Constant-Frequency, Current-Mode
Step-Up DC/DC Controller
ADP1621
Rev. A
Information furnished by Analog Devices is believed to be accurate and reliable. However, no
responsibility is assumed by Analog Devices for its use, nor for any infringements of patents or other
rights of third parties that may result from its use. Specifications subject to change without notice. No
license is granted by implication or otherwise under any patent or patent rights of Analog Devices.
Trademarks and registered trademarks are the property of their respective owners.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.
Tel: 781.329.4700 www.analog.com
Fax: 781.461.3113 ©2006 Analog Devices, Inc. All rights reserved.
FEATURES
92% efficiency (no sense resistor required)
±1.0% initial accuracy
IC supply voltage range: 2.9 V to 5.5 V
Power-input voltage as low as 1.0 V
Capable of high supply input voltage (>5.5 V)
with an external NPN or a resistor
VIN UVLO and 35 mA shunt regulator
External slope compensation with 1 resistor
Programmable operating frequency
(100 kHz to 1.5 MHz) with 1 resistor
Lossless current sensing for switch-node voltage <30 V
Resistor current sensing for switch-node voltage >30 V
Synchronizable to external clock
Current-mode operation for excellent line and load transient
responses
10 μA shutdown current
Current limit and thermal overload protection
Soft start in 2048 clock cycles
APPLICATIONS
APD bias
Portable electronic equipment
Isolated dc/dc converter
Step-up/step-down dc/dc converter
LED driver for laptop computer and navigation system
LCD backlighting
GENERAL DESCRIPTION
The ADP1621 is a fixed-frequency, pulse-width modulation
(PWM), current-mode, step-up converter controller. It drives an
external n-channel MOSFET to convert the input voltage to a
higher output voltage. The ADP1621 can also be used to drive
flyback, SEPIC, and forward converter topologies, either isolated
or nonisolated.
The ADP1621 eliminates the use of a current-sense power
resistor by measuring the voltage drop across the on resistance
of the n-channel MOSFET. This technique, allowed up to a
maximum voltage of 30 V at the switch node, maximizes
efficiency and reduces cost. For switch-node voltages higher than
30 V or for more accurate current limiting, the CS pin can be
connected to a current-sense resistor in the source of the MOSFET.
The slope compensation is implemented by an external resistor,
allowing a wide range of external components (inductors and
MOSFETs), and can be chosen for various switching frequencies
and input and output voltages.
The ADP1621 supply input voltage range is 2.9 V to 5.5 V, although
higher input voltages are possible with the use of a small-signal
TYPICAL APPLICATION CIRCUIT
ADP1621
IN
GATE
PGND
AGND
FB
SDSN
COMP
FREQ
GND
PIN
CS
R1
35.7k
1%
R2
11.5k
1%
C1
47µF
6.3V
L1
4.7µH
M1
C
COMP
1.8nF
C
OUT1
1µF
10V
f
OSC
= 600kHz
C1 = MURATA GRM31CR60J476M
C
OUT3
= SANYO POSCAP 6TPE150M
L1 = TOKO FDV0630-4R7M
M1 = VISHAY Si7882DP
D1 = VISHAY SSA33L
R
S
80C
OUT2
10µF
10V
R
FREQ
31.6k
1%
C
OUT3
150µF
6.3V
×2
V
OUT
= 5V
1A
V
IN
= 3.3V
D1
C3
1µF
10V
C4
0.1µF
10V
C2
120pF
R
COMP
9.09k
06090-001
Figure 1. High Efficiency Output Boost Converter in Lossless Mode,
3.3 V Input, 5 V Output (Bootstrapped)
100
30
0.01 10
LOAD CURRENT (A)
EFFICIENCY (%)
90
80
70
60
50
40
0.1 1
06090-042
Figure 2. Efficiency of Circuit Shown in Figure 1
NPN pass transistor or a single resistor. The voltage of the
power input can be as low as 1 V for fuel cell applications. The
switching frequency is set by an external resistor over a range of
100 kHz to 1.5 MHz and can be synchronized to an external
clock by using the SDSN pin. The shutdown quiescent current is
less than 10 μA. The ADP1621 has a thermal shutdown feature
that shuts down the gate driver when the junction temperature
reaches approximately 150°C. The internal soft start circuit limits
inrush current at startup. The ADP1621 is available in the 10-lead
MSOP lead-free package and is specified over the −40°C to +125°C
junction temperature range.
ADP1621
Rev. A | Page 2 of 32
TABLE OF CONTENTS
Features .............................................................................................. 1
Applications....................................................................................... 1
General Description......................................................................... 1
Typical Application Circuit ............................................................. 1
Revision History ............................................................................... 2
Specifications..................................................................................... 3
Absolute Maximum Ratings............................................................ 5
Thermal Resistance ...................................................................... 5
ESD Caution.................................................................................. 5
Simplified Block Diagram ............................................................... 6
Pin Configuration and Function Descriptions............................. 7
Typical Performance Characteristics ............................................. 8
Theory of Operation ...................................................................... 12
Control Loop............................................................................... 12
Current-Sense Configurations.................................................. 12
Current Limit .............................................................................. 13
Undervoltage Lockout ............................................................... 13
Shutdown..................................................................................... 13
Soft Start ...................................................................................... 13
Internal Shunt Regulators.......................................................... 13
Setting the Oscillator Frequency and Synchronization
Frequency .................................................................................... 13
Application Information: Boost Converter................................. 14
Duty Cycle ................................................................................... 14
Setting the Output Voltage........................................................ 14
Inductor Current Ripple............................................................ 14
Inductor Selection...................................................................... 14
Input Capacitor Selection.......................................................... 15
Output Capacitor Selection....................................................... 15
Diode Selection........................................................................... 15
MOSFET Selection..................................................................... 16
Loop Compensation .................................................................. 16
Slope Compensation .................................................................. 17
Current Limit.............................................................................. 18
Light Load Operation ................................................................ 18
Recommended Component Manufacturers........................... 19
Layout Considerations................................................................... 20
Efficiency Considerations ............................................................. 21
Examples of Application Circuits................................................. 22
Standard Boost Converter—Design Example ........................ 22
Bootstrapped Boost Converter................................................. 23
SEPIC Converter Circuit........................................................... 27
Low Voltage Power-Input Circuit ............................................ 27
LED Driver Application Circuits ............................................. 28
Related Parts.................................................................................... 30
Outline Dimensions ....................................................................... 31
Ordering Guide .......................................................................... 31
REVISION HISTORY
12/06—Rev. 0 to Rev. A
Changes to Table 1............................................................................ 3
Changes to Table 2............................................................................ 5
Added Table 3.................................................................................... 5
Changes to Table 5.......................................................................... 19
Changes to Ordering Guide .......................................................... 31
7/06—Revision 0: Initial Version
ADP1621
Rev. A | Page 3 of 32
SPECIFICATIONS
VIN = 5 V, RFREQ = 100 kΩ, fOSC = 200 kHz, TJ = −40°C to 125°C, unless otherwise noted.
Table 1.
Parameter Symbol Conditions Min Typ Max Unit
MAIN CONTROL LOOP
Internal Soft Start Time tSS 2048 Cycles
PIN Supply Voltage1VPIN 2.9 VSHUNT V
IN Supply Voltage1VIN 2.9 VSHUNT V
Shunt Regulation Voltage VSHUNT I
IN = 3 mA, IPIN = 3 mA, TA = 25°C 5.4 5.6 5.7 V
I
IN = 3 mA, IPIN = 3 mA 5.2 5.6 6.0 V
Shunt Resistance RSHUNT Current into IN = 8 mA to 12 mA 13 Ω
Current into PIN = 8 mA to 12 mA 7 Ω
IN Quiescent Current IIN V
IN = 2.9 V to 5.5 V, VFB = 1.215 V 1.8 3 mA
IN Shutdown Current VIN = 2.9 V to 5.5 V, SDSN = GND 1 10 μA
PIN Supply Current IPIN
Static Mode, No Switching VFB = 1.3 V, VCOMP < VCOMP,ZCT, GATE = 0 V 1 10 μA
Shutdown Mode SDSN = GND 1 10 μA
VUVLO V
UVLO rising 2.2 2.5 2.8 V
Undervoltage Lockout Threshold at
IN Pin V
UVLO hysteresis −80 mV
FB Regulation Voltage VFB T
A = 25°C 1.203 1.215 1.227 V
1.197 1.215 1.233 V
FB Input Current IFB V
FB = 1.215 V, TA = 25°C −75 +25 +75 nA
Line Regulation2∆VFB/∆VIN 2.9 V VIN ≤ 5 V, TJ = −40°C to +85°C 0.02 0.06 %/V
2.9 V ≤ VIN ≤ 5 V, TJ = −40°C to +125°C 0.02 0.072 %/V
Load Regulation3∆VFB/∆VCOMP VCOMP = 1.4 V to 1.5 V −1 −0.1 %
Error Amplifier Transconductance gm 300 μS
COMP Zero-Current Threshold VCOMP,ZCT 0.85 1.0 1.15 V
COMP Clamp High Voltage VCOMP,CLAMP T
J = −40°C to +85°C 1.9 2.0 2.1 V
T
J = −40°C to +125°C 1.9 2.0 2.2 V
Current-Sense Amplifier Gain n 7.5 9.5 11.5 V/V
Peak Slope-Compensation Current at
CS Pin4
ISC,PK VCS = 0 V to 100 mV maximum
across RS (GATE high)
55 70 85 μA
CS Pin Leakage Current ICS,LEAK V
CS = 30 V (GATE low) 5 μA
Shutdown Time tSD SDSN pin from high to low or left floating 50 μs
Thermal Shutdown Threshold5TTMSD 150 °C
Thermal Shutdown Hysteresis5 −10 °C
OSCILLATOR
Oscillator Frequency Range6fOSC 100 1500 kHz
Oscillator Frequency fOSC R
FREQ = 65 kΩ, TA = 25°C 255 325 395 kHz
Oscillator Frequency Tempco fOSC,TC ±0.06 %/°C
SDSN Input Level Threshold VSDSN,THRESH V
IN = VPIN = 5 V 1.5 1.7 1.9 V
SDSN Threshold Hysteresis −0.19 V
SDSN Internal Pull-Down Resistor RSDSN 100
Synchronization Minimum Pulse Width tSYNC,MIN V
SDSN = 0 V to VIN 45 100 ns
Synchronization Maximum Pulse Width tSYNC,MAX V
SDSN = 0 V to VIN 0.8/fSYNC ns
Synchronization Frequency fSYNC 110 1800 kHz
GATE Minimum On Time tON,MIN V
FB = 1.215 V, VCOMP = 1.0 V 180 215 ns
GATE Minimum Off Time tOFF,MIN V
FB = 1.215 V, VCOMP = 2.0 V 190 230 ns
Maximum Duty Cycle6, 7DMAX f
SW = 200 kHz, RFREQ = 100 kΩ 93 97 %
Recommended Maximum
Synchronized Frequency Ratio6, 8
fSYNC/fOSC f
OSC = 200 kHz, RFREQ = 100 kΩ, fSYNC = fSW 1.1 1.2 1.4
ADP1621
Rev. A | Page 4 of 32
Parameter Symbol Conditions Min Typ Max Unit
GATE DRIVER
GATE Rise Time9tR C
GATE = 3.3 nF 17 ns
GATE Fall Time9tF C
GATE = 3.3 nF 13 ns
1 The maximum input voltage is the shunt regulation voltage, which is typically 5.5 V and can range from 5.3 V to 6.0 V over the specified temperature range.
