LINER LTC3811EUHF

LTC3811
High Speed Dual,
Multiphase Step-Down DC/DC
Controller
DESCRIPTION
FEATURES
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
Fixed Frequency, Peak Current Mode Control
±0.5% Output Accuracy Over Temperature
Optimized for Low VOUT Applications (Up to 3.3V)
Dual or Single Output, Multiphase Operation
Wide VIN Range: 4.5V to 30V Operation
High Speed Differential Remote Sense Amplifier
Inductor DCR or Sense Resistor Capable
Adjustable Peak Current Sense Voltage:
24mV to 85mV
Very Low Duty Cycle Operation: tON(MIN) = 65ns (Typ)
Powerful Internal Gate Drivers
Output Voltage Soft-Start, Tracking and Sequencing
Programmable Load Line for Reduced COUT
Clock Input and Output for Up to 12-Phase Operation
Fixed Frequency Operation from 250kHz to 750kHz
PLL Synchronization from 150kHz Up to 900kHz
Selectable CCM or DCM Operation
Available in 5mm × 7mm QFN and G36 Packages
APPLICATIONS
■
■
■
Network Servers
High Current ASIC Supplies
Low Voltage Power Distribution
The LTC®3811 is a dual, PolyPhase® synchronous stepdown switching regulator controller optimized for output
voltages up to 3.3V. The LTC3811 includes high bandwidth
error amplifiers as well as a high speed differential remote
sense amplifier. The sense voltage range is programmable
from 24mV to 85mV, allowing the use of either the inductor
DCR or a discrete sense resistor. Multiphase operation is
made possible using the MODE/SYNC input, the CLKOUT
output and the PHASEMODE control pin, allowing 1-, 2-,
3-, 4-, 6- or 12-phase operation.
Large internal gate drivers minimize switching losses and
allow the use of multiple power MOSFETs connected in
parallel for high current applications.
The operating frequency of the LTC3811 can be programmed
from 250kHz to 750kHz and can also be synchronized to
an external clock using the internal PLL.
Tracking and sequencing are possible with the LTC3811,
and soft-start is programmed with an external capacitor.
Shutdown reduces supply current to 20μA.
, LT, LTC, LTM and PolyPhase are registered trademarks of Linear Technology
Corporation. All other trademarks are the property of their respective owners.
Protected by U.S. Patents, including 5481178, 6304066, 5929620, 6177787, 6144194,
6580258, 5705919.
TYPICAL APPLICATION
Dual Output, 2-Phase Tracking Core and I/O Supply
VIN
4.5V TO 14V
VIN
INTVCC
TG1
SW2
BG1
BG2
PGND
RUN1
LTC3811
SENSE1+
SENSE2+
SENSE1–
SENSE2–
DIFF/IN+
DIFF/IN–
DIFF/OUT
FB2
RNG1
RNG2
FB1
VOUT
100mV/DIV
BOOST2
SW1
VOUT1
2V
15A
DRVCC
TG2
BOOST1
RUN2
Load Step
COMP1
COMP2
SS/TRACK1 SGND SS/TRACK2
IL
5A/DIV
VOUT2
1.5V
15A
VIN = 12V
VOUT = 2V
IOUT = 0A to 12.5A
20μs/DIV
3811 TA01b
3811 TA01a
3811f
1
LTC3811
ABSOLUTE MAXIMUM RATINGS
(Note 1)
Input Supply Voltage (VIN) ......................... –0.3V to 30V
Topside Driver Voltages
(BOOST1, BOOST2) ............................... –0.3V to 37V
Switch Voltage (SW1, SW2) ......................... –5V to 30V
BOOST1 – SW1, BOOST2 – SW2 ................. –0.3V to 7V
DRVCC, INTVCC, EXTVCC, RUN1, RUN2, DIFF/IN+,
DIFF/IN–, PHASEMODE, PGOOD1, PGOOD2,
MODE/SYNC Voltages .................................. –0.3V to 7V
FB1, FB2, RNG1, RNG2, SS/TRACK1, SS/TRACK2,
PLL/LPF, SENSE1+, SENSE1–, SENSE2+,
SENSE2– Voltages................................. –0.3V to INTVCC
DRVCC LDO RMS Output Current .........................100mA
OperatingTemperature Range (Note 2)..... –40°C to 85°C
Junction Temperature (Note 3) ............................. 125°C
Storage Temperature Range................... –65°C to 125°C
Lead Temperature (Soldering, 10 sec)
SSOP Package .................................................. 300°C
TOP VIEW
1
36 PGOOD2
RUN1
2
35 PLL/LPF
38 37 36 35 34 33 32
RUN2
3
34 MODE/SYNC
31 BOOST1
SENSE1+
4
33 BOOST1
30 TG1
SENSE1–
5
32 TG1
SS/TRACK1 3
29 SW1
SS/TRACK1
6
31 SW1
COMP1 4
28 BG1
COMP1
7
30 BG1
FB1
8
29 EXTVCC
SGND
9
28 PGND
RUN1
PGOOD1
RUN2
CLKOUT
PLL/LPF
PGOOD2
PGOOD1
TOP VIEW
MODE/SYNC
PIN CONFIGURATION
SENSE1+ 1
SENSE1– 2
FB1 5
27 EXTVCC
SGND 6
26 PGND
39
INTVCC 7
25 DRVCC
FB2 8
INTVCC 10
24 VIN
FB2 11
23 BG2
COMP2 9
22 SW2
SS/TRACK2 10
–
21 TG2
SENSE2 11
+
20 BOOST2
SENSE2 12
PHASEMODE
RNG2
RNG1
DIFF/IN+
DIFF/IN–
CSOUT
DIFF/OUT
13 14 15 16 17 18 19
UHF PACKAGE
38-LEAD (5mm × 7mm) PLASTIC QFN
27 DRVCC
26 VIN
COMP2 12
25 BG2
SS/TRACK2 13
24 SW2
SENSE2– 14
23 TG2
SENSE2+ 15
22 BOOST2
DIFF/OUT 16
21 CLKOUT
DIFF/IN– 17
20 RNG2
DIFF/IN+ 18
19 RNG1
G PACKAGE
36-LEAD PLASTIC SSOP WIDE
TJMAX = 125°C, θJA = 34°C/W
EXPOSED PAD (PIN 39) IS PGND, MUST BE SOLDERED TO PCB
TJMAX = 125°C, θJA = 100°C/W
ORDER INFORMATION
LEAD FREE FINISH
TAPE AND REEL
PART MARKING
PACKAGE DESCRIPTION
TEMPERATURE RANGE
LTC3811EUHF#PBF
LTC3811EUHF#TRPBF
3811
38-Lead (5mm × 7mm) Plastic QFN
–40°C to 85°C
LTC3811EG#PBF
LTC3811EG#TRPBF
LTC3811EG
36-Lead Plastic SSOP Wide
–40°C to 85°C
Consult LTC Marketing for parts specified with wider operating temperature ranges.
Consult LTC Marketing for information on non-standard lead based finish parts.
For more information on lead free part marking, go to: http://www.linear.com/leadfree/
For more information on tape and reel specifications, go to: http://www.linear.com/tapeandreel/
3811f
2
LTC3811
ELECTRICAL CHARACTERISTICS
The ● denotes the specifications which apply over the full operating
junction temperature range, otherwise specifications are at TJ = 25°C. VIN = 12V, MODE/SYNC = 0V, unless otherwise specified.
SYMBOL
PARAMETER
CONDITIONS
MIN
TYP
MAX
UNITS
Input Supply
VIN
Operating Input Voltage Range
IQ
Total Quiescent Supply Current
Continuous Mode (Note 4)
Shutdown Mode
4.5
VFB1,2 = 0.7V
VRUN1,2 = 0V
●
●
10.5
20
30
V
20
40
mA
μA
1
μA
0.3
V
RUN Pin ON/OFF Control
IRUN
RUN Pin Input Leakage
VIL(RUN)
Low Level RUN Input Threshold
VIH(RUN)
High Level RUN Input Threshold
VRUN1,2 = 3.3V
–1
1.8
V
Error Amplifier Characteristics (Both Channels)
VFB1, VFB2
Feedback Voltage Accuracy
(Note 5)
●
IFB1, IFB2
Feedback Pin Input Current
VCOMP = 1.25V (Note 5)
ΔVFB/ΔVIN
Line Regulation
4.5V ≤ VIN ≤ 30V (Note 5)
●
●
598
597
600
602
603
100
nA
0.002
0.02
%/V
–100
–0.1
–0.01
mV
mV
ΔVFB/ΔVCOMP
Load Regulation
ΔVCOMP = 1.25V to 1.5V (Note 5)
f0dB(EA)
Error Amplifier Unity Gain Crossover
Frequency
(Note 6)
%
VOH(EA)
Error Amplifier Maximum Output
Voltage (Internally Clamped)
VFB = 0.54V, No Load
2.6
V
VOL(EA)
Error Amplifier Minimum Output Voltage VFB = 0.66V, No Load
10
mV
VFB(OFF)
FB Voltage Threshold to Disable Error
Amplifier Output
VINTVCC – VFB
0.3
V
ISS1, ISS2
SS/TRACK1, SS/TRACK2 Charging
Currents
VSS/TRACK1 = VSS/TRACK2 = 0.3V
–2.5
μA
RSS1, RSS2
SS/TRACK1, SS/TRACK2 Pull-Down
Resistance in Shutdown
VRUN1 = VRUN2 = 0V
1
kΩ
8
MHz
Soft-Start/Tracking
Differential Amplifier
AV
Differential Mode Gain,
ΔVDIFF/OUT/ΔVDIFF/IN
ΔVDIFF/IN = 1V to 3.5V, IDIFF/OUT = –100μA
VOS(DIFF)
Output Offset Voltage,
VDIFF/OUT – VDIFF/IN+
VDIFF/IN+ = 1.25V, VDIFF/IN– = 0V,
IDIFF/OUT = –100μA
0.995
1.000
–6
RIN
Input Resistance
Measured at VDIFF/IN+
160
PSRRDIFF
Power Supply Rejection Ratio
7V ≤ VIN ≤ 30V
100
VDM(DIFF)
Maximum Differential Mode Input
Voltage
VDIFF/IN+ – VDIFF/IN–, Measured at
VDIFF/OUT, IDIFF/OUT = –100μA
IMAX+
Maximum Sink Current
1.005
V/V
6
mV
kΩ
dB
5.5
2
IMAX–
Maximum Source Current
f0dB(DIFF)
Unity Gain Bandwidth
(Note 6)
Maximum Current Sense Threshold
(VSENSE+ – VSENSE–)
VFB = 0.575V, VCM = 1.25V
VRNG = 0V
VRNG = INTVCC
VRNG = 2V
Minimum Current Sense Threshold
(VSENSE+ – VSENSE–)
VFB = 0.625V, VCM = 1.25V
VRNG = 0V
VRNG = INTVCC
VRNG = 2V
V
mA
–2
8
mA
MHz
Current Comparators
VSENSE(MAX)
VSENSE(MIN)
14
32.5
60
24
50
85
–21
–41
–67
34
67.5
110
mV
mV
mV
mV
mV
mV
3811f
3
LTC3811
ELECTRICAL CHARACTERISTICS
The ● denotes the specifications which apply over the full operating
junction temperature range, otherwise specifications are at TJ = 25°C. VIN = 12V, MODE/SYNC = 0V, unless otherwise specified.
SYMBOL
PARAMETER
CONDITIONS
ISENSE
Total Sense Pin Current
(VSENSE+ + VSENSE–)
VCM = 1.25V
VCM(CS)
VSENSE+, VSENSE– Pin Common Mode
Input Voltage Range
MIN
TYP
MAX
–1.5
0
UNITS
μA
3.5
V
Voltage Position Amplifier (QFN Package Only)
gm
Voltage Position Transconductance,
ΔICSOUT/ΔVSENSE (Note 8)
VSENSE1+ – VSENSE1– = VSENSE2+
– VSENSE2– = ±50mV, VCM = 1.25V,
VCSOUT = 1.25V
IOS(VP)
Output Offset Current, Measured at
CSOUT
VSENSE1+ – VSENSE1– = VSENSE2+
– VSENSE2– = 1.25V, VCSOUT = 1.25V
–40
Nominal Frequency
VPLL/LPF Pin Floating, MODE/SYNC = DC
Voltage
450
fLOW
Lowest Frequency
VPLL/LPF = 0V, MODE/SYNC = DC Voltage
200
250
300
kHz
fHIGH
Highest Frequency
VPLL/LPF = INTVCC, MODE/SYNC = DC
Voltage
650
750
850
kHz
fSYNC(MIN)
Minimum Synchronizable Frequency
MODE/SYNC = External Clock
125
175
kHz
fSYNC(MAX)
Maximum Synchronizable Frequency
MODE/SYNC = External Clock
IPLL/LPF
Phase Detector Output Current
Sinking
Sourcing
5.0
mS
40
μA
550
kHz
Multiphase Oscillator and Phase-Lock Loop (Note 9)
fNOM
900
500
1000
kHz
fMODE/SYNC < fOSC
fMODE/SYNC > fOSC
–4.3
5.1
μA
μA
θ1 – θ2
Channel 1 to Channel 2 Phase
Relationship (Note 9)
VPHASMODE = 0V
VPHASMODE = 50% INTVCC
VPHASMODE = INTVCC
180
180
120
deg
deg
deg
θ1 – θCLKOUT
Channel 1 to CLKOUT Phase
Relationship (Note 9)
VPHASMODE = 0V
VPHASMODE = 50% INTVCC
VPHASMODE = INTVCC
90
60
240
deg
deg
deg
VOL(CLKOUT)
Low Level CLKOUT Output Voltage
RCLK = 50k to Ground
VOH(CLKOUT)
High Level CLKOUT Output Voltage
RCLK = 50k to Ground
RMODE/SYNC
MODE/SYNC Input Resistance
VIL(MODE/SYNC)
Low Level MODE/SYNC Input Threshold
VIH(MODE/SYNC)
High Level MODE/SYNC Input Threshold
0.2
4.0
V
5.8
V
75
kΩ
0.3
1.8
V
V
Power Good Indicators
VOL(PGOOD)
PGOOD Voltage Low
IPGOOD(OFF)
PGOOD Leakage Current
VPGOOD = 6V
ΔVFB(OV)
ΔVFB, PGOOD Overvoltage Threshold
VFB(OV) – VFB(NOM) in Percent
7
10
ΔVFB(UV)
ΔVFB, PGOOD Undervoltage Threshold
VFB(UV) – VFB(NOM) in Percent
–13
–10
–7
ΔVFB(HYST)
ΔVFB, PGOOD Comparator Hysteresis
UV or OV Comparator
tPG(FAULT)
tPG(OK)
IPGOOD = 2mA
0.12
0.30
V
1.0
μA
13
%
%
12
mV
Delay from UV/OV Condition to PGOOD
Falling
145
μs
Delay from UV/OV Fault Recovery to
PGOOD Rising
38
μs
Thermal Protection
TJSD
Thermal Shutdown Junction
Temperature
(Note 6)
165
°C
TJSD(HYST)
Thermal Shutdown Junction
Temperature Hysteresis
(Note 6)
25
°C
3811f
4
LTC3811
ELECTRICAL CHARACTERISTICS
The ● denotes the specifications which apply over the full operating
junction temperature range, otherwise specifications are at TJ = 25°C. VIN = 12V, MODE/SYNC = 0V, unless otherwise specified.
SYMBOL
PARAMETER
CONDITIONS
MIN
TYP
MAX
5.6
6.0
6.4
–2.0
–0.5
UNITS
DRVCC Linear Regulator
●
VDRVCC
LDO Regulator Output Voltage
VEXTVCC = 0V
ΔVDRVCC(LOAD)
DRVCC Load Regulation
ILOAD = 0mA to 50mA
ΔVDRVCC(LINE)
DRVCC Line Regulation
ΔVIN = 8.5V to 30V
0.01
VDRVCC(UVLO)
LDO Regulator Undervoltage Threshold
DRVCC Rising
3.7
VDRVCC(HYST)
LDO Regulator Undervoltage Hysteresis
VEXTVCC
EXTVCC Switchover Voltage
VEXTVCC(HYST)
EXTVCC Switchover Hysteresis
VEXTVCC(DROP)
EXTVCC Voltage Drop
tr (TG1, TG2)
Top Gate Rise Time
tf (TG1, TG2)
tr (BG1, BG2)
V
%
0.2
%/V
V
0.56
V
4.5
V
IDRVCC = 20mA
400
mV
IDRVCC = 20mA, VEXTVCC = 5V
100
mV
CL = 3300pF (Note 6)
20
ns
Top Gate Fall Time
CL = 3300pF (Note 6)
10
ns
Bottom Gate Rise Time
CL = 3300pF (Note 6)
20
ns
IDRVCC = 20mA, EXTVCC Rising
Gate Drivers
tf (BG1, BG2)
Bottom Gate Fall Time
CL = 3300pF (Note 6)
10
ns
RDS(ON)(TG)
TG1, TG2
Top Gate Pull-Down NMOS
On-Resistance
TG to SW
0.9
Ω
RDS(ON)(BG)
BG1, BG2
Bottom Gate Pull-Down NMOS
On-Resistance
BG to PGND
0.9
Ω
IPK(TG) TG1, TG2
Top Gate (TG) Peak Source Current
1.0
A
IPK(BG) BG1, BG2
Bottom Gate (BG) Peak Source Current
1.0
A
tDEAD1
Bottom Gate Off to Top Gate On
Deadtime
(Note 6)
30
ns
tDEAD2
Top Gate Off to Bottom Gate On
Deadtime
(Note 6)
30
ns
tON(MIN)
Minimum On-Time
VCOMP = 1.25V (Note 6, 7)
65
ns
Note 1: Stresses beyond those listed under Absolute Maximum Ratings
may cause permanent damage to the device. Exposure to any Absolute
Maximum Rating condition for extended periods may affect device
reliability and lifetime. Unless otherwise specified, all voltages are relative
to SGND and all currents are positive into a pin.
Note 2: The LTC3811E is guaranteed to meet performance specifications
from 0°C to 85°C temperature. Specifications over the –40°C to
85°C operating junction temperature range are assured by design,
characterization and correlation with statistical process controls.
