INTERSIL ISL6265AHRTZ-T

ISL6265A
®
Data Sheet
May 11, 2009
Features
• Differential Remote CPU Die Voltage Sensing
• Core Differential Current Sensing: DCR or Resistor
• Northbridge Lossless rDS(ON) Current Sensing
• Serial VID Interface
- Two Wire Clock and Data Bus
- Supports High-Speed I2C
- 0.500V to 1.55V in 12.5mV Steps
- Supports PSI_L Power-Saving Mode
• Core Outputs Feature Phase Shedding with PSI_L
• Adjustable Output-Voltage Offset
• Digital Soft-Start of all Outputs
• User Programmable Switching Frequency
• Static and Dynamic Current Sharing (Uniplane Core)
• Overvoltage, Undervoltage, and Overcurrent Protection
• Pb-Free (RoHS compliant)
Pinout
Ordering Information
PART NUMBER
(Note)
ISL6265AHRTZ
UGATE_NB
PHASE_NB
LGATE_NB
PGND_NB
OCSET_NB
RTN_NB
48 47 46 45 44 43 42 41 40 39 38 37
TEMP
RANGE
(°C)
PART
MARKING
VSEN_NB
ISL6265A
(48 LD 6X6 TQFN)
TOP VIEW
FSET_NB
A unity-gain differential amplifier is provided for remote CPU
die sensing. This allows the voltage on the CPU die to be
accurately regulated per AMD mobile CPU specifications.
Core output current sensing is realized using lossless
inductor DCR sensing. All outputs feature overcurrent,
overvoltage and undervoltage protection.
• Internal Gate Drivers with 2A Driving Capability
COMP_NB
The Serial VID Interface (SVI) allows dynamic adjustment of
the Core and Northbridge output voltages independently and
in combination from 0.500V to 1.55V. Core and Northbridge
output voltages achieve a 0.5% system accuracy
over-temperature.
• Voltage Positioning with Adjustable Load Line and Offset
FB_NB
The heart of the ISL6265A is the patented R3 Technology™,
Intersil’s Robust Ripple Regulator modulator. Compared with
the traditional buck regulator, the R3 Technology™ has a
faster transient response. This is due to the R3 modulator
commanding variable switching frequency during a load
transient.
• Precision Voltage Regulators
- 0.5% System Accuracy Over-temperature
VCC
The ISL6265A is a multi-output controller with embedded
gate drivers. A single-phase controller powers the
Northbridge (VDDNB) portion of the CPU. The two
remaining controller channels can be configured for
two-phase or individual single-phase outputs. For uniplane
CPU applications, the ISL6265A is configured as a
two-phase buck converter. This allows the controller to
interleave channels to effectively double the output voltage
ripple frequency, and thereby reduce output voltage ripple
amplitude with fewer components, lower component cost,
reduced power dissipation, and smaller area. For dual-plane
processors, the ISL6265A can be configured as independent
single-phase controllers powering VDD0 and VDD1.
• Core Configuration Flexibility
- Dual Plane, Single-Phase Controllers
- Uniplane, Two-Phase Controller
VIN
Multi-Output Controller with Integrated
MOSFET Drivers for AMD SVI Capable
Mobile CPUs
FN6884.0
PACKAGE
(Pb-Free)
PKG.
DWG. #
6265A HRTZ -10 to +100 48 Ld 6x6 TQFN L48.6x6
ISL6265AHRTZ-T* 6265A HRTZ -10 to +100 48 Ld 6x6 TQFN L48.6x6
Tape and Reel
36 BOOT_NB
OFS/VFIXEN 1
PGOOD 2
35 BOOT0
34 UGATE0
PWROK 3
SVD 4
33 PHASE0
32 PGND0
SVC 5
31 LGATE0
49
GND
[BOTTOM]
ENABLE 6
RBIAS 7
30 PVCC
*Please refer to TB347 for details on reel specifications.
OCSET 8
29 LGATE1
NOTE: These Intersil Pb-free plastic packaged products employ
special Pb-free material sets, molding compounds/die attach materials,
and 100% matte tin plate plus anneal (e3 termination finish, which is
RoHS compliant and compatible with both SnPb and Pb-free soldering
operations). Intersil Pb-free products are MSL classified at Pb-free peak
reflow temperatures that meet or exceed the Pb-free requirements of
IPC/JEDEC J STD-020
VDIFF0 9
FB0 10
28 PGND1
COMP0 11
VW0 12
26 UGATE1
1
27 PHASE1
25 BOOT1
ISN1
ISP1
VW1
COMP1
FB1
VDIFF1
VSEN1
RTN1
RTN0
VSEN0
ISP0
ISN0
13 14 15 16 17 18 19 20 21 22 23 24
CAUTION: These devices are sensitive to electrostatic discharge; follow proper IC Handling Procedures.
1-888-INTERSIL or 1-888-468-3774 | Intersil (and design) is a registered trademark of Intersil Americas Inc.
Copyright Intersil Americas Inc. 2009. All Rights Reserved
All other trademarks mentioned are the property of their respective owners.
ISL6265A
Function Block Diagram
RTN_NB
COMP_NB
FB_NB
VSEN_NB
FSET_NB
PVCC
IFSET_NB
VNB
1
PWROK
NO DROOP
PSI_L
I_OFS
VREF_NB
VREF0
SVI
INTERFACE
AND DAC
SVD
FLT
1.5kW
1.5kΩ
SVC
BOOT_NB
3.0kΩ
UGATE_NB
MODULATOR
NB
E/A
VREF_NB
PHASE_NB
SHOOT-THRU
PROTECTION
LGATE_NB
DE MODE
PGND_NB
VIN
VREF1
OFS/FIXEN
MOSFET
DRIVER
PSI_L
PVCC
VCC
OCSET_NB
FLT
VNB
V0
V1
ISEN0
ISEN1
FAULT
PROTECTION
OCSET
RBIAS
RTN1
POWER-ON
RESET AND
SOFT-START
LOGIC
ENABLE
PGOOD
GND
MODE
VW0
PVCC
IVW0
VIN
COMP0
BOOT0
FB0
FLT
E/A
I_OFS
VDIFF0
VIN
V0
1
∑
RTN0
ISP0
ISN0
NO
DROOP
ISN1
ISEN0
MODULATOR
CORE
CURRENT
SENSE
ISEN1
LGATE0
DE MODE
PGND0
PSI_L
FLT
CURRENT
BALANCE
BOOT1
UGATE1
NO
DROOP
MOSFET
DRIVER
MODE
V1
1
SHOOT-THRU
PROTECTION
PVCC
CURRENT
SENSE
VSEN1
PHASE0
MODE
MODE
ISP1
MOSFET
DRIVER
VREF0
VSEN0
UGATE0
∑
VREF1
RTN1
PHASE1
SHOOT-THRU
PROTECTION
LGATE1
DE MODE
PGND1
E/A
VDIFF1
I_OFS
PSI_L
IVW1
FB1
COMP1
VW1
FIGURE 1. SIMPLIFIED FUNCTION BLOCK DIAGRAM OF ISL6265A
2
FN6884.0
May 11, 2009
ISL6265A
Simplified Application Circuit for Dual Plane and Northbridge Support
+5V
VCC PVCC
VIN
VIN
GND
+VIN
SVI DATA
SVD
SVI CLOCK
SVC
CIN
UGATE0
ENABLE
EN
PWROK
PWROK
BOOT0
VDDPWRGD
PGOOD
PHASE0
VSEN0
LGATE0
REMOTE
SENSE
RTN0
REMOTE
SENSE
LOUT
CORE
LOAD
PGND0
VSEN1
ISP0
RTN1
ISN0
VDD0
VDD_PLANE_STRAP
RBIAS
OFS/VFIXEN
OCSET
VDIFF0
+VIN
CIN
FB0
UGATE1
COMP0
ISL6265A
BOOT1
LOUT
VDD1
PHASE1
VW0
LGATE1
CORE
LOAD
PGND1
VDIFF1
ISP1
ISN1
+VIN
FB1
CIN
COMP1
UGATE_NB
BOOT_NB
VW1
LOUT
VDDNB
PHASE_NB
FSET_NB
LGATE_NB
PGND_NB
COMP_NB
NB
LOAD
OCSET_NB
VSEN_NB
FB_NB
RTN_NB
FIGURE 2. ISL6265A BASED DUAL-PLANE AND NORTHBRIDGE CONVERTERS WITH INDUCTOR DCR CURRENT SENSING
3
FN6884.0
May 11, 2009
ISL6265A
Simplified Application Circuit for Uniplane Core and Northbridge Support
+5V
VIN
+VIN
GND
VCC PVCC
CIN
SVI DATA
SVD
SVI CLOCK
SVC
ENABLE
EN
PWROK
PWROK
VDDPWRGD
PGOOD
UGATE0
BOOT0
PHASE0
LGATE0
CORE
LOAD
PGND0
VSEN0
REMOTE
SENSE
LOUT
ISP0
RTN0
ISN0
REMOTE
SENSE
VSEN1
VDD_PLANE_STRAP
RTN1
RBIAS
OCSET
VDD0
OFS/VFIXEN
+VIN
VDIFF0
CIN
UGATE1
FB0
BOOT1
