
V PROBE
R S
C S
OPTIONAL MOV
OPTIONAL PEARSON
411 CURRENT
TRANSFORMER
MT 2
G
MT 1
56
50 Ω
TRIAC
UNDER TEST
91
3000
VMT 2-1
2 1
3 4
QUADRANT
MAP
VG
12 V
Q 1,3 Q 2,4
QUADRANT
SWITCH
1
µF
MBS4991
Figure 38. Snubber Discharge
dI
dt
Test
HORIZONTAL SCALE – 50 ms/DIV.
VERTICAL SCALE – 10 A/DIV.
C S = 0.001 µF, V Ci = 800 V, R S = 0, L = 250 mH, R TRIAC = 10 OHMS
Figure 39. Discharge Current From 0.001 µF Capacitor
Appendix B
dV
Measuring (
dt
) s
dV
Figure 40 shows a test circuit for measuring the static ------
dt
of power thyristors. A 1000 volt FET switch insures that the
voltage across the device under test (D.U.T.) rises rapidly
from zero. A differential preamp allows the use of a N-channel
device while keeping the storage scope chassis at ground
for safety purposes. The rate of voltage rise is adjusted by a
variable RC time constant. The charging resistance is low to
avoid waveform distortion because of the thyristor’s selfcapacitance
but is large enough to prevent damage to the
dI
D.U.T. from turn-on ---- . Mounting the miniature range
dt
switches, capacitors, and G-K network close to the device
under test reduces stray inductance and allows testing at
more than 10 kV/µs.
18 REV. 4.01 6/24/02

APPLICATION NOTE
AN-3008
27
DIFFERENTIAL
PREAMP
X100 PROBE
X100 PROBE
G
2
1
V DRM/V RRM SELECT
DUT
20k 2W
2W
0.33 1000V
1000
10 WATT
WIREWOUND
0.047
1000V
R GK
470pF
MOUNT DUT ON
TEMPERATURE CONTROLLED
Cµ PLATE
dV
dt
VERNIER
100
2W
82
2W
0.001
0.005
0.01
0.047
POWER
1 MEG 2W EACH
1.2 MEG
2W
TEST
1N914
0.1
20V
f = 10 Hz
PW = 100 µs
50 Ω PULSE
GENERATOR
56
2W
1000
1/4W
1N967A
18V
MTP1N100
0.47
0-1000V
10mA
ALL COMPONENTS ARE NON-INDUCTIVE UNLESS SHOWN
Figure 40. Circuit For Static
dV
dt
Measurement of Power Thyristors
Appendix C
Measuring
dV
(
dt
) c
dV
A test fixture to measure commutating ------ is shown in
dt
Figure 41. It is a capacitor discharge circuit with the load
series resonant. The single pulse test aids temperature control
and allows the use of lower power components. The limited
energy in the load capacitor reduces burn and shock hazards.
The conventional load and snubber circuit provides recovery
and damping behaviors like those in the application.
The voltage across the load capacitor triggers the D.U.T. It
terminates the gate current when the load capacitor voltage
crosses zero and the TRIAC current is at its peak.
Each V DRM , I TM combination requires different components.
Calculate their values using the equations given in Figure 41.
Commercial chokes simplify the construction of the necessary
inductors. Their inductance should be adjusted by
increasing the air gap in the core. Removal of the magnetic
pole piece reduces inductance by 4 to 6 but extends the current
without saturation.
The load capacitor consists of a parallel bank of 1500 Vdc
non-polar units, with individual bleeders mounted at each
capacitor for safety purposes.
An optional adjustable voltage clamp prevents TRIAC breakdown.
To measure ⎛dV
------ ⎞
⎝
, synchronize the storage scope on the
dt ⎠c
current waveform and verify the proper current amplitude
and period. Increase the initial voltage on the capacitor to
compensate for losses within the coil if necessary. Adjust the
snubber until the device fails to turn off after the first halfcycle.
Inspect the rate of voltage rise at the fastest passing
condition.
REV. 4.01 6/24/02 19

