LM10500
www.ti.com
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
LM10500 5A Step-Down Energy Management Unit (EMU) With PowerWise™ Adaptive
Voltage Scaling (AVS)
Check for Samples: LM10500
FEATURES
DESCRIPTION
•
•
•
•
•
•
•
•
•
•
•
The LM10500 is a 5A Energy Management Unit
(EMU) that actively reduces system-level power
consumption by utilizing a continuous, real-time,
closed-loop Adaptive Voltage Scaling (AVS) scheme.
The
LM10500
operates
cooperatively
with
PowerWise™ AVS compatible ASICs, SoCs, and
processors to optimize supply voltages adaptively
over process and temperature variations. The device
is controlled via PowerWise Interface (PWI) 1.0 or
PWI 2.0 high-speed serial interface.
1
23
Closed-Loop Adaptive Voltage Scaling (AVS)
PWI 1.0 / PWI 2.0 Compatible
Resistor-Programmable switching frequency
Frequency Synchronization
Precision Enable
Internal Soft-Start to Reduce In-Rush Current
Power Good (PWROK)
Under-Voltage Lock Out (UVLO)
Over-Voltage Protection (OVP)
Cycle-by-cycle current limiting (OCP)
Thermal Shutdown
A typical power saving of 40% can be achieved when
LM10500 is used with AVS compatible ASICs, SoCs,
and processors.
KEY SPECIFICATIONS
•
•
•
•
•
•
•
3.0 V to 18.0 V Input Voltage Range
1.5% Feedback Voltage Accuracy
0.6 Vto 1.0 V AVS Feedback Voltage Range
0.8 Vto 5.0 V Startup VOUT Range (Using
Resistor-Divider)
AVS Typical Power Saving 40% Over Fixed
Voltage
300 kHz to 1.5 MHz Switching Frequency
Range
WQFN-28 Package (5 mm x 5 mm x 0.8 mm,
0.5 mm pitch)
APPLICATIONS
•
•
•
•
•
Point-of-Load Regulation
Servers and Networking Cards
Storage Devices
Set-Top-Box Processors
Medical and Industrial Processors
1
2
3
Please be aware that an important notice concerning availability, standard warranty, and use in critical applications of
Texas Instruments semiconductor products and disclaimers thereto appears at the end of this data sheet.
PowerWise is a trademark of Texas Instruments.
All other trademarks are the property of their respective owners.
PRODUCTION DATA information is current as of publication date.
Products conform to specifications per the terms of the Texas
Instruments standard warranty. Production processing does not
necessarily include testing of all parameters.
Copyright © 2009–2013, Texas Instruments Incorporated
LM10500
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
www.ti.com
Adaptive Voltage Scaling Technology
PowerWise Adaptive Voltage Scaling (AVS) technology is an advanced closed-loop technology for reducing
active and standby energy consumption of digital processing engines and ASICs. Hardware Performance Monitor
(HPM) is designed into the digital engine together with an Advanced Power Controller (APC) to monitor the
performance of the silicon based on process and temperature variation. Information is fed back to an Energy
Management Unit (EMU) which then sets the voltage precisely according to the processor’s needs. The AVS
technology enables optimum power delivery to the processors, ASICs, and SoCs, which maximizes overall
system energy savings. AVS technology is process and architecture independent.
Figure 1. Typical Power Saving Using AVS And The LM10500
Compared To Fixed Voltage Scheme
45
POWER SAVING (%)
44
43
42
41
40
-20
0
20
40
60
80
AMBIENT TEMPERATURE (°C)
Typical Application Circuit
I/O Voltage PS
I/O Voltage
Domain
CVPWI
CIN
VIN
+3.0 ~
+18.0V
VPWI
SPWI
SCLK
PVIN
RF
Advanced
Power
Controller
(APC)
AVIN
CF
LM10500
AVS
VDD2
CVDD2
CBOOT
CBOOT
VDD1
CVDD1
Hardware
Performance
Monitor (HPM)
VOUT
L
SW
RFB1
RADDR
FB
RFRQ
COUT
AVS range: 0.6~1.0V
ADDR
SYNC
Start up to 0.8V or 1.0V
RFB2
PGND
FREQ
CFRQ
RC
CC
Core AVS
Domain
5A max
COMP
EN
PWROK
2
EN
PWROK
AGND
DGND
Submit Documentation Feedback
ASIC / SoC
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
LM10500
8
PGND
9
PGND
10
13
11
COMP
PWROK
PGND
SW
SCLK
PVIN
VDD2
PVIN
6
5
4
3
2
1
28
27
26
22
7
PVIN
SW
PVIN
SPWI
CBOOT
SW
AGND
VPWI
AVIN
21
PAD
DGND
VDD1
20
SW
FREQ
25
19
PGND
AVIN
18
EN
24
17
FB
ADDR
16
23
15
12
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
14
www.ti.com
Figure 2. WQFN-28 Package, Exposed Pad
See package Number RSG0028B
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
3
LM10500
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
www.ti.com
PIN DESCRIPTIONS
(1)
Pin #
Name
Type (1)
1,2,27,28
PVIN
P
Input voltage to the power switches integrated in the device.
3,4,5,6
SW
P
Switch node output of the power switches. It should be connected to the external
inductor.
7,8,9,10
PGND
G
Power ground for the internal power switches.
Compensation pin to connect to external compensation network.
Function
11
COMP
A
12
PWROK
OD
13
FB
A
Voltage Feedback pin. This pin can be connected to the output voltage directly or
through a resistor divider to set the output voltage range. Range of FB pin voltage is
PWI 1.0 or PWI 2.0 programmable between 0.6V and 1.0V.
14
ADDR
A
PWI Address Selection pin. An external resistor should be connected from this pin to
ground to set the PWI 1.0 or PWI 2.0 address. The voltage on this pin is only read
once by the internal register when the device is powered up.
15
EN
I
Precision enable pin. An external divider can be used to set the device turn-on
threshold. If not used, the EN pin should be connected to AVIN.
16
FREQ/SYNC
A
Frequency setting or external clock synchronization pin. This pin can be connected
to a resistor to ground to set the internal oscillator frequency. It also can be
connected to an external clock source via a capacitor, so that the switching action of
the device is synchronized to the external clock.
17
DGND
G
Digital ground for VPWI and digital interface: SPWI and SCLK.
18
VPWI
P
PowerWise Interface (PWI) supply input, 1.8 V-10% to 3.3V+10%. A bypass
capacitor of 1µF is recommended on this pin.
19
SPWI
I/O
PowerWise Interface (PWI) bi-directional data pin. This pin is internally pulled down
to ground.
20
SCLK
I
PowerWise Interface (PWI) clock input. This pin is internally pulled down to ground.
21
VDD2
P
2.5V output of internal regulator. This pin is only for bypassing the internal LDO.
Loading this pin is not recommended.
22,23
AVIN
P
Analog power input. It powers the internal 2.5V and 5.0V LDOs, which provide bias
current and internal driver power. It can be connected to PVIN through a low pass
RC filter, or can be supplied by a separate rail.
24
AGND
G
Analog ground for the internal bias circuitry.
25
VDD1
P
5.0V output of internal regulator. This pin is only for bypassing the internal LDO.
Loading this pin is not recommended.
26
CBOOT
A
Bootstrap pin to drive the high side switch. A bootstrap capacitor should be
connected between this pin to the SW pin.
PAD
PAD
Power Good or PWI Power OK, open drain output. If high, indicates the output
voltage is regulated within tolerance. A pull-up resistor (10 kΩ to 100 kΩ) is
recommended for most applications.
Exposed pad at the back of the device. The PAD should be connected to ground,
but cannot be used as primary ground connection. Use multiple vias under the PAD
for optimal thermal performance.
P: Power, A: Analog, I: Digital Input, I/O: Digital Input/Output, G: Ground, OD: Open Drain
These devices have limited built-in ESD protection. The leads should be shorted together or the device placed in conductive foam
during storage or handling to prevent electrostatic damage to the MOS gates.
4
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
LM10500
www.ti.com
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
Absolute Maximum Ratings (1) (2)
PVIN, AVIN, SW, EN, PWROK to AGND
-0.3 V to +20 V
CBOOT to AGND
-0.3 V to +25 V
CBOOT to SW
-0.3 V to +5.5 V
VDD1, VPWI, FB, COMP, ADDR, FREQ, SCLK, SPWI to AGND
-0.3 V to +6 V
VDD2 to AGND
-0.3 V to +3 V
AGND to PGND
-0.3 V to +0.3 V
Junction Temperature (TJ-MAX)
150 °C
Storage Temperature Range
-65 °C to 150 °C
Maximum Continuous Power Dissipation PD-MAX (3)
Maximum Lead Temperature Leadfree compatible
(1)
(2)
(3)
(4)
Internally limited
(4)
+260 °C
Absolute Maximum Ratings are limits beyond which damage to the device may occur. Operating Ratings are conditions under which
operation of the device is ensured. Operating Ratings do not imply ensured performance limits. For ensured performance limits and
associated test conditions, see the Electrical Characteristics tables.
If Military/Aerospace specified devices are required, please contact the TI Sales Office/Distributors for availability and specifications
The amount of Absolute Maximum power dissipation allowed for the device depends on the ambient temperature and can be calculated
using the formula P = (TJ – TA)/θJA, where TJ is the junction temperature, TA is the ambient temperature, and θJA is the junction-toambient thermal resistance. Junction-to-ambient thermal resistance is highly application and board-layout dependent. In applications
where high power dissipation exists, special care must be paid to thermal dissipation issues in PC board design. Internal thermal
shutdown circuitry protects the device from permanent damage.
For detailed soldering specifications, please refer to Application Note 1187: Leadless Leadframe Package (LLP) (AN-1187) SNOA401.
ESD Ratings
(1)
All Pins, Human Body Model (HBM)
(1)
±2 kV
The Human body model is a 100 pF capacitor discharged through a 1.5 kΩ resistor into each pin. (MIL-STD-883 3015.7)
Operating Ratings
PVIN to PGND, AGND
3 V to 18 V
AVIN to PGND, AGND
3 V to 18 V
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
5
LM10500
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
www.ti.com
Thermal Properties
−40 °C to +125 °C
Junction Temperature
Ambient Temperature
(1)
−40 °C to +85 °C
Junction-to-Ambient Thermal
(1)
(2)
Resistance (θJA)
(2)
32.4 °C/W
In applications where high power dissipation and/or poor package thermal resistance is present, the maximum ambient temperature may
have to be derated. Maximum ambient temperature (TA-MAX) is dependent on the maximum operating junction temperature (TJ-MAX-OP =
125°C), the maximum power dissipation of the device in the application (PD-MAX) and the junction-to ambient thermal resistance of the
part/package in the application (θJA), as given by the following equation: TA-MAX = TJ-MAX-OP – (θJA × PD-MAX).