COMP0
ISL6265A
LOUT
PHASE1
LGATE1
VW0
CORE
LOAD
PGND1
ISP1
OPEN
VDIFF1
OPEN
FB1
ISN1
+VIN
CIN
OPEN
COMP1
OPEN
VW1
UGATE_NB
BOOT_NB
LOUT
VDDNB
PHASE_NB
LGATE_NB
FSET_NB
PGND_NB
COMP_NB
NB
LOAD
OCSET_NB
VSEN_NB
RTN_NB
FB_NB
FIGURE 3. ISL6265A BASED UNIPLANE AND NORTHBRIDGE CONVERTERS WITH INDUCTOR DCR CURRENT SENSING
4
FN6884.0
May 11, 2009
ISL6265A
Simplified Application Circuit for Dual Layout
+5V
VIN
+VIN
GND
VCC PVCC
CIN
SVI DATA
SVD
SVI CLOCK
SVC
ENABLE
EN
PWROK
PWROK
VDDPWRGD
PGOOD
BOOT0
VDD0
LGATE0
CORE
LOAD
PGND0
ISP0
RTN0
VDD_PLANE_STRAP
+1.8V
LOUT
PHASE0
VSEN0
REMOTE
SENSE
REMOTE
SENSE
UGATE0
ISN0
RTN1
DNP UNIPLANE
VSEN1
RBIAS
DNP
DUAL
PLANE
UNIPLANE
VDD0
OCSET
OFS/VFIXEN
+VIN
VDIFF0
CIN
UGATE1
FB0
BOOT1
COMP0
ISL6265A
LOUT
PHASE1
LGATE1
VW0
VDD1
CORE
LOAD
PGND1
ISP1
POPULATION OPTIONAL IN UNIPLANE
VDIFF1
ISN1
+VIN
FB1
CIN
COMP1
UGATE_NB
BOOT_NB
VW1
LOUT
VDDNB
PHASE_NB
LGATE_NB
FSET_NB
PGND_NB
COMP_NB
NB
LOAD
OCSET_NB
VSEN_NB
RTN_NB
FB_NB
FIGURE 4. ISL6265A BASED UNIPLANE OR DUAL PLANE CORE CONVERTER WITH INDUCTOR DCR CURRENT SENSING
5
FN6884.0
May 11, 2009
ISL6265A
Absolute Maximum Ratings
Thermal Information
Supply Voltage, VCC, PVCC . . . . . . . . . . . . . . . . . . . . . . . -0.3 - +7V
Battery Voltage, VIN. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . +28V
Boot Voltage (BOOT) . . . . . . . . . . . . . . . . . . . . . . . . . -0.3V to +33V
Boot to Phase Voltage (BOOT-PHASE). . . . . . . . -0.3V to +7V(DC)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . -0.3V to +9V (<10ns)
Phase Voltage (PHASE) . . . . . . . . . -7V (<20ns Pulse Width, 10µJ)
UGATE Voltage (UGATE) . . . . . . . . . PHASE -0.3V (DC) to BOOT
LGATE Voltage (LGATE) . . . . . . . . . . . . . -0.3V (DC) to VCC + 0.3V
ALL Other Pins. . . . . . . . . . . . . . . . . . . . . . . . -0.3V to (VCC + 0.3V)
Open Drain Outputs, PGOOD . . . . . . . . . . . . . . . . . . . . . -0.3 - +7V
Thermal Resistance (Typical, Notes 1, 2) θJA (°C/W)
θJC (°C/W)
48 Ld TQFN . . . . . . . . . . . . . . . . . . . .
32
3.5
Maximum Junction Temperature . . . . . . . . . . . . . . . . . . . . . . +150°C
Maximum Storage Temperature Range . . . . . . . . . .-65°C to +150°C
Pb-Free Reflow Profile. . . . . . . . . . . . . . . . . . . . . . . . .see link below
http://www.intersil.com/pbfree/Pb-FreeReflow.asp
Recommended Operating Conditions
Supply Voltage, VCC, PVCC . . . . . . . . . . . . . . . . . . . . . . . .+5V ±5%
Battery Voltage, VIN . . . . . . . . . . . . . . . . . . . . . . . . . . . . +6V to 24V
Ambient Temperature . . . . . . . . . . . . . . . . . . . . . . .-10°C to +100°C
Junction Temperature . . . . . . . . . . . . . . . . . . . . . . .-10°C to +125°C
CAUTION: Do not operate at or near the maximum ratings listed for extended periods of time. Exposure to such conditions may adversely impact product reliability and
result in failures not covered by warranty.
NOTES:
1. θJA is measured in free air with the component mounted on a high effective thermal conductivity test board with “direct attach” features. See Tech
Brief TB379.
2. For θJC, the “case temp” location is the center of the exposed metal pad on the package underside.
Electrical Specifications
VCC = PVCC = 5V, VIN = 12V, TA = -10°C to +100°C, Unless Otherwise Specified. Parameters with MIN and/or
MAX limits are 100% tested at +25°C, unless otherwise specified. Temperature limits established by
characterization and are not production tested
PARAMETER
SYMBOL
TEST CONDITIONS
MIN
TYP
MAX
UNITS
EN = 3.3V
-
7.8
10
mA
EN = 0V
-
-
1
µA
VCC PORr
VCC Rising
-
4.35
4.5
V
VCC PORf
VCC Falling
3.9
4.1
-
V
-
-
1
µA
-0.5
-
0.5
%
-5
-
+5
mV
1.15
1.17
1.19
V
INPUT POWER SUPPLY
+5V Supply Current
IVCC
POR (Power-On Reset) Threshold
Battery Supply Current (VIN)
IVIN
EN = 0V, VIN = 24V
SYSTEM AND REFERENCES
System Accuracy
(Vcore0, Vcore1, Vcore_NB)
%Error
(VCORE)
No load, closed loop, active mode
VID = 0.75V to 1.55V
VID = 0.50V to 0.7375V
RBIAS Voltage
RRBIAS
RRBIAS = 117kΩ
Maximum Output Voltage
VCOREx
(max)
SVID = [000_0000b]
-
1.55
-
V
Minimum Output Voltage
VCOREx
(min)
SVID = [101_0100b]
-
0.500
-
V
VIN = 15.5V, VDAC = 1.55V, VFB0 = 1.60V,
force Vcomp_0 = 2V, RVW = 6.81kΩ, 2-Phase Operation
285
300
315
kHz
RFSET_NB = 22.1kΩ, CFSET_NB = 1nF, VDAC = 0.5V,
Vsen_nb = 0.51V
285
300
315
kHz
Core Frequency Adjustment Range
200
-
500
kHz
NB Frequency Adjustment Range
200
-
500
kHz
CHANNEL FREQUENCY
Nominal CORE Switching Frequency
Nominal NB Switching Frequency
fSW_core0
fSW_core_NB
AMPLIFIERS
Error Amp DC Gain (Note 3)
-
90
-
dB
GBW
CL = 20pF
-
18
-
MHz
SR
CL = 20pF
-
5.0
-
V/µs
AV0
Error Amp Gain-Bandwidth Product
(Note 3)
Error Amp Slew Rate (Note 3)
6
FN6884.0
May 11, 2009
ISL6265A
Electrical Specifications
VCC = PVCC = 5V, VIN = 12V, TA = -10°C to +100°C, Unless Otherwise Specified. Parameters with MIN and/or
MAX limits are 100% tested at +25°C, unless otherwise specified. Temperature limits established by
characterization and are not production tested (Continued)
PARAMETER
SYMBOL
TEST CONDITIONS
MIN
TYP
MAX
UNITS
Current Imbalance Threshold
-
4
-
mV
Input Bias Current
-
20
-
nA
RTN1 Threshold
-
0.8
-
V
1.25
1.875
2.50
mV/µs
5
7.5
10
mV/µs
-
1
1.5
Ω
-
2
-
A
CORE CURRENT SENSE
SOFT START/VID-ON-THE-FLY
Soft-Start Voltage Transition
VSS
VID on the Fly Transition
GATE DRIVER DRIVING CAPABILITY [CORE AND NB]
UGATE Source Resistance (Note 4)
RSRC(UGATE) 500mA Source Current
UGATE Source Current (Note 4)
ISRC(UGATE)
UGATE Sink Resistance (Note 4)
RSNK(UGATE) 500mA Sink Current
-
1
1.5
Ω
UGATE Sink Current (Note 4)
ISNK(UGATE)
-
2
-
A
LGATE Source Resistance (Note 4)
RSRC(LGATE) 500mA Source Current
-
1
1.5
Ω
LGATE Source Current (Note 4)
ISRC(LGATE)
-
2
-
A
LGATE Sink Resistance (Note 4)
RSNK(LGATE) 500mA Sink Current
-
0.5
0.9
Ω
LGATE Sink Current (Note 4)
ISNK(LGATE)
-
4
-
A
UGATE to PHASE Resistance
Rp(UGATE)
-
1
-
kΩ
VUGATE_PHASE = 2.5V
VUGATE_PHASE = 2.5V
VLGATE = 2.5V
VLGATE = 2.5V
GATE DRIVER SWITCHING TIMING (Refer to “ISL6265A Gate Driver Timing Diagram” on page 8)
UGATE Rise Time (Note 3)
tRU
PVCC = 5V, 3nF Load
-
8.0
-
ns
LGATE Rise Time (Note 3)
tRL
PVCC = 5V, 3nF Load
-
8.0
-
ns
UGATE Fall Time (Note 3)
tFU
PVCC = 5V, 3nF Load
-
8.0
-
ns
LGATE Fall Time (Note 3)
tFL
PVCC = 5V, 3nF Load