AN-3008
APPLICATION NOTE
HG = W AT LOW
LD10-1000-1000
CAPACITOR DECADE 1–10 µF, 0.01–1 µF, 100pF–0.01µF
NON-INDUCTIVE
RESISTOR DECADE
0-10k, 1Ω STEP
0.01
R S
2N3904
2N3906
51k
51k
2W
2N3904
0.1 0.1
2N3904 2N3906
C S
+5
dV
dt
SYNC
120
1/2W
2W
51 2W
51
2.2k
1/2
2N3906
120
1/2W
2W
-5
0.01
1
+CLAMP
-CLAMP
910k
TRIAD C30X
50H, 3500Ω
Q 3 Q 1
MR760
2W
2 WATT
Q 3 Q 1
910k
2W
+ 1.5kV
62µF
1kV
6.2 MEG 2W
+ – 70mA
–
150k
6.2 MEG 2W
2N3906
2N3904
Q 3 Q 1
-5
+5
PEARSON
301 X
360 1/2W 360 1/2W
1k
1k
2 CASE
CONTROLLED
2N3904
HEATSINK
– +
G 56
2N3906
TRIAC
0.22 0.22
UNDER
270k 1N5343
TEST
7.5V
270k
0-1 kV 20 mA
2.2M 2.2M, 2W
2.2M 2.2M, 2W
L L
R L
CL (NON-POLAR)
2.2M
2.2M
MR760 MR760
Appendix D
dV
Derivations
dt
Definitions
1.0 R T = R L + R S = Total Resistance
I PK I P
C L = T
6f I PK x 10 -6
W 0 V Ci 2π V L L = V Ci = T 2
I dl
=
W 0 =
=
Ci W 0 I PK 4π 2 C L
L
dt L
c
A/µs
Figure 41.
dV
(
dt
) c
Test Circuit for Power TRIACs
1.1
M
=
R S
------ = Snubber Divider Ratio
R T
1.2
ω0
=
1
-------------- = Undamped Natural Frequency
LC S
ω
=
Damped Natural Frequency
1.3
α
=
R
------ T
= Wave Decrement Factor
2L
1.4
χ 2 = --------------------
1⁄ 2LI2
=
2
1⁄
2CV
Initial Energy In Inductor
---------------------------------------------------------------------
Final Energy In Capaitor
1.5
χ
=
I
--
E
L
--- = Initial Current Factor
C
1.6
ρ
=
R
------ T
2
C
--- =
L
α
------ = Damping Factor
ω0
1.7
V OL
= E– R S I = Initial voltage drop at t = 0 across the load
20 REV. 4.01 6/24/02

APPLICATION NOTE
AN-3008
1.8
I ER
ξ = ------ – ---------- L
C S L
⎛dV
------ ⎞ = initial instantaneous dV ------ at t = 0, ignoring any instantaneous voltage step at
⎝ dt ⎠0
dt
t = 0 because of I RRM
1.9
⎛dV
------ ⎞
⎝ dt ⎠0
=
R T
V OL ------ + ξ. For All Damping Conditions
L
2.0
When I = 0,
⎛ dV ------ ⎞ =
⎝ dt ⎠0
ER
---------- S
L
⎛dV
------ ⎞ =
⎝ dt ⎠max
t max =
Maximum Instantaneous dV ------
dt
Time of maximum instantaneous dV ------
dt
t peak = Time of maximum instantaneous peak voltage across thyristor
Average dV ------ = V
dt PK ⁄ t PK = Slope of the secant line from t = 0 through V PK
V PK
=
Maximum instantaneous voltage across the thyristor.
Constants (depending on the damping factor):
2.1
No Damping ( ρ = 0 )
ω = ω 0
R T = α = ρ = 0
2.2
Underdamped ( 0 < ρ < 1)
ω = ω0 2 – α 2 = ω0 1–
ρ 2
2.3
Critial Damped ( ρ = 1)
α
ω0, ω 0R , 2 L C --- , C 2
= = = = ----------
αR T
2.4
Overdamped ( ρ > 1)
ω = α 2 – ω0 2 = ω0 ρ 2 – 1
Laplace transforms for the current and voltage in Figure 42 are:
3.0
E⁄
L+
SI E
SV OL
– ξ
i ( S)
= -------------------------------------
S 2 S R ; e = -- – -------------------------------------
T 1 S
+ ------ + ------- S 2 R T 1
+ ------S + -------
L LC
L LC
R L
L
+
–
t = 0
I
C S
R S
e
INITIAL CONDITIONS
I = I RRM
VC S = 0
Figure 42. Equivalent Circuit for Load and Snubber
REV. 4.01 6/24/02 21

