Order this document

SEMICONDUCTOR APPLICATION NOTE by AN1542/D

 

     
Prepared by: C. S. Mitter
Motorola Inc.

Input filter design has been an integral part of power supply current which can exceed far beyond the device ratings
designs. With the advent of input filters, the designer must (semiconductor devices, fuses, circuit breakers), and can
take into consideration how to control the high inrush current seriously damage or destroy the semiconductor devices, burn
due to rapid rise of voltage during the initial application of out the fuses or false trigger the circuit breakers. The high rate
power to the power supply. Depending on the input bus of rise of the voltage and fast rise of the current may activate
voltage level and the output power required by the load, the other circuitry that are dv/dt and di/dt sensitive. This high dv/dt
supply designer must also design the inductor (if used) to and di/dt introduces unwanted EMI noise.
support the DC current without saturating the core. The It is clear that a new methodology of controlling dv/dt without
inductor and capacitor is designed to meet EMI requirements. affecting the inductor size and power supply efficiency needed
Limiting initial inrush current with inductor can become very to be found.
large in size and weight, and in most cases size and weight is
a crucial requirement to the design. VARIOUS INRUSH LIMITING TECHNIQUES
In this section, a review of various active and passive
Traditionally, most of the inrush current limiting is done by
methods of inrush limiting techniques are presented. It is
using a large oversized inductor, or resistors in series with the
shown that a new and innovative method can be applied using
capacitors. These techniques do not optimally utilize the
a single MOSFET and a minimal number of components in
surface area, weight and power dissipation. In applications
many of the circuits requiring dv/dt control in order to limit the
where large DC current is required at the input of the power
high current spikes. Its design methods and simple yet
supply, the inductor not only has to be designed for low EMI,
effective equations are also presented. A variety of
but it needs to be designed to meet DC current capability
applications of this dv/dt control circuit into other areas are
without degrading the inductance value. Reduction in
proposed. The simplicity and the advantage of this technique,
inductance will mean that the EMI noise attenuation capability
as opposed to other techniques, is shown given its
is reduced. Therefore, the design of inductance becomes very
effectiveness in different applications requiring dv/dt control.
large because with increase in operating current, the core
The new inrush limiting is beneficial because dv/dt control
becomes larger. If a series resistor is used, unnecessary
reduces the EMI due to current and voltage spikes, and the
power is lost because of I2xR. This in turn degrades the power
lifetime of capacitors and the semiconductor devices
supply efficiency. In order to overcome the power dissipation
surrounding the circuitry is increased. This technique will also
of the series resistor, many designers incorporate a parallel
increase the reliability of the devices and the capacitors. And
switch with a resistor (semiconductor devices or relays).
because of its minimal parts count, the design is very cost
Depending on the operating current, the relay can become
effective.
excessively large and heavy. In addition, control circuit must
be implemented in order to control the turn–on and turn–off of
INTRODUCTION
the relay. In the cases where semiconductors are used, such
In power supply designs, the input filter design is an integral as an SCR, the device can become very bulky and dissipative.
part of the design. In most designs input filter designs This application also requires unique control circuitry in order
incorporate both inductor and capacitors. The inductor and control the SCR turn–on and turn–off.
capacitors need to behave in such way as to provide EMI Another method of inrush current limiting is done using an
reduction and provide supply hold–up requirement in case of NTC thermistor. This device has a negative temperature
short duration line dropout. This requirement along with coefficient and its resistance decreases as current is passed
derating of the capacitors for temperature variations results in through the device (current flow increases the temperature of
having to use large filter capacitors. the device and decreases the resistance) [2]. The drawback
With innovations in technology, the process for of this device is that it requires a “cool off ” time after the power
manufacturing of capacitors allows for very low equivalent is removed in order to reset to high resistive mode. Depending
series resistance (ESR), and thus they behave like nearly on the power rating of these devices, its physical size
perfect short circuits during initial power application to the becomes significant. The “cool off ” time can be overcome by
power supply [1]. The initial power applied to the power supply using an active device along with the NTC device. But this
posses a very high dv/dt. This high dv/dt interacting with the defeats the purpose of using the NTC device in first place; that
filter capacitors will introduce short–duration of high peak is, the simplicity of application and minimal parts count.