-
4.0
-
ns
UGATE Turn-on Propagation Delay
tPDHU
PVCC = 5V, Outputs Unloaded
-
36
-
ns
LGATE Turn-on Propagation Delay
tPDHL
PVCC = 5V, Outputs Unloaded
-
20
-
ns
0.43
0.58
0.67
V
BOOTSTRAP DIODE
Forward Voltage
VDDP = 5V, Forward Bias Current = 2mA
Leakage
VR = 16V
-
-
5
µA
POWER GOOD AND PROTECTION MONITOR
PGOOD Low Voltage
VOL
IPGOOD = 4mA
-
0.2
0.5
V
PGOOD Leakage Current
IOH
PGOOD = 5V
-1
-
1
µA
PGOOD High After Soft-Start
Enable to PGOOD High, VCOREx = 1.1V
570
700
1010
µs
PGOOD Low After Fault
Fault to PGOOD Low
160
208
250
µs
Undervoltage Threshold
UVH
VCOREx falls below set-point for 208µs
Overvoltage Threshold
OVHS
VO rising above threshold > 0.5µs
240
295
350
mV
1.770
1.795
1.820
V
5
6.0
7
mV
9.2
10
10.8
µA
OVERCURRENT PROTECTION VDD0 AND VDD1
OCSET Reference Voltage
(VISPx - VISNx )
VOCSET = 180mV; VIN = 15.5V
OVERCURRENT PROTECTION VDD_NB
OCSET_NB OCP Current
RBIAS pin to GND = 117kΩ; Trips after 8 PWM cycles
7
FN6884.0
May 11, 2009
ISL6265A
Electrical Specifications
VCC = PVCC = 5V, VIN = 12V, TA = -10°C to +100°C, Unless Otherwise Specified. Parameters with MIN and/or
MAX limits are 100% tested at +25°C, unless otherwise specified. Temperature limits established by
characterization and are not production tested (Continued)
PARAMETER
SYMBOL
TEST CONDITIONS
MIN
TYP
MAX
UNITS
ROFS = 240kΩ (OFS pin to GND)
1.18
1.2
1.22
V
IOFS = 10µA
9.0
9.9
10.8
µA
OFFSET FUNCTION
OFS Pin Voltage For Droop Enabling
VOFS
FB Pin Source Current
IFB
OFS Pin Voltage Threshold for VFIX
Mode and No Droop Operation
VOFS
-
1.8
-
V
OFS Pin Voltage Threshold for SVI
Mode and No Droop Operation
VOFS
-
4.0
-
V
OFS Bias
IOFS
-
4.0
-
µA
-
1.35
0.9
V
1.8V < OFS < VCC
LOGIC INPUTS
ENABLE Low Threshold
VIL(3.3V)
ENABLE High Threshold
VIH(3.3V)
2.0
1.6
-
V
Logic input is low
-1
0
-
µA
Logic input is high at 3.3V
-
0
1
µA
PWROK Input Low Threshold
-
0.65
0.8
V
PWROK Input High Threshold
-
0.9
-
V
SVC, SVD Input HIGH (VIH)
1.05
0.87
-
V
SVC, SVD Input LOW (VIL)
-
0.68
0.45
V
ENABLE Leakage Current
SVI INTERFACE
Schmitt Trigger Input Hysteresis
-
0.19
-
V
SVD Low Level Output Voltage
3mA Sink Current
-
0.1
0.285
V
SVC, SVD Leakage
EN = 0V, SVC, SVD = 0V
-
< -100
-
nA
EN = 5V, SVC, SVD = 1.8V
-
< -100
-
nA
VSEN = 0.5V to 1.55V; RTN = 0 ±0.1V
-2
-
2
mV
DIFF AMP
Accuracy
NOTES:
3. Limits should be considered typical and are not production tested.
4. Limits established by characterization and are not production tested.
ISL6265A Gate Driver Timing Diagram
PWM
tPDHU
tFU
tRU
1V
UGATE
LGATE
1V
tFL
tRL
tPDHL
8
FN6884.0
May 11, 2009
ISL6265A
UGATE_NB
PHASE_NB
PWROK
LGATE_NB
PGND_NB
OCSET_NB
RTN_NB
VSEN_NB
FSET_NB
COMP_NB
FB_NB
VCC
VIN
Functional Pin Description
48 47 46 45 44 43 42 41 40 39 38 37
36 BOOT_NB
OFS/VFIXEN 1
PGOOD 2
35 BOOT0
34 UGATE0
PWROK 3
SVD 4
33 PHASE0
32 PGND0
SVC 5
31 LGATE0
49
GND
[BOTTOM]
ENABLE 6
RBIAS 7
30 PVCC
OCSET 8
29 LGATE1
VDIFF0 9
FB0 10
28 PGND1
COMP0 11
VW0 12
26 UGATE1
27 PHASE1
25 BOOT1
ISN1
ISP1
VW1
COMP1
FB1
VDIFF1
VSEN1
RTN1
RTN0
VSEN0
ISP0
ISN0
13 14 15 16 17 18 19 20 21 22 23 24
VCC
System power good input. When this pin is high, the SVI
interface is active and I2C protocol is running. While this pin
is low, the SVC, SVD, and VFIXEN input states determine
the pre-PWROK metal VID or VFIX mode voltage. This pin
must be low prior to the ISL6265A PGOOD output going
high per the AMD SVI Controller Guidelines.
PGOOD
Controller power-good open-drain output. This pin is typically
pulled up externally by a 2.0kΩ resistor to +3.3V. During
normal operation, this pin indicates whether all output
voltages are within specified overvoltage and undervoltage
limits and no overcurrent condition is present. If any output
voltage exceeds these limits or a reset event occurs, the pin is
pulled low. This pin is always low prior to the end of soft-start.
SVC
This pin is the serial VID clock input from the AMD processor.
SVD
This pin is the serial VID data bidirectional signal to and from
the master device on the AMD processor.
The bias supply for the IC’s control circuitry. Connect this pin
to a +5V supply and decouple using a quality 0.1µF ceramic
capacitor.
ENABLE
VIN
FSET_NB
Battery supply voltage. It is used for input voltage feed-forward
to improve the input line transient performance.
A resistor from this pin to GND programs the switching
frequency of the Northbridge controller (for example,
22.1k ~ 260kHz).
PVCC
Digital input enable. A high level logic signal on this pin
enables the ISL6265A.
The power supply pin for the internal MOSFET gate drivers
of the ISL6265A. Connect this pin to a +5V power supply.
Decouple this pin with a quality 1.0µF ceramic capacitor.
FB_NB
GND
COMP_NB
The bias and reference ground for the IC. The GND
connection for the ISL6265A is through the thermal pad on
the bottom of the package.
This pin is the output of the Northbridge controller error
amplifier.
RBIAS
Remote Northbridge voltage sense input and return.
Connect isolated traces from these pins to the Northbridge
sense points of the processor.
A 117kΩ resistor from RBIAS to GND sets internal reference
currents. The addition of capacitance to this pin must be
avoided and can create instabilities in operation.
OFS/VFIXEN
A resistor from this pin to GND programs a DC current
source, which generates a positive offset voltage across the
resistor between FB and VDIFF pins. In this case, the OFS
pin voltage is +1.2V and VFIX mode is not enabled. If OFS is
pulled up to +3.3V, VFIX mode is enabled, the DAC decodes
the SVC and SVD inputs to determine the programmed
voltage, and the OFS function is disabled. If OFS is pulled up
to +5V, the OFS function and VFIX mode are disabled.
9
This pin is the output voltage feedback to the inverting input
of the Northbridge controller error amplifier.
VSEN_NB, RTN_NB
OCSET_NB
Overcurrent protection selection input for the Northbridge
controller. A resistor from this pin to PHASE_NB sets the OC
trip point.
UGATE_NB
Upper MOSFET gate signal from Northbridge controller.
LGATE_NB
Lower MOSFET gate signal from Northbridge controller.
FN6884.0
May 11, 2009
ISL6265A
PHASE_NB
Switch node of the Northbridge controller. This pin should
connect to the source of the Northbridge channel upper
MOSFET(s).