Junction-to-ambient thermal resistance (θJA) is taken from a thermal modeling result, performed under the conditions and guidelines set
forth in the JEDEC standard JESD51-7. The test board is a 4-layer standard JEDEC thermal test board or 4LJEDEC, 4" x 3" in size,
with a 3 by 3 array of thermal vias. The board has two embedded copper layers which cover roughly the same size as the board. The
copper thickness for the four layers, starting from the top one, is 2 oz./1oz./1oz./2 oz. For WQFN, thermal vias are placed between the
die attach pad in the 1st. copper layer and 2nd. copper layer. Detailed description of the board can be found in JESD 51-7. Ambient
temperature in the simulation is 22°C, still air. Power dissipation is 1W. The value of θJA of this product can vary significantly depending
on PCB material, layout, and environmental conditions. In applications with high power dissipation (e.g. high VOUT, high IOUT), special
care must be paid to thermal dissipation issues. For more information on these topics, please refer to Application Note AN-1187:
Leadless Leadframe Package (LLP) SNOA401.
General Electrical Characteristics
(1) (2)
Specifications with standard typeface are for TJ = 25 °C, and those in boldface type apply over the full Operating
Temperature Range (TJ = −40 °C to 125 °C). Unless otherwise specified, VPVIN = VAVIN = 12 V, VOUT = 1.2 V.
Symbol
Parameter
Min
Typ
Max
LM10500SQ-0.8
Remarks
0.788
0.8
0.812
LM10500SQ-1.0
0.985
1.0
1.015
Unit
VFB-default
Feedback Pin Factory
Default Voltage. All
Registers in Default States
VFB-range-top
Maximum Feedback
Voltage
Voltage Code = R0 – R9 = 7FH
1.0
VFB-range-bottom
Minimum Feedback Voltage Voltage Code = R0 – R9 = 00H
0.6
Resolution
Bit Length of The Feedback
VFB = 0.6 V to 1 V
Voltage DAC
DNL
Differential Non-Linearity of
VFB DAC
VFB = 0.6 V to 1 V
0.5
LSB
INL
Integrated Non-Linearity of
VFB DAC
VFB = 0.6 V to 1 V
1
LSB
ΔVOUT/ΔIOUT
Load Regulation
IOUT = 0.1 A to 5 A
0.02
%/A
ΔVOUT/ΔVIN
Line Regulation
VIN = 3.0 V to 18 V
0.01
%/V
RDS-ON-HS
High Side Switch On-Time
Resistance
IDS-HS = 5 A
44
RDS-ON-LS
Low Side Switch On-Time
Resistance
IDS-LS = 5 A
22
ICL-HS
High Side Switch Current
Limit
High-Side FET
5.9
7
7.87
ICL-LS
Low Side Switch Current
Limit
Low-Side FET (3)
5.9
8
10.2
INEG-CL-LS
Low Side Switch Negative
Current Limit
Low-Side FET
-7
-4.1
-1.64
Shutdown Quiescent
VAVIN = V PVIN = 5 V
Current, AVIN is Connected
VAVIN = VPVIN = 18 V
to PVIN, VEN = 0
0.1
2
ISD
1
4.1
Iq
Quiescent Current With
Switcher On, No Load,
DCM Operation
VAVIN = VPVIN = 18 V
9
9.7
IFB
Feedback Pin Input Bias
Current
VFB = 1.0 V
1
(1)
(2)
(3)
6
V
7
bit
mΩ
A
µA
mA
nA
All limits are ensured by design, test and/or statistical analysis. All electrical characteristics having room-temperature limits are tested
during production with TJ = 25°C. All hot and cold limits are ensured by correlating the electrical characteristics to process and
temperature variations and applying statistical process control.
Capacitors: Low-ESR Surface-Mount Ceramic Capacitors (MLCCs) are used in setting electrical characteristics.
The low side switch current limit is ensured to be higher than the high side switch current limit.
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
LM10500
www.ti.com
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
General Electrical Characteristics (1) (2) (continued)
Specifications with standard typeface are for TJ = 25 °C, and those in boldface type apply over the full Operating
Temperature Range (TJ = −40 °C to 125 °C). Unless otherwise specified, VPVIN = VAVIN = 12 V, VOUT = 1.2 V.
Symbol
Parameter
Remarks
Min
Typ
Max
Unit
Gm
Error Amplifier
Transconductance
AVOL
Error Amplifier Voltage
Gain
VIH-OVP
OVP Tripping Threshold
Output voltage rising threshold, % of
VOUT
VHYST-OVP
OVP Hysteresis Window
% of VOUT
VUVLO-HI-AVIN
AVIN UVLO Rising
Threshold
2.84
2.93
2.987
V
VUVLO-LO-AVIN
AVIN UVLO
FallingThreshold
2.66
2.73
2.83
V
VUVLO-HYS-AVIN
AVIN UVLO Hysteresis
Window
VVDD1
Internal LDO1 Output
Voltage
Measured At VDD1 Pin, 1 kΩ Load
COUT-VDD1
Recommended Bypass
Capacitance to VDD1 Pin
Ceramic Capacitor
IShort-VDD1
VDD1 Pin Short Circuit
Current
VVDD2
Internal LDO2 Output
Voltage
COUT-VDD2
Recommended Bypass
Capacitance to VDD2 Pin
IShort-VDD2
VDD2 Pin Short Circuit
Current
VFCBOOT-d
CBOOT Diode Forward
Voltage
ICBOOT
VFB = 1.0 V, VCOMP = 0.5 V
103.5
2400
µA/V
65
dB
109.5
115
%
-4.3
195
mV
4.88
V
1
µF
31
mA
Measured At VDD2 Pin, 1 kΩ Load
2.47
V
Ceramic Capacitor
100
nF
47
mA
Measured Between VDD1 and CBOOT
@ 10 mA
0.76
V
CBOOT Leakage Current
VCBOOT = 5.5 V, VEN = 0 V
0.65
µA
TStartup-Delay
Soft Start Delay Time
Measured From EN Rising Edge to
The Beginning of Internal Soft-Start
Ramp
160
µs
TSoft-Start
Internal Soft-Start Ramping
Time
From 10% to 90% VFB, VFB = 0.8 V
1.9
3.4
5.3
From 10% to 90% VFB, VFB = 1.0 V
2.4
4.3
6.2
695
750
795
ms
OSCILLATOR
FOSC-nom
Nominal Switching
Frequency
RFRQ = 61.9 kΩ 0.025%
FOSC-MAX
Maximum Switching
Frequency
RFRQ = 28.4 kΩ
1500
FOSC-MIN
Minimum Switching
Frequency
RFRQ = 167.5 kΩ
300
TOFF-MIN
Switch Node Minimum OFF fS = 1.5 MHz, VPVIN = 3.3 V, VFB = 1.0
Time
V, Resistor Divider Ratio = 3.3
50
TON-MIN
Switch Node Minimum ON
Time
70
kHz
ns
fS = 1.5 MHz, VFB = 0.6 V, Resistor
Divider Ratio = 1
LOGIC
VIH-EN
EN Pin Rising Threshold
1.1
1.2
1.3
V
VHYST-EN
EN Pin Hysteresis Window
130
200
302
mV
IEN-IN
EN Pin Input Current
18
23
µA
VIH-UV-PWROK
PWROK UV Rising
Threshold
93
97.5
VHYST-UV-PWROK
PWROK UV Hysteresis
Window
VEN = 12 V
87.5
% of VOUT
%
-4.2
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
7
LM10500
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
www.ti.com
General Electrical Characteristics (1) (2) (continued)
Specifications with standard typeface are for TJ = 25 °C, and those in boldface type apply over the full Operating
Temperature Range (TJ = −40 °C to 125 °C). Unless otherwise specified, VPVIN = VAVIN = 12 V, VOUT = 1.2 V.
Symbol
Parameter
Remarks
IOL- PWROK
PWROK Sink Current
VOL = 0.2 V
IOH-
PWROK Leakage Current
VOH = 18 V
PWROK
Min
Typ
Max
3
Unit
mA
460
nA
500
µs
PWI I/O: SPWI, SCLK. The VPWI pin is powered from an external source, VPWI range = 1.8V-10% to 3.3V+10%
VPWI Slew Rate
VPWI Pin Voltage Ramping
See (4)
Rate
VIH
Logic Input High
% of VPWI
VIL
Logic Input Low
% of VPWI
20
IIH
Input Leakage Current, Pin
Driven High (VPWI)
SPWI, SCLK Pins Have Internal
Pulldowns
+5
IIL
Input Leakage Current, Pin
Driven Low
VOH
Logic Output High
Sourcing 1mA, % of VPWI
VOL
Logic Output Low
Sinking 1mA, % of VPWI
RPD-PWI
Pull-down resistance for
PWI signals
SPWI, SCLK
FSCLK-MAX
Maximum PWI SCLK
Frequency
SCLK
TD
SPWI Data Change Time to
% of SCLK Cycle, FSCLK ≤ 15 MHz (5)
Valid Value
TS
SPWI Data Setup Time
% of SCLK Cycle, FSCLK ≤ 15 MHz (5)
14
TH
SPWI Data Hold Time
% of SCLK Cycle, FSCLK ≤ 15 MHz (5)
39
TXZ
SPWI Data Drive Release
Time
% of SCLK Cycle, FSCLK ≤ 15 MHz (5)
80
%
%
µA
-1
80
10
0.5
1
2
15
%
MΩ
MHz
27
%
27
THERMAL SHUTDOWN
TSD
Thermal Shutdown
See (6)
160
°C
TSD-HYS
Thermal Shutdown
Hysteresis
See (6)
10
°C
(4)
(5)
(6)
8
Recommend not to exceed 500 µs ramp time.
Ensured by design. For more details, please refer to PWI 1.0 Specifications or PWI 2.0 Specifications.
Ensured by design.
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
LM10500
www.ti.com
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
Typical Performance Characteristics
Efficiency
fs = 300 kHz, VOUT = 1.2 V, FPWM = 0
Efficiency
fs = 300 kHz, VOUT = 3.3 V, FPWM = 0
100
100
95
95
90
90
EFFICIENCY (%)
EFFICIENCY (%)
Unless otherwise specified: VPVIN = VAVIN = 12V, VOUT = 1.2V, L=2.2µH, COUT = 220µF, fs = 300kHz.
85
80
75
70
65
60
85
80
75
70
65
60
55
55
PVIN = 12V AVIN = 5V
PVIN = AVIN = 12V
50
0
1
2
3
4
LOAD CURRENT (A)
PVIN = 12V AVIN = 5V
PVIN = AVIN = 12V
50
5
0
1
2
3
4
LOAD CURRENT (A)
5
Figure 4.
Efficiency
fs = 300 kHz, PVIN = AVIN = 5 V, FPWM = 0
Efficiency
fs = 300 kHz, PVIN = AVIN = 12 V, FPWM = 0
100
100
95
95
90
90
EFFICIENCY (%)
EFFICIENCY (%)
Figure 3.
85
80
75
70
65
VOUT = 3.3V
VOUT = 1.8V
VOUT = 1.2V
VOUT = 0.8V
60
55
50
0
1
2
3
4
LOAD CURRENT (A)
85
80
75
70
65
VOUT = 5V
VOUT = 3.3V
VOUT = 1.8V
VOUT = 1.2V
VOUT = 0.8V
60
55
50
5
0
1
2
3
4
LOAD CURRENT (A)
5
Figure 6.