Note 3: TJ is calculated from the ambient temperature TA and power
dissipation PD according to the following formula:
TJ = TA + (PD • TBD°C/W)
Note 4: The dynamic input supply current is higher due to power
MOSFET gate charging (QG • fOSC). See Applications Information for more
information.
Note 5: The error amplifiers are measured in a feedback loop using an
external servo operational amplifier that drives the VFB pin and regulates
VCOMP to be equal to the external control voltage.
Note 6: Guaranteed by design, not subject to test.
Note 7: The minimum on-time condition corresponds to an inductor peakto-peak ripple current of 50% of IMAX. See Applications Information for
more details.
Note 8: The voltage positioning amplifier operates as a transconductance
amplifier, where the input voltages are the SENSE+ to SENSE– potentials
for both channels. The amplifier output current flows through an external
resistor in order to program the amount of voltage droop at full load.
Note 9: The PHASEMODE function is only available in the QFN package.
The 36-lead GW package has a fixed channel 1-to-channel 2 phase
relationship of 180°C and a channel 1-to-CLKOUT phase relationship of
90°C. The version in the 36-lead GW package is therefore optimized for
2- and 4-phase operation.
Note 10: Rise and fall times are measured at 10% and 90% levels.
3811f
5
LTC3811
TYPICAL PERFORMANCE CHARACTERISTICS
Load Step (Pulse Skip Mode)
Load Step (Forced Continuous
with Voltage Positioning)
Load Step (Forced Continuous)
VOUT
50mV/DIV
AC COUPLED
VOUT
50mV/DIV
AC COUPLED
IL
5A/DIV
IL
5A/DIV
VOUT
50mV/DIV
AC COUPLED
IL1
5A/DIV
IL2
5A/DIV
3811 G01
VIN = 12V
20μs/DIV
VOUT = 1.5V
ILOAD = 0.5A TO 8A
SW
10V/DIV
IL
2A/DIV
250ns/DIV
SW
5V/DIV
VOUT
20mV/DIV
AC COUPLED
VOUT
20mV/DIV
AC COUPLED
IL
5A/DIV
IL
5A/DIV
3811 G04
VIN = 12V
VOUT = 1.5V
ILOAD = 300mA
VIN = 12V
VOUT = 2V
95
VIN = 12V
VOUT = 1.5V
ILOAD = 300mA
EFFICIENCY (%)
80
75
70
3811 G06
IL
5A/DIV
90
85
1μs/DIV
Short-Circut Waveforms
VIN = 12V
VOUT = 2V
95
90
EFFICIENCY (%)
3811 G05
1μs/DIV
Efficiency vs Load Current with
DCR Sensing
100
3811 G03
Light Load Waveforms
(Pulse Skip Mode)
SW
5V/DIV
Efficiency vs Load Current with
RSENSE = 1.5mΩ
100
VIN = 12V
50μs/DIV
VOUT = 1.5V (2-PHASE)
ILOAD = 0A TO 15A
Light Load Waveforms
(Forced Continuous Mode)
Low Duty Cycle Waveforms
VIN = 20V
VOUT = 1.2V
ILOAD = 1.5A
3811 G02
20μs/DIV
VIN = 12V
VOUT = 1.5V
ILOAD = 0.5A TO 8A
85
PGOOD
5V/DIV
80
VOUT
1V/DIV
75
70
65
DISCONTINUOUS
FORCED CONTINUOUS
60
1
10
LOAD CURRENT (A)
100
3811 G07
65
DISCONTINUOUS
FORCED CONTINUOUS
VIN = 12V
VOUT = 1.5V
200μs/DIV
3811 G09
60
1
10
LOAD CURRENT (A)
100
3811 G08
3811f
6
LTC3811
TYPICAL PERFORMANCE CHARACTERISTICS
FB Voltage vs Temperature
600.5
1.010
0.6025
600.4
1.008
DIFFERENTIAL AMPLIFIER GAIN (V/V)
0.6030
0.6020
600.3
0.6015
0.6010
FB VOLTAGE (mV)
0.6005
0.6000
0.5995
0.5990
0.5985
600.2
600.1
600.0
599.9
599.8
599.7
0.5980
0.5975
599.6
0.5970
–50 –30 –10 10 30 50 70 90 110 130
TEMPERATURE (°C)
599.5
0
4
8
12 16 20 24 28
SUPPLY VOLTAGE (V)
3811 G10
VPLL/LPF = INTVCC
VPLL/LPF = INTVCC
OSCILLATOR FREQUECY (kHz)
700
600
VPLL/LPF = FLOATING
500
400
VPLL/LPF = 0V
300
1.000
0.998
0.996
0.994
0.992
0.990
–50 –30 –10 10 30 50 70 90 110 130
TEMPERATURE (°C)
3811 G12
Shutdown Mode Supply Current
vs Supply Voltage
800
700
1.002
Oscillator Frequency vs
Supply Voltage
800
FREQUENCY (kHz)
36
1.004
3811 G11
Oscillator Frequency vs
Temperature
200
100
600
VPLL/LPF = FLOATING
500
400
300
VPLL/LPF = 0V
200
100
0
–50 –30 –10 10 30 50 70 90 110 130 150
TEMPERATURE (°C)
0
40
5
10 15 20 25 30
SUPPLY VOLTAGE (V)
35
25°C
30
20
15
10
5
40
3.0
14
2.9
2.9
25°C
85°C
10
9
8
7
2.8
2.7
2.6
2.5
2.4
2.3
2.2
6
2.1
5
2.0
0
4
8
12 16 20 24 28
SUPPLY VOLTAGE (V)
32
36
3811 G16
SS/TRACK PIN CURRENT (μA)
SS/TRACK PIN CURRENT (μA)
3.0
11
4
8
12 16 20 24 28
SUPPLY VOLTAGE (V)
32
36
SS/TRACK Pin Current vs
Temperature
15
12
0
3811 G15
SS/TRACK Pin Current vs
Common Mode Voltage
–40°C
–40°C
25
3811 G14
Supply Current vs Supply Voltage
13
85°C
35
0
0
3811 G13
SUPPLY CURRENT (mA)
32
1.006
SHUTDOWN MODE SUPPLY CURRENT (μA)
FB VOLTAGE (mV)
Differential Amplifier Gain vs
Temperature
FB Voltage Line Regulation
2.8
2.7
2.6
2.5
2.4
2.3
2.2
2.1
0
100 200 300 400 500 600
COMMON MODE VOLTAGE (mV)
700
3811 G17
2.0
–50 –30 –10 10 30 50 70 90 110 130
TEMPERATURE (°C)
3811 G18
3811f
7
LTC3811
TYPICAL PERFORMANCE CHARACTERISTICS
DRVCC Line Regulation
DRVCC Load Regulation
6.010
DRVCC vs Temperature
6.5
6.1
6.008
6.4
6.0
6.002
6.000
5.998
5.996
6.3
5.9
6.2
DRVCC (V)
6.004
DRVCC VOLTAGE (V)
DRVCC VOLTAGE (V)
6.006
5.8
5.7
5.990
0
4
8
12 16 20 24 28
SUPPLY VOLTAGE (V)
32
5.6
5.5
–50 –30 –10 10 30 50 70 90 110 130
TEMPERATURE (°C)
5.4
36
0
50
100 150 200 250 300
DRVCC LOAD CURRENT (mA)
3811 G19
350
3811 G21
3811 G20
DRVCC Undervoltage Lockout
Thresholds vs Temperature
EXTVCC Thresholds vs
Temperature
LDO Dropout Voltage vs Current
5.0
700
4.5
5.0
TA = 25°C
600
4.0
DROPOUT VOLTAGE (mV)
DRVCC RISING
3.5
3.0
DRVCC FALLING
2.5
2.0
1.5
EXTVCC THRESHOLDS (V)
UNDERVOLTAGE LOCKOUT THRESHOLDS (V)
5.9
5.7
5.5
5.992
500
400
300
200
1.0
EXTVCC RISING
4.5
4.0
EXTVCC FALLING
3.5
100
0.5
0
–50 –30 –10 10 30 50 70 90 110 130
TEMPERATURE (°C)
0
0 10 20 30
3811 G23
Current Sense Amplifier gm
vs Temperature
2.5
RUN PIN THRESHOLD (V)
400
350
300
250
200
150
100
7.0
CURRENT SENSE AMP gm (mS)
TA = 25°C
450
3811 G24
RUN Pin Threshold vs
Supply Voltage
EXTVCC Switch IR Drop vs Current
500
3.0
–50 –30 –10 10 30 50 70 90 110 130
TEMPERATURE (°C)
40 50 60 70 80 90 100
CURRENT (mA)
3811 G22
EXTVCC SWITCH IR DROP (mV)
6.0
5.8
5.6
5.994
2.0
1.5
25°C
1.0
–40°C
85°C
0.5
6.5
6.0
5.5
5.0
4.5
4.0
3.5
50
0
6.1
0
10 20 30 40 50 60 70 80 90 100
EXTVCC CURRENT (mA)
3811 G25
0
0
4
8
12 16 20 24 28
SUPPLY VOLTAGE (V)
32
36
3811 G26
3.0
–50 –30 –10 10 30 50 70 90 110 130
TEMPERATURE (°C)
3811 G27
3811f
8
LTC3811
TYPICAL PERFORMANCE CHARACTERISTICS
Maximum SENSE Pin Voltage vs
Temperature
100
RNG = 2V
80
60
MAXIMUM SENSE PIN VOLTAGE (mV)
RNG = INTVCC
40
RNG = 0V
20
0
–20
–40
–60
0.5
1
1.5
2
COMP PIN VOLTAGE (V)
3
2.5
80
RNG = 2V
70
60
RNG = INTVCC
50
40
30
RNG = 0V
20
10
0
–50 –30 –10 10 30 50 70 90 110 130
TEMPERATURE (°C)
–80
0
90
50
45
40
35
30
25
20
15
RNG = INTVCC
5
200 300 400 500
FB VOLTAGE (mV)
50
40
30
20
10
0
0.5
600
50
40
30
20
0
1.1 1.3 1.5
VRNG VOLTAGE (V)
1.7
1.5
1.0
0
10 20 30 40 50 60 70 80 90 100
DUTY CYCLE (%)
0
0.5
1.0 1.5 2.0 2.5 3.0 3.5 4.0
COMMON MODE VOLTAGE (V)
3811 G33
Maximum SENSE Pin Threshold
vs Common Mode Voltage
MAXIMUM SENSE PIN THRESHOLD (mV)
1.9
2.0
3811 G32
SENSE Pin Current vs
Temperature
3811 G34
0.9
0.5
10
0
700
1.5
1.4
1.3
1.2
1.1
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
–50 –30 –10 10 30 50 70 90 110 130
TEMPERATURE (°C)
0.7
2.5
3811 G31
SENSE PIN CURRENT (μA)
60
SENSE Pin Current vs
Common Mode Voltage
SENSE PIN CURRENT (μA)
55
100
70
3811 G30
60
MAXIMUM SENSE PIN THRESHOLD (mV)
MAXIMUM SENSE THRESHOLD (mV)
60
0
80
Maximum SENSE Pin Threshold
vs Duty Cycle
Foldback Current Limit
10
90
3811 G29
3811 G28
Frequency vs PLL/LPF Pin
Voltage
55
1200
54
1100
1000
53
900
52
51
FREQUENCY (kHz)
CURRENT SENSE THRESHOLD (mV)
100
0
Maximum Current Sense
Threshold vs VRNG Voltage
MAXIMUM CURRENT SENSE THRESHOLD (mV)
Current Sense Threshold vs
COMP Pin Voltage
RNG = INTVCC
50
49
48
800
700
600
500
400
300
47
200
46
100
45
0
0.5
1.5 2
1
2.5 3 3.5
COMMON MODE VOLTAGE (V)
0
4
3811 G35
0
0.5
1.5
2
1
PLL/LPF PIN VOLTAGE (V)
2.5
3811 G36
3811f
9
LTC3811
TYPICAL PERFORMANCE CHARACTERISTICS
Top Gate Turn-On Waveform
Driving Renesas RJK0305DPB
Top Gate Turn-Off Waveform
Driving Renesas RJK0305DPB
TG – SW
Bottom Gate Turn-On Waveform
Driving Renesas RJK0301DPB
TG – SW
2V/DIV
2V/DIV
2V/DIV
VIN = 12V
VOUT = 1.5V
IOUT = 5A
3811 G37
10ns/DIV
VIN = 12V
VOUT = 1.5V
IOUT = 5A
Bottom Gate Turn-Off Waveform
Driving Renesas RJK0301DPB
3811 G38
10ns/DIV
Single Output, 2-Phase Current
Sharing Waveforms
IL
5A/DIV
2V/DIV
3811 G40
Start-Up Into a Pre-Biased
Output Capacitor
OV
VIN = 12V
VOUT = 1.5V
IOUT = 16A
3811 G41
1μs/DIV
VOUT = 3.3V
Ratiometric Tracking Start-Up
Sequenced Start-Up
3811 G39
1V
PRE BIAS
SW2
10V/DIV
25ns/DIV
25ns/DIV
1V/DIV
SW1
10V/DIV
VIN = 12V
VOUT = 1.5V
IOUT = 5A
VIN = 12V
VOUT = 1.5V
IOUT = 5A
RUN1
2V/DIV
500μs/DIV
3811 G42
Coincident Tracking Start-Up
VOUT1 = 2V
VOUT2 = 1.5V
RUN2
5V/DIV
500mV/DIV
500mV/DIV
VOUT1 = 2V
VOUT2 = 1.5V
500mV/DIV
VIN = 12V
10ms/DIV
RLOAD1 = RLOAD2 = 1.5Ω
RUN2 CONNECTED TO PGOOD1
3811 G43
VIN = 12V
5ms/DIV
RLOAD1 = RLOAD2 = 1.5Ω
3811 G44
5ms/DIV
3811 G45
3811f
10
LTC3811
PIN FUNCTIONS
BG1, BG2: High Current Gate Driver Outputs for the
N-Channel Lower Power MOSFETs.
BOOST1, BOOST2: Bootstrapped Supply Inputs to the
Topside Floating Drivers. A low ESR (X5R or better) ceramic
bypass capacitor should be connected between the BOOST
pin and the SW pin as close as possible to the IC.
CLKOUT: A Digital Output Used for Daisy-Chaining Multiple
LTC3811 ICs in Multiphase Systems. The PHASEMODE
pin voltage controls the phase relationship between the
channel 1 TG signal and CLKOUT.
COMP1, COMP2: Error Amplifier Output Voltages. The
error amplifiers in the LTC3811 are high bandwidth, low
offset true operational amplifiers that have low output
impedance. As a result, the outputs of two active error
amplifiers cannot be directly connected together! For
multiphase operation, connecting the FB pin of a slave
error amplifier to INTVCC will disable the output of that
amplifier. Multiphase operation can then be achieved by
connecting all of the COMP pins together and using one
channel as the master and all of the others as slaves. The
FB and COMP pins are also used for compensating the
control loop of the converter.
CSOUT (QFN Only): Output of the Voltage Positioning
gm Amplifier. This pin allows the user to program the
amount of voltage droop in the output voltage at high
load current. The output of the voltage positioning gm
amplifier is a bi-directional current proportional to the
(SENSE+ – SENSE–) voltages for both channels. The
gm is internally fixed to 5mS. Forcing the gm amplifier
output current through a low value external resistor will
program the amount of voltage droop seen at the output.
See Applications Information for more details regarding
voltage positioning.
DIFF/IN+: Remote Sense Differential Amplifier Positive
Input. A low offset, high bandwidth operational amplifier
is configured with four precision 80k resistors for a noninverting gain of one. This pin is normally connected to the
positive terminal of the decoupling capacitor at the load.
DIFF/IN–: Remote Sense Differential Amplifier Negative
Input. This pin is normally connected to the negative
terminal of the decoupling capacitor at the remote load.
The DIFF/IN+ and DIFF/IN– PCB traces should be routed
as close as possible and parallel to each other from the
IC to the output capacitor.
DIFF/OUT: Remote Sense Differential Amplifier Output
Voltage, Configured for a Noninverting Gain of One. The
voltage at the DIFF/OUT pin is normally connected through
an external resistor divider to the FB pin of one channel.
The bottom of the divider should be connected to the
SGND pin of the IC.
DRVCC: Output of the Internal 6V Low Dropout Regulator
(LDO), Supply Pin for the Bottom Gate Drivers and Output
of the PMOS EXTVCC Switch. A low ESR (X5R or better)
4.7μF ceramic bypass capacitor should be connected
between the DRVCC pin and the PGND pin, as close as
possible to the IC.
Exposed Pad (QFN Only): The Exposed Pad of the QFN
Leadframe is PGND.
EXTVCC: External Power Supply Input to an Internal PMOS
Power Switch Connected Between EXTVCC (Drain) and
DRVCC (Source). This pin allows an external supply to be
used for the high current gate drivers, thereby reducing
power dissipation in the LDO and increasing efficiency.
When EXTVCC exceeds 4.5V (rising), the high current
PMOS switch turns on and shorts EXTVCC to DRVCC,
bypassing the internal LDO. See Applications Information
for more details.
3811f
11
LTC3811
PIN FUNCTIONS
FB1, FB2: Error Amplifier Feedback Input Pins. The error
amplifiers in the LTC3811 are high bandwidth, low offset
true operational amplifiers. If differential remote sensing
is not used, the FB pin should be connected to a resistor
divider from the output of the power supply to SGND with
the resistors placed close to the IC. In normal regulation the
voltage at the FB pin is 0.6V. If remote sensing is used the
FB pin should be connected to a resistor divider from the
output of the differential amplifier to SGND. For multiphase
operation, connecting the FB pin of a slave error amplifier
to INTVCC will disable the output of that amplifier, allowing
amplifier outputs to be connected in parallel.
INTVCC: Supply Pin for All of the Internal Low Voltage
Analog and Digital Control Circuitry, Electrically Isolated
from the DRVCC Pin. The INTVCC supply is normally derived
by connecting a low value resistor (1Ω) from the output
of the LDO (DRVCC) to INTVCC and connecting a 0.1μF low
ESR (X5R or better) ceramic bypass capacitor connected
from INTVCC to SGND. This RC decoupling configuration
prevents gate driver switching noise from coupling into the
analog control circuitry. The INTVCC decoupling capacitor
should be connected as close as possible to the IC pins.