MOTOROLA
Motorola, Inc. 1995 1
AN1542
L MOSFET Switching Characteristics
Vdc MOSFETs are charge controlled devices and can be
represented with the simplified equivalent circuit shown in
Figure 4. The gate–source capacitance (Cgs) is largely
dependent on gate–oxide and source–metallization
DC – DC
C capacitance and can be measured and considered constant.
POWER
CONVERTER Gate–drain capacitance (Cgd) consists of the gate–drain
overlap oxide capacitance and gate–drain overlap depletion
capacitance. This capacitance is nonlinear due to its voltage
dependency. Drain–source capacitance is the depletion
RTN capacitance of the drain–source junction. The following
expressions can be obtained from the circuit shown in
Figure 4.
Figure 1. Inductive

R Idrain
ICgd
D1 D2

+ DC – DC Cgd
SCR
D3 C POWER Cds
D4
CONVERTER G
CONTROL
AC INPUT CIRCUIT IG Cgs

Figure 2. Passive and Active S

Figure 4. Equivalent Circuit for MOSFET

NTC

Ciss = Cgs + Cgd; Cds shorted (1)

D1 D2

DC – DC Crss = Cgd, (2)

+
D3 C POWER
D4
CONVERTER Cgs ⋅ Cgd
Coss = Cds + ; Cgs shorted (3)
AC INPUT NTC Cgs + Cgd

≈ Cds + Cgd

The expressions shown in equations 1 through 3 are

Figure 3. NTC Thermistor
parameters that are available from the MOSFET data sheets
and curves provided therein. Capacitance Ciss is equivalent
input capacitance, Crss is the reverse transfer capacitance,
INTRODUCTION TO ACTIVE CURRENT LIMITER and Coss is the equivalent output capacitance. How fast the
capacitance is charged and discharged determines how fast
In low to medium power levels which require few hundred the device will turn–on or turn–off (here the load effect on
volts of blocking capability, MOSFETs are an ideal devices switching of the device is not considered). The most effective
because they posses following characteristics: 1) fast way of obtaining and controlling the switching mechanism of
switching time due to majority carrier devices, 3) lower the MOSFET is by using the gate–charge transfer curves as
switching loss due to fast rise and fall times, 2) simple gate provided by the vendors.
drive, 3) and low RDS(on) which helps to increase the efficiency
by decreasing the voltage drop across the device during
steady state conduction. Because the active current limiting is
done by using MOSFET devices, it is essential that one
comprehend the switching characteristics of this devices. By
understanding the switching characteristics, the design
engineer will be better equipped to use the proposed circuit
without any ambiguity.

2 MOTOROLA
 AN1542
Figure 5 shows the turn–on gate–charge transfer curve. than the internal Cgd, we can rewrite equations 1 through 3
This curve contains all the necessary information for (see Figure 6).
controlling the turn–on switching of the device. Region 1 is a
pre–threshold region and the constant current is used to D
charge the input capacitance Ciss. During this period, we can
ignore the gate–drain capacitance because it is much smaller Idrain
ICgd
than the gate–source capacitance and VGS rises to device
threshold voltage Vth at a linear slope. When VGS has reached
the Vth, the drain current rises to its steady–state drain current
Cgd′ Cgd
during the region. In region 2, the drain–source voltage starts
its transition and the gate–charge transfer curve starts to Cds
G
level off.
IG Cgs
The slope of the drain–source voltage is much higher in
region 2 because the gate–drain capacitance Cgd is still
relatively small and this small capacitance can be discharged
at a faster rate with the given gate current. As the voltage
across the drain–gate is reduced even more, the dramatic
S
increase in Cgd is observed, and it is this large capacitance
which will dominate the input capacitance, and all of the gate Figure 6. Revised MOSFET Circuit with
current is used to discharge Cgd. During this flat level (region External Capacitor
3), VGS remains constant, VDS decreases to its saturation
voltage, and Vsat, and the VGS will again rise to its applied gate
voltage value. Further increase in VGS has no effect on the Given the condition: Cgd′ >> Cgd
drain–source voltage and drain current.
Ciss = Cgs + Cgd + Cgd′ ≈ Cgd′ + Cgs: Cds shorted (4)

V Crss = Cgd + Cgd′ ≈ Cgd′ (5)

VGS
Coss = Cds + Cgd + Cgd′ ≈ Cgd′; (6)
1 2 3 4

The external capacitance acts as an integrator and is used

to accurately determine the switching characteristics of the
VDS
MOSFET. Its simple expressions allow us to disregard much
90%
VGS2 of the nonlinear capacitive effects of Cgd because the
capacitance, Cgd′, along with the gate drive dominates the
drain–source voltage transition. The ability to control the
VGS1 constant linear slope of the drain voltage transition allows
accurate control of the inrush current to the capacitive load.
10% This is possible because the current flowing through the
t capacitor is dependent upon the transition of the Voltage:

Figure 5. Gate Charge Transfer Curve

ic + C dV
dt
. (7)

In observing the gate–transfer curve, it is seen that the VDS Derivation of the Design Equations for dv/dt
transition is determined during region 2 and region 3 of the Control Circuit
curve. If the drain–source voltage transition is contained in Figure 7 is a circuit used to control the dv/dt of the MOSFET
region 3, it is possible to fix and control the time during its switching cycle. RG is a series gate resistor (which
rate–of–change of the drain–source voltage, dVDS/dt, very is large in value), and RGD is a small resistor added in series
accurately. The ability for control of region 3 will allow with Cgd′ to damp out any unwanted high frequency
complete control of the dVDS/dt independent of load condition. oscillations (this resistor must be much smaller in value than
It is this ability to control dVDS/dt which will allow control of the RG: RG>>RGD. The large value of RG controls the charge rate
inrush current to the capacitive load or resistive load. of the Cgd′). The control of the dv/dt is dependent upon the
load type and is a function of gate voltage, RG, VDD, external
Proposed MOSFET Switch feedback capacitance, Cgd′ and drain current of the device.
In order to insure that the voltage transition is linear The diode, Dg, is placed in parallel with the RG to provide faster
throughout region 3, an external capacitor can be added to the turn–off process, and can be taken out if slow turn–off is of no
gate–drain connection, Cgd′. If this capacitance is much larger concern.

MOTOROLA 3
AN1542
LOAD

ICgd
L C R
Dg
RGD
Cgd′
VDD
RG NMOS
GATE
DRIVE
IG
Cgs

Figure 7. dv/dt Control Circuit

Figure 8 is a switching characteristic of dv/dt control circuit Turn–on delay is defined as the time required to charge the
of Figure 7, and this curve will be analyzed to derive all of the gate–source to threshold voltage, Vth. The time delay can be
design equations for the dv/dt circuit. found by using the following equation:
V
td = RG(Cgs + Cgd′) ln 1 – th , (9)
v VGG
VGS
1 2 3 4 Region 2
Vth
Beyond the turn–on delay, region 2, the drain current starts
to conduct. The time rate of change of drain current is
Vplt expressed as:
t dIdrain dVGS (10)
VDS = gfm

dt dt
Where dVGS/dt is a representation of the slope for regions
1 and 2, and gfm is the transconductance to support the drain
t current Idrain. During this time the drain voltage is nearly
constant.
Idrain
Region 3
In region 3, the transition of the drain–source voltage occurs
t from its blocking voltage to its saturation voltage once the
Figure 8. Gate–Source Charge Characteristic drain current has reached its maximum load current, and the
gate–source voltage remains at plateau voltage Vplt. Note the
difference between Figure 6 and Figure 7. In Figure 7, the
Region 1 transition of drain–source voltage extends linearly until the
During this period, the gate voltage is charging the end of region 3. The switching is completed when the
equivalent input capacitance which is dominated by the drain–source voltage VDS has switched to 10%. Since the
feedback capacitance, Cgd′, expressed in equation 4. This drain current is constant during this region, the gate–source
constant capacitance gives constant linear slope, and the voltage (the plateau voltage, Vplt) must be:
gate source voltage will rise exponentially:
Iinrush
VGS = VGG [1 – e –[t/(RG(Cgs+Cgd′)] ] (8) Vplt = Vth +
gfm(max)
(11)

4 MOTOROLA
 AN1542
This plateau voltage can be found by using a transfer curve Region 4
provided by the vendors (see Figure 9). In region 4, the VDS has reached its Vsat and VGS continues
to increase to its gate voltage, VGG.
40

35 VDS = 10 V dv/dt Design Steps

TJ = – 55°C 100°C The following steps are given in order to simplify some of the
I D, DRAIN CURRENT (AMPS)

30 equations, and to simplify design procedures.