BOOT_NB
This pin is the upper gate drive supply voltage for the
Northbridge controller. Connect an appropriately sized
ceramic bootstrap capacitor between the BOOT_NB and
PHASE_NB pins. An internal bootstrap diode connected to
the PVCC pin provides the necessary bootstrap charge.
PGND_NB
The return path of the Northbridge controller lower gate
driver. Connect this pin to the source of the lower
MOSFET(s).
OCSET
CORE_0 and CORE_1 common overcurrent protection
selection input. The voltage on this pin sets the (ISPx - ISNx)
voltage limit for OC trip.
VW0, VW1
A resistor from this pin to corresponding COMPx pin programs
the switching frequency (for example, 6.81k ~ 300kHz).
COMP0, COMP1
The output of the CORE_0 and CORE_1 controller error
amplifiers respectively. FBx, VDIFFx, and COMPx pins are
tied together through external R-C networks to compensate
the regulator.
FB0, FB1
These pins are the output voltage feedback to the inverting
input of the CORE_0 and CORE_1 error amplifiers.
VDIFF0, VDIFF1
Output of the CORE_0 and CORE_1 differential amplifiers.
VSEN0, RTN0
Inputs to the CORE_0 VR controller precision differential
remote sense amplifier. Connect to the sense pins of the
VDD0_FB[H, L] portion of the processor.
VSEN1, RTN1
Inputs to the CORE_1 VR controller precision differential
remote sense amplifier. Connect to the sense pins of the
VDD1_FB[H,L] portion of the processor. The RTN1 pin is
also used for detection of the VDD_PLANE_STRAP signal
prior to enable.
ISP0, ISN0, ISP1, ISN1
These pins are used for differentially sensing the corresponding
channel output current. The sensed current is used for channel
balancing, protection, and core load line regulation.
Connect ISN0 and ISN1 to the node between the RC sense
elements surrounding the inductor of their respective
channel. Tie the ISP0 and ISP1 pins to the VCORE side of
10
their corresponding channel’s sense capacitor. These pins
can also be used for discrete resistor sensing.
BOOT0, BOOT1
These pins provide the bias voltage for the corresponding
upper MOSFET drives. Connect these pins to appropriately
chosen external bootstrap capacitors. Internal bootstrap
diodes connected to the PVCC pin provide the necessary
bootstrap charge.
UGATE0, UGATE1
Connect these pins to the corresponding upper MOSFET
gate(s). These pins control the upper MOSFET gate(s) and
are monitored for shoot-through prevention.
LGATE0, LGATE1
Connect these pins to the corresponding lower MOSFET
gate(s).
PHASE0, PHASE1
Switch node of the CORE_0 and CORE_1 controllers.
Connect these pins to the sources of the corresponding
upper MOSFET(s). These pins are the return path for the
upper MOSFET drives.
PGND0, PGND1
The return path of the lower gate driver for CORE_0 and
CORE_1 respectively. Connect these pins to the
corresponding sources of the lower MOSFETs.
Theory of Operation
The ISL6265A is a flexible multi-output controller supporting
Northbridge and single or dual power planes required by
Class M AMD Mobile CPUs. In single plane applications,
both core voltage regulators operate single-phase. In
uniplane core applications, the core voltage regulators are
configured to operate as a two-phase regulator. All three
regulator outputs include integrated gate drivers for reduced
system cost and small board area. The regulators provide
optimum steady-state and transient performance for
microprocessor applications. System efficiency is enhanced
by idling a phase in uniplane configurations at low-current
and implementing automatic DCM-mode operation when
PSI_L is asserted to logic low.
The heart of the ISL6265A is the R3 Technology™, Intersil's
Robust Ripple Regulator modulator. The R3 modulator
combines the best features of fixed frequency PWM and
hysteretic PWM while eliminating many of their
shortcomings. The ISL6265A modulator internally
synthesizes an analog of the inductor ripple current and
uses hysteretic comparators on those signals to establish
PWM pulse widths. Operating on these large-amplitude,
noise-free synthesized signals allows the ISL6265A to
achieve lower output ripple and lower phase jitter than either
conventional hysteretic or fixed frequency PWM controllers.
Unlike conventional hysteretic converters, the ISL6265A has
FN6884.0
May 11, 2009
ISL6265A
PWM FREQUENCY
CONTROL
+
gmVIN
+
VW
-
VO
VR
+
gmVO
+
-
FSET
+
-
-
Modulator
VIN
-
The hysteresis window voltage is relative to the error
amplifier output such that load current transients result in
increased switching frequency, which gives the R3 regulator
a faster response than conventional fixed frequency PWM
controllers. In uniplane configurations, transient load current
is inherently shared between active phases due to the use of
a common hysteretic window voltage. Individual average
phase currents are monitored and controlled to equally
share current among the active phases.
and SVD pins to determine the soft-start target output
voltage level..
-
an error amplifier that allows the controller to maintain a
0.5% voltage regulation accuracy throughout the VID range
from 0.75V to 1.55V. Voltage regulation accuracy is slightly
wider, ±5mV, over the VID range from 0.7375V to 0.5V.
R
PWM Q
S
VCOMP
+
CR TO
PWM
CONTROL
ISL6265A
FIGURE 5. MODULATOR CIRCUITRY
The ISL6265A modulator features Intersil’s R3 technology, a
hybrid of fixed frequency PWM control and variable frequency
hysteretic control (see Figure 5). Intersil’s R3 technology can
simultaneously affect the PWM switching frequency and PWM
duty cycle in response to input voltage and output load
transients. The R3 modulator synthesizes an AC signal VR,
which is an analog representation of the output inductor ripple
current. The duty-cycle of VR is the result of charge and
discharge current through a ripple capacitor CR. The current
through CR is provided by a transconductance amplifier gm
that measures the VIN and VO voltages. The positive slope of
VR can be written as determined by Equation 1:
V RPOS = ( g m ) ⋅ ( V IN – V OUT )
RIPPLE CAPACITOR VOLTAGE CR
WINDOW VOLTAGE VW
ERROR AMPLIFIER VOLTAGE VCOMP
(EQ. 1)
PWM
The negative slope of VR can be written as determined by
Equation 2:
V RNEG = g m ⋅ V OUT
(EQ. 2)
Where gm is the gain of the transconductance amplifier.
A window voltage VW is referenced with respect to the error
amplifier output voltage VCOMP, creating an envelope into
which the ripple voltage VR is compared. The amplitude of
VW is set by a resistor connected across the FSET and GND
pins. The VR, VCOMP, and VW signals feed into a window
comparator in which VCOMP is the lower threshold voltage
and VW is the higher threshold voltage. Figure 6 shows
PWM pulses being generated as VR traverses the VW and
VCOMP thresholds. The PWM switching frequency is
proportional to the slew rates of the positive and negative
slopes of VR; it is inversely proportional to the voltage
between VW and VCOMP.
Initialization
Once sufficient bias is applied to the VCC pin, internal logic
checks the status of critical pins to determine the controller
operation profile prior to ENABLE. These pins include RTN1
which determines single vs two-phase operation and
OFS/VFIXEN for enabling/disabling the SVI interface and
core voltage droop. Depending on the configuration set by
these pins, the controller then checks the state of the SVC
11
FIGURE 6. MODULATOR WAVEFORMS DURING LOAD
TRANSIENT
Power-On Reset
The ISL6265A requires a +5V input supply tied to VCC and
PVCC to exceed a rising power-on reset (POR) threshold
before the controller has sufficient bias to guarantee proper
operation. Once this threshold is reached or exceeded, the
ISL6265A has enough bias to begin checking RTN1,
OFS/VFIXEN, ENABLE, and SVI inputs. Hysteresis between
the rising the falling thresholds assure the ISL6265A will not
inadvertently turn-off unless the bias voltage drops
substantially (see “Electrical Specifications” on page 8).
Core Configuration
The ISL6265A determines the core channel requirements of
the CPU based on the state of the RTN1 pin prior to ENABLE. If
RTN1 is low prior to ENABLE, both VDD0 and VDD1 core
planes are required. The core controllers operate as
independent single-phase regulators. RTN1 is connected to the
CPU Core1 negative sense point. For single core CPU designs
(uniplane), RTN1 is tied to a +1.8V or greater supply. Prior to
ENABLE, RTN1 is detected as HIGH and the ISL6265A drives
the core controllers as a two-phase multi-phase regulator. Dual
purpose motherboard designs should include resistor options to
FN6884.0
May 11, 2009
ISL6265A
Pre-PWROK Metal VID
open the CPU Core1 negative sense and connect the RTN1
pin to a pull-up resistor.
Assuming the OFS/VFIXEN pin is not tied to +3.3V during
controller configuration, typical motherboard start-up begins
with the controller decoding the SVC and SVD inputs to
determine the pre-PWROK metal VID setting (see Table 1).