Efficiency
fs = 500 kHz, PVIN = AVIN = 5 V, FPWM = 0
Efficiency
fs = 500 kHz, PVIN = AVIN = 12 V, FPWM = 0
100
100
95
95
90
90
EFFICIENCY (%)
EFFICIENCY (%)
Figure 5.
85
80
75
70
65
VOUT = 3.3V
VOUT = 1.8V
VOUT = 1.2V
VOUT = 0.8V
60
55
50
0
1
2
3
4
LOAD CURRENT (A)
85
80
75
70
65
VOUT = 5V
VOUT = 3.3V
VOUT = 1.8V
VOUT = 1.2V
VOUT = 0.8V
60
55
50
5
Figure 7.
0
1
2
3
4
LOAD CURRENT (A)
5
Figure 8.
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
9
LM10500
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
www.ti.com
Typical Performance Characteristics (continued)
Unless otherwise specified: VPVIN = VAVIN = 12V, VOUT = 1.2V, L=2.2µH, COUT = 220µF, fs = 300kHz.
Load Regulation (%)
Line Regulation (%)
0.20
0.10
LINE REGULATION (%)
LOAD REGULATION (%)
0.15
0.10
0.05
0.00
-0.05
-0.10
0.05
0.00
-0.05
-0.15
-0.20
-0.10
0
1
2
3
4
LOAD CURRENT (A)
5
3
Figure 10.
VOUT(%) Regulation
vs.
Temperature
Quiescent Current When IOUT = 0, FPWM = 0
10
QUIESCENT CURRENT (mA)
VOUTREGULATION (%)
18
Figure 9.
1.0
0.5
0.0
-0.5
-1.0
9
8
7
6
-40°C
25°C
85°C
5
-40
-20
0
20 40 60
TEMPERATURE (°C)
80
100
3
6
9
12
15
INPUT VOLTAGE (V)
Figure 11.
Figure 12.
High-Side and Low-Side RDSON
vs.
Temperature
Switching Frequency
vs.
RFRQ
70
18
1800
High-Side RDSon(m )
1600
FREQUENCY (kHz)
60
50
RDSON(m )
6
9
12
15
INPUT VOLTAGE, PVIN (V)
40
30
20
1400
1200
1000
800
600
400
10
200
Low-Side RDSon(m )
0
0
-40
-20
0
20
40
60
TEMPERATURE (°C)
80
100
Figure 13.
10
-40°C
25°C
125°C
0 20 40 60 80 100 120 140 160 180
RFRQ(k )
Figure 14.
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
LM10500
www.ti.com
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
Typical Performance Characteristics (continued)
Unless otherwise specified: VPVIN = VAVIN = 12V, VOUT = 1.2V, L=2.2µH, COUT = 220µF, fs = 300kHz.
Soft Start With 5 A Load, Triggered By EN Rising Edge
Soft Start With No Load, Triggered By EN Rising Edge
EN 1V/DIV
EN 1V/DIV
1.2V
1.2V
Output Voltage 0.2V/DIV
Output Voltage 0.2V/DIV
5A
Inductor Current 1A/DIV
PWROK 0.5V/DIV
Inductor Current 1A/DIV
PWROK 0.5V/DIV
1 ms/DIV
1 ms/DIV
Figure 15.
Figure 16.
Soft Start With 0.5V Pre-bias Voltage, DCM Operation,
Triggered By PWI 'Wakeup' Command
Soft Start With 0.5V Pre-bias Voltage, CCM Operation,
Triggered By PWI 'Wakeup' Command
EN 1V/DIV
EN 1V/DIV
1.2V
0.5V
Output Voltage 0.2V/DIV
1.2V
0.5V
Output Voltage 0.2V/DIV
Inductor Current 0.5A/DIV
Inductor Current 0.5A/DIV
PWROK 0.5V/DIV
PWROK 0.5V/DIV
1 ms/DIV
1 ms/DIV
Figure 17.
Figure 18.
Switching Waveform With 5 A Load
Load Transient Between 0.1 A And 5 A
Switch Node 2V/DIV
Inductor Current 1A/DIV
5A
Output Voltage 20 mV/DIV AC coupling
1.2V
2 µs/DIV
Figure 19.
Figure 20.
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
11
LM10500
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
www.ti.com
LM10500 PWI REGISTER MAPS
SUMMARY
The PWI standard supports thirty-two 8-bit registers on the PWI slave. Table 1 summarizes these registers and
shows default register bit values after reset. The following sub-sections provide additional details on the use of
each individual register. Please refer to http://www.pwistandard.com/ for more detailed PWI 1.0 and PWI 2.0
specifications.
Table 1. Slave Base Register Summary
Register
Address
Register
Name
Register Usage
(2)
Type
Default Value (1)
7
6
5
4
3
2
1
0
HEX
7F
0x0
R0
Core Voltage
R/W
0
1
1
1
1
1
1
1
0x1-0x3
R1-R3
Unused
N/A
-
-
-
-
-
-
-
-
0x4
R4
PWI Version Number /
Device Capability
R/O
-
-
-
-
-
-
0
1
01 PWI 1.0
1
0
02 PWI 2.0
0x5 - 0x8
R5 - R8
Unused
N/A
-
-
-
-
-
-
-
-
N/A
-
1
0
0
0
0
0
0
40 Startup
voltage 0.8V
-
0
0
0
0
0
0
0
00 Startup
voltage 1.0V
0x9
R9
Core Voltage Offset (3)
R/W
0xA
R10
Switcher Control (4)
R/W
-
-
-
-
-
1
1
1
07
0xB
R11
Unused
N/A
-
-
-
-
-
-
-
-
N/A
0xC-0xF
R12-R15
Reserved
N/A
-
-
-
-
-
-
-
-
N/A
0x10-0x1F
R16-R31
Not Implemented
N/A
-
-
-
-
-
-
-
-
N/A
(1)
(2)
(3)
(4)
“-” denotes unused bits. A write into unused bit position will be ignored. A read will produce ‘0’ when register is partially used and a “No
response frame” when register is completely unused. A bit in BOLD denotes a register bit that is read-only. A read will result in the
indicated value and a write will be ignored. Please refer to PWI specification version 1.0/2.0 for further information.
Factory configurable to 0x7F, 0x5F, 0x3F, or 0x1F.
Factory configurable to 0x00, 0x20, 0x40, or 0x60.
Manufacture default is Force PWM enabled.
Slave Extended Registers are not implemented in the LM10500.
Table 2. Slave Extended Register Summary
Register
Address
Register
Name
0x00-0xFF
ER0-ER255
Register Usage
Not implemented
Type
N/A
Reset Default Value
7
6
5
4
3
2
1
0
HEX
-
-
-
-
-
-
-
-
N/A
PWI Register Bit Definition
Bit definitions of R0, R4, R9 and R10 are listed in the tables below.
Table 3. R0 - Core Voltage Register
Bit
7
Field Name
Sign
Description or Comment
This bit is fixed to 0.
PWI 1.0/2.0 programmable from 7h'00 to 7h'7F. Default R0 = 7FH.
Core voltage is determined by the Voltage Code = R0 − R9.
6:0
12
Voltage code
Voltage Code [6:0]
Core voltage with no external resistor divider (V)
7h'00
0.6
7h'xx
Linear scaling from 0.6 to 1.0, 3.15 mV / LSB
7h'7F
1.0
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
LM10500
www.ti.com
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
R0 is restored to its default value when LM10500 wakes up from SLEEP state.
Table 4. R4 - PWI Version Number/Device Capability Register
Bit
Field Name
Description or Comment
7:2
Reserved
Unused
1:0
Version
Read only. Writing transactions to this register will be ignored.
Read transaction will return:
2b′01
PWI version 1.0
2b′10
PWI version 2.0
See PowerWise interface Selection section for configuring the device to PWI 1.0 or 2.0.
Table 5. R9 - Core Voltage Offset Register
Bit
Field Name
7
Reserved
6:0
Core Voltage Offset code
Description
Unused
The core voltage is determined by the Voltage Code = R0 − R9. R9 is intended for
supporting APC1 or APC2 startup voltages that are less than top of the range (0x7F).
In LM10500SQ-0.8 and LM10500SQX-0.8, default R9 = 40H
In LM10500SQ-1.0 and LM10500SQX-1.0, default R9 = 00H
R9 is NOT restored to its default value when LM10500 wakes up from SLEEP state.
Table 6. R10 - Regulator Control Register
Bit
Field Name
7:3
Reserved
2
Force PWM
(FPWM)
1
Voltage Down Slew Control
Enable
0
Voltage Up Slew Control
Enable
Description
Unused
1: Force switcher to operate in Continuous Conduction Mode (CCM) regardless of
load
0: Switcher will operate in Discontinuous Conduction Mode (DCM) at light load
1: Enable stepping slew rate control for large voltage code down step
0: Disable stepping slew rate control for down step
1: Enable stepping slew rate control for large voltage code up step
0: Disable stepping slew rate control for up step
R10 is NOT restored to its default value when LM10500 wakes up from SLEEP state.
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
13
LM10500
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
www.ti.com
Block Diagram
C6
C7
R2
R1
C5
ADDR
VDD2
CBOOT
VDD1
AVIN
VIN
VPWI
SPWI
SCLK
Address sense
LDO2
2.5V
PWI
1.0/2.0
LDO1
5V
PVIN
OCP
SPC
DGND
C1
I sense
C2
SS
TSD
Nonoverlap
UVLO
EN
Precision
Enable
DAC
REF
EA
PWROK
PWM
SW
Slope
Comp
UVLO
OVP
FREQ
SYNC
VOUT
OVP
COMP
C3
Nonoverlap
Z cross
OCL
Rev OCL
OSC/
Sync
AGND
L1
PWM control logic
FB
PGND
R1
C5
R4
C4
Operation Description
LM10500 is a PowerWise Interface (PWI) 1.0 and 2.0 compliant energy management unit (EMU). It operates
cooperatively with ASIC and FPGA cores using Advanced Power Controller (APC) to provide Adaptive or
Dynamic Voltage Scaling (AVS or DVS) which drastically reduces power consumption compared to conventional
power delivery methods. The device consists of PWI registers, logic, and a switching DC/DC buck converter to
supply the AVS or DVS voltage domain. Please refer to http://www.pwistandard.com/ for more detailed PWI 1.0
and PWI 2.0 specifications.
SWITCHING REGULATOR
The LM10500 employs a buck type (step-down) architecture. It utilizes many advanced features to achieve
excellent voltage regulation and efficiency. This easy-to-use regulator features two integrated switches and is
capable of supplying up to 5 A of continuous output current. The regulator utilizes peak current mode control with
slope compensation scaled with switching frequency to optimize stability and transient response over the entire
output voltage and switching frequency range. Peak current mode control also provides inherent line feedforward, cycle-by-cycle current limiting and easy loop compensation. The switching frequency can be adjusted
between 300 kHz and 1.5 MHz. The device can operate with a small external L-C filter and still provide very low
output voltage ripple. The precision internal voltage reference allows the output to be set as low as 0.6 V. Using
an external compensation circuit, the regulator bandwidth can be selected based on the switching frequency to
provide fast load transient response. The switching regulator is specially designed for high efficiency operation
throughout the load range. Synchronous rectification yields high efficiency for low voltage and heavy load current
situations, while optional Discontinuous Conduction Mode (DCM) operation extends high efficiency conversion to
lower load currents. Fault protection features include: high-side and low-side switch current limiting, negative
current limiting on the low-side switch, over voltage protection and thermal shutdown. The device is available in
the WQFN-28 package featuring an exposed pad to aid thermal dissipation. The LM10500 can be used in
numerous applications to efficiently step-down from a wide range of rails: 3 V to 18 V.