25°C
25
1. Use equation 7 to find the time required to meet the
20 inrush requirement:
VDD
15 dt = Cfilter. . (16)
Iinrush
10
2. Find the gate–source plateau voltage, Vplt, required to
5 supply the load current (equation 11). Use the device
transfer curve to find the plateau voltage if the data is
0 available.
2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5
VGS, GATE – TO – SOURCE VOLTAGE (VOLTS) Iinrush
Vplt = Vth + . (17)
gfm(max)
Figure 9. MOSFET Transfer Characteristics
3. Choose Cgd′ based on following condition: Cgd′ >> Cgs +
Cgd (the values for Cgs and Cgd is obtained using the
The constant Vplt allows the input current to flow through the
data sheet curves).
feedback capacitance, Cgd′, and its current is expressed as:
(VGG – Vplt) 4. Find the gate current required using equation 13 (the
Ig = . (12) feedback capacitance C gd ′ is chosen based on
RG
availability):
But since the gate current is equal to the current flowing
through feedback capacitance, Cgd′, we can rewrite the above dVDS
Igd ≈ Cgd′ . (18)
equation as a function of Cgd′, VDS, Vplt, and time, dt: dt
Given: Igd ≈ Ig, 5. Use equation 12 to find the series gate resistance:
dVDS (VGG – Vplt)
Igd ≈ Cgd′ . (13) RG = . (19)
dt Igd
Thus the rate of change of gate–drain voltage is: 6. Choose RGD: RG >> RGD.
dVGD Ig (VGG – Vplt) dv/dt Design Example. Figure 10 is a test circuit which have
= = . (14)
dt Cgd′ RG
Cgd′ been incorporated to test the effectiveness of the circuit.
The time rate of change of drain–source voltage is equal to Following parameters were used:
the time rate of change in gate–drain voltage during the region Cgs = 2000 pF
3. It is expressed as: VDD = 28 Vdc
dVDS (VGG – Vplt) Cfilter = 200 µF
= . (15) Iinrush = 2 Apk
dt RG
Cgd′
VGG = 12 V
Vth = 2.7 V
gfm(max) = 2.5 S

Cload
ICgd

Dg

Cgd VDD
SWITCH
RG NMOS

IG

VGG

Figure 10. dv/dt Control Test Circuit

MOTOROLA 5
AN1542
Step 1 Where tVgs is the time required for VGS to reach a value that
Using the expression i = C(dv/dt), the transition time for VDS will support the inrush current obtained by equation 17, and
is found: Iinrush is the maximum inrush current flowing through the
MOSFET.
VDD 28
dt = Cfilter. = 200 µF = 2.8 ms (20)
Iinrush 2
VOLTAGE

Step 2 tVgs dt VGG

During the VDS transition, VGS is constant and equation 11
is used to find the Vplt for desired peak drain current (use the
1 2 3
tranfer curve if available).
dit Vplt
Iload 2
Vplt = Vth + = 2.7 + = 3.5 V. (21)
gfm(max) 2.5
∆VGS
Step 3 td
Satisfying the condition Cgd′>>Cgs+Cgd, 0.1 µF was
TIME
arbitrarily chosen. We can either choose initial value of RG or
Cgd′, and design for the unknown. But in most cases, different Figure 11. Expanded View of Gate–Source Voltage
values of resistor is easier to obtain than the capacitors. So for
this design example, Cgd′ was chosen. The rate of rise time of the drain current of the device is
directly proportional to the rate of rise of the gate–source
Step 4 voltage (see equation 10):
Use equation 14 to find required gate current:
dIdrain dVGS
dVDS 28 ∝ (25)
Igd ≈ Cgd′ ≈ 0.1 µF ≈ 1 mA. (22) dt dt
dt 2.8 ms
The rate of change of drain current is a design requirement.
As a result, we can find the minimum time required for
Step 5 gate–source voltage to reach the value to support the initial
Using the calculated values from 21 and 22, series gate inrush current:
resistor is calculated:
Vplt
tVgs(min) = . (26)
(VGG – VGS) (12 – 3.5) (dIinrush / dt)
RG = = ≈ 8.5 kΩ. (23)
Igd 1 mA Using the minimum time required, the following must be
satisfied:
Step 6 –tVgs(min)
The damping resistor can be chosen arbitrarily if the RG
(Cgs + Cgd′) ≥ . (27)
following condition is met: Vplt
ln 1 –
VGG
RG >> RGD,
Let RGD = 100 Ω. If the above expression is not satisfied, then one of the
parameters (RG or Cgd′) must be changed to satisfy the
di/dt Design equation.
In some cases the initial rate of rise of the drain current must
di/dt Design Example. For the same circuit used for dv/dt
be limited to some fixed rate. The current slope limiting
control, the following current requirement must be met:
reduces transients associated with L(di/dt) due to stray
inductances. This current slope limiting reduces the problem dIdrain 2A
= . (28)
of loading of the line which can cause a temporary line dropout dt 100 µs
[3].
We can determine from the Figure 11 that time for VGS to
reach a voltage level given by equation 11 must be greater Step 1
than the sum of delay time, td, and the time required for The minimum time required is:
drain current to reach steady state value: 3.5
tVgs(min) = = 175 µs. (29)
tVgs > td + dit. (24) (2 A/100 µs)