2 The ADP1621 is tested in a feedback servo loop, which servos VFB to the internal reference voltage. The voltage change in FB is measured while VIN is changed from
2.9 V to 5 V. The line regulation is calculated by (∆VFB/VFB) × 100%/∆VIN.
3 The ADP1621 is tested in a feedback servo loop, which servos VFB to the internal reference voltage, and VCOMP is forced from 1.4 V to 1.5 V. The VCOMP range is
(1.0 V ≤ VCOMP ≤ 2.0 V).
4 The peak slope-compensation current at the CS pin is typically 70 μA, and effectively clamped at 116 mV. Thus, RS should not exceed 1.6 kΩ (116 mV/70 μA).
5 Guaranteed by design for thermal shutdown. When the thermal junction temperature of the ADP1621 reaches approximately 150°C, the ADP1621 goes into thermal
shutdown and the GATE voltage is pulled low. When the junction temperature drops below about 140°C, the soft start sequence is initiated and the ADP1621 resumes
normal operation.
6 fOSC is the natural oscillation frequency, fSYNC is the synchronization frequency, and fSW is the switching frequency. If synchronization is used, then fSW = fSYNC; otherwise, fSW = fOSC.
7 Guaranteed by design and bench characterization.
8 To ensure proper synchronization operation, set the synchronization frequency, fSYNC, to 1.2× of the free-running frequency, fOSC. Although the switching frequency can
be synchronized to as high as 1.8 MHz, the peak slope-compensation current decreases at higher synchronization frequencies. It is recommended that the maximum
fSYNC be less than 1.4× of fOSC and should not exceed 1.8 MHz. The slope-compensation resistor, RS, should be chosen for the synchronization frequency (see the Slope
Compensation section in the Application Information: Boost Converter section).
9 GATE rise and fall times are measured from 10% to 90% levels.
ADP1621
Rev. A | Page 5 of 32
ABSOLUTE MAXIMUM RATINGS
Table 2.
Parameter Rating
IN to GND −0.3 V to VSHUNT
FB, COMP, SDSN, FREQ, GATE to GND −0.3 V to (VIN + 0.3 V)
CS to GND −5 V to +33 V
PIN to PGND −0.3 V to VSHUNT
Supply Current into IN 25 mA
Supply Current into PIN 35 mA
Storage Temperature Range −55°C to +150°C
Junction Operating Temperature Range1 −55°C to +150°C
Junction Storage Temperature Range −55°C to +150°C
Lead Temperature (Soldering, 10 sec) 300°C
Package Power Dissipation1 (TJ,MAXTA)/θJA
1In applications where high power dissipation and poor package thermal
resistance are present, the maximum ambient temperature may need to be
derated. Maximum ambient temperature (TA,MAX) is dependent on the
maximum operating junction temperature (TJ,MAX = 150oC), the maximum
power dissipation of the device in the application (PD,MAX), and the junction-
to-ambient thermal resistance of the package in the application (θJA), is given
by the following equation: TA,MAX = TJ,MAX --- (θJA x PD,MAX).
Stresses above those listed under Absolute Maximum Ratings
may cause permanent damage to the device. This is a stress
rating only; functional operation of the device at these or any
other conditions above those indicated in the operational
section of this specification is not implied. Exposure to absolute
maximum rating conditions for extended periods may affect
device reliability.
Absolute maximum ratings apply individually only, not in
combination. Unless otherwise specified, all other voltages are
referenced to GND.
THERMAL RESISTANCE
θJA is specified for the worst-case conditions, that is, a device
soldered in a circuit board for surface-mount packages.
Table 3. Thermal Resistance
Package Type θJA Unit
10-lead MSOP on a 2-layer PCB 200 °C/W
10-lead MSOP on a 4-layer PCB 172 °C/W
Junction-to-ambient thermal resistance of the package is based
on modeling and calculation using 2-layer and 4-layer boards,
and natural convection. The junction-to-ambient thermal
resistance is application- and board-layout dependent. In
applications where high maximum power dissipation exists,
attention to thermal dissipation issues in board design is
required.
ESD CAUTION
ADP1621
Rev. A | Page 6 of 32
SIMPLIFIED BLOCK DIAGRAM
CS
PGND
FREQ
OSC
V
OSC
1.4V
V
REF
1.215V
5.5V
5.5V
100k
R
S
SET
SLOPE
COMP
PWM
COMPARATOR
ADP1621
GATE
DRIVER
ERROR
AMPLIFIER
GND
FB
GATE
PIN
COMP
IN
SDSN
g
m
n
+
+
SOFT START
(2048 CYCLES)
UVLO
06090-002
Figure 3. ADP1621 Simplified Block Diagram
ADP1621
Rev. A | Page 7 of 32
PIN CONFIGURATION AND FUNCTION DESCRIPTIONS
SDSN
1
GND
2
C
OMP
3
FB
4
FREQ
5
IN
10
CS
9
PIN
8
GATE
7
PGND
6
ADP1621
TOP VIEW
(Not to Scale)
06090-003
Figure 4. Pin Configuration
Table 4. Pin Function Descriptions
Pin No. Mnemonic Description
1 SDSN
Shutdown and Synchronization Input. Turn the ADP1621 on by driving SDSN high; turn it off by driving SDSN low.
If SDSN is left floating or when the SDSN is pulled low, the ADP1621 goes into shutdown after 50 μs. If synchronization is
needed, synchronize the switching frequency to an external clock by connecting the external clock to the SDSN
pin. An internal 100 kΩ pull-down resistor is connected from SDSN to GND.
2 GND Ground.
3 COMP
Regulation Control Compensation Node. COMP is the output of the internal transconductance error amplifier.
Connect a series RC from COMP to GND to compensate the regulator. The nominal voltage range for this pin is
1.0 V to 2.0 V.
4 FB Feedback Input. FB is the input to the internal transconductance error amplifier. Drive FB from the output voltage
through a resistive voltage divider. The ratio of the voltage divider sets the output voltage. The regulation voltage
at FB is nominally 1.215 V.
5 FREQ
Frequency Control Input. Connect a resistor from FREQ to GND to set the free-running switching frequency
between 100 kHz and 1.5 MHz. The nominal voltage of this pin is 1.4 V.
6 PGND
Power Ground Input. PGND is the ground return for the internal gate driver and the negative input of the internal
current-sense amplifier. Connect PGND to GND as close to the ADP1621 as possible.
7 GATE
Gate Driver Output. The maximum gate driver output is equal to the PIN voltage. GATE drives the gate of the
external n-channel power MOSFET. Connect GATE to the gate of the MOSFET.
8 PIN Power Input. PIN powers the gate driver output. An internal 5.5 V shunt regulator is connected to this pin. Bypass
PIN to PGND with a 0.1 μF or greater capacitor.
9 CS Current-Sense Input. CS is the positive input of the current-sense amplifier. When GATE is turned on, the voltage at
the CS pin increases linearly from 0 V to a maximum of 116 mV, and the nominal peak slope-compensation output
current is 70 μA. When GATE is off, the CS function is disabled. For current sensing in lossless mode, connect CS to
the drain of the power MOSFET. The absolute maximum voltage at CS is 33 V. For higher accuracy current sensing
or higher switch-node voltages, connect CS to a current-sense power resistor in the source of the power MOSFET.
In both sensing methods, it is required to add a slope-compensation resistor, RS, to the CS pin to achieve stability
in the inductor current for duty cycles greater than 50%. However, it is recommended to add RS for all duty cycles
because load transients can momentarily cause the duty cycle to be greater than 50%, even when the steady-
state duty cycle is less than 50%.
10 IN Input Voltage. IN powers the ADP1621 internal circuitry. An internal 5.5 V shunt regulator is connected to this pin.
Bypass IN to GND with a 0.1 μF or greater capacitor.
ADP1621
Rev. A | Page 8 of 32
TYPICAL PERFORMANCE CHARACTERISTICS
100
30
0.01 10
LOAD CURRENT (A)
EFFICIENCY (%)
90
80
70
60
50
40
0.1 1
T
A
= 25°C
f
SW
= 220kHz
V
IN
= 3.3V
V
OUT
= 5V
06090-004
Figure 5. Efficiency vs. Load Current
CH1 20mV CH2 2V M2µs A CH2 2.6V
1
2
T
A
= 25°C
V
IN
= 3.3V
V
OUT
= 5V
LOAD = 1A
V
OUT
RIPPLES @ 5V
AC-COUPLED
CH2 = GATE
0
6090-005
Figure 6. Output Voltage Ripple of the Circuit Shown in Figure 1
1.21605
1.21575
2.5 6.0
V
IN
(V)
V
FB
(V)
T
A
=25°C
1.21600
1.21595
1.21590
1.21585
1.21580
3.0 3.5 4.0 4.5 5.0 5.5
06090-006
Figure 7. VFB vs. VIN
92
84
100
SWITCHING FREQUENCY (kHz)
EFFICIENCY (%)
91
90
89
88
87
86
85
300 500 700 900 1100 1300 1500
TA = 25°C
VIN = 3.3V
VOUT = 5V
LOAD = 1A
LOAD = 0.5A
06090-007
Figure 8. Efficiency vs. Switching Frequency
100
0.00001
07
SUPPLY VOLTAGE (V)
SUPPLY CURRENT (mA)
TA = 25°C
NO SWITCHING
10
1
0.1
0.01
0.001
0.0001
123456
IIN
IPIN
06090-008
Figure 9. Supply Current vs. Supply Voltage
2.5
0
1.17 1.29
V
FB
(V)
V
COMP
(V)
T
A
= 25°C
V
IN
= 5V
2.0
1.5
1.0
0.5
1.19 1.21 1.23 1.25 1.27
06090-009
Figure 10. VCOMP vs. VFB
ADP1621
Rev. A | Page 9 of 32
45
0
0 1800
SWITCHING FREQUENCY (kHz)
PIN SUPPLY CURRENT (mA)
MOSFET Q
G
= 25nC
MOSFET Q
G
= 15nC
MOSFET Q
G
= 7nC
40
35
30
25
20
15
10
5
200 400 600 800 1000 1200 1400 1600
06090-010
Figure 11. PIN Supply Current vs. Switching Frequency
2.60
2.40
–50 150
TEMPERATUREC)
V
UVLO
(V)
SDSN = 5V
2.55
2.50
2.45
0 50 100
06090-011
Figure 12. VUVLO Threshold vs. Temperature
1.03
0.97
–50 150
TEMPERATUREC)
NORMALIZED FREQUENCY (
f
OSC
/
f
OSC,25°C
)
0 50 100
1.02
1.01
1.00
0.99
0.98
V
IN
= 5V
06090-012
Figure 13. Frequency vs. Temperature
35
0
05
GATE CAPACITANCE (nF)
GATE RISE AND FALL TIMES (ns)
0
30
25
20
15
10
5
5 1015202530354045
T
A
= 25°C
V
IN
= V
PIN
= 5V
t
R
OR
t
F
IS FROM
10% TO 90% OF
THE GATE VOLTAGE
t
F
t
R
06090-013
Figure 14. GATE Rise and Fall Times vs. CGATE
1600
0
0 200
R
FREQ
(k)
f
OSC
(kHz)
1500
1400
1300
1200
1100
1000
900
800
700
600
500
400
300
200
100
20 40 60 80 100 120 140 160 180
06090-014
Figure 15. Oscillator Frequency vs. Resistance
198
191
2
V
IN
(V)
fOSC
(kHz)
197
196
195
194
193
192
345
T
A
= 25°C
R
FREQ
= 100k
06090-015
Figure 16. Oscillator Frequency vs. VIN
ADP1621
Rev. A | Page 10 of 32
250
0
–40 160
CS LEAKAGE (nA)
TEMPERATURE (°C)
V
IN
= 5V
CS = 30V
200
150
100
50
10 60 110
06090-016
Figure 17. Temperature vs. CS Leakage
8
–16
–50 150
TEMPERATUREC)
FB BIAS CURRENT (nA)
4
0
–4
–8
–12
0 50 100
V
FB
= 1.2113V AT 25°C
FB BIAS CURRENT IS MEASURED
BY FORCING A CONSTANT 1.2113V
OVER THE TEMPERATURE RANGE.