MODE/SYNC: Mode Control and PLL Synchronization
Input. This pin programs the operating mode and serves
as the sync input to the internal phase-lock loop (PLL).
Connecting this pin to INTVCC forces continuous operation
(regardless of the load current) and connecting it to SGND
allows discontinuous mode operation at light load. Applying
an external clock between 175kHz and 900kHz will cause
the operating frequency to synchronize to the clock.
PGND: Power Supply Return Path for the Bottom Side
Gate Drivers, Connected to the Sources of the Lower
Power MOSFETs. PGND should also be connected to the
negative terminal of the DRVCC decoupling capacitor as
close as possible to the IC. PGND is electrically isolated
from the SGND pin. The Exposed Pad on the bottom of
the QFN package is PGND.
PGOOD1, PGOOD2: An Open-Drain NMOS Power Good
Output. This output turns on, pulling down the PGOOD
pin, when the FB voltage falls out of a ±10% regulation
window. The PGOOD monitor circuit contains a 130μs
nuisance filter to prevent short duration UV and OV transients from triggering the PGOOD output on, and a 30μs
filter for the recovery from a fault condition.
PHASEMODE (QFN Only): The PHASEMODE pin voltage
programs the phase relationship between the channel 1 and
channel 2 rising TG signals, as well as the phase relationship between the channel 1 TG signal and CLKOUT.
PLL/LPF: Frequency Set and PLL Lowpass Filter Input.
When not synchronized, this pin can be used to program
the operating frequency. Connecting this pin to SGND
forces 250kHz operation and connecting it to INTVCC
forces 750kHz operation. Connecting the PLL/LPF pin to
a voltage between 0.4V and 2V forces 500kHz operation.
When synchronizing to an external clock, this pin serves
as the lowpass filter input for the PLL. A series resistor and
capacitor connected from PLL/PLF to SGND compensate
the PLL feedback loop.
RNG1, RNG2: The voltage at this pin programs the sense
voltage range for peak current mode control. Connecting
this pin to SGND programs a peak sense voltage of 24mV
and connecting it to INTVCC programs a peak sense voltage of 50mV. Alternatively, the sense voltage range can be
linearly programmed by programming the RNG pin from
0.6V to 2V with a divider from INTVCC to SGND.
RUN1, RUN2: On/Off Input Pin for Each Controller.
SENSE1+, SENSE2+: Positive Inputs to the Current Comparators and Voltage Positioning gm Amplifier. The COMP
pin voltage programs the current comparator offset in
order to set the peak current trip threshold. The LTC3811
is capable of sensing current using a discrete resistor in
series with the inductor, or by indirectly sensing the voltage drop across the DCR of the inductor. See Applications
Information for more details.
3811f
12
LTC3811
PIN FUNCTIONS
SENSE1–, SENSE2–: Negative Inputs to the Current
Comparators and Voltage Positioning gm Amplifier. The
common mode input voltage range for the current comparators is 0V to 3.5V.
SW1, SW2: Bootstrapped Supply Return Paths for the
Topside Gate Drivers, Connected to the Sources of the
Upper Power MOSFETs.
SGND: Signal Ground Pin for the IC. Common to both
controllers, this pin should be connected to the negative
terminals of the VOUT and INTVCC decoupling capacitors
and should be routed separately from any high current
paths on the PC board.
SS/TRACK1, SS/TRACK2: Combined Soft-Start and Tracking Inputs. For soft-start operation, connecting a capacitor
from this pin to ground will control the voltage ramp at
the output of the power supply. An internal 2.5μA current
source will charge the capacitor and thereby control an
extra input on the reference side of the error amplifier. For
tracking operation, this input allows the start-up of a secondary output to track a primary output according to a ratio
established by a resistor divider from the primary output
to the secondary error amplifier track pin. For coincident
tracking of both outputs at start-up, a resistor divider with
values equal to those connected to the secondary FB pin
from the secondary output should be used to connect the
secondary track input from the primary output.
TG1, TG2: High Current Gate Driver Outputs for the
N-Channel Upper Power MOSFETs.
VIN: Main Supply Input. A low ESR ceramic bypass capacitor
should be connected between this pin and SGND.
3811f
13
RAVP
(OPTIONAL)
VIN
CLP
RLP
SGND
INTVCC
DRVCC
EXTVCC
VIN
DIFFOUT
DIFFIN–
DIFFIN+
CSOUT
PGOOD1
PGOOD2
PHASEMODE
CLKOUT
PLL/LPF
4.5V
80k
80k
BIAS
80k
80k
OV2
REF
3.7V
VIN
+
–
+
–
OVER
TEMP
6V
LDO
SENSE1+
SENSE2+
SENSE2–
SENSE1–
DIFFAMP
+
–
SGND
+
gm
–
CSAMP
CLK2
CLK1
MODE/SYNC
1V
PGOOD
UV2
LOGIC
AND OV1
FILTER
UV1
VCO
PHASE DET
OT
UVLO
FC
UV
OV
ICMP
0.54V
VFB
0.66V
ISET
X
+
–
–
DUPLICATE FOR SECOND
CONTROLLER CHANNEL
+
OT
R
S
BOT
+
–
–
+
+
–
EN
INTVCC
FOLDBACK
SLOPE
COMP
IR
FC
UVLO
SHDN
TOP ON
DROP
OUT
DET
V-TO-I
Q
Q
+
–
+
–
60k
EA
VFB
0.3V
–
SWITCH
LOGIC
+
–
+
–
+
–
MODE/SYNC
SHDN
COMP
VFB
SENSE–
SENSE+
RNG
SS/TRACK
INTVCC
INTVCC
2.5μA
+
PGND
BG
SW
TG
BOOST
200k RUN
5.4V
1.1V
DRVCC
SS/TRACK
0.600V
VFB
BOT
TOP
+
–
+
–
14
+
–
+
–
fCLK
CSS
CC2
CB
DB
DRVCC
L
R2
RSENSE
R1
INTVCC
CC
RC
VIN
3811 FD
COUT
CIN
VOUT
LTC3811
FUNCTIONAL DIAGRAM
3811f
LTC3811
OPERATION
(Refer to the Functional Diagram)
Main Control Loop
The LTC3811 uses a constant frequency peak current
mode control architecture. During normal operation, the
top MOSFET is turned on each cycle when the oscillator
sets the PWM latch and turned off when the main current
comparator (ICMP) resets the latch. The peak current at
which comparator ICMP resets the latch is controlled by
the voltage on the COMP pin, which is the output of the
error amplifier. The remote sense amplifier (DIFFAMP)
produces a signal equal to the differential voltage sensed
across the output capacitor and re-references it to the
local IC ground reference (SGND). The FB pin receives
a portion of this voltage feedback signal and compares
it to the internal 0.6V reference. When the load current
increases it causes a slight decrease in the FB pin voltage
relative to the 0.6V reference, which in turn causes the
COMP pin voltage to rise until the average inductor current
is equal to the load current.
The top MOSFET drivers are biased from a floating bootstrap capacitor, CB, which is normally recharged during
the off-time through an external Schottky diode. When VIN
decreases to a voltage close to VOUT, however, the loop
may enter dropout and attempt to turn on the top MOSFET
continuously. A dropout detector senses this condition and
forces the top MOSFET to turn off every 10th cycle for one
third of a cycle to recharge the bootstrap capacitor.
Differences Between the QFN and G36 Package
Options
The LTC3811 is offered in two package options, a 38-pin
QFN and a 36-pin SSOP. The full featured QFN package
option has no leads and an exposed lead frame that needs
to be soldered to the PCB, whereas the 36-pin SSOP has
leads and is therefore slightly easier to solder to a PCB
and to debug in the lab.
The primary electrical difference between the QFN and
SSOP options is the SSOP version lacks the CSOUT
and PHASEMODE pins. With no CSOUT pin, the SSOP
version has no provision for output voltage positioning.
With no PHASEMODE input (it is internally connected
to SGND), the SSOP version is limited to 2-phase and
4-phase applications.
In addition to differences in pinout, another difference
between the two package options is their thermal resistance. The QFN package, by virtue of its exposed lead
frame, has a junction-to-ambient thermal resistance of
only 34°C/W, whereas the SSOP package has a thermal
resistance of 100°C/W. The power dissipation of the IC
is a function of the input voltage, the gate charge of the
external power MOSFETs and the operating frequency.
The gate charge losses can be partially mitigated by using the EXTVCC input to supply power to the IC, but users
should beware that high input voltage applications using
very high gate charge power MOSFETs, that also need to
operate at high frequency, should only be attempted using
the QFN package option. More details covering thermal
management are given later in this data sheet.
Supplying Power to the LTC3811
The LTC3811 features several power supply input pins
and multiple ways of supplying power to the gate drivers
and low voltage analog control circuitry.
The first method of supplying power to the IC uses the
internal low dropout linear regulator (LDO) that draws
power from VIN and regulates DRVCC to 6V, as shown in
Figure 1. The DRVCC input supplies power to the internal
gate drivers, which are capable of very high peak transient
charge (1A) and discharge (5A) currents. The DRVCC supply
should be decoupled to PGND with a minimum of 4.7μF
low ESR ceramic (X5R or better) capacitance. If multiple
power MOSFETs are being driven in parallel for high current applications it is recommended that this capacitance
VIN
VIN
EXTVCC
4.5V
+
–
6V
LDO
DRVCC
GATE DRIVER SUPPLY
INTVCC
ANALOG SUPPLY
BIAS
SGND
VFB
SS/TRACK
0.600V
–
+EA
3811 F01
Figure 1. Supplying Power to the LTC3811 from VIN
3811f
15
LTC3811
OPERATION
(Refer to the Functional Diagram)
be increased to 10μF. Because of the high peak current
capability of the gate driver, it is essential that this capacitor
be placed as close as possible to DRVCC and PGND pins,
and on the same PCB layer as the IC.
The INTVCC pin supplies power to all of the low voltage
analog circuitry and is electrically isolated from DRVCC. The
INTVCC supply is normally derived from DRVCC through
an RC filter, in order to prevent gate driver supply noise
from coupling into sensitive analog control circuitry. Typical values for this RC filter consist of a 1Ω resistor from
DRVCC to INTVCC and a 0.1μF low ESR ceramic capacitor
from INTVCC to SGND. The INTVCC capacitor should be
placed as close as possible to the INTVCC and SGND pins
and on the same PCB layer as the IC.
A third power supply pin, EXTVCC, serves as an auxiliary
input for applications where the power dissipation in the
internal LDO is excessive, or where maximum efficiency
is essential. This configuration is shown in Figure 2. When
the EXTVCC pin is left open or is connected to a voltage
less than 4.5V, the internal 6V LDO supplies DRVCC power
from VIN. If EXTVCC is tied to an external power supply
greater than 4.5V, however, the 6V LDO is turned off and
power is supplied to DRVCC through a 5Ω PMOS switch
from EXTVCC. For 4.5V < EXTVCC < 7V this PMOS switch
is on and DRVCC is approximately equal to EXTVCC. Using
the EXTVCC pin allows the gate driver and control power to
be derived from a high efficiency external source, dramatically reducing power dissipation on the IC.
VIN
4.5V
AUX 5V
SUPPLY
+
–
Once the total quiescent current for the application is
known, the power dissipation, PD, on the IC will be approximately IEXTVCC times VIN, since the gate drive current
and control circuitry quiescent current would be required
to flow through the VIN pin. The junction temperature of
the IC can then be estimated using the following wellknown formula:
TJ = TA + (PD • RθJA)
If the maximum junction temperature is close to the Absolute Maximum Rating for the particular device being used,
the use of an auxiliary supply and the EXTVCC pin may be
required. Alternatively, lower gate charge MOSFETs should
be used or the switching frequency should be reduced.
The LTC3811 control circuit has a minimum input voltage of 4.5V, making it a good choice for applications that
experience low supply conditions. However, care should
be taken to determine the minimum gate drive supply
voltage in order to choose the optimum power MOSFETs.
Important parameters that can affect the minimum gate
drive voltage are the minimum input voltage (VIN), the
LDO dropout voltage, and the EXTVCC supply voltage, if
an external gate drive supply is being used.
6V
LDO
DRVCC
GATE DRIVER SUPPLY
INTVCC
ANALOG SUPPLY
BIAS
SGND
When a voltage above 4.5V is applied to the EXTVCC pin, the
internal LDO in the LTC3811 is switched off and the power
is supplied by the external 5V power supply as shown in
Figure 2. Under these conditions, the quiescent current
at the VIN pin of the IC is very low (less than 1mA), and
most of the current required to power the analog control
circuitry and the gate drivers flows into the EXTVCC pin. As
a result, this auxiliary supply can be used as a diagnostic
tool in order to measure the total current for thermal
calculations. In order to match the actual condition when
the internal LDO is on, the voltage applied to EXTVCC when
the measurements are taken should be 6V (the same as
the regulated LDO output voltage).
Operation at Low Supply Voltage
VIN
EXTVCC
Using an External 5V Supply to Measure Dynamic
Quiescent Current
VFB
SS/TRACK
0.600V
–
+EA
3811 F02
Figure 2. Supplying Power to the LTC3811 from EXTVCC
3811f
16
LTC3811
OPERATION
(Refer to the Functional Diagram)
If the internal LDO is supplying power to the gate driver
and the input voltage is low enough for the LDO to be in
dropout, then the minimum gate drive supply voltage is:
VDRVCC = VIN(MIN) – VDROPOUT
The LDO dropout voltage is a function of the total gate
drive current and the quiescent current of the IC (typically
10mA). A curve of dropout voltage vs output current for the
LDO is shown in the Typical Performance Characteristics.
The temperature coefficient of the LDO dropout voltage is
approximately +4000ppm/ºC.
The total Q-current (IQ(TOT)) flowing in the LDO is the sum
of the controller quiescent current (typically 10mA) and
the total gate charge drive current.
IQ(TOT) = IQ + QG(TOT) • f
If an external supply is being used to supply power to the
gate driver through the EXTVCC pin, then the minimum
gate drive supply voltage is:
VDRVCC = VEXTVCC – IQ(TOT) • REXTVCC
The resistance of the internal EXTVCC PMOS switch is
typically 5Ω at 25°C and has a temperature coefficient of
approximately 3400ppm/°C.
After the calculations have been completed, it is important
to measure the gate drive waveforms (BG-to-PGND and
TG-to-SW) and the gate driver supply voltage (DRVCC-toPGND) over all operating conditions (low VIN, mid VIN,
and high VIN, as well as from light load-to-full load) to
ensure adequate power MOSFET enhancement. Consult
the power MOSFET data sheet to determine the actual
RDS(ON) for the measured VGS, and verify your thermal
calculations by measuring the component temperatures
using an infrared camera.
The RUN pins may be externally pulled up or driven directly by logic. Be careful not to exceed the 7V absolute
maximum rating on this pin.
Soft-Start and Tracking Using the SS/TRACK Pin
The start-up of each controller’s output voltage VOUT
is normally controlled by the voltage on the SS/TRACK
pin for that channel. The SS/TRACK pin represents a
2nd noninverting input to the error amplifier, as shown
in Figure 3. The error amplifier is configured so that the
lower of the two noninverting inputs (the SS/TRACK pin
or the 0.6V reference) controls the feedback loop. That
is, when the voltage on the SS/TRACK pin is less than the
0.6V internal reference, the LTC3811 regulates the FB pin
voltage to be approximately equal to the SS/TRACK pin
voltage instead of the internal 0.6V reference. This allows
the user to connect a capacitor from the SS/TRACK pin to
SGND to program the soft-start of the power supply output.
An internal 2.5μA current source charges this capacitor,
creating a voltage ramp on the SS/TRACK pin. As the SS/
TRACK pin voltage rises from 0V to 0.6V, the output voltage, VOUT, rises smoothly from 0V to its final value. Once
the soft-start interval is over, the internal 2.5μA current
source will continue to charge the SS/TRACK capacitor
up to a maximum voltage equal to INTVCC.
Alternately, the SS/TRACK pin can be used to force the
start-up of VOUT to track the voltage of another supply.
Typically, this requires connecting the SS/TRACK pin to
an external divider from the other supply to ground (see
Applications Information).
INTVCC
INTVCC
Q2 Q4
On/Off Control Using the RUN Pin
The two channels of the LTC3811 can be independently
turned on and off using the RUN1 and RUN2 pins. Pulling either of these pins low shuts down the main control
loop for that channel. Pulling both pins low disables both
controllers and most of the internal circuitry, including the
DRVCC low dropout regulator (LDO). In shutdown mode
(both RUN pins low) the LTC3811 typically draws only
20μA of current.
VFB
Q3
INTVCC
2.5μA
0.6V
SS/TRACK
Q1 Q5
SHDN
M1
3811 F03
Figure 3. Simplified LTC3811 Error Amplifier Input Stage
3811f
17
LTC3811
OPERATION
(Refer to the Functional Diagram)
The SS/TRACK pin has an internal open-drain NMOS pulldown transistor that turns on when the corresponding RUN
pin is pulled low to disable that controller, when the voltage
on the DRVCC pin is below 3.7V (the rising undervoltage
lockout threshold), or during an overtemperature condition.
During an undervoltage lockout, UVLO, or overtemperature, OT, condition, both controllers are disabled and the
external MOSFETs are held off.
In multiphase applications, one master error amplifier is
used to control all of the phase current comparators. The
FB pins for the unused error amplifiers are connected to
INTVCC in order to three-state these amplifier outputs. As a
result, the SS/TRACK pins for the unused error amplifiers
should be left open.
Programming the Operating Mode
The MODE/SYNC pin serves to either program the operating mode or to synchronize the operating frequency to
an external clock using the internal PLL. Connecting the
MODE/SYNC pin to ground programs pulse-skip mode
operation and connecting the pin to INTVCC programs
forced continuous operation, as shown in Table 1. In pulseskip mode the inductor current is not allowed to reverse,
resulting in discontinuous mode, DCM, operation at light
load. Pulse-skip mode is ideal for applications where light
load efficiency is a higher priority than transient response.
In forced continuous mode, the synchronous switch turns
on after the primary switch turns off and remains on for the
duration of the clock cycle, regardless of the load current.
Forced continuous mode is ideal for applications needing optimized transient response, or for systems where
constant frequency operation is important.