Once the enable input (EN) exceeds the rising enable
threshold, the ISL6265A decodes and locks the decoded
value in an on-board hold register.
Mode Selection
The OFS/VFIXEN pin selects between the AMD defined
VFIX and SVI modes of operation and enables droop if
desired in SVI mode only. If OFS/VFIXEN is tied to VCC,
then SVI mode with no droop on the core output(s) is
selected. Connected to +3.3V, VFIX mode is active with no
droop on the core output(s). SVI mode with droop is enabled
when OFS/VFIXEN is tied to ground through a resistor sized
to set the core voltage positive offset. Further information is
provided in “Offset Resistor Selection” on page 17.
TABLE 1. PRE-PWROK METAL VID CODES
Serial VID Interface
The on-board Serial VID Interface (SVI) circuitry allows the
processor to directly control the Core and Northbridge voltage
reference levels within the ISL6265A. The SVC and SVD
states are decoded according to the PWROK and VFIXEN
inputs as described in the following sections. The ISL6265A
uses a digital-to-analog converter (DAC) to generate a
reference voltage based on the decoded SVI value. See
Figure 7 for a simple SVI interface timing diagram.
2
1
3
4
5
SVC
SVD
OUTPUT VOLTAGE (V)
0
0
1.1
0
1
1.0
1
0
0.9
1
1
0.8
The internal DAC circuitry begins to ramp Core and
Northbridge planes to the decoded pre-PWROK metal VID
output level. The digital soft-start circuitry ramps the internal
reference to the target gradually at a fixed rate of
approximately 2mV/µs. The controlled ramp of all output
6
7
8
9
10
11
12
VCC
SVC
SVD
ENABLE
PWROK
METAL_VID
V_SVI
METAL_VID
V_SVI
VDD AND VDDNB
VDDPWRGD
(PGOOD)
FIXEN
Interval 1 to 2: ISL6265A waits to POR.
Interval 2 to 3: SVC and SVD are externally set to pre-Metal VID code.
Interval 3 to 4: EN locks core output configuration and pre-Metal VID code. All outputs soft-start to this level.
Interval 4 to 5: PGOOD signal goes HIGH indicating proper operation.
Interval 5 to 6: CPU detects VDDPWRGD high and drives PWROK high to allow ISL6265A to prepare for SVI code.
Interval 6 to 7: SVC and SVD data lines communicate change in VID code.
Interval 7 to 8: ISL6265A responds to VID-ON-THE-FLY code change.
Interval 8 to 9: PWROK is driven low and ISL6265A returns all outputs to pre-PWROK Metal VID level.
Interval 9 to 10: PWROK driven high once again by CPU and ISL6265A prepares for SVI code.
Interval 10 to 11: SVC and SVD data lines communicate new VID code.
Interval 11 to 12: ISL6265A drives outputs to new VID code level.
Post 12 : Enable falls and all internal drivers are tri-stated and PGOOD is driven low.
FIGURE 7. SVI INTERFACE TIMING DIAGRAM: TYPICAL PRE-PWROK METAL VID STARTUP
12
FN6884.0
May 11, 2009
ISL6265A
voltage planes reduces in-rush current during the soft-start
interval. At the end of the soft-start interval, the PGOOD
output transitions high indicating all output planes are within
regulation limits.
If the EN input falls below the enable falling threshold, the
ISL6265A tri-states all outputs. PGOOD is pulled low with
the loss of EN. The Core and Northbridge planes will decay
based on output capacitance and load leakage resistance. If
bias to VCC falls below the POR level, the ISL6265A
responds in the same manner previously described. Once
VCC and EN rise above their respective rising thresholds,
the internal DAC circuitry re-acquires a pre-PWROK metal
VID code and the controller soft-starts.
VFIX MODE
In VFIX Mode, the SVC and SVD levels fixed external to the
controller through jumpers to either GND or VDDIO. These
inputs are not expected to change. In VFIX mode, the IC
decodes the SVC and SVD states per Table 2.
TABLE 2. VFIXEN VID CODES
SVC
SVD
OUTPUT VOLTAGE (V)
0
0
1.4
0
1
1.2
1
0
1.0
1
1
0.8
Once enabled, the ISL6265A begins to soft-start both Core
and Northbridge planes to the programmed VFIX level. The
internal soft-start circuitry slowly ramps the reference up to the
target value. The same fixed internal rate of approximately
2mV/µs results in a controlled ramp of the power planes.
Once soft-start has ended and all output planes are within
regulation limits, the PGOOD pin transitions high.
the PGOOD signal remains asserted. The Northbridge
voltage plane must remain active during this time.
If the PWROK input is deasserted, then the controller steps
both Core and Northbridge planes back to the stored
pre-PWROK metal VID level in the holding register from
initial soft-start. No attempt is made to read the SVC and
SVD inputs during this time. If PWROK is reasserted, then
the on-board SVI interface waits for a set VID command.
If EN goes low during normal operation, all internal drivers
are tri-stated and PGOOD is pulled low. This event clears
the pre-PWROK metal VID code and forces the controller to
check SVC and SVD upon restart.
A POR event on VCC during normal operation will shutdown
all regulators and PGOOD is pulled low. The pre-PWROK
metal VID code is not retained.
VID-on-the-Fly Transition
Once PWROK is high, the ISL6265A detects this flag and
begins monitoring the SVC and SVD pins for SVI
instructions. The microprocessor will follow the protocol
outlined in the following sections to send instructions for
VID-on-the-Fly transitions. The ISL6265A decodes the
instruction and acknowledges the new VID code. For VID
codes higher than the current VID level, the ISL6265A
begins stepping the required regulator output(s) to the new
VID target with a typical slew rate of 7.5mV/µs, which meets
the AMD requirements.
In the same manner described in “Pre-PWROK Metal VID”
on page 12, the POR circuitry impacts the internal driver
operation and PGOOD status.
When the VID codes are lower than the current VID level,
the ISL6265A begins stepping the regulator output to the
new VID target with a typical slew rate of -7.5mV/µs. Both
Core and NB regulators are always in CCM during a down
VID transition. The AMD requirements under these
conditions do not require the regulator to meet the minimum
slew rate specification of -5mV/µs. In either case, the slew
rate is not allowed to exceed 10mV/µs. The ISL6265A does
not change the state of PGOOD (VDDPWRGD in AMD
specifications) when a VID-on-the-fly transition occurs.
SVI MODE
SVI WIRE Protocol
Once the controller has successfully soft-started and
PGOOD transitions high, the processor can assert PWROK
to signal the ISL6265A to prepare for SVI commands. The
controller actively monitors the SVI interface for set VID
commands to move the plane voltages to start-up VID
values. Details of the SVI Bus protocol are provided in the
AMD Design Guide for Voltage Regulator Controllers
Accepting Serial VID Codes specification.
The SVI wire protocol is based on the I2C bus concept. Two
wires (serial clock (SVC) and serial data (SVD)), carry
information between the AMD processor (master) and VR
controller (slave) on the bus. The master initiates and
terminates SVI transactions and drives the clock, SVC,
during a transaction. The AMD processor is always the
master and the voltage regulators are the slaves. The slave
receives the SVI transactions and acts accordingly. Mobile
SVI wire protocol timing is based on high-speed mode I2C.
See AMD Griffin (Family 11h) processor publications for
additional details.