6 MOTOROLA
 AN1542
Step 2 active dv/dt and di/dt control is useful due to the simple gate
Once total time required is calculated, the time constant can drive requirements and low RDS(on). With just a single
be checked (the effect will be discarded): MOSFET and its corresponding components, dv/dt and di/dt
transition can be accurately controlled. This control circuit
(30)
results in precise control of inrush current magnitude, and
RG
(Cgs + Cgd′) = 8.5K
(2000 pF + 0.1 µF) = 867 µs. reduces much of high frequency noise associated with high
voltage and current transitions. Not only is the circuit precise,
–tVgs(min) –175 µs but it also reduces weight and surface area as opposed to
and, = = 507 µs. (31) bulky inductors.
Vplt 3.5 The simple and yet effective equations have been derived
ln 1 – ln 1 –
VGG 12 for dv/dt and di/dt control circuit. These simple equations with
appropriate components can overcome even some of the
most stringent inrush current limiting requirements without
Thus sacrificing weight and area. The following are the design
–tVgs(min) objectives which have been discussed:
RG
(Cgs + Cgd′) ≥ = 867 µs ≥ 507 µs. (32)
Vplt 1. Add a feedback capacitance Cgd′ to negate nonlinear
ln 1 –
VGG voltage dependent capacitor Cgd.
2. The following requirement must be met: Cgd′>>Cgs+Cgd
The condition is satisfied and nothing needs to be changed. (maximum Cgd can be obtained from the vendor data).
If equation 32 was not satisfied, a different value of RG (or 3. Linearize rate of change of drain source voltage during all
Cgd′) can be chosen. of the active region (region 3).
4. Control dVDS/dt by controlling the charge rate of Cgd′ .
CONCLUSION 5. Fix the charge rate by controlling the charging current.
It is shown in this section that by using a low RDS(on) 6. Gate current IG, is fixed during active region by choosing
MOSFET in current limiting, the shortcomings due to passive correct RG.
methods can be overcome. Using an MOSFET device in an

MOTOROLA 7
AN1542
APPLICATIONS USING DV/DT CONTROL CIRCUIT
INTRODUCTION
The active dv/dt circuit is applicable to many applications V
which require slow ramp up of voltage across the load or
current through the load. Detailed analysis was done in dt
previous section of the paper which showed all of the required VGS
design equations. Using this same equations, and with slight Vplt
modifications, dv/dt circuit can be implemented to numerous
applications. In this section some examples will be shown and
discussed where it is appropriate.
t

Series Pass Switch Vout = VDD

There are requirements where the power to the load must
be switched on at a controlled rate either 1) due to inrush
current requirement to capacitive load or 2) due to power
source type (batteries). Using a MOSFET to switch on the t
power to the load allows the flexibility of accurately controlling Figure 2. PMOS Switching Transitions
the dv/dt and di/dt to the load. Using a controlled ramp–up of
power to the load also relaxes some of the load transient
requirements for primary power source (DC–DC SMPS). NMOS Application
The following examples are circuits used for switching When NMOS is used to switch the power to the load, the
power to the load. The examples are given to show how the gate voltage of the device is constantly increasing nonlinearly
different device types can be implemented in a switching load due to the output voltage level shifting the gate–source
application using the dv/dt concepts discussed previously. voltage, VGS (see Figure 4).
The same design steps can be followed as in previous
sections. But for simplicity, all the steps will be rewritten for the Idrain
convenience of the reader. NMOS Vout
VDD
D S
PMOS Application
Given: output voltage transition time dt, G
Cgd′
Given: output voltage, Vout
Given: gate voltage, VGG RGD
LOAD
Given: gate drain capacitance, Cgd′
Given: gate–source plateau voltage, Vplt ICgd′

1. Choose Cgd′ >> Cgs + Cgd

RG
VGG
dVDS Vout
2. Igd′ ≈ Cgd′ = Cgd′ (1) IG
dt dt

VDD – Vplt Figure 3. NMOS Load Switch

3. RG = or (2)
Igd′

VDD – Vth VOLTAGE

RG = : if very low output current (in mA). (3)
Igd′ t

Choose RGD << RG. VT

ACTUAL
CURVE VG
PMOS Vout
VDD
S D
LINEARIZED
G
Cgd′ 0
dt VGS
Vplt
ICgd′
LOAD
0
RGD
RG Vout
VGG

IG
0 TIME
Figure 1. dv/dt Control Circuit Used As Load Switch
Figure 4. Voltage Transitions of NMOS Switch