Certain applications can result in the startup of the converter into a non-zero load voltage, where residual charge
is stored on the output capacitor at the onset of converter
switching. In order to prevent a reversal of current in the
inductor under these conditions, pulse-skip operation
is asserted at startup until the FB pin reaches the lower
PGOOD threshold of 0.54V. Once the FB pin voltage exceeds
0.54V, the operating mode is determined by the voltage
on the MODE/SYNC pin.
When the operating frequency of the converter is synchronized to an external clock using the MODE/SYNC pin, the
operating mode will always be forced continuous. Forcing
continuous mode operation results in constant frequency
operation and a more predictable noise spectrum from
the converter.
Table 1
MODE/SYNC
OPERATING MODE
DESCRIPTION
SGND
Pulse-Skip
DCM Operation at Light Load
INTVCC
Forced Continuous
CCM from No Load to Full Load
External Clock
Forced Continuous
Operating Frequency Synchronized
Using Internal PLL (CCM)
Frequency Selection and the Phase-Lock Loop
The selection of the switching frequency is a tradeoff between efficiency, transient response and component size.
Low frequency operation increases efficiency by reducing
MOSEFT switching losses, but requires a larger inductor
and output capacitor to maintain low output ripple.
The switching frequency of the LTC3811’s controllers can
be selected using the PLL/LPF pin. If the MODE/SYNC pin is
not being driven by an external clock, the PLL/LPF pin can be
tied to SGND, left open or tied to INTVCC to select 250kHz,
500kHz or 750kHz, respectively, as shown in Table 2.
Table 2
PLL/LPF
MODE/SYNC
FREQUENCY
SGND
0V or INTVCC (DC)
250kHz
Floating
0V or INTVCC (DC)
500kHz
INTVCC
0V or INTVCC (DC)
750kHz
RC Filter to SGND
Connected to External
Clock
Phase Locked to
External Clock
A phase-lock loop is available on the LTC3811 to synchronize the internal oscillator to an external clock source
connected to the MODE/SYNC pin. In this case, a series
RC network connected from the PLL/LPF pin to SGND
serves as the PLL’s loop filter. The PLL/LPF pin is both the
output of the phase detector and the input to the voltage
controlled oscillator, VCO. The LTC3811 phase detector
adjusts the voltage on the PLL/LPF pin to align the rising edge of TG1 to the leading edge of the external clock
signal. The turn-on of the second channel TG2 will depend
upon the voltage on the PHASEMODE pin as shown in the
Electrical Characteristics.
3811f
18
LTC3811
OPERATION
(Refer to the Functional Diagram)
The typical capture range of the LTC3811’s PLL is approximately 125kHz to 1.1MHz, with a guarantee over
all manufacturing variations to be between 175kHz and
900kHz. The amplitude of the sync pulse to the LTC3811
should be greater than 1.8V and the minimum pulse width
should be greater than 200ns.
Using the CLKOUT and PHASEMODE Pins in
Multiphase Applications
The LTC3811 features two pins (CLKOUT and PHASEMODE)
that allow multiple LTC3811 ICs to be daisy-chained together in multiphase applications. The clock output signal
on the CLKOUT pin can be used to synchronize additional
power stages in a multiphase power supply solution feeding
a single high current output or even separate outputs. The
PHASEMODE pin is used to adjust the phase relationship
between channel 1 and channel 2, as well as the phase
relationship between channel 1 and CLKOUT, as summarized in Table 3. The phases are calculated relative to the
zero degrees, defined as the rising edge of the top gate
driver output of channel 1, TG1.
The PHASEMODE input comparators are referenced to an
internal divider from INTVCC that has 33% and 67% INTVCC
thresholds. For 6-phase operation, connect PHASEMODE
to an external divider from INTVCC with equal value resistors (e.g., 100k), so that PHASEMODE is always 50% of
INTVCC.
Table 3
# PHASES
2
3
4
6
12
IC #
1
1
2
1
2
1
2
3
1
2
3
4
5
6
PHASEMODE
0V
INTVCC
0V
0V
0V
50% INTVCC
50% INTVCC
50% INTVCC
50% INTVCC
50% INTVCC
0V
50% INTVCC
50% INTVCC
50% INTVCC
CLKOUT CONNECTS TO
N/A
MODE/SYNC of IC # 2
N/A
MODE/SYNC of IC # 2
N/A
MODE/SYNC of IC # 2
MODE/SYNC of IC # 2
N/A
MODE/SYNC of IC # 2
MODE/SYNC of IC # 3
MODE/SYNC of IC # 4
MODE/SYNC of IC # 5
MODE/SYNC of IC # 6
N/A
Remote Sensing Using the Differential Amplifier
The LTC3811 has a differential amplifier for true remote
sensing of the output voltage. The sensing connections
should be returned from the load back to the differential
amplifier’s inputs through a common, tightly coupled pair
of PCB traces. The differential amplifier rejects common
mode signals capacitively or inductively radiated into the
feedback PCB traces, as well as ground loop disturbances.
The differential amplifier output signal is typically divided
down and compared with the internal precision 0.6V voltage reference by the error amplifier.
The differential amplifier utilizes four precision internal
resistors to enable instrumentation-type measurement
of the output voltage. The amplifier has a gain of 1.000,
contains a CMOS rail-to-rail output stage, and is optimized
for low input offset and high bandwidth.
The output voltage is set by an external resistive divider
according to the following formula:
⎡ R2⎤
VOUT = 0.6 • ⎢1+ ⎥
⎣ R1⎦
where R2 and R1 are the upper and lower divider resistors,
respectively. The differential amplifier was optimized for
divider currents in the range of 100μA to 600μA, meaning
that R1 in the equation above should be 1k to 6k.
Using the LTC3811 Operational Error Amplifiers in
Multiphase Applications
The LTC3811 error amplifiers are true operational amplifiers, meaning that they have high DC gain and low output
impedance. In previous generations of multiphase controllers, such as the LTC1628 family, the error amplifiers were
transconductance amplifiers, meaning that they could be
connected in parallel for multiphase applications.
Multiphase applications using the LTC3811 will use one
operational error amplifier as the master and will disable
all of the slave phase error amplifiers. Typically, the channel 1 amplifier for phase = 0º will be used as the master
and phases 2 through n (up to 12 phases) will serve as
slaves. To disable the slave error amplifiers but still use
their current comparators and power stages, connect the
3811f
19
LTC3811
OPERATION
(Refer to the Functional Diagram)
ONLY PGOOD PIN FOR
MASTER CHANNEL
IS USED (FLOAT SLAVE
CHANNEL PGOOD PINS)
INDIVIDUAL INTVCC AND DRVCC
PINS LOCALLY DECOUPLED
(DRVCC NOT SHOWN)
MASTER
VOUT+
VOUT–
MASTER
DIFFERENTIAL
AMPLIFIER
USED TO DRIVE
CHANNEL 1
ERROR
AMPLIFIER
LTC3811
INTVCC
PGOOD1
PGOOD2
FB2
DIFF/IN+
DIFF/IN–
RUN1
DIFF/OUT
RUN2
FB1
COMP1 SS/TRACK1
COMP2 SS/TRACK2
CLKOUT SGND
ON/OFF
CONTROL
ALL
SS/TRACK
PINS
CONNECTED
TOGETHER
SLAVE
SLAVE
CHANNEL
FB PINS ALL
CONNECTED
TO LOCAL
INTVCC PINS
TO DISABLE
ERROR
AMPLIFIERS
SYNC LTC3811
INTVCC
PGOOD1
FB2
PGOOD2
FB1
RUN1
COMP1
RUN2
COMP2 SS/TRACK1
SS/TRACK2
CLKOUT SGND
ALL RUN
PINS
CONNECTED
TOGETHER
SLAVE
SLAVE
CHANNEL
COMP
PINS ALL
CONNECTED
TO MASTER
CHANNEL
COMP PIN
SYNC LTC3811
PGOOD1
INTVCC
PGOOD2
FB2
RUN1
FB1
RUN2
COMP1
COMP2 SS/TRACK1
SS/TRACK2
SGND
the amplitude of those for one regulator to be drawn from
the input capacitor. These large amplitude current pulses
increased the total RMS current flowing from the input
capacitor, requiring the use of more expensive input capacitors and increasing both EMI and losses in the input
capacitor.
With multiphase operation, the two channels of the dualswitching regulator are operated 180 degrees out of phase.
This effectively interleaves the current pulses drawn by the
switches, greatly reducing the overlap time where they add
together. The result is a significant reduction in total RMS
input current, which in turn allows less expensive input
capacitors to be used, reduces shielding requirements for
EMI and improves real world operating efficiency.
Figure 5 illustrates the benefits of multiphase operation.
Current ripple at the input is reduced by a factor of 1.41
(square root of 2), reducing the size and cost of the input
capacitor. In addition, since power losses are proportional
to IRMS2, significant efficiency improvements in the input
power path components (batteries, switches, protection
circuitry and PCB traces) can be achieved. Improvements
in both conducted and radiated EMI also directly accrue
as a result of the reduced RMS input current.
3811 F04
SGND BUS ISOLATED FROM
PGND AND INDEPENDENTLY
ROUTED TO NEGATIVE
TERMINAL OF OUTPUT CAPACITOR
Figure 4. LTC3811 Error Amplifier Configuration
for Multiphase Operation
FB pin of a slave phase to INTVCC. As shown in the Functional Diagram, a comparator detects when the FB pin is
shorted to INTVCC and three-states this amplifier’s output
and input. The COMP pins for all of the phases can then
be shorted together in order to provide compensation for
the feedback loop, as shown in Figure 4.
Theory and Benefits of Multiphase Operation
Why the need for multiphase operation? Up until the
multiphase family, constant frequency dual switching
regulators operated both channels in phase (i.e., singlephase operation). This means that both switches turned
on at the same time, causing current pulses of up to twice
IL
5A/DIV
SW1
10V/DIV
SW2
10V/DIV
VIN = 12V
VOUT = 1.5V
IOUT = 16A
1μs/DIV
3811 G41
Figure 5. 2-Phase, Single Output Current Sharing Waveforms
Of course, the improvement afforded by 2-phase operation is a function of the dual switching regulator’s relative
duty cycles which, in turn, are dependent upon the input
voltage VIN (Duty Cycle = VOUT/VIN).
Figure 6 shows the net ripple current seen by the output
capacitors for the different phase configurations. The
output ripple current is plotted for a fixed output voltage
as the duty factor is varied between 10% and 90% on the
3811f
20
LTC3811
OPERATION
(Refer to the Functional Diagram)
x-axis. The output ripple current is normalized against
the inductor ripple current at zero duty factor. The graph
can be used in place of tedious calculations. As shown in
Figure 6, the zero output ripple current is obtained
when:
VOUT k
=
where k = 1, 2, ..., N – 1
VIN
N
So the number of phases used can be selected to minimize
the output ripple current and therefore the output ripple
voltage at the given input and output voltages. In applications having a highly varying input voltage, additional
phases will produce the best results.
Accepting larger values of ΔIL allows the use of low inductances, but can result in higher output voltage ripple.
A reasonable starting point for setting ripple current is
ΔIL = 0.4(IOUT)/N, where N is the number of channels and
IOUT is the total load current. Remember, the maximum
ΔIL occurs at the maximum input voltage. The individual
inductor ripple currents are constant determined by the
inductor, input and output voltages.
1.0
1-PHASE
2-PHASE
3-PHASE
4-PHASE
6-PHASE
0.9
0.8
0.6
VO/fL
ΔIO(P-P)
0.7
lockout or overtemperature condition. When the FB pin
voltage is within the ±10% window, the internal PGOOD
MOSFET is turned off and the pin is normally pulled up by
an external resistor. The absolute maximum voltage rating
of the PGOOD pins is 7V.
The PGOOD logic contains separate filters depending on
whether the controller is entering or exiting a fault condition. When the FB pin is exiting a fault condition (such as
during normal output voltage start-up, prior to regulation),
the PGOOD pin will remain low for an additional 30μs. This
allows the output voltage to reach steady-state regulation
and prevents the enabling of a heavy load from re-triggering
a UVLO condition. When the FB pin is entering either an
undervoltage, UV, or overvoltage, OV, fault condition, the
PGOOD pin will remain high 130μs after the onset of the
fault. This non-integrating filter prevents noise or short
duration overload conditions from triggering the PGOOD
outputs and causing a false system reset. Figure 7 illustrates
the timing diagram for a hypothetical undervoltage event
on the FB pin, and the resulting PGOOD waveform.
In multiphase applications, one error amplifier is used to
control all of the phase current comparators. In addition,
since the FB pins for the unused error amplifiers are connected to INTVCC (in order to three-state these amplifiers),
the PGOOD outputs for these amplifiers will be asserted.
In order to prevent falsely reporting a fault condition, the
0.5
0.4
0.66V = OV
THRESHOLD
0.7
0.3
0.6
0.2
0.1
0.2
0.3 0.4 0.5 0.6 0.7
DUTY FACTOR (VOUT/VIN)
0.8
0.9
VFB (V)
0
0.54V = UV
THRESHOLD
0.5
0.1
0.4
0.3
3811 F06
Figure 6. Normalized Peak Output Current vs
Duty factor [IRMS ≈ 0.3(ΔIO(P-P))]
0.2
0.1
0
Each PGOOD pin is connected to the open drain of an
internal N-channel pull-down MOSFET. The MOSEFT turns
on and pulls the PGOOD pin low when the corresponding
FB pin is outside a ±10% window around the 0.6V reference voltage. The PGOOD pin is also pulled low when the
corresponding RUN pin is low, or during an undervoltage
5
VPGOOD (V)
Power Good Pins (PGOOD1, PGOOD2)
TIME
t < 30μs
130μs
0
30μs
TIME
3811 F07
Figure 7. PGOOD Filter Timing Diagram
3811f
21
LTC3811
OPERATION
(Refer to the Functional Diagram)
PGOOD outputs for the unused error amplifiers should be
left open. Only the PGOOD output for the master control
error amplifier should be connected to the fault monitor.
Fault Conditions: Current Limit and Foldback
One of the main advantages of the LTC3811 is the fact that
the maximum inductor current is inherently limited due
to the use of peak current mode control. The maximum
sense voltage is controlled by the voltage on the RNG pins
and the maximum DC output current is:
ILIMIT =
VSENSE(MAX)
RSENSE
–
ΔIL(SC) = tON(MIN) •
1
• ΔIL
2
The current limit value should be checked to ensure that
ILIMIT(MIN) > IOUT(MAX). The minimum value of the current
limit generally occurs with the largest VIN at the highest
ambient temperature, conditions that cause the largest
power losses in the converter.
To further limit current in the event of an output short-circuit
to ground, the LTC3811 includes foldback current limiting.
When the FB pin falls below 0.3V (50% of its nominally
regulated value), the foldback circuit is activated, progressively lowering the peak current limit in proportion to the
severity of the overcurrent or short circuit condition. If the
FB pin reaches 0V, the peak current sense threshold will
be reduced to 30% of its maximum value. The foldback
current limit transfer function is shown in Figure 8.
60
MAXIMUM SENSE THRESHOLD (mV)
Under short-circuit conditions with very low duty cycles,
the LTC3811 may begin to skip pulses in order to limit the
maximum current. In this situation the bottom MOSFET
will be dissipating most of the power; however this will
be less than in normal operation at maximum load. In this
case the short circuit ripple current is determined by the
minimum on-time, tON(MIN), of the LTC3811 (about 65ns),
the input voltage, and the inductor value, according to the
following equation:
55
VIN
L
The resulting short-circuit current is
0.3 • VSENSE(MAX) 1
ISC =
– • ΔIL(SC)
RSENSE
2
Depending upon the ratio of the DC value of the current
limit to the maximum load current and the percentage
ripple current in the inductor, it is possible that the converter will operate in discontinuous mode when VFB = 0V
(a so-called “dead short”). In this case, the short-circuit
current of the converter will be:
ISC =
ΔIL(SC)
2
= tON(MIN) •
VIN
2•L
In order for the converter to start up properly with a nonlinear load, the foldback current limiting circuit in the
LTC3811 is disabled during the initial soft-start interval.
When the FB pin voltage reaches 0.54V, the soft-start interval is terminated and the foldback circuit is enabled.
50
In the event the converter is turned on into a shorted load,
the foldback circuit will be disabled until the SS/TRACK
pin reaches 0.54V. This ensures that the converter will still
limit the maximum current to a safe level and reduce the
peak power dissipated with a shorted load.
45
40
35
30
25
20
15
RNG = INTVCC
10
5
0
0
100
200 300 400 500
FB VOLTAGE (mV)
600
700
3811 G31
Figure 8. Current Foldback Charactistic
3811f
22
LTC3811
APPLICATIONS INFORMATION
Duty Cycle Considerations
COMP
The duty cycle for a buck converter is well known:
LTC3811
V
D = OUT = tON • f
VIN
VOUT
FB
R1
Rearranging, the minimum on-time for a given application
can be calculated:
tON(MIN) =
VOUT
VIN(MAX ) • f
For a given input and output voltage, it is important to know
how close the minimum on-time of the application comes
to the minimum on-time of the control IC. The LTC3811
has a typical minimum on-time of 65ns, allowing both high
input to output ratios and high frequency operation.
In an application circuit, if the IC’s minimum on-time
exceeds the value required in the duty cycle equations,
the converter will begin to skip pulses and operate at a
fraction of the intended frequency. This frequency division
will result in higher current and voltage ripple and is of
particular concern in forced continuous applications with
low ripple currents at light loads.
Setting the Output Voltage
The LTC3811 output voltages are each set by external
feedback resistor dividers, according to the following
equation:
⎡ R2⎤
VOUT = 0.6V • ⎢1+ ⎥
⎣ R1⎦
Care should be taken to place the output divider resistors
and the compensation components as close as possible
to the IC FB and SGND pins, in order to prevent switching
noise from coupling into the signal path. This configuration
is shown in Figure 9. The top of R2 is normally routed to
the top of the output capacitor, or to the output of the differential amplifier, if remote sensing is being employed.
Because the common mode range of the current comparator input stages is 0V to 3.5V, the output voltage range is
limited from 0.6V to 3.3V.