Once a set VID command is received, the ISL6265A decodes
the information to determine which output plane is affected
and the VID target required (see Table 3).The internal DAC
circuitry steps the required output plane voltage to the new
VID level. During this time, one or more of the planes could be
targeted. In the event either core voltage plane, VDD0 or
VDD1, is commanded to power-off by serial VID commands,
13
FN6884.0
May 11, 2009
ISL6265A
TABLE 3. SERIAL VID CODES
SVID[6:0]
VOLTAGE
(V)
SVID[6:0]
VOLTAGE
(V)
SVID[6:0]
VOLTAGE
(V)
SVID[6:0]
VOLTAGE
(V)
000_0000b
1.5500
010_0000b
1.1500
100_0000b
0.7500
110_0000b
0.3500*
000_0001b
1.5375
010_0001b
1.1375
100_0001b
0.7375
110_0001b
0.3375*
000_0010b
1.5250
010_0010b
1.1250
100_0010b
0.7250
110_0010b
0.3250*
000_0011b
1.5125
010_0011b
1.1125
100_0011b
0.7125
110_0011b
0.3125*
000_0100b
1.5000
010_0100b
1.1000
100_0100b
0.7000
110_0100b
0.3000*
000_0101b
1.4875
010_0101b
1.0875
100_0101b
0.6875
110_0101b
0.2875*
000_0110b
1.4750
010_0110b
1.0750
100_0110b
0.6750
110_0110b
0.2750*
000_0111b
1.4625
010_0111b
1.0625
100_0111b
0.6625
110_0111b
0.2625*
000_1000b
1.4500
010_1000b
1.0500
100_1000b
0.6500
110_1000b
0.2500*
000_1001b
1.4375
010_1001b
1.0375
100_1001b
0.6375
110_1001b
0.2375*
000_1010b
1.4250
010_1010b
1.0250
100_1010b
0.6250
110_1010b
0.2250*
000_1011b
1.4125
010_1011b
1.0125
100_1011b
0.6125
110_1011b
0.2125*
000_1100b
1.4000
010_1100b
1.0000
100_1100b
0.6000
110_1100b
0.2000*
000_1101b
1.3875
010_1101b
0.9875
100_1101b
0.5875
110_1101b
0.1875*
000_1110b
1.3750
010_1110b
0.9750
100_1110b
0.5750
110_1110b
0.1750*
000_1111b
1.3625
010_1111b
0.9625
100_1111b
0.5625
110_1111b
0.1625*
001_0000b
1.3500
011_0000b
0.9500
101_0000b
0.5500
111_0000b
0.1500*
001_0001b
1.3375
011_0001b
0.9375
101_0001b
0.5375
111_0001b
0.1375*
001_0010b
1.3250
011_0010b
0.9250
101_0010b
0.5250
111_0010b
0.1250*
001_0011b
1.3125
011_0011b
0.9125
101_0011b
0.5125
111_0011b
0.1125*
001_0100b
1.3000
011_0100b
0.9000
101_0100b
0.5000
111_0100b
0.1000*
001_0101b
1.2875
011_0101b
0.8875
101_0101b
0.4875*
111_0101b
0.0875*
001_0110b
1.2750
011_0110b
0.8750
101_0110b
0.4750*
111_0110b
0.0750*
001_0111b
1.2625
011_0111b
0.8625
101_0111b
0.4625*
111_0111b
0.0625*
001_1000b
1.2500
011_1000b
0.8500
101_1000b
0.4500*
111_1000b
0.0500*
001_1001b
1.2375
011_1001b
0.8375
101_1001b
0.4375*
111_1001b
0.0375*
001_1010b
1.2250
011_1010b
0.8250
101_1010b
0.4250*
111_1010b
0.0250*
001_1011b
1.2125
011_1011b
0.8125
101_1011b
0.4125*
111_1011b
0.0125*
001_1100b
1.2000
011_1100b
0.8000
101_1100b
0.4000*
111_1100b
OFF
001_1101b
1.1875
011_1101b
0.7875
101_1101b
0.3875*
111_1101b
OFF
001_1110b
1.1750
011_1110b
0.7750
101_1110b
0.3750*
111_1110b
OFF
001_1111b
1.1625
011_1111b
0.7625
101_1111b
0.3625*
111_1111b
OFF
NOTE:
* Indicates a VID not required for AMD Family 10h processors.
14
FN6884.0
May 11, 2009
PSI_L
ISL6265A
6
5
4
3
2
1
7
0
(See Table 3)
SVID
6
5
4
3
2
1
0
SVC
STOP
ACK
DATA PHASE
ACK
SLAVE ADDRESS PHASE
WRITE
START
SVD
FIGURE 8. SEND BYTE EXAMPLE
SVI Bus Protocol
The AMD processor bus protocol is compliant with SMBus
send byte protocol for VID transactions (see Figure 8). During a
send byte transaction, the processor sends the start sequence
followed by the slave address of the VR for which the VID
command applies. The address byte must be configured
according to Table 4. The processor then sends the write bit.
After the write bit, if the ISL6265A receives a valid address
byte, it sends the acknowledge bit. The processor then sends
the PSI-L bit and VID bits during the data phase. The Serial VID
8-bit data field encoding is outlined in Table 5. If ISL6265A
receives a valid 8-bit code during the data phase, it sends the
acknowledge bit. Finally, the processor sends the stop
sequence. After the ISL6265A has detected the stop, it can
then proceed with the VID-on-the-fly transition.
TABLE 4. SVI SEND BYTE ADDRESS DESCRIPTION
BITS
DESCRIPTION
6:4 Always 110b
3
Reserved by AMD for future use
2
VDD1, if set then the following data byte contains the VID for
VDD1
1
VDD0, if set then the following data byte contains the VID for
VID0
0
VDDNB, if set then the following data byte contains the VID
for VIDNB
implement dynamic VID changes, and shutdown individual
outputs.
The ISL6265A controls the no-load output voltage of core and
Northbridge output to an accuracy of ±0.5% over-the-range of
0.75V to 1.5V. A fully differential amplifier implements core
voltage sensing for precise voltage control at the
microprocessor die.
Switching Frequency
The R3 modulator scheme is a variable frequency PWM
architecture. The switching frequency increases during the
application of a load to improve transient performance. It
also varies slightly due to changes in input and output
voltage and output current. This variation is normally less
than 10% in continuous conduction mode.
CORE FREQUENCY SELECTION
A resistor connected between the VW and COMP pins of the
Core segment of the ISL6265A adjusts the switching window
and therefore adjusts the switching frequency. The RFSET
resistor that sets up the switching frequency of the converter
operating in CCM can be determined using Equation 3,
where RFSET is in kΩ and the switching period is in ms.
Designs for 300kHz switching frequency would result in a
RFSET value of 6.81kΩ.
R FSET ( kΩ ) = ( Period ( μs ) – 0.4 ) × 2.33
(EQ. 3)
In discontinuous conduction mode (DCM), the ISL6265A
runs in period stretching mode.
TABLE 5. SERIAL VID 8-BIT DATA FIELD ENCODING
BITS
7
6:0
DESCRIPTION
PSI_L:
= 0 means the processor is at an optimal load for the
regulator(s) to enter power-savings mode
= 1 means the processor is not at an optimal load for the
regulator(s) to enter power-saving mode
SVID[6:0] as defined in Table 3.
NORTHBRIDGE FREQUENCY SELECTION
The Northbridge switching frequency to programmed by a
resistor connected from the FSET_NB pin to the GND pin.
The approximate PWM switching frequency is written as
shown in Equation 4:
1
F SW = ----------------------------------K ⋅ R FSETNB
(EQ. 4)
Estimating the value of RFSET_NB is written as shown in
Equation 5:
Operation
After the start-up sequence, the ISL6265A begins regulating
the core and Northbridge output voltages to the pre-PWROK
metal VID programmed. The controller monitors SVI
commands to determine when to enter power-savings mode,
15
1
R FSET = --------------------K ⋅ F SW
(EQ. 5)
Where FSW is the PWM switching frequency, RFSET_NB is
the programming resistor and K = 1.5 x 10-10.
FN6884.0
May 11, 2009
ISL6265A
It is recommended that whenever the control loop
compensation network is modified, the switching frequency
should be checked and adjusted by changing RFSET_NB if
necessary.
Current Sense
Core and Northbridge regulators feature two different types
of current sense circuits.
CORE CONTINUOUS CURRENT SENSE
The ISL6265A provides for load current to be measured
using either resistors in series with the individual output
inductors or using the intrinsic series resistance of the
inductors as shown in the applications circuits in Figures 2
and 3. The load current in a particular output is sampled
continuously every switching cycle. During this time, the
current-sense amplifier uses the current sense inputs to
reproduce a signal proportional to the inductor current. This
sensed current is a scaled version of the inductor current.
IL
VIN
UGATE
L
LGATE
DCR
VOUT
+
VL(s)
R1
COUT
VC(s)
-
+
INDUCTOR
-
MOSFET
DRIVER
C1
R2
ISL6265A INTERNAL CIRCUIT
RNTC
ISP
CURRENT
SENSE
R3
OPTIONAL
NTC
NETWORK
Inductor windings have a characteristic distributed
resistance or DCR (Direct Current Resistance). For
simplicity, the inductor DCR is considered as a separate
lumped quantity, as shown in Figure 9. The inductor current,
IL, flowing through the inductor, passes through the DCR.
Equation 6 shows the s-domain equivalent voltage, VL,
across the inductor.
(EQ. 6)
A simple R-C network across the inductor (R1, R2 and C)
extracts the DCR voltage, as shown in Equation 7. The
voltage across the sense capacitor, VC, can be shown to be
proportional to the output current IL, shown in Equation 7.
16
(EQ. 8)
Sensing the time varying inductor current accurately
requires that the parallel R-C network time constant match
the inductor L/DCR time constant. If the R-C network
components are selected, such that the R-C time constant
matches the inductor L/DCR time constant (see Equation 9),
then VC is equal to the voltage drop across the DCR
multiplied by the ratio of the resistor divider, K.
R1 ⋅ R2
L
- ⋅ C1
------------- = -------------------R1 + R2
DCR
(EQ. 9)
The inductor current sense information is used for current
balance in dual plane applications, overcurrent detection in
core outputs and output voltage droop depending on
controller configuration.