8 MOTOROLA
 AN1542
The gate voltage is changing at a rate: INRUSH CURRENT LIMITER FOR
DC–DC CONVERTERS
VG = VGG(1–e –(t/RGDCgd′)) (4)
Note that gate voltage has to be greater than the input dc Figures 5a and 5b shows the dv/dt circuit used as a inrush
voltage; VGG > dd. Taking this assumption we can than current limiter in a DC–DC power converter (The operation of
linearize the gate voltage curve and the following design the circuit is not discussed in this section, but is presented in
requirements be implemented: the following section). In Figure 5a, the MOSFET is placed in
the return path of the power supply. This will cause the SMPS
1. Choose Cgd′ >> Cgd + Cgs RTN path to rise to the VDD value, and then it is brought down
to ground potential at a fixed rate by the dv/dt control circuit.
2. Find the gate voltage at which the output will have This configuration is beneficial where the RDS(on) of the
ramped up to its input voltage: MOSFET does not introduce an additional ESR to the filter
capacitor.
VT = Vplt + Vout (5) Figure 5b shows the dv/dt circuit placed in series with the
3. Find the total time required for gate voltage to reach VT: input filter capacitor. This configuration is beneficial because
the MOSFET can never be shorted due to improper ground
Vout dt
= , (6) connection, and the RDS(on) can act as an damping resistor.
VT t

VT VDD
and t = dt , (7)
Vout
DC – DC
where, t, is equal to time needed for gate to charge Cfilter SMPS
up to VT.
4. Calculate the series gate resistor using the time, t:
Rch RGD SMPS
–t 1 RTN
RG = . . (8)
Cgd′ VT Cgd′
ln 1 –
VGG RG NMOS

Cch
Choose RGD << RG. Dgs
DC Dg
RTN
Sample Calculations:
Figure 5a. MOSFET Placed at Return Path
Given:
VDD =5V
Vout = VDD = 5 V
VGG = 12 V L
dt = 5 ms Vdc
Vplt = 2.7
DC – DC
Cfilter
1. Cgd′ = 0.1 µF (9) SMPS

2. VT = Vplt + Vout = 2.7 + 5 = 7.7 (10) RGD

Rch SMPS
RTN
VT 7.7 Cgd′
3. t = dt = 5 ms = 7.7 ms (11)
Vout 5 RG NMOS

–t 1 –7.7 ms Cch
4. RG = . = . (12) Dgs
Cgd′ VT 0.1 µF DC Dg
ln 1 – RTN
VGG
Figure 5b. MOSFET Placed Series With
1
= 75 kΩ Filter Capacitor
7.7
ln 1 –
12

5. Let RGD = 100

MOTOROLA 9
AN1542
INRUSH CONTROLLER FOR DC LIGHT BULBS the inrush current can be controlled accurately. Initially the
light bulb has very low cold resistance, but once the current is
In order to prolong the lifetime of incandescent lamps, the conducting the filament warms up and its resistance increases
inrush current must be controlled (it can be any light emitting accordingly. The dv/dt control circuit is used so that the voltage
device). If many light bulbs are paralleled, the peak current to across the lamp will increase very slowly, and this slow rise of
the load must be controlled because it may exceed the fuse the voltage will allow the lamp to heat up before the full supply
or circuit breaker rating. By using the dv/dt controlled circuit, voltage is across the lamp.

VDD CONCLUSION
Active dv/dt control circuit can be applied to numerous
applications where voltage and current must be controlled at
initial turn–on. It was shown in this section that the circuit can
be implemented in series switch to the load, control input filter
inrush current, and control inrush current to the dc light bulbs.
Rch RGD Using dv/dt control circuit is beneficial because it reduces the
EMI and prolongs the lifetime of the electronic components in
Cgd′ the circuit.
RG NMOS