R2
COUT
SGND
3811 F09
DIVIDER AND COMPENSATION
COMPONENTS PLACED NEAR
FB, SGND AND COMP PINS
Figure 9. Output Divider and Compensation
Component Placement
Sensing the Output Voltage with a Differential
Amplifier
The LTC3811 includes a low offset, unity gain, high bandwidth differential amplifier for applications that require true
remote sensing. Sensing both SENSE+ and SENSE– greatly
benefits regulation in high current, low voltage applications,
where board interconnection losses can be a significant
portion of the total error budget.
The LTC3811 differential amplifier has a typical output slew
rate of 8V/μs and has rail-to-rail output drive capability.
The amplifier is configured for unit gain, meaning that the
difference between SENSE+ and SENSE– is translated to
DIFFOUT, relative to SGND.
Care should be taken to route the SENSE+ and SENSE– PCB
traces parallel to each other all the way to the terminals
of the output capacitor or remote sensing points on the
board. In addition, avoid routing these sensitive traces near
any high speed switching nodes in the circuit. Ideally, the
SENSE+ and SENSE– traces should be shielded by a low
impedance ground plane to maintain signal integrity.
Choosing the Inductor Value and Saturation Current
Rating
The operating frequency and inductor value are interrelated
in that higher operating frequencies allow the use of smaller
inductors and capacitors. Higher frequency operation also
results in higher switching and gate drive losses, so a basic
tradeoff exists between size and efficiency.
3811f
23
LTC3811
APPLICATIONS INFORMATION
L=
⎤
VOUT ⎡
V
⎢1– OUT ⎥
f • ΔIL ⎢⎣ VIN(MAX) ⎥⎦
Choosing a larger value of ΔIL allows the use of a lower
value inductor, but results in higher output voltage ripple
and greater core losses. A reasonable starting point for
setting the ripple current is 40% to 50% of the maximum
output current, or:
ΔIL = 0.4 • IOUT(MAX)
The inductor saturation current rating needs to be higher
than the peak inductor current during an overload condition. If IOUT(MAX) is the maximum rated load current, then
the maximum overload current, IMAX, would normally
be chosen to be some factor (e.g., 30%) greater than
IOUT(MAX):
IMAX = 1.3 • IOUT(MAX)
IL(PK) = IMAX +
1
• ΔIL
2
For a 40% ripple application, the minimum saturation
current rating of the inductor would therefore be:
IL(PK) = 1.5 • IO(MAX)
In other words, for an application with 40% inductor ripple
current and a maximum output current 30% greater than
the full load current, the inductor’s saturation current rating needs to be at least 1.5 times the maximum output
current.
Inductor Core Selection
Once the value of L is known, the type of inductor must be
selected. High efficiency converters generally cannot afford
the core losses found in low cost powdered iron cores,
forcing the use of more expensive ferrite or molypermalloy
cores. Actual core loss is independent of core size for a fixed
inductor value, but is very dependent on the inductance
selected. As inductance increases, core losses go down.
Unfortunately, increased inductance requires more turns
of wire and therefore copper losses will increase.
Ferrite designs have very low core loss and are preferred
at high switching frequencies, so design goals can concentrate on copper loss and preventing saturation. Ferrite
core materials exhibit “hard” saturation, meaning that
the inductance collapses abruptly when the peak current
capability is exceeded. This results in an abrupt increase
in inductor ripple current and output voltage ripple. Do
not allow the core to saturate!
Programming the Maximum Sense Voltage Using the
RNG Pin
The RNG pin can be used in two different ways in order to
program the maximum peak current sense voltage. The
easiest way to program the peak sense voltage is to tie
the RNG pin to either ground or INTVCC. Connecting the
RNG pin to ground results in a 24mV peak sense voltage
and connecting it to INTVCC programs in a 50mV peak
sense voltage. Alternately, an external resistor divider
from INTVCC to ground can be used to set the RNG pin
between 0.6V and 2V, resulting in a nominal peak sense
voltage range of 24mV to 85mV. Figure 10 illustrates the
transfer function from the RNG pin to the peak sense
voltage, which closely follows the following equation for
0.6V < VRNG < 2V:
VSENSE(MAX) = 0.0436 • VRNG – 0.0022
In general, the accuracy of the SENSE pin threshold will
scale with the peak sense voltage defined by the RNG
MAXIMUM CURRENT SENSE THRESHOLD (mV)
For CCM operation, the inductor value can be chosen using
the following equation:
90
80
70
60
50
40
30
20
10
0
0.5
0.7
0.9
1.1 1.3 1.5
VRNG VOLTAGE (V)
1.7
1.9
3811 G30
Figure 10. Maximum Current Sense Threshold
vs RNG Pin Voltage
3811f
24
LTC3811
APPLICATIONS INFORMATION
In multiphase applications, only one error amplifier is used
to control all of the phase current comparators. As a result,
in multiphase applications all of the RNG pins should all be
tied to the same potential, in order to program the same
power stage gm for each phase.
SENSE+ and SENSE– Pins
The common mode input voltage range of the current
comparators is 0V to 3.5V. Continuous linear operation is
provided throughout this range, allowing output voltages
between 0.6V (the reference input to the error amplifiers)
and 3.3V. The SENSE+ and SENSE– pins are also the inputs
to the voltage positioning current sense gm amplifier.
Under normal operation, a small current of about 1.5μA
flows out of the SENSE inputs and represents the total
base current of the two vertical PNP input stages (one in
the current comparator and one in the voltage positioning
current sense amplifier). When the common mode voltage
is lower than about 0.4V, the current flowing out of the
SENSE pins increases, up to about 2.2μA at VSENSE = 0V.
Figure 11 illustrates the change in the SENSE pin current
as a function of common mode voltage.
with operation up to 1MHz are becoming more common.
Under these conditions the voltage drop across the sense
resistor’s parasitic inductance is no longer negligible.
A typical sensing circuit using a discrete resistor is shown
in Figure 12. In previous generations of controllers, a small
RC filter placed near the IC was commonly used to reduce
the effects of capacitive and inductive noise coupled in
the sense traces on the PCB. A typical filter consists of
two series 10Ω resistors connected to a parallel 1000pF
capacitor, resulting in a time constant of 20ns.
2.5
SENSE PIN CURRENT (μA)
pin. For applications requiring maximum current limit accuracy, a higher peak sense voltage (e.g., 85mV) should
be chosen. An additional benefit of a higher peak SENSE
pin threshold is a slight reduction in the minimum ontime of the controller. That is, for a given ripple current
in the inductor, a higher peak sense voltage results in
higher SENSE pin dV/dt, speeding up the input stage of
the current comparator slightly. For applications where
high efficiency and tight current limit accuracy are both
important, the peak current sense voltage can be reduced
to as low as 24mV.
2.0
1.5
1.0
0.5
0
0
0.5
1.0 1.5 2.0 2.5 3.0 3.5 4.0
COMMON MODE VOLTAGE (V)
3811 G33
Figure 11. SENSE Pin Input Bias Current vs
Common Mode (Output) Voltage
VIN
VIN
DRVCC
LTC3811
SENSE RESISTOR
PLUS PARASITIC
INDUCTANCE
BOOST
TG
RS
SW
BG
Sensing Techniques Using Low Value Resistors
For previous generation current mode controllers, the
maximum sense voltage was high enough (e.g., 75mV
for the LTC1628 family) that the voltage drop across the
parasitic inductance of the sense resistor represented a
relatively small error. For today’s highest current density
solutions, however, the value of the sense resistor can be
less than 1mΩ and the peak sense voltage can be as low as
24mV. In addition, inductor ripple currents greater than 50%
PGND
SENSE+
SENSE–
SGND
RF
ESL
VOUT
CF • 2RF = ESL/RS
POLE-ZERO
CANCELLATION
CF
RF
3811 F12
FILTER COMPONENTS
PLACED NEAR SENSE PINS
Figure 12. Using a Resistor to Sense Current with the LTC3811
3811f
25
LTC3811
APPLICATIONS INFORMATION
This same RC filter, with minor modifications, can be used
to extract the resistive component of the current sense
signal in the presence of parasitic inductance. For example,
Figure 13 illustrates the voltage waveform across a 1.5mΩ
Panasonic metal strip resistor (ERJ-M1WTJ1M5U). The
waveform is the superposition of a purely resistive component and a purely inductive component and was measured
with a single low impedance scope probe through a BNC
connected directly across the sense resistor terminals.
Based on additional measurements of the inductor ripple
current and the on- and off-times of the primary switch,
the value of the parasitic inductance was determined to
be 0.5nH using the equation:
⎡ ΔI
ΔI ⎤
VESL(STEP) = ESL • ⎢ L + L ⎥
⎣ tON tOFF ⎦
If the R∙C time constant is chosen to be exactly the same
as the parasitic inductance divided by the sense resistor
(L/R), the resulting waveform looks resistive again, as
shown in Figure 14. For applications using low maximum
10mV/DIV
sense voltages, check the sense resistor manufacturer’s
data sheet for information about parasitic inductance. In
the absence of data, measure the voltage drop directly
across the sense resistor using a low impedance connection to extract the magnitude of the ESL step and
use the equation above to determine the proper filter
time constant, keeping the two filter resistor values equal
and less than about 200Ω each. Finally, place these filter
components close to the IC and run the positive and
negative sense traces parallel to each other all the way
to the sense resistor.
Inductor DCR Sensing
For applications requiring the highest possible efficiency,
the LTC3811 is capable of sensing the voltage drop across
the inductor DCR, as shown in Figure 15. The DCR of
the inductor represents the small amount of DC winding
resistance of the copper, which can be less than 1mΩ for
today’s low value, high current inductors. If the external
RC time constant is chosen to be exactly equal to the
L/DCR time constant, the voltage drop across the external
capacitor is equal to the voltage drop across the inductor
DCR. Check the manufacturer’s data sheet for specifications
regarding the inductor DCR in order to properly dimension
the external filter components. The DCR of the inductor
can also be measured using a good RLC meter.
VIN
250ns/DIV
3811 F13
Figure 13. Current Sense Waveform for the Circuit in Figure 33
VIN DRVCC
LTC3811
BOOST
INDUCTOR
TG
L
SW
SW
DCR
VOUT
BG
R1
PGND
SENSE+
10mV/DIV
SENSE–
R1 = R2
R2
C1
R1 • C1 = L/DCR
SGND
3811 F15
250ns/DIV
3811 F14
Figure 14. Waveform at the SENSE+ and SENSE– Pins
Using RC – L/R Time Constant Cancellation
FILTER COMPONENTS
SHOULD BE PLACED NEAR
SENSE+, SENSE– PINS
PLACE R1 NEAR
INDUCTOR TO
MINIMIZE SW NODE
NOISE COUPLING
Figure15: Current Mode Control Using the Inductor DCR
3811f
26
LTC3811
APPLICATIONS INFORMATION
The value of the resistors in the RC filter is a tradeoff
between power dissipation and DC accuracy. The power
loss on R1 is:
PR1 =
(VIN – VO) • VO
R1
for a buck converter.
If the value of the filter resistor is too low, its power dissipation will rise, resulting in a larger package size and
decreased efficiency at light load. If the value of the filter
resistor is too high, the input bias current flowing out of
the SENSE+ pin (approximately 1.5μA) could cause the
voltage drop across the resistor to be the same order
of magnitude of the peak sense voltage, which is also
undesirable. A good balance is to use a resistor value of
about 1k. An additional 1k resistor (R2) in the SENSE– path
is used to compensate for the drop in the SENSE+ path,
and ideally these two resistors (R1 and R2) should match
one another.
In general, the larger the sense voltage range is, the smaller
the percentage error due to a mismatch in the filter resistor IR drops. The current comparators were designed for
low offset and high speed, specifically for applications
requiring a small peak sense voltage.
Gate Drive Power Supply Considerations
The LTC3811 user has a choice of how to supply power to
the gate drivers and low voltage analog control circuitry.
The first of these is to use the internal low dropout linear
regulator, LDO, to draw power from VIN and regulate DRVCC
to 6V. The second way of supplying power to the gate
drivers and analog control circuitry is through the EXTVCC
pin. The choice of which supply path to use depends upon
system flexibility, power dissipation and the maximum
junction temperature in the application.
The internal DRVCC LDO is capable of sourcing up to
100mA, allowing the user to connect multiple power
MOSFETs in parallel on both channels for the high power
density applications. High input voltage applications in
which multiple large MOSFETs are being driven at high
frequencies, however, may cause the maximum junction
temperature rating for the LTC3811 to be exceeded.
In general, there are three potential sources of power
dissipation in the LTC3811:
1. The quiescent current consumed by all of the analog
control circuitry connected to INTVCC
2. Gate drive losses
3. Losses in the LDO when power is being supplied from
VIN
The steady-state quiescent current of the IC is typically
10mA and flows into the INTVCC pin, either through the
LDO from VIN or through an auxiliary power supply connected to the EXTVCC pin.
The second source of power dissipation is the gate drivers
connected to DRVCC. The lower MOSFET gate drivers are
directly connected to DRVCC and the upper ones are connected to DRVCC through the bootstrap diode and floating
supply capacitor CB (refer to Functional Diagram). The gate
driver current requirement depends upon the number of
MOSFETs being driven, their total gate charge, QG(TOT),
and the operating frequency, f, of the converter. The total
current required by the low voltage circuitry is the sum of
the DC quiescent current and the gate drive current.
IVCC = 10mA + QG(TOT) • f
If the internal LDO in the LTC3811 is used to supply power
to DRVCC and INTVCC, care should be taken to ensure that
the total low voltage current doesn’t exceed the 100mA
limit for the LDO.
Assuming that DRVCC = EXTVCC = INTVCC = 6V, power
dissipation due to the quiescent current and gate drive
losses is:
PVCC = 6V • (10mA + QG(TOT) • f)
The third source of power dissipation occurs in the LDO,
which supplies power to the DRVCC pin when EXTVCC is
less than 4.7V. When power is being drawn from VIN the
power dissipated in the LDO is:
PLDO = (VIN – VDRVCC) • (10mA + QG(TOT) • f)
3811f
27
LTC3811
APPLICATIONS INFORMATION
The total power dissipation is the sum of these two and
the junction temperature can then be estimated using the
following equation:
TJ = TA + (PVCC + PLDO) • RθJA
As an example, consider a 2-phase, single-output application with a 12V input voltage and a 1.2V output at up to
30A (15A/phase), using the QFN version of the LTC3811.
The upper power MOSFETs are the Renesas RJK0305DPB
(one per phase) and the lower power MOSFETs are the
RJK0301DPB (one per phase). The upper MOSFETs have
a typical RDS(ON) = 10mΩ at VGS = 4.5V and a typical QG =
8nC. The lower MOSFETs have a typical RDS(ON) = 3mΩ at
VGS = 4.5V and a typical QG = 32nC. The total gate charge
is therefore 80nC and the operating frequency is 500kHz.
With a maximum ambient temperature of 70°C and a
thermal resistance of 34°C/W for the QFN package,
IDRVCC = 10mA + 500kHz • 80nC = 50mA
PDRVCC = 6V • (10mA + 500kHz • 80nC) = 300mW
PLDO = (12V – 6V) • (10mA + 500kHz • 80nC)
= 300mW
TJ = 70°C + (0.3 + 0.3) • 34°C/W = 90°C
A 20°C rise in the junction temperature and a maximum LDO
current of 50mA are acceptable numbers but could be improved upon by using the EXTVCC pin to supply power to the
gate drivers. The use of an auxiliary supply connected to the
EXTVCC pin would reduce the junction temperature rise by
a factor of 2, resulting in a max junction temperature of:
TJ = 70°C + 0.3 • 34°C/W = 80°C
For applications where the internal LDO is being used to
supply power to the IC, to prevent the maximum junction
temperature from being exceeded the input supply current should be monitored at maximum VIN in continuous
conduction mode (i.e., with MODE/SYNC connected to
INTVCC).
Using the EXTVCC Pin to Supply Power to the LTC3811
The LTC3811 contains an internal P-channel MOSFET
switch connected between the EXTVCC and DRVCC pins.
When the voltage applied to EXTVCC exceeds 4.5V, the
internal LDO is turned off and the PMOS switch turns on,
connecting the EXTVCC pin to the DRVCC pin and thereby
supplying the internal analog and digital circuitry and
MOSFET gate drive power. Do not apply greater than 7V
to the EXTVCC pin (its absolute maximum rating) and ensure that EXTVCC < VIN + 0.3V when using the application
circuits shown. If an external voltage source is applied to
the EXTVCC pin when the VIN supply is not present, a diode
can be placed in series with the LTC3811’s VIN pin and a
Schottky diode between the EXTVCC pin and the VIN pin,
to prevent current from backfeeding into VIN through the
PMOS body diodes.
Significant energy gains can be realized by powering DRVCC and INTVCC from an auxiliary supply, since
the VIN current resulting from the driver and analog
control circuitry currents will be scaled by the ratio:
Duty Cycle/Efficiency
The following list summarizes the three possible connections for EXTVCC:
1. EXTVCC left open (or grounded). This will cause DRVCC
and INTVCC to be powered from the internal 6V LDO,
resulting in a significant efficiency penalty and excess
power dissipation at high input voltages.
2. EXTVCC connected to an external supply. If an external
supply is available in the 5V to 7V range it may be used
to power EXTVCC, provided it is capable of satisfying
the gate drive and control IC current requirements. VIN
must be greater than or equal to the voltage applied to
the EXTVCC pin.
3. EXTVCC connected to an output-derived boost network.
For 3.3V and other low voltage regulators, efficiency
gains can still be realized by connecting EXTVCC to
an output-derived voltage which has been boosted to
greater than 4.5V but less than 7V. This can be done
with a capacitive charge pump shown in Figure 16.
Power MOSFET and Schottky Diode (Optional)
Selection
Two external power MOSFETs must be selected for each
controller in the LTC3811: one N-channel MOSFET for the
top (main) switch, and one N-channel MOSFET for the
bottom (synchronous) switch.
3811f
28
LTC3811
APPLICATIONS INFORMATION
CIN
+
VIN
+
VIN
0.22μF
BAT85
PMAIN =
BAT85
TG1
LTC3811
BAT85
VN2222LL
EXTVCC
RSENSE
SW1
VOUT
L1
+
BG1
COUT
D1
⎤
(VIN)2 ⎡⎢IMAX
(R )(C
)•
2 ⎥ DR MILLER
⎣
PSYNC =
Figure 16. Capacitive Charge Pump for EXTVCC
The peak-to-peak drive levels are set by the INTVCC
voltage. This voltage is typically 6V during start-up
(see EXTVCC Pin Connection). Consequently, logic-level
threshold MOSFETs should be used in most applications.