CORE DCR TEMPERATURE COMPENSATION
It may also be necessary to compensate for changes in
inductor DCR due to temperature. DCR shifts due to
temperature cause time constant mismatch, skewing inductor
current accuracy. Potential problems include output voltage
droop and OC trip point, both shifting significantly from
expected levels. The addition of a negative temperature
coefficient (NTC) resistor to the R-C network compensates for
the rise in DCR due to temperature. Typical NTC values are in
the 10kΩ range. A second resistor, R3, in series with the NTC
allows for more accurate time-constant and resistor-ratio
matching as the pair of resistors are placed in parallel with R2
(Figure 9). The NTC resistor must be placed next to the
inductor for good heat transfer, while R1, R2, R3, and C1 are
placed close to the controller for interference immunity.
By adjusting the ratio between inductor DCR drop and the
voltage measured across the sense capacitor, the load line
can be set to any level, giving the converter the correct
amount of droop at all load currents.
FIGURE 9. DCR SENSING COMPONENTS
s⋅L
⎛ ------------+ 1⎞
⎝ DCR
⎠
---------------------------------------------------------- ⋅ K ⋅ DCR ⋅ I L
VC ( s ) =
(
⋅
R
)
R
⎛
⎞
1
2
⎜ s ⋅ ------------------------ ⋅ C 1 + 1⎟
R1 + R2
⎝
⎠
R2
K = --------------------R2 + R1
CORE DCR COMPONENT SELECTION FOR DROOP
ISN
V L ( s ) = I L ⋅ ( s ⋅ L + DCR )
Where:
(EQ. 7)
Equation 10 shows the relation between droop voltage,
I MAX
V DROOP = -------------- ⋅ 5 ⋅ V C, OC
I OC
(EQ. 10)
maximum output current (IMAX), OC trip level and current
sense capacitor voltage at the OC current level, VC(OC).
AMD specifications do not require droop and provide no load
line guidelines. Tight static output voltage tolerance limits
push acceptable level of droop below a useful level for Griffin
applications. Care must be taken in applications which
implement droop to balance time constant mismatch, sense
capacitor resistor ratio, OC trip and droop equations.
Temperature shifts related to DCR must also be addressed,
as outlined in the previous section.
FN6884.0
May 11, 2009
ISL6265A
NORTHBRIDGE CURRENT SENSE
During the off-time following a PHASE transition low, the
Northbridge controller samples the voltage across the lower
MOSFET rDS(ON). A ground-referenced amplifier is
connected to the PHASE node through a resistor,
ROCSET_NB. The voltage across ROCSET_NB is equal to the
voltage drop across the rDS(ON) of the lower MOSFET while
it is conducting. The resulting current into the OCSET_NB
pin is proportional to the inductor current. The sensed
inductor current is used for overcurrent protection and
described in the “Fault Monitoring and Protection” on
page 18. The Northbridge controller does not support output
voltage droop.
Selecting RBIAS For Core Outputs
To properly bias the ISL6265A, a reference current is
established by placing a 117kΩ, 1% tolerance resistor from the
RBIAS pin to ground. This will provide a highly accurate, 10µA
current source from which OC reference current is derived.
Care must be taken in layout to place the resistor very close
to the RBIAS pin. A good quality signal ground should be
connected to the opposite end of the RBIAS resistor. Do not
connect any other components to this pin as this would
negatively impact performance. Capacitance on this pin
could create instabilities and is to be avoided.
A resistor divider off this pin is used to set the Core side OC
trip level. Additional direction on how to size is provided in
“Fault Monitoring and Protection” on page 18 on how to size
the resistor divider.
Offset Resistor Selection
If the OFS pin is connected to ground through a resistor, the
ISL6265A operates in SVI mode with droop active. The
resistor between the OFS pin and ground sets the positive
Core voltage offset per Equation 11.
1.2V ⋅ R FB
R OFS = ---------------------------V OFS
(EQ. 11)
Where VOFS is the user defined output voltage offset.
Typically, VOFS is determined by taking half the total output
voltage droop. The resulting value centers the overall output
voltage waveform around the programmed SVID level. For
example, RFB of 1kΩ and a total output droop of 24mV would
result in an offset voltage of 12mV and a ROFS of 100kΩ.
Internal Driver Operation
The ISL6265A features three internal gate-drivers to support
the Core and Northbridge regulators and to reduce solution
size. The drivers include a diode emulation mode, which
helps to improve light-load efficiency.
MOSFET Gate-Drive Outputs
The ISL6265A has internal gate-drivers for the high-side and
low-side N-Channel MOSFETs. The low-side gate-drivers
are optimized for low duty-cycle applications where the
17
low-side MOSFET conduction losses are dominant,
requiring a low r DS(ON) MOSFET. The LGATE pull-down
resistance is low in order to strongly clamp the gate of the
MOSFET below the VGS(th) at turn-off. The current transient
through the gate at turn-off can be considerable because the
gate charge of a low r DS(ON) MOSFET can be large.
Adaptive shoot-through protection prevents a gate-driver
output from turning on until the opposite gate-driver output
has fallen below approximately 1V.
The high-side gate-driver output voltage is measured across
the UGATE and PHASE pins while the low-side gate-driver
output voltage is measured across the LGATE and PGND
pins. The power for the LGATE gate driver is sourced
directly from the PVCC pin. The power for the UGATE
gate-driver is sourced from a “boot” capacitor connected
across the BOOT and PHASE pins. The boot capacitor is
charged from a 5V bias supply through a “boot diode” each
time the low-side MOSFET turns on, pulling the PHASE pin
low. The ISL6265A has an integrated boot diode connected
from the PVCC pin to the BOOT pin.
Diode Emulation
The ISL6265A implements forced
continuous-conduction-mode (CCM) at heavy load and
diode-emulation-mode (DE) at light load, to optimize
efficiency in the entire load range. The transition is
automatically achieved by detecting the inductor current
when PSI_L is low. If PSI_L is high, the controller disables
DE and forces CCM on both Core and NB regulators.
Positive-going inductor current flows either from the source
of the high-side MOSFET, or into the drain of the low-side
MOSFET. Negative-going inductor current flows into the
drain of the low-side MOSFET. When the low-side MOSFET
conducts positive inductor current, the phase voltage is
negative with respect to the GND and PGND pins.
Conversely, when the low-side MOSFET conducts negative
inductor current, the phase voltage is positive with respect to
the GND and PGND pins. The ISL6265A monitors the phase
voltage when the low-side MOSFET is conducting inductor
current to determine the direction of the inductor current.
When the output load current is less than half the inductor
ripple current, the inductor current goes negative. Sinking the
negative inductor through the low-side MOSFET lowers
efficiency by preventing DCM period stretching and allowing
unnecessary conduction losses. In DE, the ISL6265A Core
regulators automatically enter DCM after the PHASE pin has
detected positive voltage and LGATE was allowed to go high.
The NB regulator enters DCM after the PHASE pin has
detected positive voltage and LGATE was allowed to go high
for eight consecutive PWM switching cycles. The ISL6265A
turns off the low-side MOSFET once the phase voltage turns
positive, indicating negative inductor current. The ISL6265A
returns to CCM on the following cycle after the PHASE pin
detects negative voltage, indicating that the body diode of the
low-side MOSFET is conducting positive inductor current.
FN6884.0
May 11, 2009
ISL6265A
Efficiency can be further improved with a reduction of
unnecessary switching losses by reducing the PWM
frequency. It is characteristic of the R3 architecture for the
PWM frequency to decrease while in diode emulation. The
extent of the frequency reduction is proportional to the
reduction of load current. Upon entering DCM, the PWM
frequency makes an initial step-reduction because of a 33%
step-increase of the window voltage V W.
Power-Savings Mode
The ISL6265A has two operating modes to optimize
efficiency based on the state of the PSI_L input from the
AMD SVI control signal. When this input is low, the controller
expects to deliver low power and enters a power-savings
mode to improve efficiency in this low power state. The
controller’s operational modes are designed to work in
conjunction with the AMD SVI control signal to maintain the
optimal system configuration for all conditions.
Northbridge And Dual Plane Core
While PSI_L is high, the controller operates all three
regulators in forced CCM. If PSI_L is asserted low by the SVI
interface, the ISL6265A initiates DE in all three regulators.
This transition allows the controller to achieve the highest
possible efficiency over the entire load range for each output.
A smooth transition is facilitated by the R3 technology™, which
correctly maintains the internally synthesized ripple current
throughout mode transitions of each regulator.
Uniplane Core
In uniplane mode, the ISL6265A Core regulator is in 2-phase
multiphase mode. The controller operates with both phases
fully active, responding rapidly to transients and delivering the
maximum power to the load. When the processor asserts
PSI_L low under reduced load levels, the ISL6265A sheds one
phase to eliminate switching losses associated with the idle
channel. Even with the regulator operating in single-phase
mode, transient response capability is maintained.
While operating in single-phase DE with PSI_L low, the lower
MOSFET driver switches the lower MOSFET off at the point of
zero inductor current to prevent discharge current from
flowing from the output capacitor bank through the inductor. In
DCM, switching frequency is proportionately reduced, thus
greatly reducing both conduction and switching loss. In DCM,
the switching frequency is defined by Equation 12.