06090-017
Figure 18. FB Bias Current vs. Temperature
90
0
1.0
f
SYNC
/f
OSC
PEAK SLOPE COMPENSATION CURRENT (µA)
2.2
f
OSC
= 200kHz
f
OSC
= 550kHz
80
70
60
50
40
30
20
10
1.2 1.4 1.6 1.8 2.0
06090-018
Figure 19. Slope-Compensation Current vs. fSYNC/fOSC
1.6
0
–50 150
TEMPERATUREC)
SHUTDOWN IN CURRENT (µA)
0 50 100
1.4
1.2
1.0
0.8
0.6
0.4
0.2
V
IN
= 5V
SDSN = 0V
06090-019
Figure 20. Shutdown IN Current vs. Temperature
1.2165
1.2120
–50 150
TEMPERATURE (°C)
V
FB
(V)
050100
V
IN
= 5V
SDSN = 0V
1.2160
1.2155
1.2150
1.2145
1.2140
1.2135
1.2130
1.2125
06090-020
Figure 21. FB Voltage vs. Temperature
CH1 5V CH2 5V
CH4 500mA
M2µs A CH1 2.9V
1
2
4
T
A
= 25°C
V
IN
= 3.3V
V
OUT
= 5V
LOAD = 0.1A
DCM OPERATION
CH1 = GATE
CH4 = INDUCTOR CURRENT
CH2 = DRAIN VOLTAGE
06090-021
Figure 22. DCM Switching Waveform
ADP1621
Rev. A | Page 11 of 32
CH1 5V CH2 5V
CH4 500mA
M2µs A CH1 2.9V
1
2
4
T
A
= 25°C
V
IN
= 3.3V
V
OUT
= 5V
LOAD = 0.3A
CCM OPERATION
CH1 = GATE
CH4 = INDUCTOR CURRENT
CH2 = DRAIN VOLTAGE
06090-022
CH1 50mV
CH4 1A
M200µs A CH4 700V
1
4
TA = 25°C
VIN = 3.3V
VOUT = 5V
LOAD CURRENT
FROM 0.2A TO 1.2A
OUTPUT, AC-COUPLED
06090-025
Figure 23. CCM Switching Waveform Figure 26. Load Transient Response of the Circuit Shown in Figure 1
CH1 1V
CH3 5V
CH2 5V M2ms A CH1 4.5V
1
2
3
TA = 25°C
VIN = 3.3V
VOUT = 5V
f
SW = 220kHz
SOFT-START = 9.3ms
CH3 = GATE
CH1 = VOUT
CH2 = SDSN
06090-023
CH1 50mV CH2 2V M400µs A CH2 3.8V
2
1
TA = 25°C
VOUT = 5V
NO LOAD AT VOUT
CH1 = VOUT, AC-COUPLED
CH2 = VIN FROM 3V TO 4V
0
6090-026
Figure 24. Soft Start Waveform Figure 27. Line Transient Response of the Configuration Shown in Figure 1
with No Load
CH1 50mV CH2 2V M400µs A CH2 3.8V
2
1
TA = 25°C
VOUT = 5V
LOAD AT VOUT = 1A
CH1 = VOUT, AC-COUPLED
CH2 = VIN FROM 3V TO 4V
0
6090-024
Figure 25. Line Transient Response of the Configuration Shown in Figure 1
with a 1 A Load
ADP1621
Rev. A | Page 12 of 32
THEORY OF OPERATION
The ADP1621 is a fixed-frequency, current-mode, step-up dc/dc
converter controller. It drives an external n-channel MOSFET
to step the input voltage up to a higher output voltage. It can be
used for SEPIC, flyback, boost, buck-boost, forward, and other
converter topologies. It operates at a fixed switching frequency that
is set by an external resistor over a range of 100 kHz to 1.5 MHz,
and it can be synchronized to an external clock by connecting
the SDSN pin to the clock.
The input supply current to the ADP1621 is less than 3 mA
during normal operation and less than 10 μA during shutdown.
The ADP1621 can drive large external MOSFETs, allowing it to
support load currents in excess of 10 A.
CONTROL LOOP
The ADP1621 uses a current-mode architecture to regulate the
output voltage. The output voltage is monitored at FB through
a resistive voltage divider. The voltage at FB is compared to the
internal 1.215 V reference voltage by the internal transconductance
error amplifier to create an error current at COMP. A resistor-
capacitor compensation impedance connected from COMP to
GND converts the error current to an error voltage.
At the beginning of the switching cycle, the MOSFET is turned
on and the inductor current ramps up. The MOSFET current is
measured and converted to a voltage using RCS or RDSON and is
added to the stabilizing slope-compensation ramp. The resulting
voltage sum passes through the current-sense amplifier to generate
the current-sense voltage. When the current-sense voltage is
greater than the COMP error voltage, the MOSFET is turned off
and the inductor current ramps down until the internal clock
initiates the next switching cycle. The duty-cycle of the PWM
modulator is thus adjusted to provide the necessary load current
at the desired output voltage. Because the output voltage ultimately
controls the peak inductor current through the COMP error
voltage, this scheme is referred to as peak current-mode control.
With light loads, the converter can also operate under discon-
tinuous conduction mode and pulse-skipping modulation to
maintain output-voltage regulation. These two forms of operation
are discussed in detail in the Light Load Operation section.
Note that the converter can also be designed to operate in
discontinuous conduction mode at full load if desired.
Overall, the current-mode regulation system of the ADP1621
allows fast transient responses while maintaining a stable output
voltage. By selecting the proper resistor-capacitor network from
COMP to GND, the regulator response can be optimized for a
wide range of input voltages, output voltages, and load currents.
CURRENT-SENSE CONFIGURATIONS
The ADP1621 can sense the current across the on resistance of
the MOSFET to minimize external component count and improve
efficiency by eliminating the power that would be lost in a current-
sense resistor. This lossless technique eliminates the need for an
expensive current-sense resistor. In the lossless mode configuration,
the voltage at the CS pin (or the switch-node voltage at the drain of
the MOSFET) must not exceed 30 V (see Figure 28). This technique
maximizes efficiency and reduces cost. In practice, when the
calculated VSW approaches 30 V, one should build the board and
measure the actual VSW before committing to the lossless mode
design. Because of the parasitic inductance in the diode, output
capacitor, and PCB traces, VSW typically has narrow peaks that
exceed the theoretical maximum voltage at VSW—the sum of
VOUT and the forward-voltage drop of Diode D1. If the measured
peak voltage exceeds 30 V, or if a more accurate current limit is
desired, then the CS pin can be connected to an external current-
sense resistor in the source of the MOSFET (Figure 29). The
maximum power output is limited by the selection of the
external components.
IN
GATE
SDSN
GND
PIN
CS
PGND
L
R
S
D1
V
IN
V
OUT
V
SW
C
O
06090-027
ADP1621
Figure 28. CS Pin Connection for VSW < 30 V, Lossless Mode
(No Current-Sense Resistor Needed)
IN
CS
SDSN
GND
PIN
GATE
PGND
L
R
S
D1
V
IN
V
OUT
R
CS
V
SW
C
O
06090-028
ADP1621
Figure 29. CS Pin Connection for VSW > 30 V, Resistor Sense Mode
with a Current-Sense Resistor, RCS
ADP1621
Rev. A | Page 13 of 32
CURRENT LIMIT SETTING THE OSCILLATOR FREQUENCY AND
SYNCHRONIZATION FREQUENCY
The current limit is achieved by the COMP voltage clamp, owing
to the current-mode operation of the ADP1621. A detailed
explanation of how the current limit is determined can be found
in the Current Limit section of the Application Information:
Boost Converter section.
The free-running oscillator frequency, fOSC, is set by a resistor
from FREQ to GND. A 100 kΩ resistor sets the typical oscillator
frequency to 200 kHz, a 65 kΩ resistor sets it to 325 kHz, a 32 kΩ
resistor sets it to 600 kHz, and a 10 kΩ resistor sets it to 1.5 MHz.
Figure 30 shows a typical relationship between fOSC and RFREQ.
UNDERVOLTAGE LOCKOUT
1600
0
0 200
R
FREQ
(k)
f
OSC
(kHz)
1500
1400
1300
1200
1100
1000
900
800
700
600
500
400
300
200
100
20 40 60 80 100 120 140 160 180
06090-029
An internal undervoltage lockout (UVLO) circuit at the IN pin
holds the GATE voltage low when the IN voltage is below the
UVLO voltage, which is typically 2.5 V.
SHUTDOWN
The ADP1621 goes into shutdown approximately 50 μs after the
SDSN pin is pulled low or left floating. There is an internal 100 kΩ
resistor connected between SDSN and GND.
When the junction temperature of the ADP1621 reaches about
150°C, the ADP1621 goes into thermal shutdown and the GATE
voltage is pulled low. When the junction temperature drops below
about 140°C, the ADP1621 resumes normal operation after the
soft start sequence.