The only exception is if low input voltage is expected
(VIN < 5V); then, sub-logic level threshold MOSFETs
(VGS(TH) < 3V) should be used. Pay close attention to the
BVDSS specification for the MOSFETs as well; most of the
logic-level MOSFETs are limited to 30V or less.
Selection criteria for the power MOSFETs include the “ON”
resistance RDS(ON), Miller capacitance CMILLER, input
voltage and maximum output current. Miller capacitance,
CMILLER, can be approximated from the gate charge curve
usually provided on the MOSFET manufacturers’ data
sheet. CMILLER is equal to the increase in gate charge
along the horizontal axis while the curve is approximately
flat divided by the specified change in VDS. This result is
then multiplied by the ratio of the application applied VDS
to the Gate charge curve specified VDS. When the IC is
operating in continuous mode the duty cycles for the top
and bottom MOSFETs are given by:
VOUT
VIN
Synchronous Switch Duty Cycle =
VIN – VOUT
VIN
The MOSFET power dissipations at maximum output
current are given by:
⎦
⎡
1 ⎤
1
+
⎢
⎥ (f)
⎣ VINTVCC – VTHMIN VTHMIN ⎦
3811 F16
PGND
Main Switch Duty Cycle =
VOUT
2
IMAX ) RDS(ON) (1+ δ) +
(
VIN
VIN – VOUT
2
IMAX ) (1+ δ)RDS(ON)
(
VIN
where δ is the temperature dependency of RDS(ON) and
RDR (approximately 2Ω) is the effective top gate driver
resistance at the MOSFET’s Miller threshold voltage. VTHMIN
is the typical MOSFET minimum threshold voltage.
Both MOSFETs have I2R losses while the topside N-channel
equation includes an additional term for transition losses,
which are highest at high input voltages. For VIN < 20V
the high current efficiency generally improves with larger
MOSFETs, while for VIN > 20V the transition losses rapidly
increase to the point that the use of a higher RDS(ON) device
with lower CMILLER actually provides higher efficiency. The
synchronous MOSFET losses are greatest at high input
voltage when the top switch duty factor is low or during
a short-circuit when the synchronous switch is on close
to 100% of the period.
The term (1+δ) is generally given for a MOSFET in the
form of a normalized RDS(ON) vs Temperature curve, but
δ = 0.005/°C can be used as an approximation for low
voltage MOSFETs.
The optional Schottky diode, D1, shown in Figure 16 conducts during the dead-time between the conduction of the
two power MOSFETs. This prevents the body diode of the
bottom MOSFET from turning on, storing charge during
the dead-time and requiring a reverse recovery period that
could cost as much as 1% in efficiency at high VIN. A 1A
to 3A Schottky is generally a good compromise for both
regions of operation due to the relatively small average
current. Larger diodes result in additional transition losses
due to their larger junction capacitance.
3811f
29
LTC3811
APPLICATIONS INFORMATION
CIN and COUT Selection
VOUT k
= where k = 1, 2, ..., N – 1
VIN
N
So the phase number can be chosen to minimize the input
capacitor size for the given input and output voltages.
In the graph of Figure 17, the local maximum input RMS
capacitor currents are reached when:
VOUT 2k − 1
=
where k = 1, 2, ..., N
2N
VIN
These worst-case conditions are commonly used for design
because even significant deviations do not offer much relief.
Note that capacitor manufacturer’s ripple current ratings
are often based on only 2000 hours of life. This makes
it advisable to further derate the capacitor, or to choose
a capacitor rated at a higher temperature than required.
Several capacitors may also be paralleled to meet size or
height requirements in the design. Always consult the
capacitor manufacturer if there is any question.
The graph shows that the peak RMS input current is
reduced linearly, inversely proportional to the number, N
of stages used. It is important to note that the efficiency
loss is proportional to the input RMS current squared and
therefore a 2-stage implementation results in 75% less
power loss when compared to a single phase design. Battery/input protection fuse resistance (if used), PC board
trace and connector resistance losses are also reduced
by the reduction of the input ripple current in a PolyPhase
0.5
DC LOAD CURRENT
RMS INPUT RIPPLE CURRNET
In continuous mode, the drain current of each top N-channel
MOSFET is a square wave of duty cycle VOUT/VIN. A low ESR
input capacitor sized for the maximum RMS current must
be used. The details of a close form equation can be found
in Application Note 77. Figure 17 shows the input capacitor
ripple current for different phase configurations with the
output voltage fixed and input voltage varied. The input
ripple current is normalized against the DC output current.
The graph can be used in place of tedious calculations. The
minimum input ripple current can be achieved when the
product of phase number and output voltage, N(VOUT), is
approximately equal to the input voltage VIN or:
0.6
1-PHASE
2-PHASE
3-PHASE
4-PHASE
6-PHASE
0.4
0.3
0.2
0.1
0
0.1
0.2
0.3 0.4 0.5 0.6 0.7
DUTY FACTOR (VOUT/VIN)
0.8
0.9
3811 F17
Figure 17. Normilized Input RMS Ripple Current vs
Duty Factor for 1 to 6 Output Stages
system. The required amount of input capacitance is further
reduced by the factor, N, due to the effective increase in
the frequency of the current pulses.
The selection of COUT is driven by the required effective
series resistance (ESR). Typically once the ESR requirement has been met, the RMS current rating generally far
exceeds the IRIPPLE(P-P) requirements. The steady-state
output ripple (ΔVOUT) is determined by:
⎡
1 ⎤
ΔVOUT ≈ ΔIRIPPLE ⎢ESR +
⎥
8NfCOUT ⎦
⎣
where f = operating frequency of each stage, N is the
number of phases, COUT = output capacitance and
ΔIRIPPLE = combined inductor ripple currents.
The output ripple varies with input voltage since ΔIL is a
function of input voltage. The output ripple will be less than
50mV at max VIN with ΔIL = 0.4IOUT(MAX)/N assuming:
COUT required ESR < 2N(RSENSE) and
COUT > 1/(8Nf)(RSENSE)
The emergence of very low ESR ceramic capacitors in
small, surface mount packages makes very physically
small implementations possible. The ability to externally
compensate the switching regulator loop using the
LTC3811’s true operational error amplifier allows a much
wider selection of output capacitor types. The ability to
3811f
30
LTC3811
APPLICATIONS INFORMATION
use type III compensation effectively removes constraints
on output capacitor ESR. The impedance characteristics
of each capacitor type are significantly different than an
ideal capacitor and therefore require accurate modeling
and bench evaluation during design.
Manufacturers such as Nichicon, United Chemicon and
Sanyo should be considered for high performance throughhole capacitors. The OS-CON semiconductor dielectric
capacitor available from Sanyo and the Panasonic SP
surface mount types have the lowest (ESR)(size) product
of any aluminum electrolytic at a somewhat higher price.
An additional ceramic capacitor in parallel with OS-CON
type capacitors is recommended to reduce inductance
effects.
In surface mount applications, multiple capacitors may
have to be paralleled to meet the ESR or RMS current
handling requirements of the application. Aluminum
electrolytic and dry tantalum capacitors are both available
in surface mount configurations. New special polymer
surface mount capacitors offer very low ESR also but
have much lower capacitive density per unit volume. In
the case of tantalum, it is critical that the capacitors are
surge tested for use in switching power supplies. Several
excellent choices are the AVX TPS, AVX TPSV or the KEMET
T510 series of surface mount tantalums, available in case
heights ranging from 2mm to 4mm. Other capacitor types
include Sanyo OS-CON, Nichicon PL series and Sprague
595D series. Consult the manufacturer for other specific
recommendations. A combination of capacitors will often
result in maximizing performance and minimizing overall
cost and size.
Tracking and Soft-Start (SS/TRACK Pins)
The start-up of each VOUT is controlled by the voltage on
the respective SS/TRACK pin. When the voltage on the
SS/TRACK pin is less than the internal 0.6V reference,
the LTC3811 regulates the VFB pin voltage to the voltage
on the SS/TRACK pin instead of 0.6V. The SS/TRACK pin
can be used to program an external soft-start function or
to allow VOUT to “track” another supply during start-up.
Soft-start is enabled by simply connecting a capacitor from
the SS/TRACK pin to ground, as shown in Figure 18. An
internal 2.5μA current source charges up the capacitor,
providing a linear ramping voltage at the SS/TRACK pin.
The LTC3811 will regulate the VFB pin (and hence VOUT)
according to the voltage on the SS/TRACK pin, allowing
VOUT to rise smoothly from 0V to its final regulated value.
The total soft-start time will be approximately:
t SS = C SS •
0.6V
2.5μA
Alternatively, the SS/TRACK pin can be used to track two
(or more) supplies during start-up, as shown qualitatively
in Figures 19a and 19b. To do this, a resistor divider
should be connected from the master supply (VX) to the
SS/TRACK pin of the slave supply (VOUT), as shown in
Figure 20. During start-up VOUT will track VX according
to the ratio set by the resistor divider:
VOUT R TRACKA
R A + RB
=
•
VX
RA
R TRACKA + R TRACKB
For coincident tracking (VOUT = VX during start-up),
RA = RTRACKA
RB = RTRACKB
Note that the small SS/TRACK charging current is always
flowing, producing a small offset error. To minimize this
error, select the tracking resistor divider values to be small
enough to make this error negligible.
Topside MOSFET Driver Supply (CB, DB)
External bootstrap capacitors CB connected to the BOOST
and SW pins supply the gate drive voltages for the topside
MOSFETs. Capacitor CB in the Functional Diagram is
charged though external diode DB from DRVCC when the
SW pin is low. When one of the topside MOSFETs is to
be turned on, the driver places the CB voltage across the
gate-source of the desired MOSFET. This enhances the
MOSFET and turns on the topside switch. The switch
LTC3811
SS/TRACK
CSS
SGND
3811 F18
Figure 18. Using the SS/TRACK pin to Program Soft-Start
3811f
31
LTC3811
APPLICATIONS INFORMATION
node voltage, SW, rises to VIN and the BOOST pin follows.
With the topside MOSFET on, the boost voltage is above
the input supply: VBOOST = VIN + VDRVCC. The value of the
boost capacitor CB needs to be 100 times that of the total
input capacitance of the topside MOSFET(s). The reverse
breakdown of the external Schottky diode must be greater
than VIN(MAX). When adjusting the gate drive level, the
final arbiter is the total input current for the regulator. If
OUTPUT VOLTAGE
VX (MASTER)
VOUT (SLAVE)
TIME
3811 F19a
(19a) Coincident Tracking
OUTPUT VOLTAGE
VX (MASTER)
VOUT (SLAVE)
TIME
3811 F19b
(19b) Ratiometric Tracking
Figure 19. Two different Modes of Output Votlage Tracking
Vx VOUT
RB
LTC3811
VFB
RA
RTRACKB
SS/TRACK
RTRACKA
2.5μA
3811 F20
Figure 20. Using the SS/TRACK Pin for Tracking
a change is made and the input current decreases, then
the efficiency has improved. If there is no change in input
current, then there is no change in efficiency.
Overvoltage Protection
The LTC3811 contains a comparator that monitors the
FB pin voltage for potential overvoltage conditions. This
comparator (OV in the Functional Diagram) detects when
the FB pin voltage exceeds 0.66V, or is 10% above nominal
regulation. When this condition is sensed, the top MOSFET
is turned off and the bottom MOSFET is turned on. For an
overvoltage condition that persists, the inductor current will
reverse until the negative current limit of the converter is
reached. If the OV condition terminates VOUT will return to
regulation and normal operation automatically resumes.
The OV signal that controls the top and bottom MOSFET
switching does not propagate through the PGOOD filter
before action is taken. The OV comparator is capable of
sensing a fault condition within 100ns to 200ns, after
which the top MOSFET is turned off. The PGOOD filter
will delay the signal to the open-drain NMOS transistor
connected to the PGOOD pin, however, preventing OV (and
UV) transients of less than about 130μs from forcing a
system reset.
Phase-Locked Loop and Frequency Synchronization
The LTC3811 has a phase-locked loop (PLL) comprised of
an internal voltage-controlled oscillator (VCO) and a phase
detector. This allows the turn-on of the top MOSFET of
controller 1 to be locked to the rising edge of an external
clock signal applied to the MODE/SYNC pin. The turn-on
phase of controller 2’s top MOSFET is controlled by the
voltage on the PHASEMODE pin. The phase detector is
an edge sensitive digital type that provides zero degrees
phase shift between the external and internal oscillators.
This type of phase detector does not exhibit false lock to
harmonics of the external clock.
The output of the phase detector is a pair of complementary current sources that charge or discharge the
external filter network connected to the PLL/LPF pin. The
relationship between the voltage on the PLL/LPF pin and
operating frequency, when there is a clock signal applied
3811f
32
LTC3811
APPLICATIONS INFORMATION
to MODE/SYNC, is shown in Figure 21 and specified in the
Electrical Characteristics table. Note that the LTC3811 can
only be synchronized to an external clock whose frequency
is within range of the LTC3811’s internal VCO, which is
nominally 125kHz to 1.1MHz. This is guaranteed to be
between 175kHz and 900kHz. A simplified block diagram
is shown in Figure 22.
If the external clock frequency is greater than the internal
oscillator’s frequency, fOSC, then current is sourced continuously from the phase detector output, pulling up the
PLL/LPF pin. When the external clock frequency is less
than fOSC, current is sunk continuously, pulling down
the PLL/LPF pin. If the external and internal frequencies
are the same but exhibit a phase difference, the current
sources turn on for an amount of time corresponding to
the phase difference. The voltage on the PLL/LPF pin is
adjusted until the phase and frequency of the internal and
external oscillators are identical. At the stable operating
point, the phase detector output is high impedance and
the filter capacitor CLP holds the voltage.
The loop filter components, CLP and RLP, smooth out
the current pulses from the phase detector and provide a
stable input to the voltage-controlled oscillator. The filter
components CLP and RLP determine how fast the loop
acquires lock. Typically RLP = 10k and CLP is 2200pF
to 0.01μF.
1200
1000
900
FREQUENCY (kHz)
RLP
CLP
MODE/
SYNC
EXTERNAL
OSCILLATOR
PLL/LPF
DIGITAL
PHASE/
FREQUENCY
DETECTOR
OSCILLATOR
3811 F22
Figure 22. Phase-Locked Loop Block Diagram
Typically, the external clock (on MODE/SYNC pin) input
high threshold is 1.1V, while the input low threshold
is 1.0V.
Efficiency Considerations
The percent efficiency of a switching regulator is equal to
the output power divided by the input power times 100%.
It is often useful to analyze individual losses to determine
what is limiting the efficiency and which change would
produce the most improvement. Percent efficiency can
be expressed as:
%Efficiency = 100% – (L1 + L2 + L3 + ...)
where L1, L2, etc. are the individual losses as a percentage of input power.
Although all dissipative elements in the circuit produce
losses, four main sources usually account for most of the
losses in LTC3811 circuits: 1) IC VIN current, 2) DRVCC
regulator current, 3) I2R losses, 4) Topside MOSFET
transition losses.
1100
800
700
600
500
1. The VIN current has two components: the first is the
DC supply current given in the Electrical Characteristics
table, and the second is the MOSFET driver and control
currents.
400
300
200
100
0
2.4V
0
0.5
1.5
2
1
PLL/LPF PIN VOLTAGE (V)
2.5
3811 G36
Figure 21. Relationship Between Oscillator Frequency
and Voltage at the PLL/LPF Pin When Synchronizing to
an External Clock
2. DRVCC current is the sum of the MOSFET driver and
control currents. The MOSFET driver current results
from switching the gate capacitance of the power
MOSFETs. Each time a MOSFET gate is switched
from low to high to low again, a packet of charge dQ
3811f
33
LTC3811
APPLICATIONS INFORMATION
moves from DRVCC to ground. The resulting dQ/dt is
a current out of DRVCC that is typically much larger
than the control circuit current. In continuous mode,
IGATECHG = f(QT + QB), where QT and QB are the gate
charges of the topside and bottom side MOSFETs.
Supplying DRVCC and INTVCC power through the
EXTVCC switch input from an output-derived source
will scale the VIN current required for the driver and
control circuits by a factor of (Duty Cycle)/(Efficiency).
For example, in a 20V input to 2.5V output application,
40mA of DRVCC current results in approximately 5mA
of VIN current. This reduces the mid-current efficiency
loss from 10% or more (if the driver was powered
directly from VIN) to only a few percent.
3. I2R losses are predicted from the DC resistances of
the fuse (if used), MOSFET, inductor, current sense
resistor, and input and output capacitor ESR. In
continuous mode the average output current flows
through L and RSENSE, but is “chopped” between the
topside MOSFET and the synchronous MOSFET. If the
two MOSFETs have approximately the same RDS(ON),
then the resistance of one MOSFET can simply be
summed with the resistances of L, RSENSE and ESR to
obtain I2R losses. For example, if each RDS(ON) = 5mΩ,
RL = 1mΩ, RSENSE = 1.5mΩ and RESR = 4mΩ (sum
of both input and output capacitance losses), then the
total resistance is 16mΩ. This results in losses ranging
from 5.6% to 8.4% as the output current increases
from 10A to 15A for a 2.5V output. Efficiency varies
as the inverse square of VOUT for the same external
components and output power level. The combined
effects of increasingly lower output voltages and higher
currents required by high performance digital systems
is not doubling but quadrupling the importance of loss
terms in the switching regulator system!
4. Transition losses apply only to the topside MOSFET(s),
and become significant only when operating at high
input voltages (typically 15V or greater). Transition
losses can be estimated from:
Translation Loss = (VIN)
2 IMAX
• RDR • CMILLER
2
Other “hidden” losses such as copper trace and internal
battery resistances can account for an additional 5% to
10% efficiency degradation in portable systems. It is very
important to include these “system” level losses during
the design phase. The internal battery and fuse resistance
losses can be minimized by making sure that CIN has adequate charge storage and very low ESR at the switching
frequency. A 25W supply will typically require a minimum of
20μF to 40μF of capacitance having a maximum of 20mΩ to
50mΩ of ESR. The LTC3811 2-phase architecture typically
halves this input capacitance requirement over competing
solutions. Other losses including Schottky conduction
losses during dead-time and inductor core losses generally
account for less than 1% total additional loss.