Cch
Dgs
DC Dg
RTN

Figure 6. Inrush Limiter for Incandescent Light Bulbs

10 MOTOROLA
 AN1542
AVOIDING FALSE TURN–ON DUE TO STATIC DV/DT FOR MOSFET–BASED
ACTIVE INRUSH CURRENT LIMITER
Figure 1 is a active inrush current limiter incorporated into We will assume the worst case (step function). When VDD
DC–DC power converter. Initially the NMOS switch is turned is applied, the instantaneous gate–source voltage will charge
off until the dc buss voltage VDD is applied. In many up to the following:
applications, the input voltage VDD is switched. The rate of rise
Cgd 200
of the supply voltage is a function of the switch speed and the VGS = VDD = 100 = 9.1 V. (1)
parasitic components within the circuit. When VDD is fully Cgs + Cgd 200 + 2000
applied to the circuit, the drain of the MOSFET will see all of
the applied voltage because of its high impedance during off This voltage is high enough to turn the device fully on and
state. When high dv/dt is applied to the drain of the MOSFET, cause a large inrush current to flow through the input filter
the voltage transient can be fed back to the gate via drain–gate capacitance. In Figure 1 the capacitance Cch is inserted in
feedback capacitance. If there is enough charge present in the order to keep the VGS at a level below the threshold voltage,
gate, the switch will turn on and the dv/dt control will be lost. and keep the device off during step voltage application.
The magnitude of the gate–source voltage will depend upon Figure 3 shows the representation of the time varying
the gate–impedance of the device. voltage and current waveforms for the circuit shown in Figure
1. During the initial application of the voltage, the gate voltage
will try to ramp up because of high dv/dt seen via the feedback
VDD capacitance Cgd′, but as soon as the gate–source voltage is
high enough to turn the diode Dg on, it causes all of the charge
DC – DC
Cfilter SMPS
to be transferred to capacitance Cch. The voltage across the
charge capacitance Cch, is determined by the voltage division
between Cch and Cgd′. This voltage across the charge
Rch RGD SMPS capacitance is expressed as:
RTN
Cgd′
Cgd′ Vch = VDD. . (2)
RG NMOS Cgd′ + Cch
The capacitance Cch must be large enough in order to
Cch maintain the device in the off state. The gate–source voltage
Dgs
DC Dg will be:
RTN VGS = Vch + VDG, (3)
Figure 1. Inrush Current Limiter where VDG is a junction potential of the diode Dg.
Initially, the current flowing through the RGD, and Cgd′ is:
Figure 2 is a equivalent circuit for MOSFET showing the
internal parasitic capacitance. In order to show how much VDD
io = , (4)
impact the parasitic capacitance has during initial voltage RGD
applied to the drain of the device, consider the following and this current decays exponentially at a rate:
example.
iCgd′ = iRgd = ioe –t/(RGD
Cgd′) (5)
Given:
Cgd = 200 pF
Cgs = 2000 pF
VDD = 100 V
VDD
D

Idrain
ICgd

VGS = Vch + VDG

Cgd

Cds Cgd
G Vch = VDD
Cgd + Cch
IG Cgs io

e –t / (RGD Cgd)
t
S
Figure 2. MOSFET Equivalent Circuit Figure 3. Step Function

MOTOROLA 11
AN1542
During the step voltage application, the voltage across Cch Design Equations
is set at: The total decay time required for the initial current is found
Vch = Vthmin – VDG = Vthmin – 1. (6) by taking equation 5 and solving for time, t :

This is to insure that the MOSFET does not turn on during iCgd′
tdelay = RGD
Cgd′
abs ln , (7)
the initial VDD application. io
In Figure 3, it is shown that the initial current during the step
function is determined by the RGD and VDD. This current than where iCgd′ is the current through the feedback capacitance at
decays at some exponential rate. In order for the VGS not to given time. For this purpose, the desired current level is set at
overshoot, the time delay must be provided by the RchCch time 0.5% of initial current io. The time delay is than expressed as
constant such that the gate–source voltage must not rise to a function of feedback resistor RGD, and capacitance Cgd′:
Vplateau before the current has decayed to near zero. If VGS tdelay ≈ 5.3 RGD
Cgd′ (8)
has reached plateau voltage Vplt, before the current has
decayed to zero, a overshoot will be observed at the gate and where the constant 5.3 is obtained by:
the device will turn on rapidly for a short duration and may
iCgd′
cause excessive inrush current to flow through the input filter abs ln = abs ln |0.005| ≈ 5.3. (9)
capacitor, Cfilter. io

The time constant Rch⋅Cch must meet the following

Operation of Charge Control Circuit condition:
Figure 1 is a equivalent circuit for inrush current limiter.
When VDD is applied, the charge capacitance maintains the tdelay
Rch
Cch ≥ (10)
VGS to a safe level so that MOSFET does not turn–on. The (Vplt – Vch – VDG)
capacitance Cch is then charged at a rate dependent on abs ln 1 –
VDD
(Rch⋅Cch) time constant, and VGS charges at the same rate (at
diode drop greater) until it reaches a voltage to support the where, Vch is found by using equation 6.
inrush current. The drain voltage decreases at a linear slope The capacitance Cch is found using the following
which is determined by the (RG⋅Cgd′) time constant. This slope expression:
then determines the maximum magnitude of the inrush current
to the capacitors. Examples of the theoretical waveforms are Cgd′
(VDD – Vch)
Cch ≈ (11)
shown in Figure 4. Vch
and charge resistance, Rch is:
1 tdelay
Rch ≥ . (12)
VDD Cch (Vplt – Vch – VDG)
abs ln 1 –
VDD