SOFT START Figure 30. fOSC vs. RFREQ
The ADP1621 has an internal soft start circuit that ramps
the FB regulation voltage from 0 V to 1.215 V in 64 steps over
2048 clock oscillator cycles. This soft start ramp allows the
output voltage to slowly rise to the steady-state output voltage,
preventing input inrush current at startup.
The switching frequency can be synchronized to an external clock
by driving the SDSN pin with that clock signal. The SDSN pin
serves the two functions of shutdown control and frequency
synchronization input. If the SDSN input detects a low-to-high
transition within 10 μs of a high-to-low transition, it resets the
oscillator to synchronize to the frequency of the signal at SDSN.
The ADP1621 only synchronizes to frequencies greater than the
free-running switching frequency. To ensure proper synchronization
operation, set the synchronization frequency, fSYNC, to 1.2× the free-
running frequency, fOSC. The switching frequency, fSW, is equal to
fSYNC. Although the switching frequency can be synchronized to as
high as 1.8 MHz, the peak slope-compensation current decreases at
higher fSYNC. It is recommended that the maximum fSYNC be less than
1.4× of fOSC. The slope-compensation resistor, RS, should be chosen
for the synchronization frequency (see the Slope Compensation
section). For SDSN to detect a high input, the high state must
remain high for at least 100 ns.
INTERNAL SHUNT REGULATORS
The IN and PIN pins each have an internal shunt regulator that
allows the ADP1621 to operate over a wide input voltage range.
The shunt regulators limit the voltages at IN and PIN to about
5.5 V, allowing the use of logic-level MOSFETs independent of
the input and/or output voltage. The shunt regulator voltage can
reach 5.7 V at 10 mA. See Figure 9 for the I-V characteristics of
these shunt regulators.
The internal power is derived from the IN pin, whereas the
MOSFET gate driver (GATE) current comes from the power
input, PIN. By separating the two inputs, PIN can be driven
with an external small-signal NPN transistor to limit the power
loss in the PIN shunt regulator when the input voltage is higher
than 5.5 V. See Figure 37 for an example. The maximum currents
going into PIN and IN should not exceed 35 mA and 25 mA,
respectively.
ADP1621
Rev. A | Page 14 of 32
APPLICATION INFORMATION: BOOST CONVERTER
In this section, an analysis of a boost converter is presented,
along with guidelines for component selection. A typical boost-
converter application circuit is shown in Figure 1.
DUTY CYCLE
To determine the worst-case inductor current ripple, output voltage
ripple, and slope-compensation factor, it is first necessary to
determine the system duty cycle. The duty cycle in continuous
conduction mode (CCM) is calculated by the equation
D
OUT
IND
OUT
VV
V
V
V
D+
+
= (1)
where VOUT is the desired output voltage, VIN is the input
voltage, and VD is the forward-voltage drop of the diode. A
typical Schottky diode has a forward-voltage drop of 0.5 V.
The GATE minimum on and off times determine the minimum
and maximum duty cycles, respectively. The minimum on and
off times are typically 180 ns and 190 ns, respectively. The
minimum and maximum duty cycles are given by
SWMINON
SW
MINON
MIN ft
t
t
D×== ,
, (2)
)(11 ,
,
SWMINOFF
SW
MINOFF
MAX ft
t
t
D×== (3)
where DMIN is the minimum duty cycle, DMAX is the maximum duty
cycle, tON,MIN is the minimum on time, tOFF,MIN is the minimum off
time, tSW is the switching period, and fSW is the switching frequency.
Note that when the converter tries to operate at a duty cycle
lower than DMIN, pulse-skipping modulation occurs to maintain
the output voltage regulation (see the Light Load Operation
section).
SETTING THE OUTPUT VOLTAGE
The output voltage is set through a voltage divider from the
output voltage to the FB input. The feedback resistor ratio sets
the output voltage of the system. The regulation voltage at FB is
1.215 V. The output voltage is given by (see Figure 1)
+×= R2
R1
VOUT 1V215.1 (4)
The input bias current into FB is 25 nA typical, 70 nA
maximum. For a 0.1% degradation in regulation voltage and
with 70 nA bias current, R2 must be less than 18 kΩ, which
results in 68 μA of divider current. Choose the value of R1 to set
the output voltage. Using higher values for R2 results in reduced
output voltage accuracy due to the input bias current at the FB
pin, whereas lower values cause increased quiescent current
consumption.
INDUCTOR CURRENT RIPPLE
Choose a peak-to-peak inductor ripple current between 20%
and 40% of the average inductor current. A good starting point
for a design is to choose the peak-to-peak ripple current to be
30% of 1/(1 − D) times the maximum load current:
D
I
IMAXLOAD
L
×=Δ 1
3.0 , (5)
where ΔIL is the peak-to-peak inductor ripple current, and ILOAD,MAX
is the maximum load current required by the application.
INDUCTOR SELECTION
The inductor value choice is important because it dictates
the inductor current ripple and therefore the voltage ripple
at the output.
The average inductor current, IL,AVE, is given by
D
I
ILOAD
AVEL
=1
, (6)
and the peak-to-peak inductor ripple current is inversely
proportional to the inductor value:
Lf
DV
I
SW
IN
L×
×
=Δ (7)
where fSW is the switching frequency, and L is the inductor value.
Assuming continuous conduction mode (CCM) operation, the
peak inductor current is given by
Lf
DV
D
I
I
D
I
I
SW
INLOAD
L
LOAD
PKL ××
×
+
=
Δ
+
=2121
, (8)
Smaller inductor values are typically smaller in size and usually
less expensive, but increase the ripple current. Larger ripple current
also increases the power loss in the inductor core. Too large an
inductor value results in added expense and may impede load
transient responses because it reduces the effect of slope
compensation.
Assuming the ripple current is 30% of 1/(1 − D) times the max-
imum load current, a reasonable choice for the inductor value is
(
)
MAXLOADSW
IN
If
DDV
L
,
3.0
1
××
= (9)
From this starting point, modify the inductance to obtain the
right balance of size, cost, and output voltage ripple while
maintaining the inductor ripple current between 20% and 40%
of 1/(1 − D) times the maximum load current. Keep in mind
that the inductor saturation current must be greater than the
peak inductor current. Magnetically shielded inductors are
generally recommended, although they cost slightly more than
unshielded inductors.
Also, losses due to the inductor winding resistance reduce the
efficiency of the boost converter. This power loss is given by
W
LOAD
WL R
D
I
P×
=
2
,1 (10)
where PL,W is the power dissipation in the winding of the
inductor, and RW is the winding resistance.
ADP1621
Rev. A | Page 15 of 32
INPUT CAPACITOR SELECTION
The bulk input capacitor provides a low impedance path for the
inductor ripple current. Capacitor C1 in Figure 1 represents a
bulk input capacitor. Choose a bulk input capacitor whose
impedance at the switching frequency is lower than the
impedance of the voltage source VIN.
The preferred bulk input capacitor is a 10 μF to 100 μF ceramic
capacitor because it has low equivalent series resistance (ESR) and
low impedance. Aluminum electrolytic and aluminum polymer
capacitors can also be used as the bulk input capacitors. The bulk
input capacitor does not need to be placed very close to the IN
and PIN pins. Aluminum electrolytic capacitors are the cheapest
and generally have high ESR values, which increase dramatically at
temperatures less than 0°C. Some aluminum electrolytic capacitors
have ESR less than 20 mΩ, but their capacitances are generally
greater than 800 μF. Aluminum polymer capacitors are more
expensive than the aluminum electrolytic ones, but are generally
cheaper than the ceramic capacitors for the same amount of
capacitance. Polymer capacitors have relatively low ESR, with
some models having less than 10 mΩ.
Regardless of the type of capacitor used, make sure the ripple
current rating of the bulk input capacitor, ICIN,RMS, is greater than
2
3
1
,
L
RMSCIN
I
IΔ
×= (11)
where ΔIL is the peak-to-peak inductor ripple current.
In addition to the bulk input capacitor, a bypass input capacitor is
required. The function of the bypass capacitor is to locally filter the
input voltage to the ADP1621 and maintain the input voltage at a
steady value during switching transitions. The bypass capacitor is
typically a 0.1 μF or greater ceramic capacitor and should be placed
as close as possible to the IN and PIN pins of the ADP1621.
Capacitors C3 and C4 in Figure 1 represent the bypass capacitors.
OUTPUT CAPACITOR SELECTION
The output capacitor maintains the output voltage and supplies
current to the load while the external MOSFET is on.
The value and characteristics of the output capacitor greatly
affect the output voltage ripple and stability of the converter.
The amount of peak-to-peak output voltage ripple, ΔVOUT, can
be approximated by
×
Δ
+
Δ 21
L
LOAD
OUT
I
D
I
V
()
22
2
2
2
1ESLfESR
Cf SW
OUT
SW
××π++
××π (12)
where ΔIL is the peak-to-peak inductor ripple current, fSW is the
switching frequency, COUT is the output capacitance, ESR is the
effective ESR of COUT, and ESL is the effective equivalent series
inductance of COUT.
Because the output capacitor is typically greater than 40 μF, the
ESR dominates the output capacitance impedance and thus the
output voltage ripple. The use of low ESR, ceramic dielectric
capacitors is preferred, although aluminum electrolytic,
tantalum, OS-CON™ (from Sanyo), and aluminum polymer
capacitors can be used. At higher switching frequencies, the ESL
of the output capacitor may also be a factor in determining the
output voltage ripple. Multiple capacitors can be connected in
parallel to reduce the effective ESR and ESL. Keep in mind that
the capacitance of a given capacitor typically degrades with
increased temperature and bias voltage. Consult the capacitor
manufacturers data sheet when determining the actual
capacitance of a capacitor under certain conditions.
Ensure that the output capacitor ripple current rating, ICOUT,RMS,
is greater than
D
D
II LOAD
RMSCOUT
×= 1
, (13)
DIODE SELECTION
The diode conducts the inductor current to the output capacitor
and load while the MOSFET is off. The average diode current is
the load current:
LOADAVEDIODE II
=
, (14)
The rms diode current in continuous conduction mode is given by
D
D
I
ILOAD
RMSDIODE ×
=1
1
, (15)
where D is the duty cycle.
The power dissipated in the diode is
LOAD
D
DIODE IVP
×
=
(16)
where VD is the forward-voltage drop of the diode.
ADP1621
Rev. A | Page 16 of 32
The total power dissipation determines the diode junction
temperature, which is given by
JADIODEADIODEJ θPTT ×+=
, (17)
where TJ,DIODE is the junction temperature, TA is the ambient tem-
perature, and θJA is the junction-to-ambient thermal resistance
of the diode package. The diode junction temperature must not
exceed its maximum rating at the given power dissipation level.
For high efficiency, Schottky diodes are recommended. The low
forward-voltage drop of a Schottky diode reduces the power losses
during the MOSFET off time, and the fast switching speed reduces
the switching losses during the MOSFET transitions. However,
for high voltage, high temperature applications where the reverse
leakage current of the Schottky diode can become significant
and degrade efficiency, use an ultrafast-recovery junction diode.