Feedback Loop Compensation
The LTC3811 incorporates a peak current mode control
topology. Peak current mode control provides excellent
line and load transient response, and inherently provides
the best possible phase-to-phase current sharing in
multiphase applications.
The LTC3811 incorporates a true operational error amplifier in the feedback loop, enabling the user the flexibility
to place poles and zeros at well defined frequencies in
the transfer function, thereby optimizing the loop’s AC
response.
The control-to-output transfer function has a pole at the
origin in order to provide DC regulation, and a pole due
to the load resistance and capacitance at:
fP(LOAD) =
1
2π • RL • CL
The output decoupling capacitor ESR contributes a zero
to the transfer function at:
fZ(ESR) =
1
2π • ESR • CL
The transfer function also has a mathematical double pole at
half the switching frequency due to the sampling nature of
current mode control, although the pole-splitting behavior
of the LTC3811’s internal slope compensation reduces the
phase shift for frequencies below fSW/2.
For most systems, the simple 2-pole, single-zero response
of a Type-II compensation network (shown in Figure 23)
3811f
34
LTC3811
APPLICATIONS INFORMATION
will provide adequate phase margin at the unity-gain
frequency of the loop.
In a Type-II compensation scheme, the zero is typically
placed below the target unity-gain frequency, depending
upon the desired settling time of the converter, and the pole
is placed no higher than half the switching frequency in
order to attenuate the switching frequency from the loop.
The gain between the zero and pole is typically adjusted
until the desired phase margin is achieved.
In general, the output capacitor is chosen based on cost
and size considerations, given a certain error budget
due to output ripple voltage and load transient response.
Oftentimes, multiple capacitor types (such as ceramic
and special polymer) are connected in parallel in order
to achieve a good combination of bulk capacitance and
low ESR. In general, the output capacitor is not normally
chosen to optimize the bode response.
Due to their small case size and low ESR, ceramic output
capacitors are well suited to very low voltage, high current
applications. Their low ESR and relatively high RMS current
capability make them a good choice for today’s demanding processor-based loads. A fully ceramic output stage,
however, will result in very low ESR, pushing the ESR zero
frequency relatively close to the unity-gain frequency of the
loop. In this case a Type-III compensation network using
3 poles and 2 zeros may be necessary (see Figure 24). For
particularly demanding applications requirements, please
consult Linear Technology’s Applications department.
C1
VOUT
f P1 =
f P2 =
GAIN (dB)
fZ =
fZ
+
EA
–20dB
DECADE
+
0.6V
1
2π • R2 (C1+ C2)
fP1 =
1
⎛ C1• C2 ⎞
2π • R3 ⎜
⎝ C1+ C2 ⎟⎠
fP2 =
1
2π • R3 • C2
fP3 =
fP2
AV = 20 log •
R3
R1
1
2π • R2 (C1+ C2)
1
⎛ C1• C2 ⎞
2π • R3 ⎜
⎝ C1+ C2 ⎟⎠
EA
f Z1 =
1
2 π • R3 • C2
f Z2 =
1
2 π • (R2 + R4) • C3
1
2 π • R4 • C3
fZ1
30
fZ2
fP2
fP3
20
10
–20dB/DECADE
C2
–
R1
20
10
R3
C3
R2
GAIN (dB)
0.6V
30
R4
–
R1
C1
C2
R3
R2
VOUT
f
–20dB/DECADE
–20dB
DECADE
f
f
PHASE (DEG)
PHASE (DEG)
90
–45
–90
3811 F23
Figure 23. Type-II Compensation Network and
Frequency Response
45
f
0
–45
–90
3811 F24
Figure 24. Type-III Compensation Network and
Frequency Response
3811f
35
LTC3811
APPLICATIONS INFORMATION
Measuring the Loop’s Transient Response
Once the compensation components have been chosen, the
AC performance of the power supply should be verified in
the lab. The two most common ways of checking the AC
response of the circuit are with load and line steps, and
by measuring the loop gain using a network analyzer or
Venable measurement system. Both of these measurement
techniques should be performed on the final design to ensure adequate correlation between the two, and to identify
and correct potential regions of marginal stability. These
measurements should be performed over all of the load,
line, temperature and components tolerance variations the
system will experience in a practical application.
Figure 25 illustrates a typical load step response for the
LTC3811. When a positive load step occurs, the output
voltage immediately drops by ΔILOAD • ESR, where ESR
is the equivalent series resistance of the output capacitor.
The increased load current then begins to discharge the
output capacitor, generating a feedback error signal that
forces the regulator to adapt to the current change and
return VOUT to its steady-state regulated value. During
this recovery time VOUT can be monitored for excessive
overshoot or ringing which would indicate a stability
problem. Assuming a second order system, the phase
margin and/or damping factor can be estimated using the
percentage overshoot seen at the output.
An output current pulse of 20% to 100% of full load having
a rise time of 0.1μs to 1μs will produce an output voltage
waveform that will give an indication of the loop stabil-
ity without having to break the feedback loop. Placing a
power MOSFET directly across the output capacitor and
driving the gate with an appropriate signal generator or
gate driver is a practical way to produce a realistic load
step condition.
Voltage Positioning for Single Output,
Multiphase Applications
The output voltage load line can be programmed with the
LTC3811 using one external resistor, allowing the user to
reduce the total output capacitance required for a given
error budget. The inductor current information is sensed
using the SENSE+ and SENSE– inputs for both channels
and fed into a transconductance amplifier with two input
stages, as shown in Figure 26. The output current of the
transconductance amplifier, along with one external resistor (RAVP), allows the user to inject a load-current-related
error signal into the voltage feedback loop. Please note that
because the gm amplifier mixes the signals from both channels, voltage positioning is only possible for multiphase,
single output applications; dual output applications with
voltage positioning are not possible.
The internal mixing of the current sense signals within
the voltage positioning amplifier, combined with the fact
that the gm amplifier output signal is a current, allows the
user to connect the CSOUT pins of several LTC3811 chips
together in multiphase applications. The transconductance
(gm) of the voltage positioning amplifier is 2.5mS/phase,
and the load slope is:
R
⎤
⎡ R2⎤ ⎡
VOUT = 0.6 • ⎢1+ ⎥ – ⎢IOUT • SENSE • 5m • R AVP ⎥
n
⎣ R1⎦ ⎣
⎦
VOUT
50mV/DIV
AC COUPLED
where n is the number of phases.
IL
5A/DIV
VIN = 12V
20μs/DIV
VOUT = 1.5V
ILOAD = 0.5A TO 8A
3811 G02
Figure 25. Load Step Response for the
LTC3811 Circuit in Figure 33
The input common mode range of the voltage position
gm amplifier is 0.6V to 3.5V, comfortably allowing output
voltages up to 3.3V. In addition, the output voltage range
of the gm amplifier for linear operation is limited to voltages above 0.6V, due to the headroom requirements of
the NMOS sink transistors in the output stage. And finally,
the maximum differential input voltage for linear operation
is ±100mV.
3811f
36
LTC3811
APPLICATIONS INFORMATION
L1
L2
VOUT
50mV/DIV
AC COUPLED
SENSE1+
SENSE2+
RS1
RS2
SENSE1–
IL1
5A/DIV
+
gm
–
IL2
5A/DIV
CSAMP
SENSE2–
VIN = 12V
50μs/DIV
VOUT = 1.5V (2-PHASE)
ILOAD = 0A TO 15A
CSOUT
RAVP
DIFFIN+ 80k
80k
DIFFIN– 80k
DIFFOUT
80k
SGND
+
–
COUT
DIFFAMP
COMP
Figure 27. Load Step Response for the Circuit in Figure 30
To program 500kHz operation, float the PLL/LPF pin.
The inductor value in this design is chosen assuming
50% ripple current. The highest ripple current occurs at
maximum input voltage.
C2
R2
R3
C1
FB
0.6V
+
–
EA
L=
3811 F26
R1
VOUT
f • ΔIL
⎤
⎡
V
• ⎢1– OUT ⎥
⎢⎣ VIN(MAX) ⎥⎦
1.5V
⎡ 1.5V ⎤
• ⎢1–
= 0.36μH
500kHz • 0.5 • 15A ⎣ 14V ⎥⎦
A 0.4μH inductor will result in 6.7A of ripple current, or
45%. Assuming a value for the current limit value 30%
greater than the maximum load current, then IMAX = 1.3 •
15A = 19.5A. The peak inductor current will be the maximum DC value plus one half the ripple current, or:
=
SGND
Figure 26. Simplified LTC3811 Voltage Positionig Block Diagram
Figure 27 illustrates the load step response for the circuit
in Figure 30, with voltage positioning.
A Design Example
As a design example consider one channel of a 2-phase
single output supply. Assume VIN = 4.5V to 14V, VOUT =
1.5V, IMAX = 30A(15A per phase) and f = 500kHz.
In order to achieve the best output accuracy, 1% resistors
(or better) should be used in the divider that programs the
output voltage. Choosing 600μA for the divider current,
R2 = 1.5k and R1 = 1k. The nominal output voltage will
therefore be 1.50V.
For the input and output conditions given above, the steady
state minimum on-time for this application at VIN = 14V
will be approximately:
tON(MIN) =
3811 G03
VOUT
VIN(MAX ) • f
=
1.5V
= 214ns
14V • 500kHz
IL(PK ) = 19.5A +
1
• 6.7 A = 22.9 A
2
This represents the minimum saturation current rating for
the inductor. For this application, a Vitec 59P9875 inductor
was chosen. This inductor has a room temp saturation
current rating of 23A and a DCR of 0.32mΩ.
To maintain good cycle-by-cycle control of the inductor
current and still have good efficiency, a 1.5mΩ, 1W sense
resistor from Panasonic (ERJ-M1WTJ1M5U) connected
in series with the inductor is used for current sensing.
With a maximum peak inductor current of 22.9A, the peak
sense voltage will be:
VSENSE(PK) = 22.9A • 0.0015Ω = 34.4mV
3811f
37
LTC3811
APPLICATIONS INFORMATION
The maximum power dissipation in the sense resistor
will be:
PR(SENSE) = 22.9A2 • 0.0015Ω = 0.79W
To ensure that the maximum current can be delivered
over all of the power component and IC tolerances, the
maximum sense voltage for the LTC3811 is chosen to be
50mV. This is programmed by connecting the RNG pin
to INTVCC.
Due to the use of a 400nH inductor and 500kHz operation, the magnitude of the inductive voltage drop across
the sense resistor should be calculated and compared to
the maximum sense voltage (50mV). First calculate the
nominal switch on-time:
tON =
VOUT
1.5V
=
= 250ns
VIN • f 12V • 500kHz
The inductor ΔIL/dt is therefore:
The power MOSFETs chosen for this application are the
Renesas RJK0305DPB (top) and RJK0301DPB (bottom).
The upper MOSFET, which is optimized for low switching
losses, has a typical RDS(ON) of 10mΩ at VGS = 4.5V, a
total gate charge of 8nC, and a minimum BVDSS of 30V.
The bottom MOSFET, which is zero-voltage switched and
is optimized for low on-resistance, has a typical RDS(ON)
of 3mΩ at VGS = 4.5V, a total gate charge of 32nC, and a
minimum BVDSS of 20V.
From the datasheet of the RJK0305DPB upper MOSFET,
the Miller capacitance is calculated to be:
CMILLER =
Assuming a top MOSFET junction temperature of 75°C,
δ = 0.25 and the power dissipated in this MOSFET is:
PMAIN =
ΔIL
6.7A 26.8A
=
=
dt 250ns
μs
The Panasonic sense resistor has a typical parasitic series
inductance (ESL) of 0.5nH, meaning that the inductive
voltage drop across the resistor is:
VL(SENSE) = ESL •
ΔIL
26.8A
= 0.5nH •
= 13.4mV
dt
μs
The ESL/R time constant for the sense resistor is
therefore:
τ=
ESL
RSENSE
ΔQG 2nC
=
= 167pF
ΔVDS 12V
VOUT
I
• IMAX 2 • RDS(ON) • (1+ δ) + VIN 2 • MAX T
VIN
2
⎡
1
1 ⎤
• RDR • CMILLER • ⎢
+
⎥•f
⎢⎣ VINTVCC – VTH(MIN) VTH(MIN) ⎥⎦
PMAIN =
1.5V
15A
• 15A 2 • 0.01• (1+ 0.25) + 12V 2 •
12V
2
1⎤
⎡ 1
• 2Ω • 167pF • ⎢
+ ⎥ • 500kHz
6V
–
1V
1V
⎣
⎦
PMAIN = 0.351W + 0.216W = 0.567W
For the synchronous MOSFET the power dissipation is:
= 333ns
The sense pins need an RC filter with the same time constant
in order for the waveform at the SENSE+ and SENSE– pin
to accurately represent the inductor current. Choosing a
value of 1000pF for the filter capacitor, the total resistance
should therefore be 333Ω. Split between the SENSE+
and SENSE– pins, each resistor should be 165Ω. These
components should be placed adjacent to the SENSE+ and
SENSE– pins on the LTC3811, and the PCB traces from
the 165Ω filter resistors should be minimum width and
run parallel to each other all the way to the sense resistor
location on the board.
PSYNC =
VIN – VOUT
• IMAX 2 • RDS(ON) • (1+ δ)
VIN
PMAIN =
12V – 1.5V
• 15A 2 • 0.003 • (1+ 0.25)
12V
= 0.738W
To determine the RMS current rating of the input capacitor,
we need to first determine the minimum and maximum
duty cycle. For an output voltage of 1.5V and an input range
of 4.5V to 14V, the duty cycle range is 12.5% to 33.3%.
We then use Figure 17 to determine the percentage of
3811f
38
LTC3811
APPLICATIONS INFORMATION
the maximum load current that represents the minimum
RMS current rating of the input capacitor. The worst-case
condition occurs when only one output is operational. The
output with the highest (VOUT)(IOUT) product should be
used to determine the minimum RMS current rating of the
input capacitor. From Figure 17 we can see that the worst
case condition for this output occurs at the maximum duty
cycle of 33.3%, and that the minimum RMS current rating
of the input capacitor needs to exceed 7A (47% of 15A).
The ceramic output capacitors chosen have an effective
ESR of 5mΩ and a bulk capacitance of 660μF. The peakto-peak output ripple for this configuration is:
⎡
⎤
1
ΔVOUT = ΔIL • ⎢ESR +
⎥
8 • n • f • COUT ⎦
⎣
1
⎡
⎤
= 6.7A • ⎢0.005Ω +
8 • 2 • 500kHz • 660μF ⎥⎦
⎣
ΔVOUT = 33.5mV + 1.3mV = 34.8mV
This represents a ripple voltage of 2.3%. As can be seen
from the calculation, the biggest portion of the output ripple
comes from the ESR of the capacitor. This is why low ESR
ceramic capacitors are so important in low voltage, high
current applications.
PC Board Layout Checklist
When laying out the printed circuit board, the following
checklist should be used to ensure proper operation of
the converter:
1. The connection between the SGND pin on the LTC3811
and all of the small-signal components surrounding
the IC should be isolated from the power ground and
PGND pin, and should be Kelvin-connected to the main
ground node near the bottom terminal of the output
capacitors.
2. Place the small-signal components away from high
frequency switching nodes on the board. The pinout of
the LTC3811 was carefully designed in order to make
component placement as easy and noise free as possible. All of the power components can be placed on
one side of the IC, away from all of the small-signal
components.
3. The bottom terminals of the input and output capacitors
should be placed as close as possible to one another,
with the small-signal ground connection in between
them. This component arrangement will reduce differential mode noise due to the two high di/dt loops
in the power circuit.
4. If the output capacitor is located far away from the
IC and the remote sense differential amplifier is being
used to level-shift the output voltage back to the local
IC ground, the small-signal ground around the LTC3811
should be Kelvin-connected to the main ground node
near the bottom terminal of the input capacitor.
5. The PGND pin should be connected to the sources
of the bottom MOSFETs using a wide, short trace on
the top layer of the board. The MOSFETs should also
be placed on the top layer of the board. The exposed
area on the bottom of the QFN package is internally
connected to the PGND node of the IC.
6. Place the INTVCC analog supply decoupling capacitor
and resistor right next to the INTVCC and SGND pins
on the same layer as the IC. A low ESR 0.1μF to 1μF
ceramic capacitor should be used.
7. Place the DRVCC gate driver supply decoupling capacitor right next the DRVCC and PGND pins, on the same
layer as the IC. This capacitor carries high di/dt MOSFET
gate drive currents. A low ESR (X5R or better) 4.7μF
to 10μF ceramic capacitor should be used.
8. The floating gate driver supply decoupling capacitor
should be located right next the BOOST and SW pins,
on the same layer as the IC. This capacitor carries high
di/dt currents to drive the upper power MOSFETs. A
low ESR (X5R or better) ceramic capacitor at least 100
times the total input capacitance of the upper power
MOSFETs for that channel should be used.
9. The resistor divider connected to the FB pin to program
the output voltage should be located as near as possible to the IC, with the bottom resistor connecting to
the isolated signal ground node near the SGND pin.
The PCB trace connecting the top resistor to the upper
terminal of the output capacitor should avoid any high
3811f
39
LTC3811
APPLICATIONS INFORMATION
frequency switching nodes in the circuit and should
ideally be shielded (both laterally and vertically) by
ground planes.
especially from the opposite channel’s voltage- and
current-sensing feedback signals. The SW, TG and
BOOST nodes can have slew rates in excess of 1V/ns
relative to ground and should therefore be kept on
the “output side” of the LTC3811.
10. If the differential remote sense amplifier is being used,
the PCB traces connecting DIFF/IN+ and DIFF/IN– to
the output capacitor should avoid any high frequency
switching nodes in the circuit and should ideally be
shielded (both laterally and vertically) by ground
planes. In addition, the DIFF/IN+ and DIFF/IN– PCB
traces should be routed parallel to one another with
minimum spacing in between. Due to the 160kΩ
input impedance of these pins, it is critical that these
traces avoid any high frequency switching nodes in
the circuit.