14
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
LM10500
www.ti.com
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
POWERWISE INTERFACE
The LM10500 is programmable via the low-power, 2-wire PowerWise Interface (PWI). This serial interface
controls the various voltages and states of the regulator in the device. The switching regulator voltage at the
feedback pin can be set between 0.6 V and 1.0 V in 127 steps (linear scaling, 3.15 mV per step). This highresolution voltage control enables accurate temperature and process compensation in AVS.
The LM10500 supports the full command set as described in PWI 1.0 / 2.0 specifications:
• Core Voltage Adjust
• Reset
• Sleep
• Shutdown
• Wake-up
• Register Read
• Register Write
• Authenticate
• Synchronize
DEVICE OPERATIONAL STATES
This device has four operating states: STARTUP, ACTIVE, SHUTDOWN and SLEEP.
The STARTUP state is the default state after PWI Reset Command. The STARTUP state is entered from any
other state if the EN pin is pulled to logic zero, or the PWI Reset Command is issued. When EN pin is in logic
low, the regulator is off, and PWROK output is ‘0’. The DC-DC regulator will be enabled when the EN pin is
pulled high. After the internal soft start, the device enters the ACTIVE state.
In the ACTIVE state, the voltage regulator is in normal operation and the PWROK output is ‘1’. Immediately after
soft start, the output voltage is at the default level. The output voltage can then be scaled by programming the
corresponding PWI control registers. If a PWI Shutdown Command is issued, the device will enter the
SHUTDOWN state. Or if a PWI Sleep Command is issued, it will enter the SLEEP state.
In the SLEEP state, the voltage regulator is off, but the PWROK output is still ‘1’. The PWI registers can be
programmed in the SLEEP state. The device can be activated from the SLEEP state to the ACTIVE state by the
PWI Wake-Up Command. The register R0 will be reset to its default value (7FH) and all other register contents
will remain unchanged. The device will enter the SHUTDOWN state by the PWI Shutdown Command.
In the SHUTDOWN state, the voltage regulator is off and PWROK signal is ‘0’ as well. The device will exit the
SHUTDOWN state to the STARTUP state if the EN pin is pulled low or a Reset Command is issued.
The diagram below summarized the four states and transitions. This figure assumes that the supply voltage to
the regulator IC is in the valid range.
EN = 0
SHUTDOWN
Reset Command
All regulators off
PWROK = µ0'
Shutdown
Command
STARTUP
All regulators off
PWROK = µ0'
Reset
Command
Shutdown
Command
EN = 0
EN = 0
Reset
Command
EN = 1
(Soft-start)
Sleep Command
ACTIVE
SLEEP
Core regulator off
PWROK = µ1'
Wakeup Command
(soft-start)
All regulators on
Voltage scaling active
PWROK = µ1'
Figure 21. The LM10500 PWI Operating States
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
15
LM10500
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
www.ti.com
VOLTAGE SCALING
The device is designed to be used in a voltage scaling (AVS or DVS) enabled system to lower the power
dissipation of SoCs or ASICs. By scaling the supply voltage with process variations, temperature variations,
aging or the clock frequency, dramatic power savings can be achieved. Two types of voltage scaling are
supported, dynamic voltage scaling (DVS) and adaptive voltage scaling (AVS). DVS systems switch between
pre-characterized voltages which are paired to clock frequencies used for frequency scaling in the ASIC. AVS
systems track the ASIC’s performance and optimize the supply voltage adaptively to the required performance.
AVS is a closed loop system that provides process and temperature compensation such that for any given
process, temperature, or clock frequency, the minimum supply voltage that can achieve desired performance is
delivered.
The output voltage of the switching regulator is programmed via the Core Voltage Adjust command sent by APC.
PWI commands adjust the content of the PWI registers to adjust the AVS reference voltage (output of the
internal DAC). The programmable reference voltage range is between 0.6 V and 1.0 V. The output voltage is the
same as the AVS reference voltage when the FB pin is connected to the output voltage directly, otherwise, it will
be scaled up by the output resistor divider. Please refer to the Design Guide section for design guidelines.
DIGITALLY ASSISTED VOLTAGE SCALING
The switching regulator in the LM10500 is designed to work in a voltage scaling system. This requires that the
regulator has a well-controlled large signal transient response. The device delivers fast, controlled voltage scaling
transients with the help of a digital state machine. The state machine automatically optimizes large signal
transients providing minimal over / undershoot and maintaining settling times less than 100 µsec. This is an
important characteristic for voltage scaling systems that rely on minimal undershoot to set voltages as low as
possible to save more energy.
When a large voltage up or down step is requested by the PWI command, the straight forward way is to change
the AVS voltage reference to the final value right away, as shown by the dash line in the figure below. The
control loop will regulate the output voltage to follow the reference. Also, the inductor current needs to charge or
discharge the output capacitors. The output voltage will have overshoot or undershoot and the inductor current
will also show large current spikes.
In the LM10500, the large single step in AVS voltage reference is divided in a binary manner, as shown in by the
solid line in the figure. By doing so, the large-signal voltage scaling response is smoothed out over time and the
transients show almost no overshoot / undershoot. The slew rate control for voltage up and down steps can be
enabled or disabled independently by register R10.
...
Figure 22. AVS Reference Voltage Large Step With Slew Rate Control (Solid Line)
And Without Slew Rate Control (Dash Line)
PEAK CURRENT MODE CONTROL
In most applications, the peak current mode control architecture used in the LM10500 only requires two external
components to achieve a stable design. The external compensation allows the user to set the crossover
frequency and phase margin, thus optimizing the transient performance of the device. For duty cycles above
50%, all current mode control buck converters require the addition of an artificial ramp to avoid sub-harmonic
oscillation. This linear ramp is commonly referred to as slope compensation. The amount of slope compensation
in the LM10500 will automatically change depending on the switching frequency: higher the switching frequency,
larger the slope compensation. This allows smaller inductors to be used with higher switching frequency to
increase power density.
16
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
LM10500
www.ti.com
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
SWITCHING FREQUENCY SETTING AND SYNCHRONIZATION
The switching regulator in the LM10500 device can operate at frequencies ranging from 300 kHz to 1.5 MHz.
The switching frequency can be set / controlled in two ways. One is by selecting an external resistor RFRQ
attached to the FREQ pin to set the internal oscillator frequency, which controls the switching frequency. An
external 100 pF capacitor, CFRQ, should also be connected from the FREQ pin to the analog ground as a noise
filter, as shown in Figure 23.
LM10500
CFRQ
FREQ
RFRQ
Figure 23. Switching Frequency Set By An External Resistor
The other way is to synchronize the switching action to an external clock or other fixed frequency signal in the
range of 300 kHz to 1.5 MHz. The external clock should be applied through a 100 pF coupling capacitor, CFRQ,
as shown in Figure 24:
LM10500
CFRQ
FREQ
External
Clock
RFRQ
Figure 24. Switching Frequency Synchronized To External Clock
The recommendations for the external clock include peak-to-peak voltage above 1.5 V, duty cycle between 20%
and 80%, and the edge rate faster than 100 ns. Circuits that use an external clock should still have a resistor,
RFRQ, connected from the FREQ pin to the analog ground. The external clock frequency should be within −10%
to +50% of the frequency set by RFRQ. This allows the regulator to continue operating at approximately the same
switching frequency if the external clock fails and the coupling capacitor on the clock side is grounded or pulled
to logic high.
If the external clock fails low, timeout circuits will prevent the high-side FET from staying off for longer than 1.5
times the switching period (switching period = 1 / switching frequency). At the end of this timeout period, the
regulator will begin to switch at the frequency set by RFRQ.
If the external clock fails high, timeout circuits will again prevent the high-side FET from staying off longer than
1.5 times the switching period. After this timeout period, the internal oscillator takes over and switches at a fixed
1 MHz until the voltage on the FREQ pin has decayed to approximately 0.6V. This decay follows the time
constant of CFRQ and RFRQ, and once it is completed, the regulator will switch at the frequency set by RFRQ.
LIGHT LOAD OPERATION
The LM10500 offers increased efficiency at light loads when the FPWM bit (bit 2 in PWI register R10) is ‘0’,
where Discontinuous Conduction Mode (DCM) is enabled. When the load current is reduced to a point where half
of the inductor ripple current is greater than the load current, the device will enter DCM, thus preventing negative
inductor current. The point at which this occurs is the critical conduction boundary and can be calculated by the
following equation:
VOUT x (1 ± D)
IBOUNDARY =
2LfS
(1)
where D is the duty cycle of the high side switch, equal to (high side switch on time / switching period). Please
refer to CALCULATING THE DUTY CYCLE under Design Guide for more details. Several diagrams are shown in
Figure 25 illustrating Continuous Conduction Mode (CCM), Discontinuous Conduction Mode (DCM), and the
boundary condition.
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
17
LM10500
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
www.ti.com
Switchnode Voltage
It can be seen that in DCM, whenever the inductor current reaches zero, the SW node will become high
impedance. Ringing will occur on this pin as a result of the LC tank circuit formed by the inductor and the
effective parasitic capacitance at the switch node. If this ringing is of concern, an additional RC snubber circuit
can be added from the switch node to the power ground. At very light loads, usually below 100 mA, several
pulses may be skipped in between switching cycles, effectively reducing the switching frequency and further
improving light-load efficiency.
Continuous Conduction Mode (CCM)
VIN
Inductor Current
Time (s)
Continuous Conduction Mode (CCM)
IAVERAGE
Inductor Current
Time (s)
DCM - CCM Boundary
IAVERAGE
Switchnode Voltage
Time (s)
Discontinuous Conduction Mode (DCM)
VIN
Inductor Current
Time (s)
Discontinuous Conduction Mode (DCM)
IPeak
Time (s)
Figure 25. CCM and DCM Operation
18
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
LM10500
www.ti.com
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
FPWM OPERATION (FPWM BIT)
The regulator in the LM10500 can be programmed to operate in Continuous Conduction Mode (CCM) regardless
of load, by setting the FPWM bit (bit 2 of PWI register R10) to ‘1’. When FPWM=1, the inductor current can be
negative when load current is lower than critical conduction boundary (refer to LIGHT LOAD OPERATION
section for more details). A negative current limit is implemented in the LM10500. If the inductor current goes
below the limit, the low side switch will be turned off. When FPWM = 0, the regulator will operate in
Discontinuous Conduction Mode (DCM) at light load, where inductor current will not be negative. At light load,
DCM provides higher power conversion efficiency, while CCM operation provides smaller output voltage ripple
and better transients.