Make sure that the diode is rated to handle the average output
load current. Many diode manufacturers derate the current
capability of the diode as a function of the duty cycle. Verify
that the diode is rated to handle the average output load current
with the minimum duty cycle. Also, ensure that the peak inductor
current is less than the maximum rated current of the diode.
MOSFET SELECTION
When turned on, the external n-channel MOSFET allows
energy to be stored in the magnetic field of the inductor. When
the MOSFET is turned off, this energy is delivered to the load to
boost the output voltage.
The choice of the external power MOSFET directly affects the
boost converter performance. Choose the MOSFET based on
the following: threshold voltage (VT), on resistance (RDSON),
maximum voltage and current ratings, and gate charge.
The minimum operating voltage of the ADP1621 is 2.9 V.
Choose a MOSFET with a VT that is at least 0.3 V less than the
minimum input supply voltage at PIN used in the application.
Ensure that the maximum VGS rating of the MOSFET is at least
a few volts greater than the maximum voltage that is applied to
PIN. Ensure that the maximum VDS rating of the MOSFET
exceeds the maximum VOUT by at least 5 V to 10 V. Depending
on parasitics, the MOSFET may be exposed to voltage spikes that
exceed the sum of VOUT and the forward-voltage drop of the diode.
Estimate the rms current in the MOSFET under continuous
conduction mode by
D
D
I
ILOAD
RMSMOSFET ×
=1
, (18)
where D is the duty cycle. Derate the MOSFET current at least
20% to account for inductor ripple and changes in the forward-
voltage drop of the diode.
The MOSFET power dissipation due to conduction is thus
(
KRD
D
I
PDSON
LOAD
C+×××
=1
1
2
)
(19)
where PC is the conduction power loss, and RDSON is the MOSFET
on resistance. The variable K is a factor that models the increase
of RDSON with temperature:
(
)
C25C/005.0 oo ×= J,MOSFET
TK (20)
where TJ,MOSFET is the MOSFET junction temperature. Note that
multiple n-channel MOSFETs can be placed in parallel to reduce
the effective RDSON.
The power dissipation due to switching transition loss is
approximated by
()
()
2
1SW
FR
LOAD
D
OUT
SW
ftt
D
I
VV
P
×+×
×+
= (21)
where PSW is the switching power loss, tR is the MOSFET rise
time, and tF is the MOSFET fall time. The MOSFET rise and fall
times are functions of both the gate drive circuitry and the
MOSFET used in the application.
The total power dissipation of the MOSFET is the sum of the
conduction and transition losses:
SWC
MOSFET PPP
+
=
(22)
where PMOSFET is the total MOSFET power dissipation. Ensure
that the maximum power dissipation is significantly less than
the maximum power rating of the MOSFET.
The total power dissipation also determines the MOSFET
junction temperature, which is given by
JA
MOSFET
A
MOSFETJ θPTT ×
+
=
, (23)
where TJ,MOSFET is the junction temperature, TA is the ambient
temperature, and θJA is the junction-to-ambient thermal
resistance of the MOSFET package. The MOSFET junction
temperature must not exceed its maximum rating at the given
power dissipation level.
If lossless current sensing is not used, there will also be power
dissipation in the external current-sense resistor, RCS. The power
dissipation, PCS, in the external resistor due to conduction losses
is given by
CS
LOAD
CS RD
D
I
P××
=
2
1 (24)
LOOP COMPENSATION
The ADP1621 uses external components to compensate the
regulator loop, allowing optimization of the loop dynamics for
a given application.
The step-up converter produces an undesirable right-half plane
(RHP) zero in the regulation feedback loop. This RHP zero
requires compensating the regulator such that the crossover
ADP1621
Rev. A | Page 17 of 32
frequency occurs well below the frequency of the RHP zero. The
location of the RHP zero is determined by the following equation:
()
L
R
Df LOAD
RHPZ ×π
×= 2
12
, (25)
where fZ,RHP is the RHP zero frequency, and RLOAD is the equivalent
load resistance or the output voltage divided by the load current.
To stabilize the regulator, ensure that the regulator crossover
frequency is less than or equal to one-fifth of the RHP zero
frequency and less than or equal to one-fifteenth of the switching
frequency. For an initial practical design, choose the crossover
frequency fC to be the lower of
15
SW
C
f
f= (26)
and
5
,RHPZ
f
C
f= (27)
where fC is the crossover frequency, and fSW is the switching
frequency.
The regulator loop gain is
()
||
1
||1 OUT
CS
COMP
m
OUT
FB
VL Z
Rn
ZgD
V
V
A×
×
××××= (28)
where AVL is the loop gain, VFB is the feedback regulation
voltage (typically 1.215 V), VOUT is the regulated output voltage,
D is the duty cycle, gm is the error amplifier transconductance
gain (typically 300 μS), ZCOMP is the impedance of the RC network
from COMP to GND, n is the current-sense amplifier gain
(typically 9.5), RCS is the current-sense resistance, and ZOUT is
the impedance of the load and output capacitor. In the case of
lossless current sensing, as shown in Figure 28, RCS is equal to the
on resistance, RDSON, of the external power MOSFET. Otherwise,
RCS represents the external current-sense resistor, as shown in
Figure 29.
To determine the crossover frequency, it is important to note
that at that frequency the compensation impedance, ZCOMP, is
dominated by Resistor RCOMP, and the output impedance, ZOUT,
is dominated by the impedance of the output capacitor, COUT.
When solving for the crossover frequency, the equation is
simplified to
=|| VL
A
()
1
2
11
1=
××π
×
×
××××
OUT
CCS
COMP
m
OUT
FB
CfRn
RgD
V
V
(29)
where fC is the crossover frequency, RCOMP is the compensation
resistor, and COUT is the output capacitance.
Solving for RCOMP gives
()
m
FB
OUT
CS
OUT
C
COMP gDV
VRnCf
R××
×××××π
=1
2 (30)
Once the compensation resistor, RCOMP, is known, set the zero
formed by the resistor and compensation capacitor, CCOMP, to
one-fourth of the crossover frequency, or
COMPC
COMP Rf
C××π
=2 (31)
Capacitor C2 is chosen to cancel the zero introduced by the output
capacitance ESR. Thus, C2 should be set to (see Figure 31)
COMP
OUT
R
CESR
C
×
=2 (32)
where ESR represents the ESR of COUT.
For low ESR output capacitors, such as ceramic capacitors, C2
is small, generally in the range of 10 pF to 400 pF. Because of the
parasitic inductance, resistance, and capacitance of the PCB layout,
the RCOMP, CCOMP, and C2 values might need to be adjusted by
observing the load transient response of the ADP1621 to establish a
stable operating system and achieve optimal transient performance.
For most applications, RCOMP is in the range of 5 kΩ to 100 kΩ,
and CCOMP is in the range of 100 pF to 30 nF.
COMP
C
COMP
R
COMP
C2
REF g
m
2
3
06090-030
Figure 31. Compensation Components
SLOPE COMPENSATION
The ADP1621 includes a circuit that allows adjustable slope
compensation. Slope compensation is required by current-
mode regulators to stabilize the current-control loop when
operating in continuous conduction and the switching duty
cycle is greater than 50%.
Slope compensation is achieved by internally forcing a ramping
current source out of the CS current-sense pin. By placing a resistor
between the CS pin and the current sensing device (the drain of
the external MOSFET in the case of lossless current sensing or
the source of the MOSFET if a current-sense resistor is used), a
voltage is developed across the resistor that is proportional to
the slope-compensation current.
To ensure stability of the current-mode control loop, use a
compensation voltage slope that is equal to or greater than one-
half of the current-sense representation of the inductor current
downslope. Therefore, it follows that
L
VVV
R
ft
fI
RIND
OUT
CS
SWMINOFF,
SWSC,PK
S
+
×>
×
×
×× 1
2 (33)
where RS is the slope-compensation resistor, ISC,PK is the peak slope-
compensation current, fSW is the switching frequency, RCS is the
current-sense resistor, VOUT is the regulated output voltage, VD is the
forward-voltage drop of the diode, VIN is the input voltage, tOFF,MIN is
the minimum off time, and L is the power-stage inductor. In the
case of lossless current sensing, RCS is equal to the on resistance,
ADP1621
Rev. A | Page 18 of 32
RDSON, of the external power MOSFET. Otherwise, RCS
represents the external current-sense resistor.
Solving for RS gives the slope-compensation criterion:
(
)
(
)
LfI
ftVVVR
R
SWPKSC
SWMINOFF
IND
OUT
CS
S×××
××+×
>
,
,
2
1 (34)
Keep in mind that the above inequality is a function of both
ADP1621 parameters and off-chip components, the values of
which vary from part to part and with temperature. Select RS to
ensure current-loop stability for all possible variations.
After accounting for parameter variations, use values of RS that
are as close to the calculated limit as possible because excessive
slope compensation reduces the benefits of current-mode control
and increases the “softness” of the current limit, as discussed in the
Current Limit section. Given a typical peak slope-compensation
current of 70 μA, RS should not exceed 1.6 kΩ because the voltage
at the CS pin is typically clamped at 116 mV. It is also recom-
mended that RS be greater than 20 Ω. If the calculated RS is greater
than 1.6 kΩ, the parameters in Equation 34, such as RCS, fSW, and L,
can be adjusted such that RS is less than 1.6 kΩ.
In conclusion, the value of RS should be 20 Ω ≤ RS ≤ 1.6 kΩ.
CURRENT LIMIT
The current limit in the ADP1621 limits the peak inductor
current and is achieved by the COMP voltage clamp. The peak
inductor current, IL,PK, is given by
CS
SWMINOFF
SPKSCZCTCOMPCLAMPCOMP
PKL R
ft
DRI
n
VV
I×
××
=,
,,,
,
1 (35)
where VCOMP,CLAMP is the COMP clamp voltage (typically 2.0 V),
VCOMP,ZCT is the COMP zero-current threshold (typically 1.0 V),
n is the current-sense amplifier gain (typically 9.5), ISC,PK is the
peak slope-compensation current (typically 70 μA), RS is the
slope-compensation resistor, D is the duty cycle, fSW is the
switching frequency, tOFF,MIN is the minimum off time (typically
190 ns), and RCS is the current-sense resistor. In the case of
lossless current sensing, RCS is equal to the on resistance, RDSON,
of the external power MOSFET. Otherwise, RCS represents the
external current-sense resistor.
The current limit in the ADP1621 is a “soft” current limit.
When the inductor current reaches the IL,PK limit given in
Equation 35, the duty cycle decreases, and the output voltage
drops below the desired voltage. The IL,PK limit in Equation 35
then increases in response to the smaller duty cycle, D. The
larger the slope-compensation resistor, RS, the larger the effect
on IL,PK for an incremental decrease in D. This behavior results
in a “soft” current limit for the ADP1621. Use values of RS that are
as close as possible to the calculated limit derived from
Equation 34. If high-precision current limiting is required,
consider inserting a fuse in series with the inductor.