AN-3008
APPLICATION NOTE
The inverse laplace transform for each of the conditions gives:
Underdamped (Typical Snubber Design)
4.0
4.1
α
αt ξ
e = E–
V OL
Cos( ωt)
– --- sin( ωt)
e – + --- sin ( ωt)
e – αt
ω
ω
de
----- = V
dt OL
2αCos( ωt)
( ω 2 – α 2 )
+ ----------------------- sin( ωt)
e – αt
α
+ ξ Cos( ωt)
– --- sin( ωt)
e – αt
ω
ω
4.2
1
t PK ---tan – 1 2αV OL + ξ
= – -----------------------------------------------
ω
ω 2 – α 2
V OL
------------------
⎝
⎛ ω ⎠
⎞ –
ξα ------
ω
When M
= 0, R S = 0I , = 0 ; ωt PK = π
4.3
α
2 2
V PK = E + ----- – α t
ω PK ω 0 VOL + 2αξV OL
+ ξ 2
0
When I = 0, R L = 0M , = 1:
4.4
V
---------- PK
= ( 1+
e – α t
E
PK )
Average dV V
------ = ---------- PK
dt t PK
4.5
1
t max = ---ATN
ω
2
ω( 2 αξ–
V OL
( ω – 3α 2 ))
--------------------------------------------------------------------------
V OL
( α 3 – 3αω 2 ) + ξ( α 2 – ω 2 )
4.6
⎛dV
------ ⎞ =
⎝ dt ⎠max
2 2
V OL
ω 0 + 2αξ V OL
+ ξ 2 αt max
e –
No Damping
5.0
5.1
I
e = E( 1 – Cos( ω 0 t))
+ ---------- sin( ω
Cω 0 t)
0
de
I
----- = Eω
dt 0 sin( ω 0 t)
+ ---cos( ω
C 0 t)
5.2
⎛dV
------ ⎞
⎝ dt ⎠0
I
= --- = 0 when I = 0
C
5.3
5.4
5.5
5.6
5.7
π – tan – 1 ⎛--------------
I ⎞
⎝CEω 0
⎠
t PK = ------------------------------------------
ω 0
V PK = E+
E 2 + ---------------
I2
2 2
ω 0 C
⎛dV
------ ⎞
V
= ---------- PK
⎝ dt ⎠AVG
t PK
I
t max ----- tan – 1 ⎛--------------
ω EC 0 ⎞ I
=
= ----- π -- when I = 0
ω 0
⎝ I ⎠ ω 0 2
⎛dV
------ ⎞ I
--- E 2 2 2
= ω
⎝ dt ⎠max
C 0 C + I 2 = ω 0 E when I = 0
22 REV. 4.01 6/24/02

APPLICATION NOTE
AN-3008
Critical Damping
6.0
6.1
6.2
6.3
e = E– V OL
( 1 – αt )e – αt + ξte – αt
de
----- = [ αV
dt OL
( 2 – αt) + ξ( 1 – αt)
]e – αt
ξ
2 + -------------
2V
t OL
PK = -----------------------
ξ
α + ---------
V OL
V PK = E–
[ V OL
( 1 – α t PK )–
ξ t PK ]e – α t PK
6.4
Average dV ------
dt
=
V
---------- PK
t PK
When I = 0,
R S = 0M , = 0
e(t) rises asymptotically to E. t PK and average dV ------ do not exist.
dt
6.5
3αV OL
+ 2ξ
t max = -----------------------------
α 2 V OL
+ αξ
When I = 0,
t max = 0
For R S
------ ≥ 3⁄
4
R T
6.6
then dV dV
------ = ⎛------
⎞
dt max ⎝ dt ⎠0
⎛dV
------ ⎞
⎝ dt ⎠max
–
[ αV OL
( 2 – α t max ) + ξ ( 1–
α t max )]e αt max
=
Appendix E
Snubber Discharge
Overdamped
dI
dt
Derivations
1.0
i
=
V
---------- CS
α – αt sinh( ωt)
ωL S
1.1
1.2
C
i PK = V S
CS
------ e – α t
L PK
S
1
t PK = --- tanh – 1 ω ω ᾱ --
Critical Damped
2.0
i
=
V CS
-------- te – αt
L S
2.1
V CS
i PK = 0.736--------
R S
REV. 4.01 6/24/02 23