The FPWM bit can be configured on the fly. However, when the load current is lower than the critical conduction
boundary, an small over- or under-shoot may appear when toggling the FPWM bit. The reason is that duty cycle
is different in CCM and DCM for a certain output voltage and load condition. When toggling the FPWM bit, the
feedback loop needs time to adjust the duty cycle to the new value, resulting a small spike on VOUT. If the FPWM
bit is toggled when the current is above the critical conduction boundary, there will be NO spike at all. It is
recommended to toggle the FPWM bit when the load current is above the critical conduction boundary, or when
the device is in . The same reason explains a small voltage undershoot seen after soft start with light load and
FPWM = 1. The regulator operates in diode emulation mode during soft start, the same behavior as in DCM.
After desired VOUT is reached, PWROK is released to high and the regulator will operate in CCM. The transition
will result a small undershoot if load current is below the critical conduction boundary. The undershoot will NOT
appear if FPWM = 0, or load current is above the critical conduction boundary.
PRECISION ENABLE
The enable (EN) pin allows the output of the device to be enabled or disabled by an external control signal. This
pin is a precision analog input that enables the device when the voltage exceeds 1.2 V (typical). The EN pin has
200 mV (typical) of hysteresis and will disable the output when the enable voltage falls below 1.0 V (typical). If
the EN pin is not used, it should be pulled up to AVIN via a 10 kΩ to 100 kΩ resistor. Since the enable pin has a
precise turn on threshold, it can be used along with an external resistor divider network from AVIN or an external
voltage to configure the device to turn on at a precise voltage. The precision enable circuitry will remain active
even when the device is disabled. The turn on voltage with a divider can be found by:
REN1
REN2
§
©
§
©
VEN-EXT = 1.2 1 +
(2)
VEN-EXT
LM10500
REN1
EN
REN2
Figure 26. Use External Resistor To Set
The EN Threshold
DEVICE SHUTDOWN AND ENABLE
The device output can be turned off by turning off AVIN, pulling the EN pin low, or issuing a PWI Shutdown
Command with supply power ON and EN pin high.
To enable the device, EN pin must be high with the presence of AVIN and PVIN. Once enabled, the device
engages the internal soft start and output voltage goes to its default value. The soft start feature allows the
regulator output to gradually reach the steady state operating point, thus reducing stresses on the input supply
and controlling start up current. Soft start begins when at the rising edge of EN with AVIN above UVLO level. It is
important to make sure PVIN is high when soft start begins. The LM10500 allows AVIN to be higher than PVIN,
or PVIN higher than AVIN, as long as both of them are within their operating voltage ranges.
When using PWI to issue a Shutdown Command, the PWI will be disabled along with the regulator in the device.
To restart the part, EN pin must be toggled (high → low → high). The part will then begin soft start and output
voltage will ramp to its default value. Please refer to the DEVICE OPERATIONAL STATES for more details.
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
19
LM10500
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
www.ti.com
The EN pin provides flexibility for system control. In larger systems, it can be advantageous to enable/disable a
subsystem independently. The EN pin also allows the system controller to issue a global reset command to all
the subsystems.
Figure 27 illustrates a shutdown and enable sequence of the LM10500.
xx
xx
Shutdown Command
PWI
EN and
RESETN
OFF
OFF
Voltages
On
Off
Figure 27. Shutdown And Enable Methods
UVLO
The LM10500 has a built-in Under-Voltage Lock-Out (UVLO) protection circuit that prevents the device from
switching until the AVIN voltage reaches 2.93V (typical). The UVLO threshold has typically 195 mV of hysteresis
that keeps the device from responding to power-on glitches during startup.
SOFT-START AND PRE-BIAS STARTUP CAPABILITY
Soft start of the LM10500 is controlled internally. It typically takes 4.3 ms to finish the soft start sequence in
LM10500SQ-1.0 and 3.4 ms in LM10500SQ-0.8. PWROK will be high after soft start is finished.
The LM10500 is in a pre-biased state when the device starts up with an output voltage greater than zero. This
often occurs in many multi-rail applications such as when powering an FPGA, ASIC, or DSP. In these
applications the output can be pre-biased through parasitic conduction paths from one supply rail to another.
Even though the LM10500 is a synchronous converter it will not pull the output low when a pre-bias condition
exists. During start up the LM10500 will be in diode emulation mode: the low side switch is turned off when
inductor current zero crossing is detected. After soft start, the operation mode programmed in PWI register R10
(FPWM = 0 or 1) is resumed.
PEAK CURRENT PROTECTION AND NEGATIVE CURRENT LIMITING
The switching regulator in the LM10500 detects the peak inductor current and limits it to the value of 7A typical.
To determine the average current limit from the peak current limit, the inductor size, input and output voltages,
and the switching frequency must be known. The average current limit can be found by:
(3)
When the peak inductor current sensed from the high-side switch reaches the current limit threshold, an overcurrent event is triggered and the internal high-side switch is turned off and the low-side FET is turned on,
allowing the inductor current to ramp down until the next switching cycle. When the high-side over-current
condition persists, the output voltage of the device will drop due to the reduced high side switch ON time.
In cases such as short circuit or when the high-side switch minimum on time conditions are reached, the highside switch current limiting may not be sufficient to limit the inductor current. The LM10500 features an additional
low-side switch current limit to prevent the inductor current from running away. The low-side switch current limit is
set higher than the high-side current limit: 8 A typical. When the low-side switch current is higher than the limit
level, PWM pulses will be skipped until the low-side over current event is not detected during the entire low-side
switch conduction time. Normal PWM switching subsequently occurs when the condition is removed.
20
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
LM10500
www.ti.com
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
High-side and low-side current protections result in a current limit that does not aggressively fold back for brief
over-current events, while at the same time providing frequency and voltage fold back protection during hard
short circuit conditions.
The low-side switch also has negative current limit, -4.1 A typical. If the negative current limit is triggered, the
low-side switch will be turned off. The negative current will be forced to go through the high-side switch body
diode and it will quickly reduce to zero. The switch node will have a voltage pulse at (PVIN + 0.7 V) level due to
the turning on of the high-side switch diode.
PWROK AND OVER-/UNDER-VOLTAGE HANDLING
The PWROK pin is an open-drain output. It should be pulled high with an external resistor (10 kΩ to 1 MΩ
recommended). The PWROK pin will be high when the FB voltage is within −7% to +9.5% (typical) of the AVS
reference voltage. Otherwise, an internal pull-down device will pull the PWROK pin low, as defined in the PWI
specification.
The LM10500 has built-in under- and over-voltage comparators that control the high-side and low-side switches.
Whenever there is an excursion in output voltage above the set Over Voltage Protection (OVP) threshold, +9.5%
typical, the device will terminate high-side ON pulse if present, turn on the low-side switch, and pull the PWROK
pin low. The low-side switch will remain ON until either the FB voltage falls back into regulation or the inductor
current zero-cross is detected which in turn tri-states the switches. OVP is disabled during soft start to prevent
false triggering.
If the output reaches the Under Voltage Protection (UVP) threshold, −7% typical, the device will continue
switching and the PWROK pin will be pulled low. To avoid false tripping during transient glitches, the PWROK pin
has 16 μs of built in deglitch time to both rising and falling edges.
INTERNAL REGULATORS
The LM10500 contains two internal Low DropOut (LDO) regulators to produce internal driving and bias voltage
rails from AVIN. One LDO produces 5.0 V to power the internal MOSFET drivers, the other produces 2.5 V to
power the internal bias circuitry. The 5.0 V LDO should be bypassed to analog ground through VDD1 pin with an
external ceramic capacitor (1 µF recommended). The 2.5 V LDO should be bypassed to analog ground through
VDD2 pin with an external ceramic capacitor (0.1 µF recommended). Good bypassing is necessary to supply the
large transient currents required by the power MOSFET gate drivers.
Applications with high input voltage and high switching frequency will increase die temperature because of the
higher power dissipation within the LDOs. Connecting a load to the VDD1 or VDD2 pin is not recommended
since it will degrade their driving capability to the internal circuitry, further pushing the LDOs into their RMS
current ratings, and increasing power dissipation and die temperature.
The LM10500 allows AVIN to be as low as 3 V which makes the voltage at the VDD1 LDO lower than 5 V. Low
supply voltage at the MOSFET drivers will increase on-time resistance of the high-side and low-side MOSFETs
and reduce efficiency of the regulator. When AVIN is between 3 V to 5 V, the best practice is to short the VDD1
pin to AVIN to avoid the voltage drop on the internal LDO. However, the device can be damaged if the VDD1 pin
is pulled to a voltage higher than 5.5 V.
For efficiency considerations, it would be the best to use AVIN = 5 V if possible. When AVIN is above 5 V,
reduced efficiency can be observed at light load due to the power loss on the LDOs. When AVIN is close to 3 V,
increased MOSFET on-time resistance can reduce efficiency at high load current operations.
MINIMUM ON-TIME CONSIDERATIONS
Minimum on-time, TON-MIN, is the smallest duration of time that the high-side MOSFET can be on. This time is
typically 70 ns in the LM10500. In CCM operation, the minimum on-time limit imposes a minimum duty cycle of
DMIN = fS x TON-MIN
(4)
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
21
LM10500
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
www.ti.com
With a given output voltage, minimum on-time limits the switching regulator when operating with high input
voltage and high switching frequency at the same time. As the equation shows, reducing the operating frequency
will alleviate the minimum duty cycle constraint. With a given switching frequency and desired output voltage, the
maximum allowed PVIN can be approximated by
VPVIN-max =
VOUT
fS
x
1
TON-MIN
(5)
Similarly, if the input rail is fixed, the maximum switching frequency without imposing minimum on-time can be
found by
(6)
In rare cases, steady-state operation at minimum duty cycle is unavoidable. If the regulator is operating with
FPWM = 1, the output voltage will increase with supply voltage or switching frequency, until OVP is triggered.
OVP response will skip PWM cycles until output voltage drops back to the regulation band (below 105% of the
regulated VOUT). VOUT will be regulated slightly above target with bigger ripple. This is a safe operating condition
in most AVS applications. If FPWM = 0, DCM mode is allowed. The regulator will automatically skip PWM cycles
to keep VOUT in regulation, similar to light load DCM operation.
THERMAL PROTECTION
Internal thermal shutdown circuitry is provided to protect the integrated circuit in the event that maximum junction
temperature is exceeded. When activated, typically at 160 °C, the LM10500 tri-states the high-side and low-side
power switches and resets soft-start. After the junction temperature cools down to approximately 150 °C, the
regulator starts up again using normal startup routine. This feature is provided to prevent catastrophic failures
from accidental device overheating.
Design Guide
This section walks the designer through the steps necessary to select the external components to build a fully
functional PWI energy management unit (EMU). As with any DC-DC converter, numerous tradeoffs are possible
to optimize the design for efficiency, size, and performance. These will be taken into account and highlighted
throughout this discussion. To facilitate component selection discussions, the typical application circuit below
may be used as a reference. Unless otherwise indicated, all formulas assume units of Amps (A) for current,
Farads (F) for capacitance, Henries (H) for inductance, Volts (V) for voltages and Hertz (Hz) for frequencies.