Design Steps for Charge Control

When designing for the Cch and RGD, following steps must
Vch be done in order:
1. Find the voltage Vch:
Vch = Vthmin – VDG = Vthmin – 1 (13)
VGS 2. Using the calculated Vch, find the capacitance Cch:
Cgd′
(VDD – Vch)
Cch ≈ (14)
Vch
3. Find tdelay: tdelay ≈ 5.3 RGD
Cgd′ (15)
Vdrain
Find Rch:

1 tdelay
Rch ≥ . (16)
Cch (Vplt – Vch – VDG)
Idrain
abs ln 1 –
VDD
t

Figure 4. Switching Waveforms

12 MOTOROLA
 AN1542
Design Example: 3. tdelay = 5.3 (0.01 µF
1K) = 53 µs
Given: Vplt = 3.75 V
Vthmin = 2 V 1 53 µs
VDD = 50 V 4. Rch ≥ . ≥ 3K
VDG = 1 V 0.49 µF (3.75 – 1 – 1)
abs ln 1 –
Cgd′ = 0.01 µF 50
RGD = 1K.

1. Vch = 2 – 1 = 1 V

0.01 µF (50 – 1)
2. Cch = = 0.49 µF
1

CONCLUSION REFERENCES

Active inrush current limiting can be accomplished with a [1] Mattingly, David. “Increasing Reliability of SMD Tantalum
single MOSFET and few external passive components. But in Capacitors In Low Impedance Applications,” Technical
an environment where high dv/dt is observed, the designer Information, AVX (1994).
must make sure that initially the MOSFET remains off. In order [2] Clow, Dave, Loomba, Jeet, Check, Ken. “NTC
to divert the charge at the gate of the device, a small Thermistors versus Active Circuits for Inrush Current
capacitance Cch can be added. Using the appropriate Suppression,” PCIM, (November 1993), pp. 18–24.
equations, correct values for Cch and Rch can be obtained [3] Humbert, Donald L., Martin, Hubert, Rainwater, Sam L.,
which will prevent false turnon due to high dv/dt. In a noisy Wittenbreder, Ernest H. “Active Inrush Current and
environment, the charge capacitance will divert much of the Current Slope Limiting,” in Conference Rec HFPC.,
noise away from the gate of the MOSFET. (1992), pp. 286–296.
[4] Baliga, B. Jayant. Modern Power Devices, John Wiley &
Sons, (1987), pp. 310–314.

MOTOROLA 13
AN1542

Motorola reserves the right to make changes without further notice to any products herein. Motorola makes no warranty, representation or guarantee regarding
the suitability of its products for any particular purpose, nor does Motorola assume any liability arising out of the application or use of any product or circuit,
and specifically disclaims any and all liability, including without limitation consequential or incidental damages. “Typical” parameters can and do vary in different
applications. All operating parameters, including “Typicals” must be validated for each customer application by customer’s technical experts. Motorola does
not convey any license under its patent rights nor the rights of others. Motorola products are not designed, intended, or authorized for use as components in
systems intended for surgical implant into the body, or other applications intended to support or sustain life, or for any other application in which the failure of
the Motorola product could create a situation where personal injury or death may occur. Should Buyer purchase or use Motorola products for any such
unintended or unauthorized application, Buyer shall indemnify and hold Motorola and its officers, employees, subsidiaries, affiliates, and distributors harmless
against all claims, costs, damages, and expenses, and reasonable attorney fees arising out of, directly or indirectly, any claim of personal injury or death
associated with such unintended or unauthorized use, even if such claim alleges that Motorola was negligent regarding the design or manufacture of the part.
Motorola and are registered trademarks of Motorola, Inc. Motorola, Inc. is an Equal Opportunity/Affirmative Action Employer.

Literature Distribution Centers:

USA/EUROPE: Motorola Literature Distribution; P.O. Box 20912; Phoenix, Arizona 85036.
JAPAN: Nippon Motorola Ltd.; 4-32-1, Nishi-Gotanda, Shinagawa-ku, Tokyo 141, Japan.
ASIA PACIFIC: Motorola Semiconductors H.K. Ltd.; Silicon Harbour Center, No. 2 Dai King Street, Tai Po Industrial Estate, Tai Po, N.T., Hong Kong.

*AN1542/D*
14 ◊ MOTOROLA
AN1542/D