AN-3008
APPLICATION NOTE
2.2
1
t PK = --
α
Underdamped
3.0
3.1
3.2
V CS
i = ---------- e – αt sin (ωt )
ωL S
C
i PK = V S
CS
------ e – α t
L PK
S
1
t PK = ---tan – 1 ω ω ᾱ --
R S
L S
t = 0
VC S
C S
i
INITIAL CONDITIONS:
i = 0, VC S = INITIAL VOLTAGE
Figure 43. Equivalent Circuit for Snubber Discharge
No Damping
4.0
i
=
V CS
---------- sin( ωt)
ωL S
4.1
4.2
C
i PK = V S
CS
------
L S
t PK
π
= ------
2ω
Bibliography
Bird, B. M. and K. G. King. An Introduction To Power Electronics.
John Wiley & Sons, 1983, pp. 250–281.
Blicher, Adolph. Thyristor Physics. Springer-Verlag, 1976.
Gempe, Horst. “Applications of Zero Voltage Crossing Optically
Isolated TRIAC Drivers,” AN982, Motorola Inc., 1987
“Guide for Surge Withstand Capability (SWC) Tests,” ANSI
337.90A-1974, IEEE Std 472–1974.
“IEEE Guide for Surge Voltages in Low-Voltage AC Power
Circuits,” ANS1/lEEE C62.41 -1980, IEEE Std 587–1980
dI
Ikeda, Shigeru and Tsuneo Araki. “The ---- Capability of
dt
Thyristors,” Proceedings of the IEEE, Vol.53, No.8, August
1967.
Kervin, Doug. "The MOC3011 and MOC3021," EB-101,
Motorola Inc., 1982.
McMurray, William. “Optimum Snubbers For Power Semicondutors,"
IEEE Transactions On Industry Applications,
Vol.1A-8, September/October 1972.
Rice, L. R. “Why R-C Networks And Which One For Your
Converter,” Westinghouse Tech Tips 5-2.
“Saturable Reactor For Increasing Turn-On Switching Capability,”
SCR Manual Sixth Edition, General Electric, 1979.
Zell, H. P. “Design Chart For Capacitor-Discharge Pulse
Circuits," EDN Magazine, June 10, 1968.
24 REV. 4.01 6/24/02

AN-3008
APPLICATION NOTE
DISCLAIMER
FAIRCHILD SEMICONDUCTOR RESERVES THE RIGHT TO MAKE CHANGES WITHOUT FURTHER NOTICE TO ANY
PRODUCTS HEREIN TO IMPROVE RELIABILITY, FUNCTION OR DESIGN. FAIRCHILD DOES NOT ASSUME ANY
LIABILITY ARISING OUT OF THE APPLICATION OR USE OF ANY PRODUCT OR CIRCUIT DESCRIBED HEREIN; NEITHER
DOES IT CONVEY ANY LICENSE UNDER ITS PATENT RIGHTS, NOR THE RIGHTS OF OTHERS.
LIFE SUPPORT POLICY
FAIRCHILD’S PRODUCTS ARE NOT AUTHORIZED FOR USE AS CRITICAL COMPONENTS IN LIFE SUPPORT DEVICES
OR SYSTEMS WITHOUT THE EXPRESS WRITTEN APPROVAL OF THE PRESIDENT OF FAIRCHILD SEMICONDUCTOR
CORPORATION. As used herein:
1. Life support devices or systems are devices or systems
which, (a) are intended for surgical implant into the body,
or (b) support or sustain life, or (c) whose failure to perform
when properly used in accordance with instructions for use
provided in the labeling, can be reasonably expected to
result in significant injury to the user.
2. A critical component is any component of a life support
device or system whose failure to perform can be
reasonably expected to cause the failure of the life support
device or system, or to affect its safety or effectiveness.
www.fairchildsemi.com
6/24/02 0.0m 001
Stock#AN300000xx
© 2002 Fairchild Semiconductor Corporation