CIN
VIN
3.0 V to
18.0V
PVIN
SPWI
SCLK
RF
VPWI
CVPWI
AVIN
CF
LM10500
VDD2
CVDD2
CBOOT
CBOOT
VDD1
CVDD1
L
VOUT
SW
RFB1
FB
RADDR
RFRQ
SYNC
Start up to 0.8V or 1.0V
AVS range: 0.6~1.0V
ADDR
RFB2
COUT
ROUT
PGND
FREQ
CFRQ
RC
CC
COMP
EN
PWROK
EN
PWROK
AGND
DGND
Figure 28. LM10500 Typical Application Circuit
22
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
LM10500
www.ti.com
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
PROGRAMMING THE OUTPUT VOLTAGE
The output voltage of LM10500 can be adjusted in two ways: VFB is run-time programmable by PWI command
through PWI interface, and a resistor divider can be used on the board to scale the (VOUT / VFB) ratio.
The FB pin can be connected to VOUT directly or through a resistor divider. With external resistor divider, the
output voltage can be scaled up from VFB. Figure 29 shows the connection of the divider to the FB pin.
VOUT
LM10500
RFB1
FB
RFB2
Figure 29. Setting the Output Voltage By Resistor Divider
The output voltage can be found by:
(7)
VFB power up voltage is 0.8 V in LM10500SQ-0.8 and 1.0 V in LM10500SQ-1.0. The output voltage programming
range is scaled up by a factor of (1 + RFB1 / RFB2). For example, if the desired power up voltage is 1.2 V, RFB1 = 2
kΩ, RFB2 = 10 kΩ can be used with LM10500SQ-1.0. The output voltage programming range becomes 0.72 V to
1.2 V. The maximum allowed output voltage is 5 V in the LM10500.
Unlike non-AVS power supplies, the reference voltage VFB in the LM10500 is programmable by the PWI
registers. The range of the AVS reference voltage is from 0.6 V to 1.0 V, corresponding linearly to voltage code
0x0 to 0x7F. The voltage code is determined by the codes in PWI register R0 (Core voltage register) and R9
(core voltage offset) as:
-R0 - R9, R0 - R9 t 0
Voltage Code = ®
R0 - R9 0
¯0,
Voltage Code = (R0 - R9) [6:0]
Reference Voltage (V)
0x00
0.6
0x00 < code < 0x7F
Linear scaling from 0.6 to 1.0, 3.15
mV/LSB
0x7F
1.0
The R0 default code after power up is always 0x7F. The difference between LM10500SQ-0.8 and LM10500SQ1.0 is shown in the table below. The default AVS reference voltage is usually the maximum reference voltage
used in AVS operation to ensure safe operation for all conditions at start up.
R0 default
R9 default
VFB default
VFB range
LM10500SQ-0.8
0x7F
0x40
0.8V
0.6 to 1.0V
LM10500SQ-1.0
0x7F
0x00
1.0V
0.6 to 1.0V
CALCULATING THE DUTY CYCLE
The first value to calculate for any buck converter is the duty cycle. In an ideal (no loss) buck converter, the ideal
duty cycle can be found by:
Dideal =
VOUT
VPVIN
(8)
In applications with low output voltage (3 A), which is typical in AVS operation,
the losses should not be ignored when calculating the duty cycle. Considering the effect of conduction losses
associated with the MOSFETs and inductor, the duty cycle can be approximated by:
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
23
LM10500
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
www.ti.com
(9)
Rdson-HS and Rdson-LS are the ON-time parasitic resistances of the high-side and low-side MOSFETs, respectively.
Rdcr is the equivalent DC resistance of the inductor used in the output filter. Other parasitics, such as PCB trace
resistance, can be included in RDCR if desired. IOUT is the load current. It is also equal to the average inductor
current. The duty cycle will increase slightly with the increase of load current.
SUPPLY POWER AND INPUT CAPACITORS
PVIN is the supply voltage for the regulator's power conversion. It is the supply that delivers the output power to
the load. The input capacitors on PVIN rail supply the large AC switching current drawn by the switching action of
the internal MOSFETs. The input current of a buck converter is discontinuous and the ripple current supplied by
the input capacitors can be quite large. The input capacitors must be rated to handle this current, as well as the
voltage. To prevent large voltage transients from occurring, a low ESR input capacitor sized for the maximum
RMS current should be used. The maximum RMS current is given by:
IRMS_CIN = IOUT
VOUT (VIN - VOUT)
PVIN
(A)
(10)
The power dissipation in the input capacitors can be found by:
PD_CIN = I2RIM_CINRESR_CIN(W)
(11)
where RESR_CIN is the ESR of the input capacitor.
This equation has a maximum at PVIN = 2×VOUT, where IRMS≊ IOUT/2 and D=50%. This simple worst-case
condition is commonly used for design purposes because even significant deviations from the worst case duty
cycle operating point do not offer much relief. Note that ripple current ratings from capacitor manufacturers are
often based on only 2000 hours of life which makes it advisable to further derate the capacitor, or 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. For low-input voltage applications, sufficient bulk input capacitance is needed
to minimize transient effects during output load changes. A 1µF ceramic bypass capacitor is also recommended
directly adjacent the PVIN and PGND pins. Please refer to the PCB LAYOUT CONSIDERATIONS section
provided later in this document.
AVIN FILTER AND VPWI BYPASS
AVIN is the supply voltage for the internal control circuitry and the MOSFET drivers. An RC filter should be
added to prevent any switching noise on PVIN from interfering with the internal analog circuitry connected to
AVIN. These can be seen on the schematic as components RF and CF. There is a practical limit to the size of the
resistor RF as the AVIN pin will draw a short 60 mA burst of current during startup, and if RF is too large the
resulting voltage drop can trigger the UVLO comparator. A recommended 1 µF CF capacitor coupled with a 1Ω
resistor provides approximately 10 dB of attenuation at the 500 kHz switching frequency.
VPWI is the supply voltage to the PWI interface. The VPW voltage range is 1.8 V to 3.3 V, allowing ±10%
voltage variation. It is recommended that VPWI voltage ramp up time is less than 500 µs. A 1 µF bypass
capacitor is recommended to bypass the VPWI pin to ground.
SWITCHING FREQUENCY
The LM10500 supports a wide range of switching frequencies: 300 kHz to 1.5 MHz. The choice of switching
frequency is usually a compromise between conversion efficiency and the size of the circuit. Lower switching
frequency implies reduced switching losses (including gate charge losses, switch transition losses, etc.) and
usually result in a higher overall efficiency. However, higher switching frequency allows use of smaller LC output
filters and hence a more compact design. Lower inductance also helps transient response (higher large signal
slew rate of inductor current), and reduces the DCR loss. The optimal switching frequency is usually a trade-off
in a given application and thus meeds to be determined on a case-by-case basis. It is related to the input
voltage, output voltage, most frequent load current level(s), external component choices, and circuit size
requirement.
24
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
LM10500
www.ti.com
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
The choice of switching frequency may also be limited if an operating condition triggers TON-MIN or TOFF-MIN.
Please refer to the aforementioned MINIMUM ON-TIME CONSIDERATIONS section.
The following equation or figure can be used to calculate the resistance to obtain a desired frequency of
operation: .
fS[kHz] = 31000 x RFRQ-0.9[kΩ]
(12)
1800
1600
FREQUENCY (kHz)
1400
1200
1000
800
600
400
200
0
0
20
40
60
80 100 120 140 160 180
RFRQ (kŸ)
Figure 30. External Resistor Selection To
Set The Switching Frequency
INDUCTOR
A general recommendation for the filter inductor in an LM10500 application is to keep a peak-to-peak ripple
current between 20% and 40% of the maximum DC load current of 5A.
The peak-to-peak current ripple can be calculated by:
'iLp-p =
(1 ± D) x VOUT
VOUT
fS x L
fS x L
(1 ±
VOUT
VPVIN
)
(13)
The current ripple is larger with smaller inductance and/or lower switching frequency. It is recommended to
choose L such that:
(1 ± D) x VOUT
fS x 0.4 x IL(MAX)
7L7
(1 ± D) x VOUT
fS x 0.2 x IL(MAX)
(14)
The inductor should be rated to handle the maximum load current plus the ripple current:
(15)
An inductor with saturation current higher than the over current protection limit is a safe choice.
In general, it is preferable to choose lower inductance in switching power supplies, because it usually
corresponds to faster transient response, smaller DCR, and reduced size for more compact designs. But too low
of an inductance can generate too large of an inductor current ripple such that over current protection at the full
load could be falsely triggered. It also generates more conduction loss, since the RMS current is slightly higher
relative that with lower current ripple at the same DC current. Larger inductor current ripple also implies larger
output voltage ripple with the same output capacitors. With peak current mode control, it is not recommended to
have too small of an inductor current ripple, so that the peak current comparator has enough signal-to-noise
ratio.
Once the inductance is determined, the type of inductor must be selected. Ferrite designs have very low core
losses and are preferred at high switching frequencies, so design goals can concentrate on copper loss and
preventing saturation. Ferrite core material saturates hard, which means that inductance collapses abruptly when
the peak design current is exceeded. The ‘hard’ saturation results in an abrupt increase in inductor ripple current
and consequent output voltage ripple. Do not allow the core to saturate!
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
25
LM10500
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
www.ti.com
OUTPUT CAPACITOR
The device is designed to be used with a wide variety of LC filters. It is generally desired to use as little output
capacitance as possible to keep cost and size down. The output capacitor(s), COUT, should be chosen with care
since it directly affects the steady state output voltage ripple, loop stability and the voltage over/undershoot
during load current transients.
The output voltage ripple is essentially composed of two parts. One is caused by the inductor current ripple going
through the Equivalent Series Resistance (ESR) of the output capacitors:
(16)
The other is caused by the inductor current ripple charging and discharging the output capacitors:
'iLp-p
ÂVOUT-C =
8fSCOUT
(17)
The two components in the voltage ripple are not in phase, so the actual peak-to-peak ripple is smaller than the
sum of the two peaks:
ÂVOUT < 'iLp-p x (
1
8fSCOUT
+ ESR
)
(18)
Figure 31 shows an illustration of two ripple components.
IL
ÂiLp-p
time
Âvout-ESR
time
Âvout-C
time
Figure 31. Two Components of VOUT Ripple
Output capacitance is usually limited by transient performance specifications if the system requires tight voltage
regulation with presence of large current steps and fast slew rates. When a fast large load transient happens,
output capacitors provide the required charge before the inductor current can slew to the appropriate level. The
initial output voltage step is equal to the load current step multiplied by the ESR. VOUT continues to droop until
the control loop response increases or decreases the inductor current to supply the load. To maintain a small
over- or undershoot during a transient, small ESR and large capacitance are desired. But these also come with
higher cost and size. Thus, the motivation is to seek a fast control loop response to reduce the output voltage
deviation.
One or more ceramic capacitors are generally recommended because they have very low ESR and remain
capacitive up to high frequencies. The capacitor dielectric should be X5R, X7R, or comparable material to
maintain proper tolerance. Other types of capacitors also can be used, particularly if large bulk capacitance is
needed (such as tantalum, POSCAP and OSCON). Such capacitors have lower ESR zero {1/(2πESR×C)}
frequency than ceramic capacitors. The lower ESR zero frequency can influence the control loop, particularly if it
occurs close to the desired crossover frequency. If high switching frequency and high crossover frequency are
desired, an all ceramic design would be more appropriate.