16. Check the stress on the power MOSFETs by independently measuring the drain-to-source voltages directly
across the devices terminals. Beware of inductive
ringing that could exceed the maximum voltage rating
of the MOSFET. If this ringing cannot be avoided and
exceeds the maximum rating of the device, choose a
higher voltage rated MOSFET.
11. The high di/dt loops formed by the input capacitor
and the power MOSFETs should be kept as small as
possible to avoid EMI and differential mode switching
noise. The upper power MOSFETs should be located
close to one another and as close as possible to the
positive terminal of the input decoupling capacitor. Do
not attempt to split the input decoupling for the two
channels as it can cause a large resonant loop.
18. Minimize the capacitive load on the CLKOUT pin to
minimize excess phase shift. Buffer the CLKOUT signal
with an emitter follower if necessary.
12. The bottom MOSFETs sources should also be located
close to one another and as close as possible to the
negative terminal of the input capacitor. Since the
inductor can be modeled as a current source, its
placement on the board is less critical than the high
di/dt components.
13. The switch node area should be kept small, with
the upper power MOSFET sources and lower power
MOSFET drains in close proximity.
14. The filter capacitor between the SENSE+ and SENSE–
pins, as well as the filter resistors, should be located
as close as possible to the IC. In addition, the connections between the SENSE+ and SENSE– filter resistors
and the sense resistor should be routed parallel to one
another with minimum spacing in between. These
traces should avoid any high frequency switching
nodes in the circuit.
17. When synchronizing the LTC3811 to an external clock,
use a low impedance source such as a logic gate to
drive the MODE/SYNC pin and keep the lead as short
as possible.
The diagram in Figure 28 illustrates all the branch currents in
a 2-phase single output switching regulator. After studying
the waveforms it is clear why it is critical to reduce the area
of the high dV/dt nodes as much as possible. High electric
and magnetic fields will radiate from these “loops,” just as
a radio station broadcasts a signal. The output capacitor
ground should return to the negative terminal of the input
capacitor and not share a common ground path with any
switched current paths. The left half of the circuit gives
rise to the “noise” generated by the switching regulator.
The ground terminations of the synchronous MOSFETs and
Schottky optional diodes should return to the bottom plate
of the input capacitor with a short, isolated PC trace since
very high di/dt currents are present. A separate, isolated
path from the negative terminals of the input and output
capacitors should be used to connect the IC signal ground
pin (SGND). This technique keeps inherent signals generated by the high di/dt current pulses from taking alternate
current paths that have finite impedances during the total
period of the switching regulator.
15. Keep the switch nodes (SW1, SW2), the top gate
nodes (TG1, TG2) and the boost nodes (BOOST1,
BOOST2) away from sensitive small-signal nodes,
3811f
40
LTC3811
APPLICATIONS INFORMATION
PC Board Layout Debugging
Start with one controller on at a time. It is helpful to use
a DC-50MHz current probe to monitor the current in the
inductor while testing the circuit. Monitor the output
switching node (SW pin) to synchronize the oscilloscope
to the internal oscillator and probe the actual output voltage
as well. Check for proper performance over the operating
voltage and current range expected in the application. The
frequency of operation should be maintained over the
input voltage range.
The duty cycle percentage should be maintained from cycle
to cycle in a well-designed, low noise PCB implementation.
SW1
Variation in the duty cycle at a subharmonic rate can suggest noise pickup at the current or voltage sensing inputs
or inadequate loop compensation. Overcompensation of
the loop can be used to tame a poor PC layout if regulator bandwidth optimization is not required. Only after
each controller is checked for its individual performance
should both controllers be turned on at the same time.
A particularly difficult region of operation is when one
controller channel is nearing its current comparator trip
point when the other channel is turning on its top MOSFET.
This occurs around 50% duty cycle on either channel due
to the phasing of the internal clocks and may cause minor
duty cycle jitter.
L1
RSENSE1
D1
VIN
VOUT
RIN
CIN
+
+
SW2
BOLD LINES INDICATE
HIGH, SWITCHING
CURRENT LINES.
KEEP LINES TO A
MINIMUM LENGTH.
L2
COUT
RL
RSENSE2
D2
3811 F28
Figure 28. Instantaneous Current Path Flow in a Multiple Phase Switching Regulator
3811f
41
LTC3811
APPLICATIONS INFORMATION
Reduce VIN from its nominal level to verify operation of
the regulator at low VIN. Check the operation of the under-voltage lockout circuit by further lowering VIN while
monitoring the outputs to verify operation.
Investigate whether any problems exist only at higher output currents or only at higher input voltages. If problems
coincide with high input voltages and low output currents,
look for capacitive coupling between the BOOST, SW, TG,
and possibly BG connections and the sensitive voltage
and current pins. The capacitor placed across the current
sensing pins needs to be placed immediately adjacent to
the pins of the IC. This capacitor helps to minimize the
effects of differential noise injection due to high frequency
capacitive coupling. If problems are encountered with
high current output loading at lower input voltages, look
for inductive coupling between CIN, Schottky and the top
MOSFET components to the sensitive current and voltage
sensing traces. In addition, investigate common ground
path voltage pickup between these components and the
SGND pin of the IC.
An embarrassing problem, which can be missed in an
otherwise properly working switching regulator, results
when the current sensing leads are hooked up backwards.
The output voltage under this improper hookup will still
be maintained but the advantages of current mode control
will not be realized. Compensation of the voltage loop will
be much more sensitive to component selection. This
behavior can be investigated by temporarily shorting out
the current sensing resistor—don’t worry, the regulator
will still maintain control of the output voltage.
3811f
42
LTC3811
TYPICAL APPLICATIONS
1Ω
47μF
X5R
0.1μF
BAT54
RJK0305DPB
EXTVCC VIN INTVCC DRVCC
TG1
0.1μF
SW2
BG1
1.21k
BG2
CLKOUT
PGND
PLL/LPF
PGOOD1
MODE/SYNC
PGOOD2
PHASMODE
0.4μH
0.5mΩ
CS/OUT
RJK0301DPB
100k
1.21k
SENSE1+
SENSE2+
SENSE1–
SENSE2–
RNG2
DIFF/IN+
RNG1
DIFF/IN–
11.8k
1%
680nF
330μF
×2
47μF
X5R
4.99k
1%
100pF
DIFF/OUT
FB2
FB1
1.21k
VOUT2
1.5V
15A
33pF
COMP2
COMP1
56k
0.4μH
0.5mΩ
7.5k
1%
RUN1
33pF
1.21k
100k
RUN2
VOUT1
2V
15A
INTVCC
LTC3811
680nF
RJK0305DPB
0.1μF
BOOST2
SW1
RJK0301DPB
VIN
4.5V TO 14.5
10μF
X5R
BAT54
TG2
BOOST1
330μF
16V
SS/TRACK1 SGND SS/TRACK2
5nF
100pF
56k
4.99k
1%
47μF
X5R
330μF
×2
3811 F29
2.43k
1.21k
Figure 29. High Efficiency Core-I/O Power Supply with Differential Remote Sensing and Tracking
3811f
43
LTC3811
TYPICAL APPLICATIONS
1μF
AUX 5V GATE
DRIVE SUPPLY
100k
TO SUPPLY
MONITOR
5nF
1.0Ω
EXTVCC VIN INTVCC DRVCC
BOOST1
PGOOD1
PGOOD2
TG1
SS/TRACK1
SW1
RNG1
PGND
RNG2
SENSE1+
MODE/SYNC
LTC3811
SENSE1–
BOOST2
PLL/LPF
TG2
PHASEMODE
SW2
SGND
BG2
SENSE2+
SENSE2–
3.32kΩ
1%
INTVCC
33pF
4.99k
1%
100pF
285k
10μF
X5R
RJK0305DPB
0.4μH
FB2
CSOUT
FB1
DIFF/IN+
COMP1
DIFF/IN–
COMP2
DIFF/OUT
0.0015Ω
RJK0301DPB
50Ω
2200pF
50Ω
0.33μF
RJK0305DPB
0.4μH
0.0015Ω
RJK0301DPB
RUN1
RUN2
BAT54
BG1
SS/TRACK2
CLKOUT
BAT54
0.33μF
VIN
12V
100μF
X5R
×4
50Ω
100μF
X5R
330μF
×6
VOUT
1V
30A
2200pF
50Ω
500Ω
3811 F30
100pF
Figure 30. 2-Phase, 12V Input, 1V/30A Output ASIC Supply with Voltage Positioning
3811f
44
LTC3811
TYPICAL APPLICATIONS
1μF
AUX 5V GATE
DRIVE SUPPLY
100k
TO SUPPLY
MONITOR
5nF
1.0Ω
EXTVCC VIN INTVCC DRVCC
PGOOD1
BOOST1
PGOOD2
TG1
SS/TRACK1
SW1
SS/TRACK2
BG1
RNG1
MODE/SYNC
SENSE1+
SENSE1
PHASEMODE
–
BOOST2
FB2
TG2
PLL/LPF
SW2
CLKOUT
BG2
RUN1
BAT54
10μF
X5R
RJK0305DPB
0.4μH
0.0015Ω
RJK0301DPB
50Ω
2200pF
50Ω
0.33μF
RJK0305DPB
0.4μH
0.0015Ω
SENSE2+
SENSE2–
7.5k
1%
VOUT1
1.5V
45A
RJK0301DPB
200μF
X5R
CSOUT
RUN2
50Ω
2000μF
2200pF
50Ω
100pF
DIFF/IN+
FB1
33pF
4.99k
1%
BAT54
0.33μF
470μF
PGND
LTC3811
RNG2
OUTPUT 1
ON/OFF
CONTROL
VIN
12V
47μF
X5R
DIFF/IN–
COMP1
285k
COMP2
SGND
DIFF/OUT
100pF
1μF
AUX 5V GATE
DRIVE SUPPLY
100k
TO SUPPLY
MONITOR
1.0Ω
EXTVCC VIN INTVCC DRVCC
PGOOD2
PGOOD1
TG1
RNG1
SW1
FB1
10k
BAT54
10μF
X5R
RJK0305DPB
0.4μH
PGND
SENSE1+
SS/TRACK2
SENSE1–
SS/TRACK1
BOOST2
CLKOUT
TG2
PLL/LPF
SW2
MODE/SYNC
0.0015Ω
RJK0301DPB
BG1
LTC3811
PHASEMODE
10nF
BAT54
0.33μF
BOOST1
RNG2
5nF
22μF
X5R
50Ω
2200pF
50Ω
0.1μF
RJK0305DPB
0.4μH
0.0015Ω
RJK0301DPB
BG2
100μF
RUN1
CSOUT
COMP1
SENSE2+
RUN2
SENSE2–
FB2
DIFF/OUT
22μF
X5R
VOUT2
3.3V
5A
50Ω
2200pF
50Ω
22.6k
1%
33pF
100pF
4.99k
28.5k
DIFF/IN+
COMP2
SGND
DIFF/IN–
Figure 31. 3-Phase, 12V Input, 1.5V/45A Output with a 3.3V/5A Auxiliary Output
3811f
45
LTC3811
PACKAGE DESCRIPTION
UHF Package
38-Lead Plastic QFN (5mm × 7mm)
(Reference LTC DWG # 05-08-1701)
0.70 ± 0.05
5.50 ± 0.05
(2 SIDES)
4.10 ± 0.05
(2 SIDES)
3.15 ± 0.05
(2 SIDES)
PACKAGE
OUTLINE
0.25 ± 0.05
0.50 BSC
5.15 ± 0.05 (2 SIDES)
6.10 ± 0.05 (2 SIDES)
7.50 ± 0.05 (2 SIDES)
RECOMMENDED SOLDER PAD LAYOUT
5.00 ± 0.10
(2 SIDES)
3.15 ± 0.10
(2 SIDES)
0.75 ± 0.05
0.00 – 0.05
PIN 1 NOTCH
R = 0.30 TYP OR
0.35 × 45° CHAMFER
37 38
0.40 ±0.10
PIN 1
TOP MARK
(SEE NOTE 6)
1
2
5.15 ± 0.10
(2 SIDES)
7.00 ± 0.10
(2 SIDES)
0.40 ± 0.10
0.200 REF 0.25 ± 0.05
0.200 REF
0.00 – 0.05
0.75 ± 0.05
NOTE:
1. DRAWING CONFORMS TO JEDEC PACKAGE
OUTLINE M0-220 VARIATION WHKD
2. DRAWING NOT TO SCALE
3. ALL DIMENSIONS ARE IN MILLIMETERS
0.50 BSC
R = 0.115
TYP
(UH) QFN 0205
BOTTOM VIEW—EXPOSED PAD
4. DIMENSIONS OF EXPOSED PAD ON BOTTOM OF PACKAGE DO NOT INCLUDE
MOLD FLASH. MOLD FLASH, IF PRESENT, SHALL NOT EXCEED 0.20mm ON ANY SIDE
5. EXPOSED PAD SHALL BE SOLDER PLATED
6. SHADED AREA IS ONLY A REFERENCE FOR PIN 1 LOCATION
ON THE TOP AND BOTTOM OF PACKAGE
3811f
46
LTC3811
PACKAGE DESCRIPTION
G Package
36-Lead Plastic SSOP (5.3mm)
(Reference LTC DWG # 05-08-1640)
12.50 – 13.10*
(.492 – .516)
1.25 ±0.12
7.8 – 8.2
36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19
5.3 – 5.7
7.40 – 8
(.291 – .
0.42 ±0.03
0.65 BSC
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
RECOMMENDED SOLDER PAD LAYOUT
2.0
(.079)
MAX
5.00 – 5.60**
(.197 – .221)
0° – 8°
0.09 – 0.25
(.0035 – .010)
0.55 – 0.95
(.022 – .037)
NOTE:
1. CONTROLLING DIMENSION: MILLIMETERS
MILLIMETERS
2. DIMENSIONS ARE IN
(INCHES)
0.65
(.0256)
BSC
0.22 – 0.38
(.009 – .015)
TYP
0.05
(.002)
MIN
G36 SSOP 0204
3. DRAWING NOT TO SCALE
*DIMENSIONS DO NOT INCLUDE MOLD FLASH. MOLD FLASH
SHALL NOT EXCEED .152mm (.006") PER SIDE
**DIMENSIONS DO NOT INCLUDE INTERLEAD FLASH. INTERLEAD
FLASH SHALL NOT EXCEED .254mm (.010") PER SIDE
3811f
Information furnished by Linear Technology Corporation is believed to be accurate and reliable.
However, no responsibility is assumed for its use. Linear Technology Corporation makes no representation that the interconnection of its circuits as described herein will not infringe on existing patent rights.
47
LTC3811
RELATED PARTS
PART NUMBER
DESCRIPTION
COMMENTS
LTC1628/LTC1628-PG/
LTC1628-SYNC
2-Phase, Dual Output Synchronous Step-Down
DC/DC Controller
Reduces CIN and COUT, Power Good Output Signal, Synchronizable,
3.5V ≤ VIN ≤ 36V, IOUT Up to 20A, 0.8V ≤ VOUT ≤ 5V
LTC1735
High Efficiency Synchronous Step-Down Switching
Regulator
Output Fault Protection, 16-Pin SSOP Package
LTC1778/LTC1778-1
No RSENSETM Current Mode Synchronous Step-Down
Controllers
Up to 97% Efficiency, 4V ≤ VIN ≤ 36V, 0.8V ≤ VOUT ≤ (0.9)(VIN),
IOUT Up to 20A
LTC3708
Dual, 2-Phase, DC/DC Controller with Output Tracking
Current Mode, No RSENSE, Up/Down Tracking, Synchronizable
LTC3727/LTC3727A-1
2-Phase Dual Synchronous Controller
0.8V ≤ VOUT ≤ 14V; 4V ≤ VIN ≤ 36V
LTC3728
Dual, 550kHz, 2-Phase Synchronous Step-Down
Controller
Dual 180° Phased Controllers, VIN 3.5V to 35V, 99% Duty Cycle,
5mm × 5mm QFN and SSOP-28 Packages
LTC3729
20A to 200A, 550kHz PolyPhase Synchronous
Controller
Expandable from 2-Phase to 12-Phase, Uses All Surface Mount
Components, VIN Up to 36V
LTC3731
3- to 12-Phase Step-Down Synchronous Controller
60A to 240A Output Current, 0.6V ≤ VOUT ≤ 6V, 4.5V ≤ VIN ≤ 32V
LTC3773
Triple Output DC/DC Synchronous Controller
3-Phase Step-Down DC/DC Controller,
3.3V ≤ VIN ≤ 36V, Fixed Frequency 160kHz to 700kHz
LTC3826/ LTC3826-1
Ultralow IQ, Dual Output Synchronous Step-Down
DC/DC Controller
50μA IQ, 0.8V ≤ VOUT ≤ 10V, 4V ≤ VIN ≤ 36V
LTC3827/LTC3827-1
Dual, Selectable 140kHz to 650kHz 2-Phase,
Synchronous Step-Down Controller
Low IQ, VIN from 4V to 36V, VOUT from 0.8V to 10V, 5mm × 5mm
QFN32
LTC3828
Dual, 2-Phase Synchronous Step-Down Controller with Up to Six Phases, 0.8V ≤ VOUT ≤ 7V, 4.5V ≤ VIN ≤ 28V
Tracking
LTC3835/LTC3835-1
Low IQ Synchronous Step-Down Controller
Single Channel LTC3827/LTC3827-1
LT3845
Low IQ, High Voltage Single Output Synchronous
Step-Down DC/DC Controller
1.23V ≤ VOUT ≤ 36V, 4V ≤ VIN ≤ 60V, 120μA IQ
LTC3850
Dual, 550kHz, 2-Phase Synchronous Step-Down
Controller
Dual 180° Phased Controllers, VIN 4V to 24V, 97% Duty Cycle,
4mm × 4mm QFN-28, SSOP-28 Packages
No RSENSE is a trademark of Linear Technology Corporation.
3811f
48 Linear Technology Corporation
LT 0807 • PRINTED IN USA
1630 McCarthy Blvd., Milpitas, CA 95035-7417
(408) 432-1900 ● FAX: (408) 434-0507
●
www.linear.com
© LINEAR TECHNOLOGY CORPORATION 2007