Also, keep in mind that the current limit is a function of both
ADP1621 parameters and off-chip components, the values of
which vary from part to part and with temperature. If lossless
current sensing is used, consider that the on resistance of a
MOSFET typically increases with increasing junction temperature.
The peak inductor current limit also limits the maximum load
current at a given output voltage. The maximum load current,
assuming CCM operation, is given by
(
)
×
=
DI MAXLOAD 1
,
××
×
×
××
Lf DV
R
ft
DRI
n
VV
SW
IN
CS
SWMINOFF
SPKSCZCTCOMPCLAMPCOMP
2
1,
,,,
(36)
If the load current exceeds ILOAD,MAX, the output voltage drops
below the desired voltage.
LIGHT LOAD OPERATION
Discontinuous Conduction Mode
With light loads, the average inductor current is small, and,
depending on the converter design, the instantaneous inductor
current may reach 0 during the time when the MOSFET is off.
This mode of operation is termed discontinuous conduction
mode. The condition for entering discontinuous conduction
mode in a boost converter is
(
)
SW
IN
LOAD fL
DDV
I××
×
×
<2
1 (37)
When the instantaneous inductor current reaches 0 during the
cycle, the inductor ceases to be a current source, and ringing
can be observed in the waveforms of the MOSFET drain voltage
and the inductor current. The frequency of the ringing is the
resonant frequency of the inductor and the total capacitance
from the SW node to GND, which includes the capacitances of
the MOSFET and diode, and any parasitic capacitances from
the PCB. While adding a resistive element, such as a snubber, to
the system further dampens the resonance, it also decreases the
efficiency of the regulator.
Pulse-Skipping Modulation
The ADP1621 features circuitry that improves the converter
efficiency and minimizes power consumption with no load or
very light loads. When the COMP voltage drops below VCOMP,ZCT
(typically 1.0 V), which can occur at sufficiently light loads, the
MOSFET is powered off until the FB voltage drops below 1.215 V.
Then, the error amplifier drives the COMP voltage higher, and
the converter resumes switching when the COMP voltage rises
above the VCOMP,ZCT voltage. While the MOSFET is powered off,
the output capacitor supplies current to the load.
With light loads, the COMP voltage hovers around 1.0 V, and
short periods of switching are followed by long periods of the
MOSFET being powered off. This pulse-skipping modulation
operation improves converter efficiency by reducing the number of
switching cycles and therefore reducing the gate drive current and
the switching transition power loss.
ADP1621
Rev. A | Page 19 of 32
Given the minimum on time of the ADP1621, pulse-skipping
modulation is also a requirement to maintain output voltage
regulation with light loads. During the short switching periods
of pulse-skipping modulation, the MOSFET is turned on for the
minimum on time each cycle, storing just enough energy in the
inductor to charge the output capacitor. During the long period
when the MOSFET is off, no current flows through the inductor,
and the light load current is supplied by the output capacitor.
RECOMMENDED COMPONENT MANUFACTURERS
Table 5.
Vendor Components
AVX Corporation Capacitors
Central Semiconductor Corp. Diodes
Coilcraft, Inc. Inductors
Diodes, Inc. Diodes
International Rectifier Diodes, MOSFETs
Murata Manufacturing Co., Ltd. Capacitors, inductors
ON Semiconductor Diodes, MOSFETs
Rubycon Corporation Capacitors
Sanyo Capacitors
Sumida Inductors
Taiyo Yuden, Inc. Capacitors, inductors
Toko America, Inc. Inductors
United Chemi-Con, Inc. Capacitors
Vishay Siliconix Diodes, MOSFETs, resistors, capacitors
ADP1621
Rev. A | Page 20 of 32
LAYOUT CONSIDERATIONS
Layout is important for all switching regulators, but is par-
ticularly important for regulators with high switching frequencies.
To achieve high efficiency, good regulation, and stability, a well-
designed printed circuit board layout is required. A sample PCB
layout for the standard boost converter circuit shown in Figure 33
is given in Figure 32.
Follow these guidelines when designing printed circuit boards:
Keep the low ESR bypass input capacitor of 0.1 μF or higher
close to IN/PIN and GND.
Keep the high current path from Bulk Input Capacitor C1
through Inductor L1 and MOSFET M1 to PGND as short
as possible.
Keep the high current path from Bulk Input Capacitor C1
through Inductor L1, Diode D1, and Output Capacitor COUT
to PGND as short as possible. Place COUT as close to PGND
as possible to reduce ground bouncing.
Keep high current traces as short and wide as possible to
minimize parasitic series inductance, which causes spiking
and electromagnetic interference (EMI).
To minimize switching noise, the drain of the power MOSFET
should be placed very close to the inductor, and the source
of the MOSFET (or the bottom side of the sense resistor)
should be connected directly to the power GND plane. Use
wide copper traces on the drain and on the source of the
MOSFET to minimize parasitic inductance and resistance.
Parasitic inductance can lead to excessive ringing during
switching transitions, and parasitic resistance reduces the
converter efficiency. Make sure that the MOSFET selected
is capable of handling the total power loss (conduction plus
transition losses) in the application circuit.
Avoid routing high impedance traces near any node con-
nected to the switch node (the MOSFET drain) or near
Inductor L1 to prevent radiated switching-noise injection.
Add an extra copper plane at the connection of the MOSFET
drain and the anode of the diode to help dissipate the heat
generated by losses in those components.
Avoid ground loops by having one central ground node on the
PCB. If this is impractical, place the power ground with high
current levels physically closer to the PCB ground terminal.
The analog, low current-level ground should be placed farther
from the PCB ground terminal.
Minimize the length of the PCB trace between the GATE
pin and the MOSFET gate. The parasitic inductance in this
PCB trace can give rise to excessive voltage ringing at the
MOSFET gate and drain, as well as the regulator output. It
is recommended to add 5 Ω of resistance for every inch of
PCB trace. This helps to reduce the overshoot and ringing at
the drain and the output. However, this added resistance
increases the rise and fall times of the MOSFET; thus, the
switching loss in the MOSFET is increased.
Place the feedback resistors as close to FB as possible to
prevent high frequency switching-noise injection.
Place the top of the upper feedback resistor, R1, as close
as possible to the top of COUT for optimum output voltage
sensing.
If a current-sense resistor is connected between the source
of the MOSFET and PGND, ensure that the capacitance from
CS to PGND is minimized.
Place the compensation components as close as possible
to COMP.
V
IN
V
OUT
C
OUT1
C
OUT2
VIAS TO GND PLANE
VIAS TO 2ND LAYER
REMOTE OUTPUT
SENSING
C
OUT3
L1 GND
C1
GND
SDSN ADP1621
C3
C2
R1
R2
C4
R
FREQ
R
COMP
R
S
C
COMP
GATE
D1
M1
0
6090-031
Figure 32. PCB Layout of the Circuit Shown in Figure 33 (2-layer PCB)
ADP1621
Rev. A | Page 21 of 32
EFFICIENCY CONSIDERATIONS
The efficiency, η, of a dc/dc converter is given by
%100×=
IN
OUT
P
P
η (38)
where POUT is the output power, and PIN is the input power to the
converter. While switching regulators are ideally lossless converters
of power, the nonideal characteristics of regulator components
degrade the efficiency of the regulator.
The primary sources of power dissipation in the regulator include
The power dissipation in the external power MOSFET due
to conduction and switching losses.
SWC
MOSFET PPP += (39)
+
+×××
=)1(
1KRD
D
I
DSON
LOAD
×+×
×+
2
)(
1
)( SW
FR
LOAD
D
OUT ftt
D
I
VV
The power dissipation in the external current-sense
resistor if lossless current sensing is not used.
CS
LOAD
CS RD
D
I
P××
=
2
1 (40)
The power dissipation in the external diode.
LOAD
D
DIODE IVP ×= (41)
The power dissipation in the winding resistance of the
power stage inductor.
W
LOAD
WL R
D
I
P×
=
2
,1 (42)
The supply current to the ADP1621 IC, which includes the
quiescent current and the gate driver charging current. The
power dissipation due to gate charging loss is approximated by
SWG
PIN
GfQVP
×
×
=
(43)
where PG is the gate charging power loss, VPIN is the voltage at
the PIN pin, QG is the MOSFET total gate charge, and fSW is
the converter switching frequency. Therefore, the total power
dissipation in the IC itself is given by
(
)
Q
IN
GIC IVPP
×
+
=
(44)
(
)
(
)
Q
IN
SWG
PIN IVfQV ×+×
×
=
where PIC is the total power dissipated in the IC, IQ is the
quiescent current, and VIN is the voltage at the IN pin.
The secondary sources of power dissipation in the regulator include
The power dissipation in the ESR of the input and output
capacitors.
Inductor core losses due to hysteresis and eddy currents.
ADP1621
Rev. A | Page 22 of 32
EXAMPLES OF APPLICATION CIRCUITS
STANDARD BOOST CONVERTER—
DESIGN EXAMPLE
The example covered here is for the ADP1621 configured as a
standard boost converter, as shown in Figure 33, where lossless
current sensing is employed. The design parameters are VIN =
3.3 V, VOUT = 5 V, and a maximum load current of 1 A.
To begin this design, a switching frequency of 600 kHz is chosen
(by setting RFREQ to 32 kΩ, see Figure 30) so that a small inductor
and small output capacitors can be used. The duty cycle is cal-
culated from Equation 1 to be 0.4, given a forward-voltage drop of
0.5 V for the Schottky diode. The feedback resistors are calculated
to be R1 = 35.7 kΩ and R2 = 11.5 kΩ from Equation 4.
Assuming that the inductor ripple is 30% of 1/(1 − D) times
the maximum load current, the inductor size is calculated to be
about 4.4 μH, according to Equation 9. The small, magnetically
shielded 4.7 μH Toko FDV0630-4R7M inductor is selected.
Because ceramic capacitors have very low ESR (a few milliohms),
a 47 μF/6.3 V Murata GRM31CR60J476M ceramic capacitor is
chosen for the input capacitor. The output voltage ripple for a
given COUT, ESR, and ESL can be found by solving Equation 12.
By choosing an output voltage ripple equal to 1% of the output
voltage, Equation 12 yields that the minimum COUT required is
100 μF and the maximum ESR required is 25 mΩ. Other com-
binations of capacitance and ESR are possible by choosing a
much larger COUT and a larger ESR. In this case, a small 1 μF
ceramic capacitor and two 150 μF Sanyo POSCAP™ capacitors
are selected. The low ESR ceramic capacitor helps to suppress
the high frequency overshoot at the output. POSCAP has low
ESR and high capacitance in a relatively small package. Ceramic
capacitors can also be used. Generally, bigger ceramic capacitors
are more expensive.