26
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
LM10500
www.ti.com
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
EFFICIENCY CONSIDERATIONS
The efficiency of a switching regulator is defined as the output power divided by the input power times 100%.
Efficiency also can be found by
Total Power Loss
K=1Input Power
(19)
It is often useful to analyze individual losses to determine what is limiting the efficiency and which change would
produce the most improvement. Although all dissipative elements in the circuit produce losses, three main
sources usually account for most of the losses in the LM10500-based converters: 1) conduction losses, 2)
switching and gate driving losses, 3) bias losses.
Conduction losses are the I2R losses on parasitic resistances in the path of the output current, including ON-time
resistances of the internal switches (RDS-ON), equivalent inductor DC resistance Rdcr, and PC board trace
resistances Rtrace. The conduction loss can be approximated by:
Wcond-loss=I2OUT(D×RDS-ON-HS+(1–D)×RDS-ON-LS+DCR+Rtrace)
(20)
The total conduction loss can be reduced by reducing these parasitic resistances. For example, the LM10500 is
designed to have low RDS-ON internal MOSFET switches. The inductor DCR should be small. The traces that
conduct the current should be wide, thick and as short as possible. The conduction losses affect the efficiency
more at heavier load.
Switching losses include all the losses generated by the switching action of two power MOSFETs. Each time the
switch node swings from low to high or vice versa, charges are applied or removed from the parasitic
capacitance from the SW node to GND. Each time a power MOSFET gate is switched from low to high and to
low again, a packet of charge moves from VDD1 to ground. Further more, each time a power MOSFET is turned
on or off, a transition loss is generated relative to the overlapping area of voltage and current. MOSFET parasitic
diodes generate reverse recovery loss and dead time conduction loss. RMS currents through the input and
output capacitor ESR also generates loss. All of these losses should be evaluated and carefully considered to
design a high efficiency switching power converter. Since these losses only occur during ‘switching’, reducing the
switching frequency always help to reduce the switching loss. The resultant improvement in efficiency is more
pronounced at lighter load.
AVIN provides MOSFET driving voltage and control circuit bias voltage. One part of loss on AVIN is from
providing the current to the drivers, equals to VAVIN x idrive. The other portion of AVIN power loss is the bias
current through VDD2, equals to VAVIN x ibias. idrive does not change for AVIN above 5.0 V. If AVIN is below 5.0V,
idrive will reduce slightly, but RDS-ON of the switches will increase and reduce efficiency more. ibias is constant with
AVIN between 3.0 V and 18.0 V. Powering AVIN from a 5 V system rail provides an optimal tradeoff between
bias power loss and switching loss (RDS-ON loss).
LOAD CURRENT DE-RATING AT HIGHER DUTY CYCLE OPERATION
The LM10500 is optimized for lower duty cycle operation, e.g. high input to output voltage ratio. The high-side
MOSFET is designed to be half the size of the low-side MOSFET, thus optimizing the relative levels of switching
loss in the high-side switch and the conduction loss in the low-side switch. The continuous current rating in the
low-side switch is the maximum load current of 5 A, while the high-side switch is rated at 2.5 A. If the LM10500 is
operating with duty cycle higher than 50%, the maximum output current should be derated.
IOUT-max = 5 × min[ (1.5 - D) , 1]
(21)
Derating of maximum load current when D > 50% is also illustrated in Figure 32:
IOUT-MAX
5.0A
2.5A
0
50%
100%
D
Figure 32. LM10500 Maximum Load Current Derating When D > 50%
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
27
LM10500
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
www.ti.com
CONTROL LOOP COMPENSATION
VC
VREF +
VO
+
-
Fc(s)
Compensator
Fp(s) x Fh(s)
Power Stage
Loop (T)
H
Feedback
Figure 33. Control Block Diagram Of A
Current Mode Controlled Buck Converter
The LM10500 employs a current mode controller and therefore the control block diagram representation involves
two feedback loops (see Figure 33). The inner feedback loop derives its feedback from the sensed inductor
current, while the outer loop monitors the output voltage. This section will not provide a rigorous analysis of
current mode control, but rather a simplified yet relatively accurate method to determine the control loop
compensation network.
The LM10500 compensation components from COMP pin to AGND are shown in Figure 34. The purpose of the
compensator block is to stabilize the control loop and achieve high performance in terms of the transient
response, audio susceptibility and output impedance. The LM10500 will typically require only a single resistor Rc
and capacitor Cc1 for compensation. However, depending on the location of the power stage ESR zero, a second
(small) capacitor, Cc2, may be required between COMP and AGND to create a high-frequency pole.
LM10500
COMP
RC
CC1
Figure 34. LM10500 Compensation Network
The overall loop transfer function is a product of the power stage transfer function, internal amplifier gains and
the feedback network transfer function, and can be express by
T = Gain0Fp(s)Fh(s)Fcomp(s)
(22)
where Gain0 includes all the DC gains in the loop, Fp(s) represents the power stage pole and zero (including the
inner current loop), Fh(s) represents the sampling effect in such a switch-mode converter and Fcomp(s) is the
transfer function of the external compensator. Figure 35 shows an asymptotic approximation plot of the loop gain.
Gain0 /sCc1
fp
fzcomp
0 dB
fc
fs/2
fESR
(fpcomp)
Figure 35. LM10500 Loop Gain Asymptote Approximation
The loop gain determines both static and dynamic performance of the converter. The power stage response is
fixed by the selection of the power components and the compensator is therefore designed around the power
stage response to achieve the desired loop response. The goal is to design a control loop with high crossover
frequency (or loop bandwidth) and adequate phase margin under all operation conditions.
28
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
LM10500
www.ti.com
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
SELECT COMPENSATION COMPONENTS
To select the compensation components, a desired crossover frequency, fc, needs to be selected. It is
recommended to select fc ≤1/6 fS(switching frequency). The effect of Fh(s) can be ignored in this range to simplify
the design. The capacitor ESR zero is also assumed to be at least 3 times higher than fc. The compensation
resistor can be found by
(23)
Cc1 does not affect the crossover frequency fc, but it sets the compensator zero fzcomp and affects the phase
margin of the loop. For a fast design, Cc1 = 10 nF gives adequate performance in most LM10500 applications.
Larger Cc1 gives larger phase margin but in the expense of longer transient response settling time. Lower Cc1
gives higher gain at lower frequency thus faster transient response. It is recommended to set the compensation
zero no higher than fc/3 to ensure enough phase margin, implying
Cc1 8
3
2SRcfc
(24)
PLOTTING THE LOOP GAIN
The include the effect of Fh(s) and ESR zero, the complete loop gain can be plotted using a software tool, such
as Matlab, Mathcad, or Excel. The components in the loop gain can by found as follows. The constant Gain0 in
the loop, as shown in Figure 35, can be found by
(25)
where fs is the switching frequency; mc can be found by
mc = 1 +
4 x fS x L
VIN - VOUT
(26)
and
D' = 1−D
(27)
Minimum ROUT should be used in the calculation:
ROUT = VOUT/IOUT
(28)
Fp(s) can be expressed by:
Fp(s) =
1 + sCOUT * ESR
s
1+
2S ´
¶p
(29)
where the power stage pole considering the slope compensation effect is:
fp =
1
1
1
(mcD' - 0.5))
(
+
2SCOUT ROUT fsL
(30)
The high frequency behavior Fh(s) can be expressed by:
1
Fh(s) =
1+
s2
s
+ 2
wnQp wn
(31)
where
wn = Sfs and Qp =
1
S(mcD' ± 0.5)
(32)
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
29
LM10500
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
www.ti.com
The compensation network transfer function is:
(33)
With above equations, the loop gain
T = Gain0Fp(s)Fh(s)Fcomp(s)
(34)
can be plotted and more accurate loop performance metrics (crossover frequency and phase margin) can be
found.
HIGH-FREQUENCY CONSIDERATIONS
Fh(s) represents the additional magnitude and phase drop around fs/2 caused by the switching behavior of the
current mode converter. Fh(s) contains a pair of double poles with quality factor Qp at half of the switching
frequency. It is a good idea to check that Qp is between 0.15 and 2, ideally around 0.6. If Qp is too high, the
resonant peaking at fs/2 could become severe and coincide with subharmonic oscillations in the duty cycle and
inductor current. If Qp is too low, the two complex poles split, the converter begins to act like a voltage mode
controlled converter and the compensation scheme used above should be changed.
Fp(s) also contains the ESR zero of the output capacitors:
fESR =
1
2SCOUTESR
(35)
In a typical ceramic capacitor design, fESR is at least three times higher than the desired crossover frequency fc. If
fESR is lower tha fs/2, an additional capacitor Cc2 can be added between the COMP pin and AGND to give a highfrequency pole.
Cc2 =
1
2SRcfESR
(36)
CC2 should be and usually is much smaller than CC1 to avoid affecting the compensation zero.
BOOTSTRAP CAPACITOR
A ceramic capacitor is needed between the CBOOT pin to the SW node to supply the gate drive charge when
the high-side switch is turning ON. The capacitance should be large enough to supply the charge without
significant voltage drop. A 0.1 µF bootstrap capacitor is recommended in LM10500 applications.
POWERWISE INTERFACE ADDRESS SELECTION
External 1% resistor connecting between the ADDR pin to AGND sets the PWI address.
PWI Standard
Description
Typ
Unit
Address selection resistor for PWI-1.0
≤20
kΩ
RPWI2.0-0
Address selection resistor for PWI-2.0, address 0
40.2
kΩ
RPWI2.0-1
Address selection resistor for PWI-2.0, address 1
60.4
kΩ
RPWI2.0-2
Address selection resistor for PWI-2.0, address 2
80.6
kΩ
RPWI2.0-3
Address selection resistor for PWI-2.0, address 3
100
kΩ
RPWI1.0
The external resistance is only sensed one time when the part is powered up. If the address selection resistor is
modified after power up, it won’t take effect until a power cycling is performed.
VDD1 AND VDD2 BYPASS CAPACITORS
VDD1 and VDD2 pins are internal LDO outputs. As previously mentioned, the two LDOs are used for internal
circuits only and should not be substantially loaded.
Bypass capacitors are needed to stabilize the LDOs. Ceramic capacitors within a specified range should be used
to meet stability requirements. The dielectric should be X5R, X7R, or better and rated for the required operating
temperature range. Use the following table to choose suitable LDO bypass capacitor.
30
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
LM10500
www.ti.com
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
Output voltage
NOMINAL (V)
Output Capacitance Range
(recommended typical value)
VDD1
4.88
1 µF ± 20% 16V
VDD2
2.47
0.1 µF ± 20% 10V
PCB LAYOUT CONSIDERATIONS
PC board layout is an important part of DC/DC converter design. Poor PC board layout can disrupt the
performance of a DC/DC converter and surrounding circuitry by contributing to EMI, noise coupling, ground
bounce, and resistive voltage loss in the traces, and thermal problems. Erroneous signals can reach the DC-DC
converter, possibly resulting in poor regulation or in instability.
Good PCB layout for an LM10500-based converter can be implemented by following a few simple design rules.
1. Provide adequate device heat sinking by utilizing the PAD of the IC as the primary thermal path. Use a 3 by
3 array of 10 mil thermal vias to connect the PAD to the system ground plane heat sink. The vias should be
evenly distributed under the PAD. Use a four-layer board with the copper thickness for the four layers,
starting from the top one, 2 oz / 1 oz / 1 oz / 2 oz. Four layer boards with enough copper thickness provides
low current conduction impedance and proper shielding.
2. It is imperative that the input capacitors be located as close as possible to the PVIN pins; the inductor should
be placed as close as possible to the SW pins and the output capacitors. This is to minimize the area of
switching current loops to reduce EMI, and reduce the resistive loss on the high current path. The copper
area of the switch node should be thick and short to provide a good conduction path for the switch node
current to the inductor. Make input and output power bus connections as wide and short as possible. This
reduces any voltage drops on the input or output of the converter and can improve efficiency. Use copper
plates/planes on the top layer to connect the multiple PVIN pins together and PGND pins together.
3. All bypass capacitors should be placed as close as possible to the corresponding pin and ground. Based on
the LM10500 pinout, a 1 µF to 10 µF capacitor can be placed right by pins 1, 2 and pin 7, across the SW
node trace, as an addition to the bulk input capacitors. Using a size 1206 or larger capacitor allows enough
copper width for the switch node to be routed underneath the capacitor for good conduction (see evaluation
board layout in application node AN-2080).
4. It is recommended to use one of the middle layers as a solid ground plane. Ground plane provides shielding
for sensitive circuits and traces. It also provides a quiet reference potential for the control circuitry. The
AGND and DGND pins should be connected to the ground plane using vias right next to the bypass
capacitors. DGND should also be connected to the source where VPWI is provided. PGND pins are
connected to the source of the internal low-side switch. They should be connected directly to the grounds of
the input and output capacitors. The PGND net contains noise at switching frequency and may bounce due
to load variations. PGND trace, as well as PVIN and SW traces, should be constrained to one side of the
ground plane. The other side of the ground plane contains much less noise and should be used for sensitive
routes.
5. To reduce noise sensitivity of the output voltage feedback path, it is important to place the resistor divider
close to the FB pin, rather than close to the load. The FB pin is the input to the error amplifier, so it is a high
impedance node and very sensitive to noise. Placing the resistor divider closer to the FB pin reduces the
trace length of FB signal and reduces noise coupling. The output node is a low impedance node, so the trace
from Vout to the resistor divider can be long if short path is not available. The COMP is also a noise sensitive
node and the compensation components should be located as close as possible to the IC.
6. If voltage accuracy at the load is important, make sure voltage sense is made at the load. Doing so will
correct for voltage drops along the traces and provide the best output accuracy. The voltage sense trace
from the load to the feedback resistor divider should be routed away from the SW node path and the inductor
to avoid contaminating the feedback signal with switch noise, while also minimizing the trace length. This is
most important when high value resistors are used to set the output voltage. It is recommended to route the
voltage sense trace and place the resistor divider on a different layer than the inductor and SW node path
such that there is a ground plane in between the feedback trace and inductor/SW node polygon. This
provides further shielding for the voltage feedback path from EMI noise.
7. The 0.1 µF boot capacitor connected between the CBOOT pin and the SW node should be placed as close
as possible to the CBOOT pin and SW pins.
8. The frequency set resistor and its associated capacitor should be placed as close as possible to the FREQ
pin.
9. The PWI address set resistor should be place as close as possible to the ADDR pin.
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
31
LM10500
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
www.ti.com
10. The traces to SCLK and SPWI pins should be routed parallel to each other and as short as possible. If this
is a multi-master and/or multi-slave system, care should be taken in matching the trace lengths of all
segments of the PWI bus. Additionally, the designer must ensure that the electrical characteristics of
interconnect do not violate the restrictions in the PowerWise Interface 1.0 / 2.0 specifications.
THERMAL CONSIDERATIONS
The thermal characteristics of the LM10500 are specified using the parameter θJA, which relates junction
temperature to ambient temperature in a particular application. Although the value of θJA is dependant on many
variables, it still can be used to approximate the operating junction temperature of the device. To obtain an
estimate of the device junction temperature, one may use the following relationship:
TJ= PDθJA+TA
where
•
•
•
•
•
•
•
TJ = Junction temperature in °C
PD = PIN x (1 − Efficiency) − 1.1 x IOUT x DCR
PIN = Input power in Watts (PIN = VIN x IIN)
IOUT = Output load current
DCR = Inductor DC parasitic resistance.
θJA = Junction-to-ambient thermal resistance of the LM10500 in °C/W
TA = Ambient temperature in °C
(37)
It is important to always keep the LM10500 operating junction temperature (TJ) below 125 °C to ensure reliable
operation. If the junction temperature exceeds 160 °C, the device will cycle in and out of thermal shutdown. If
thermal shutdown occurs, it is a sign of inadequate heat-sinking and/or excessive power dissipation in the
device. PC Board heat-sinking can be improved by using more thermal vias, a larger board, or more heatspreading layers within the board.
Application Example
AVS FB 0.6V to 1V
C4
PWROK
R2
C5
8
PGND
PGND
9
10
PGND
11
COMP
13
12
PWROK
15
EN
FB
ADDR
14
R3
7
EN
SW
FREQ
SPWI
SW
SCLK
PVIN
2
PVIN
3V to 18V
1
VDD2
PVIN
VDD1
AGND
24
25
AVIN
23
22
AVIN
C7
C9
C8
C1
PVIN
21
C3
28
20
PVIN
SPWI
SCLK
4
SW
VOUT
0.6 to 5V
5A
3
CBOOT
19
VPWI
L1
SW
27
18
C6
6
5
U1
LM10500
DGND
26
17
R4
R5
PGND
16
SYNC
R6
C2
R1
Figure 36. Application Circuit Example
32
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
LM10500
www.ti.com
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
Bill Of Materials
The table below shows BOMs for a VOUT = 0.8 V design and for a VOUT = 1.2 V design. In both applications,
PVIN = AVIN = 12 V, fs = 300 kHz, IOUT-MAX = 5 A, and PWI 1.0 are used. For information on the LM10500
evaluation boards, please refer to AN-2080.
Table 7. Bill of Materials for LM10500 0.8 V and 1.2 V Applications
Designator
For 0.8V design
For 1.2V design
Manufacture
U1
LM10500SQ-0.8
LM10500SQ-1.0
TI
1
TANT 47µF 25V
TANT 47µF 25V
KEMET
1
CERAMIC 10µF 50V X7S
CERAMIC 10µF 50V X7S
TAIYO YUDEN
1
CERAMIC 1.0 µF 35V X5R
CERAMIC 1.0 µF 35V X5R
TAIYO YUDEN
1
CERAMIC 0.1 µF 50V X7R
CERAMIC 0.1 µF 50V X7R
TAIYO YUDEN
2
CERAMIC 47 µF10V X5R
CERAMIC 47 µF10V X5R
MURATA
1
C1
C2, C7
C3
Qty
220 µF POLYMER 6.3V
220 µF POLYMER 6.3V
PANASONIC
1
C4
CERAMIC 10000 pF 25V
CERAMIC 10000 pF 25V
TDK
1
C5
CERAMIC 100 pF 100V
CERAMIC 100 pF 100V
PANASONIC
1
C6, C8, C9
CERAMIC 1.0 µF 35V X5R
CERAMIC 1.0 µF 35V X5R
TAIYO YUDEN
3
L1
1.2 µH SMD INDUCTOR
2.2 µH SMD INDUCTOR
COILCRAFT
1
R1
1Ω 0603 1%
1Ω 0603 1%
YAGEO
1
R2
1.75 KΩ 0603 1%
2kΩ 0603 1%
YAGEO
1
R3
0 Ω 0603 1%
0 Ω 0603 1%
YAGEO
1
R4
169 KΩ 0603 1%
169 kΩ 0603 1%
YAGEO
1
R5
0 RΩ 0603 1%
2kΩ 0603 1%
YAGEO
1
R6
open
10.0 kΩ 0603 1%
YAGEO
1
PCB Layout Example
An example of an LM10500 PCB layout is shown in Figure 37. Only the top layer and the silk screen are shown.
For more details, please refer to application note AN-2080 SNVA453.
Figure 37. PCB Layout Example: Top Layer and Silk Screen
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
33
LM10500
SNVS630G – SEPTEMBER 2009 – REVISED MARCH 2013
www.ti.com
REVISION HISTORY
Changes from Revision F (March 2013) to Revision G
•
34
Page
Changed layout of National Data Sheet to TI format .......................................................................................................... 33
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Product Folder Links: LM10500
PACKAGE OPTION ADDENDUM
www.ti.com
10-Dec-2020
PACKAGING INFORMATION
Orderable Device
Status
(1)
Package Type Package Pins Package
Drawing
Qty
Eco Plan
(2)
Lead finish/
Ball material
MSL Peak Temp
Op Temp (°C)
Device Marking
(3)
(4/5)
(6)
LM10500SQ-0.8/NOPB
ACTIVE
WQFN
RSG
28
1000
RoHS & Green
SN
Level-2A-260C-4
WEEK
-40 to 85
1050008
LM10500SQ-1.0/NOPB
ACTIVE
WQFN
RSG
28
1000
RoHS & Green
SN
Level-2A-260C-4
WEEK
-40 to 85
1050010
LM10500SQE-0.8/NOPB
ACTIVE
WQFN
RSG
28
250
RoHS & Green
SN
Level-2A-260C-4
WEEK
1050008
LM10500SQE-1.0/NOPB
ACTIVE
WQFN
RSG
28
250
RoHS & Green
SN
Level-1-260C-UNLIM
1050010
LM10500SQX-0.8/NOPB
ACTIVE
WQFN
RSG
28
4500
RoHS & Green
SN
Level-2A-260C-4
WEEK
-40 to 85
1050008
LM10500SQX-1.0/NOPB
ACTIVE
WQFN
RSG
28
4500
RoHS & Green
SN
Level-2A-260C-4
WEEK
-40 to 85
1050010
(1)
The marketing status values are defined as follows:
ACTIVE: Product device recommended for new designs.
LIFEBUY: TI has announced that the device will be discontinued, and a lifetime-buy period is in effect.
NRND: Not recommended for new designs. Device is in production to support existing customers, but TI does not recommend using this part in a new design.
PREVIEW: Device has been announced but is not in production. Samples may or may not be available.
OBSOLETE: TI has discontinued the production of the device.
(2)
RoHS: TI defines "RoHS" to mean semiconductor products that are compliant with the current EU RoHS requirements for all 10 RoHS substances, including the requirement that RoHS substance
do not exceed 0.1% by weight in homogeneous materials. Where designed to be soldered at high temperatures, "RoHS" products are suitable for use in specified lead-free processes. TI may
reference these types of products as "Pb-Free".
RoHS Exempt: TI defines "RoHS Exempt" to mean products that contain lead but are compliant with EU RoHS pursuant to a specific EU RoHS exemption.
Green: TI defines "Green" to mean the content of Chlorine (Cl) and Bromine (Br) based flame retardants meet JS709B low halogen requirements of