The next step is to choose a Schottky diode. The average
and rms diode currents are calculated to be 1.0 A and 1.3 A,
respectively, using Equations 14 and 15. A Vishay SSA33L
Schottky diode meets the current and thermal requirements
and is an excellent choice.
The power MOSFET must be chosen based on threshold voltage
(VT), on resistance (RDSON), maximum voltage and current ratings,
and gate charge. The rms current through the MOSFET is given
by Equation 18 as 1.1 A. The Vishay Si7882DP is a 20 V n-channel
power MOSFET that meets the current and thermal requirements.
It comes in a PowerPAK® package and offers low RDSON and gate
charge. At VGS = 2.5 V, the on resistance, RDSON, is 8 mΩ.
The loop-compensation components are chosen to be RCOMP =
9.1 kΩ and CCOMP = 1.7 nF from Equations 30 and 31, respectively.
A roll-off capacitor of C2 = 120 pF is also added. The slope-
compensation resistor is set to be RS = 80 Ω from Equation 34.
Lastly, given the chosen components, the peak inductor current
as set by the current limit circuitry is given by Equation 35 as
IL,PK = 12 A. Thus, the maximum load current, assuming CCM
operation, is given by Equation 36 as ILOAD,MAX = 8 A, which is
safely above the 1.0 A load current requirement for this design
example. Note that the current limit is a strong function of RCS,
which can vary part to part and with temperature. In addition,
note that RCS can be implemented with an external current-
sense resistor or with the RDSON of a MOSFET. Variations in RCS
and the other parameters in Equations 35 and 36 must be taken
into account if precise current limiting is necessary. Due to the
parasitic resistance of PCB traces, RS might need to be adjusted
on the actual circuit board to achieve the desired current limit.
Keep in mind that RS must be less than 1.6 kΩ. Using a MOSFET
with a different RDSON or adjusting RCS can also set the current
limit to the desired level.
ADP1621
IN
GATE
PGND
AGND
FB
SDSN
COMP
FREQ
GND
PIN
CS
R1
35.7k
1%
R2
11.5k
1%
C1
47µF
6.3V
L1
4.7µH
M1
C
COMP
1.8nF
C
OUT1
1µF
10V
f
OSC
= 600kHz
C1 = MURATA GRM31CR60J476M
C
OUT3
= SANYO POSCAP 6TPE150M
L1 = TOKO FDV0630-4R7M
M1 = VISHAY Si7882DP
D1 = VISHAY SSA33L
R
S
80
C
OUT2
10µF
10V
R
FREQ
31.6k
1%
C
OUT3
150µF
6.3V
×2
D1
C3
1µF
10V
C4
0.1µF
10V
C2
120pF
R
COMP
9.09k
V
OUT
= 5V
1A
V
IN
= 3.3
V
06090-032
Figure 33. Typical Boost Converter Application Circuit
ADP1621
Rev. A | Page 23 of 32
BOOTSTRAPPED BOOST CONVERTER
The inputs of the ADP1621 can be driven from the step-up
converter output voltage to improve efficiency for low input
voltages. For low input voltages, bootstrapped operation improves
efficiency with heavy loads by increasing the available gate drive
voltage, thus reducing the on resistance of the MOSFET. However,
because the internal circuitry is driven from IN, the ADP1621
quiescent current and gate drive current supplied from the input
increases due to the step-up ratio and the conversion efficiency loss.
The circuit shown in Figure 1 shows a bootstrapped boost con-
verter, where VIN = 3.3 V and VOUT = 5 V. To ensure that the circuit
starts, make sure that the input voltage minus the forward-voltage
drop of the diode is greater than the UVLO voltage and the gate
threshold voltage of the MOFSET. In this example, the MOSFET
has a gate threshold voltage of 2.5 V. The regulator shown in
Figure 1 is very similar to that shown in Figure 33, which is a
standard boost without bootstrapping. Because the same MOSFET
and inductor are used in both circuits and the input and output
conditions are the same, the compensation components remain
unchanged.
Figure 34 shows a bootstrapped application circuit for output
voltages greater than 5.5 V. In this case, the output is 12 V.
Notice that a resistor, R3, of 700 Ω is placed between VOUT and
the IN and PIN pins to limit the input currents because the IN
and PIN pins are regulated to 5.5 V. A diode, D2, is placed between
VIN and the IN/PIN pins to supply the necessary quiescent current
to start the ADP1621. Once the ADP1621 starts and the output
voltage reaches 12 V, the quiescent current stops flowing
through D2 and is supplied by the output. Keep in mind that the
dynamic supply current to PIN increases as the switching fre-
quency increases because more gate drive is needed for a higher
switching frequency. Therefore, R3 needs to be set appropriately.
The PIN supply current can be approximated by
GSW
PIN QfI
×
=
(45)
where IPIN is the PIN supply current, fSW is the switching frequency,
and QG is the gate charge of a particular MOSFET.
An alternative implementation to Figure 34 is shown in Figure 35,
where an NPN transistor is used to supply the necessary current
to the input PIN at various loads, but the gate drive voltage is
limited to approximately 4.8 V (one diode drop below the
voltage at IN). Signal Diodes D2 and D3 help to provide the
necessary quiescent current to start the ADP1621. Once the
ADP1621 starts, the current stops flowing through these two
diodes because the voltages at PIN and IN are approximately
4.8 V and 5.5 V, respectively. One advantage of this technique
is that Q1 provides enough current to the gate driver at any
switching frequency with a wide range of MOSFETs that have
different gate charge specifications.
Notice that the output capacitor, COUT2 in Figure 34 and Figure 35,
is a large aluminum electrolytic capacitor, both in physical size
and capacitance. Such capacitors are very cheap relative to
ceramic capacitors (such as Sanyo POSCAP) or aluminum
polymer capacitors. The ADP1621 can work with a wide range
of capacitor types.
ADP1621
Rev. A | Page 24 of 32
ADP1621
IN
GATE
PGND
AGND
FB
SDSN
COMP
FREQ
GND
PIN
CS
R1
88.7k
1%
R2
10k
1%
C1
47µF
6.3V
L1
10µH
M1
C
COMP
330pF
C
OUT1
10µF
16V
×2
f
OSC
= 600kHz
C1 = MURATA GRM31CR60J476M
C
OUT2
= RUBYCON 25ZL330M8x16
L1 = COILCRAFT MSS1260-103ML
M1 = IRF7470
D1 = VISHAY SSC53L
D2 = SIGNAL DIODE
R
S
200
R3
700
R
FREQ
31.6k
1%
C
OUT2
330µF
25V
×2
V
OUT
= 12V
1A
D1
D2
C3
1µF
10V
C4
0.1µF
10V
C2
220pF
R
COMP
51.5k
V
IN
= 3.3
V
06090-033
Figure 34. Bootstrapped Application Circuit for VOUT > 5.5 V
ADP1621
IN
GATE
PGND
AGND
FB
SDSN
COMP
FREQ
GND
PIN
CS
R1
88.7k
1%
R2
10k
1%
C1
47µF
6.3V
L1
10µH
M1
C
COMP
330pF
C
OUT1
10µF
16V
×2
f
OSC
= 600kHz
C1 = MURATA GRM31CR60J476M
C
OUT2
= RUBYCON 25ZL330M8x16
L1 = COILCRAFT MSS1260-103ML
Q1 = SIGNAL NPN TRANSISTOR
M1 = IRF7470
D1 = VISHAY SSC53L
D2, D3 = SIGNAL DIODE
R
S
200
R3
1.5k
R
FREQ
31.6k
1%
C
OUT2
330µF
25V
×2
V
OUT
= 12V
1A
D1
D2
Q1
C3
1µF
10V
C4
0.1µF
10V
C2
220pF
R
COMP
51.5k
D3
V
IN
= 3.3
V
06090-034
Figure 35. Bootstrapped Application Circuit for VOUT > 5.5 V
ADP1621
Rev. A | Page 25 of 32
Low Input and High Output Boost Converter
Figure 36 shows a typical application boost converter circuit
that operates at a switching frequency of 200 kHz with VIN = 5 V
and VOUT = 30 V with a 1 A load. The duty cycle for this circuit
is about 83%. A higher switching frequency can be selected, but
the switching power loss in the MOSFET increases and a bigger
MOSFET is needed. For switch-node voltages greater than 30 V,
a sense resistor, RCS, is needed because the absolute maximum
voltage at CS is 33 V.
High Input Voltage Boost Converter Circuit
Input voltages higher than 5.5 V are possible with the addition
of a resistor and an NPN transistor, as shown in Figure 37, or just
with a single resistor, as shown in Figure 38. When there is a
wide input voltage range, it is sometimes desirable to use the
pass NPN transistor, as shown in Figure 37. If the input voltage
range is narrow, a single resistor connecting to the IN and PIN
pins is sufficient, as shown in Figure 38. In Figure 37, Resistor R3
limits the current going into IN, and there is power loss in this
resistor. The voltages at IN and PIN are both clamped to about
5.5 V, which can rise to as high as 5.9 V when the shunt current
is 30 mA. Refer to Figure 9 for the I-V characteristics of the
shunt regulators. Ensure that Resistor R3 is physically large
enough to handle the power dissipation. For switch-node
voltages higher than 30 V, a current-sense resistor is needed and
the CS pin senses the voltage across the sense resistor.
ADP1621
PIN
GATE
CS
AGND
PGND
SDSN
COMP
FREQ
GND
IN
FB
R1
115k
1%
R2
4.87k
1%
C1
47µF
6.3V
×2
L1
7.8µH
M1
C
COMP
20pF
f
OSC
= 200kHz
C1 = MURATA GRM31CR60J476M
C
OUT1
= MURATA GRM31CR72A10
C
OUT2
= MURATA GRM55ER71H475K
C
OUT3
= RUBYCON 50ZL330M10x23
M1 = VISHAY SUD50N06-07L
D1 = IRF 15TQ060
L1 = COILCRAFT DO501DH-782ML
R
S
909
R
CS
3m
R
FREQ
100k
1%
D1
C3
1µF
10V
C4
0.1µF
10V
C2
120pF
R
COMP
1.6M
V
IN
= 5
V
C
OUT1
1µF
100V
C
OUT2
4.7µF
50V
C
OUT3
330µF
50V
×2
V
OUT
= 30V
1A
06090-035
Figure 36. Low Input, High Output Boost Converter
ADP1621
PIN
GATE
CS
AGND
PGND
SDSN
COMP
FREQ
GND
IN
FB
R1
115k
1%
R3
700
R2
4.87k
1%
C1
22µF
16V
×2
L1
8.2µH
M1
CCOMP
220pF
f
OSC = 560kHz
C1 = MURATA GRM32ER61C226K
COUT1 = MURATA GRM31CR72A105K
COUT2 = MURATA GRM55ER71H475K
COUT3 = RUBYCON 50ZL220M10x23
M1 = IRF7470
Q1 = SIGNAL NPN TRANSISTOR
D1 = MBRB7H50
L1 = COILCRAFT MSS1260-822ML
RS
402
RCS
3m
RFREQ
34.8k
D1
Q1
C4
0.1µF
10V
C3
1µF
10V
C2
120pF
RCOMP
2M
V
IN = 8V TO 15
V
COUT1
1µF
100V
COUT2
4.7µF
50V
COUT3
330µF
50V
×2
VOUT = 30V
1A
06090-036
Figure 37. High Input Voltage and High Output Voltage Converter
ADP1621
Rev. A | Page 26 of 32
ADP1621
PIN
GATE
CS
AGND
PGND
SDSN
COMP
FREQ
GND
IN
FB
R1
324k
1%
R2
10.2k
1%
R3
649
C1
22µF
16V
×2
L1
15µH
M1
CCOMP
18pF
f
OSC = 560kHz
C1 = MURATA GRM32ER61C226K
COUT1 = MURATA GRM31CR72A105K
COUT2 = MURATA GRM55ER71H475K
COUT3 = RUBYCON 63ZL220M10x23
M1 = VISHAY Si7478DP
D1 = MBRB7H50
L1 = COILCRAFT MSS1278-153ML
RS
442
RCS
0.01
RFREQ
34.8
D1
C3
1µF
10V
C4
0.1µF
10V
C2
120pF
RCOMP
2M
V
IN = 12
V
COUT1
1µF
100V
COUT2
4.7µF
50V
COUT3
220µF
63V
×2
VOUT = 40V
1A
06090-037
Figure 38. High Input Voltage and High Output Voltage Converter
ADP1621
Rev. A | Page 27 of 32
SEPIC CONVERTER CIRCUIT
A single-ended primary inductance converter (SEPIC) topology
is shown in Figure 39. This topology is useful for an unregulated
input voltage, where the regulated output voltage falls within the
input voltage range.
The input and output are dc-isolated by a coupling capacitor,
C5. L1 and L2 are coupled inductors with a 1:1 turn ratio, which
saves space on the PCB. In steady state, the average voltage across
C5 is the input voltage. When the MOSFET turns on and the
diode turns off, the input voltage provides energy to L1, and C5
provides energy to L2. The output capacitor, COUT, supplies the
load current during this time. When the MOSFET turns off and
the diode turns on, the energy in L1 and L2 is released to charge
the output capacitor, COUT, and the coupling capacitor, C5, as
well as to supply current to the load.
LOW VOLTAGE POWER-INPUT CIRCUIT
The ADP1621 can be configured to run from a low voltage
(as low as 1 V) power input. The power source generally needs
to have a high current capability, such as a fuel cell. Figure 40
illustrates such an application, where the voltage of the power
input is 1 V and the voltage of the chip supply to the IN and
PIN pins is provided by an auxiliary low power source.
ADP1621
IN
GATE
PGND
AGND
FB
SDSN
COMP
FREQ
GND
PIN
CS
R1
17.4k
1%
R2
10k
1%
C1
22µF
10V
×2
C5
10µF
10V
X5R
L2
2.4µH
L1
2.4µH
M1
CCOMP
1.2nF
f
OSC = 325kHz
C1 = MURATA GRM332ER61A226K
COUT2 = SANYO POSCAP 6TPE150MI
C5 = MURATA GRM21BR61A106K
L1, L2 = COUPLED INDUCTORS, 1:1 RATIO, BH ELECTRONICS BH510-1006
M1 = VISHAY Si7882DP
D1 = VISHAY SSC53L
RS
80
COUT1
1µF
10V
RFREQ
65k
COUT2
150µF
6.3V
×3
VOUT = 3.3V
2A
V
IN = 3V TO 5.5
V
D1
C3
1µF
10V
C4
0.1µF
10V
C2
33pF
RCOMP
26k
06090-038
Figure 39. A SEPIC DC/DC Converter
ADP1621
IN
GATE
PGND
AGND
FB
SDSN
COMP
FREQ
GND
PIN
CS
R1
35.7k
1%
R2
11.5k
1%
C1
100µF
X5R
6.3V
L1
2.2µH
M1
C
COMP
56nF
C
OUT1
1µF
10V
f
OSC
= 600kHz
C1 = MURATA GRM32ER60J107ME20
C
OUT2
= MURATA GRM21BR60J106K
C
OUT3
= SANYO POSCAP 6TPE150MI
M1 = VISHAY Si7882DP
D1 = MBRD835L
L1 = TOKO FDV0630-2R2M
R
S
249
C
OUT2
10µF
6.3V
R
FREQ
31.6k
1%
C
OUT3
150µF
6.3V
×2
V
OUT
= 5V
1A
V
IN
= 1
V
V
CC
= 2.9V TO 5.5V
D1
C3
1µF
10V
C4
0.1µF
10V
C2
260pF
R
COMP
9.4k
06090-039
Figure 40. Low Voltage Power-Input Application Circuit
ADP1621
Rev. A | Page 28 of 32
LED DRIVER APPLICATION CIRCUITS
The ADP1621 can be used as an LED driver. Two LED application
circuits are shown in Figure 41 and Figure 42, where each circuit
is driving 20 white LEDs in series. Each white LED has a typical
current of 150 mA at a typical forward voltage of 4.0 V, with a
maximum voltage of 4.5 V over the temperature range of −40°C
to +125°C.
Two methods for dimming the brightness of the LEDs are
shown in Figure 41 and Figure 42. In Figure 41, a PWM signal
is fed to the SDSN pin to turn the ADP1621 controller on and
off. As a result, the LED current is turned on and off, and the
average LED current is dependent on the PWM duty cycle. The
advantage of this method is that no current flows through the
LEDs during the PWM off cycle. In addition, when the ADP1621 is
on, the forward current through the LEDs is constant, which
guarantees constant color emission across the entire dimming
range. Because the soft start period is fixed at 2048 oscillator
cycles, the PWM frequency range is limited.
As shown in Figure 41, because the natural switching frequency
chosen is 400 kHz, the useful PWM frequency range is 90 Hz to
195 Hz. However, when driving fewer LEDs, the ADP1621 can
be set to run at a faster frequency, increasing the maximum PWM
frequency. The PWM duty cycle can be between 5% and 95%. A
higher PWM duty cycle produces a higher average LED current.
Another method for driving the LEDs is shown in Figure 42,
where the PWM signal is filtered by an RC low-pass filter and is
fed to the FB node. The effective FB voltage at the bottom of the
LED string is modulated in an analog manner by the PWM
duty cycle. Thus, the average current through the LEDs is
modulated accordingly. Unlike the case depicted in Figure 41, a
higher duty cycle produces a lower average LED current using
the filtered PWM scheme in Figure 42. The advantage of this
circuit is that the PWM frequency can be in the range between
90 Hz and 100 kHz, and the duty cycle can be between 5% and
95%. The disadvantage of this method is that the forward
current through the LEDs is directly modified to control the
brightness of the LEDs. Because the wavelength of the light
emitted from an LED is a weak function of its forward current,
perfect color purity across the entire dimming range cannot be
guaranteed.
If PCB space is a constraint, smaller inductors can be selected
for the circuits shown in Figure 41 and Figure 42. For example,
a 4.7 μH inductor can be used, and a 200 kHz switching fre-
quency can be selected. However, with this small inductor, the
system operates in DCM, which is slightly less efficient than
operating in CCM.
ADP1621
Rev. A | Page 29 of 32
ADP1621
IN
CS
FB
AGND
PGND
SDSN
COMP
FREQ
GND
PIN
PWM
GATE
R1
8
1/4W
150mA
20
LEDS
R
B
800
C1
2.2µF
25V
L1
33µH
M1
100V
C
COMP
390nF
f
OSC
= 400kHz
C1 = MURATA GRM31MR71E225K
C
OUT
= MURATA GRM31CR72A105K
L1 = COILCRAFT MSS1038-333NL
M1 = VISHAY Si4482DY
D1 = IRF 10MQ100
R
S
800
C
OUT
1µF
100V
×3
R
FREQ
50k
1%
V
OUT
V
IN
= 10V TO 16
V
D1
100V
C3
0.1µF
C4
0.1µF
C2
18pF
R
COMP
101k
R
CS
3m
06090-040
Figure 41. 20-Series LED Driver with PWM at SDSN
ADP1621
IN
CS
FB
AGND
PGND
SDSN
COMP
FREQ
GND
PIN
PWM =
0V TO 4V
GATE
R1
8
1/4W
R5
18k
150mA
20
LEDS
R
B
800
C1
2.2µF
25V
L1
33µH
M1
100V
C5
0.1µF
6.3V
C
COMP
390nF
f
OSC
= 400kHz
C1 = MURATA GRM31MR71E225K
C
OUT
= MURATA GRM31CR72A105K
L1 = COILCRAFT MSS1038-333NL
M1 = VISHAY Si4482DY
D1 = IRF 10MQ100
R
S
800
R2
10k
R3
22.9k
R4
10k
C
OUT
1µF
100V
×3
R
FREQ
50k
1%
V
OUT
V
IN
= 10V TO 16
V
D1
100V
C3
0.1µF
C4
0.1µF
C2
10pF
R
COMP
101k
R
CS
3m
06090-041
Figure 42. 20-Series LED Driver with Filtered PWM
ADP1621
Rev. A | Page 30 of 32
RELATED PARTS
Table 6.
Part Number Description Comments
ADP1610 Current-mode PWM step-up controller Maximum output = 12 V; PWM frequency = 700 kHz or 1.2 MHz;
integrated 1.2 A, 0.2 Ω MOSFET power switch
ADP1611 Current-mode PWM step-up controller Maximum output = 20 V; PWM frequency = 700 kHz or 1.2 MHz;
integrated 1.2 A, 0.2 Ω MOSFET power switch
ADP1621
Rev. A | Page 31 of 32
OUTLINE DIMENSIONS
COMPLIANT TO JEDEC STANDARDS MO-187-BA
0.23
0.08
0.80
0.60
0.40
0.15
0.05
0.33
0.17
0.95
0.85
0.75
SEATING
PLANE
1.10 MAX
10 6
5
1
0.50 BSC
PIN 1
COPLANARITY
0.10
3.10
3.00
2.90
3.10
3.00
2.90
5.15
4.90
4.65
Figure 43. 10-Lead Mini Small Outline Package [MSOP]
(RM-10)
Dimensions shown in millimeters
ORDERING GUIDE
Model Temperature Range Package Description
Package
Option
Ordering
Quantity Branding
ADP1621ARMZ1−40°C to +125°C 10-Lead Mini Small Outline Package [MSOP] RM-10 50 L3M
ADP1621ARMZ-R71−40°C to +125°C 10-Lead Mini Small Outline Package [MSOP] RM-10 1,000 L3M
ADP1621-EVAL Evaluation Board 1
1 Z = Pb-free part.
ADP1621
Rev. A | Page 32 of 32
NOTES
©2006 Analog Devices, Inc. All rights reserved. Trademarks and
registered trademarks are the property of their respective owners.
D06090-0-12/06(A)