2
2 ⋅ L ⋅ IO
F CCM
F DCM = ------------------- ⋅ ------------------------------------2
VO ⎞
⎛
1.33
V O ⋅ ⎜ 1 – ---------⎟
V
⎝
IN⎠
These fault monitors trigger protective measures to prevent
damage to the processor. One common power good
indicator is provided for linking to external system monitors.
Power-Good Signal
The power-good pin (PGOOD) is an open-drain logic output
that signals if the ISL6265A is not regulating Core and
Northbridge output voltages within the proper levels or
output current in one or more outputs has exceeded the
maximum current setpoint.
This pin must be tied to a +3.3V or +5V source through a
resistor. During shutdown and soft-start, PGOOD is pulled low
and is released high only after a successful soft-start has raised
Core and Northbridge output voltages within operating limits.
PGOOD is pulled low when an overvoltage, undervoltage, or
overcurrent (OC) condition is detected on any output or when
the controller is disabled by a POR or forcing enable (EN) low.
Once a fault condition is triggered, the controller acts to protect
the processor. The controller latches off and PGOOD is pulled
low. Toggling EN or VCC initiates a soft-start of all outputs. In
the event of an OV, the controller will not initiate a soft-start by
toggling EN, but requires VCC be lowered below the falling
POR threshold to reset.
Overcurrent Protection
Core and Northbridge outputs feature two different methods
of current sensing. Core output current sensing is achieved
via inductor DCR or discrete resistor sensing. The
Northbridge controller uses lower MOSFET rDS(ON) sensing
to detect output current.
CORE OC DETECTION
Core outputs feature an OC monitor which compares a
voltage set at the OCSET pin to the voltage measured
across the current sense capacitor, VC. When the voltage
across the current sense capacitor exceeds the programmed
trip level, the comparator signals an OC fault. Figure 10
shows the basic OC functions within the IC.
CURRENT
SENSE
ISP
5x
5 x VC(OC) @
OC TRIP CURRENT
Fault Monitoring and Protection
ISN
BIAS
CKT
(EQ. 12)
Where FCCM is equivalent to the Core frequency set by
Equation 3.
SEE FIGURE 9 FOR
ADDITIONAL DETAIL
RBIAS
+
Vc
_
1.17V
10µA
OC
+
6
OCSET
VOCSET
6
RBIAS
VOCSET
ROCSET
ISL6265A
FIGURE 10. OC TRIP CIRCUITRY
The ISL6265A actively monitors Core and Northbridge
output voltages and currents to detect fault conditions.
18
FN6884.0
May 11, 2009
ISL6265A
The sense capacitor voltage, VC, will increase as inductor
current rises per Equation 7. When the inductor current rises
to the OC trip level, the voltage across the sense capacitor
will reach a maximum based on the resistor ratio K. This
maximum value, VC(OC), is gained up by a factor of 5 and
compared to the static OC trip level set by the OCSET pin.
The recommended voltage range for VC,OC is 6mV to 25mV,
which sets the resistor divider ratio K, where IOC is the
user-defined OC trip level (see Equation 13). Typical
inductor DCR values are on the order of 1mΩ which result in
more than enough voltage drop to support this VC,OC range.
V C ( OC )
K = ---------------------------I OC ⋅ DCR
(EQ. 13)
The resistor divider components also impact time-constant
matching, these components need to meet the parallel
combination requirements of Equation 9.
Based on the selected VC(OC) level, the required OC monitor
trip level is set. The recommended VC(OC) level range will
result in an OC monitor trip level range of 30mV to 125mV
based on the internal gain of 5.
This OC monitor trip level sets the voltage level required at
the OCSET pin to create an OC fault at the user-defined OC
trip level. A resistor divider from the RBIAS pin to ground
with the mid-point connected to OCSET sets the voltage at
the pin (see Figure 10). This voltage is internally divided by 6
and compared with VC(OC). Working backwards, the voltage
required at the OCSET pin to achieve this OC trip level
ranges from 180mV to 0.750mV as defined in Equation 14.
(EQ. 14)
V OCSET = V C ( OC ) ⋅ 30
The resistor divider ratio used to determine the RBIAS and
ROCSET values is shown in Equation 15.
V OCSET
R OCSET
----------------------------------------------- = ----------------------R OCSET + R BIAS
1.17V
(EQ. 15)
The resistor values must also meet the RBIAS requirement
that the total series resistance to ground equal 117kΩ. An
OC condition must be sustained for 100µs before action is
taken by the controller in response to the OC fault.
A short-circuit OC loop is also active based on the same
sense elements outlined above with a threshold set to 2.25
times the OCSET threshold set. The controller takes
immediate action when this fast OC fault is detected.
NORTHBRIDGE OC DETECTION
Northbridge OC sensing is achieved via rDS(ON) sensing
across the lower MOSFET. An internal 10µA current source
develops a voltage across ROCSET_NB, which is compared
with the voltage developed across the low-side MOSFET as
measured at the PHASE pin. When the voltage drop across
the MOSFET exceeds the voltage drop across the resistor,
19
an OC event occurs. The OCSET_NB resistor is selected
based on the relationship in Equation 16.
I OC ⋅ r DS ( ON )
R OCSETNB = ------------------------------------10μA
(EQ. 16)
Where IOC is the OC trip level selected for the Northbridge
application and rDS(ON) is the drain-source ON-resistance of
the lower MOSFET.
OC FAULT RESPONSE
When an OC fault occurs on any combination of outputs,
both Core and Northbridge regulators shutdown and the
driver outputs are tri-stated. The PGOOD signal transitions
low indicating a fault condition. The controller will not attempt
to restart the regulators and the user must toggle either EN
or VCC to clear the fault condition.
Overvoltage Protection
The ISL6265A monitors the individual Core and Northbridge
output voltages using differential remote sense amplifiers.
The ISL6265A features a severe overvoltage (OV) threshold
of 1.8V. If any of the outputs exceed this voltage, an OV fault
is immediately triggered. PGOOD is latched low and the
low-side MOSFETs of the offending output(s) are turned on.
The low-side MOSFETs will remain on until the output
voltage is pulled below 0.85V at which time all MOSFETs are
turned off. If the output again rises above 1.8V, the
protection process repeats. This offers protection against a
shorted high-side MOSFET while preventing output voltage
from ringing below ground. The OV is reset by toggling EN
low. OV detection is active at all times that the controller is
enabled including after one of the other faults occurs so that
the processor is protected against high-side MOSFET
leakage while the MOSFETs are commanded off.
Undervoltage Protection
Undervoltage protection is independent of the OC limit. A
fault latches if any of the sensed output voltages are less
than the VID set value by a nominal 295mV for 205µs. The
PWM outputs turn off both Core and Northbridge internal
drivers and PGOOD goes low.
General Application Design Guide
This design guide is intended to provide a high-level
explanation of the steps necessary to design a single-phase
power converter. It is assumed that the reader is familiar with
many of the basic skills and techniques referenced in the
following section. In addition to this guide, Intersil provides
complete reference designs that include schematics, bills of
materials, and example board layouts.
Selecting the LC Output Filter
The output inductor and output capacitor bank form a
low-pass filter responsible for smoothing the pulsating
voltage at the phase node. The output filter also must
support the transient energy required by the load until the
controller can respond. Because it has a low bandwidth
FN6884.0
May 11, 2009
ISL6265A
The duty cycle of an ideal buck converter is a function of the
input and the output voltage. This relationship is written as
Equation 17:
VO
D = --------V IN
(EQ. 17)
The output inductor peak-to-peak ripple current is written as
Equation 18:
VO • ( 1 – D )
I P-P = -----------------------------f SW • L
(EQ. 18)
For this type of application, a typical step-down DC/DC
converter has an IP-P of 20% to 40% of the maximum DC
output load current. The value of IP-P is selected based upon
several criteria such as MOSFET switching loss, inductor core
loss, and the resistive loss of the inductor winding. The DC
copper loss of the inductor can be estimated by Equation 19:
P COPPER = I LOAD
2
•
DCR
(EQ. 19)
Where ILOAD is the converter output DC current.
The copper loss can be significant so attention must be given to
the DCR selection. Another factor to consider when choosing
the inductor is its saturation characteristics at elevated
temperature. A saturated inductor could cause destruction of
circuit components as well as nuisance OCP faults.
A DC/DC buck regulator must have output capacitance CO
into which ripple current IP-P can flow. Current IP-P develops
a corresponding ripple voltage VP-P across CO, which is the
sum of the voltage drop across the capacitor ESR and of the
voltage change stemming from charge moved in and out of
the capacitor. These two voltages are written as shown in
Equation 20:
ΔV ESR = I PP • E SR
(EQ. 20)
Selection of the Input Capacitor
The input capacitors are responsible for sourcing the AC
component of the input current flowing into the upper
MOSFETs. Their RMS current capability must be sufficient to
handle the AC component of the current drawn by the upper
MOSFETs, which is related to duty cycle and the number of
active phases.
The important parameters for the bulk input capacitance are
the voltage rating and the RMS current rating. For reliable
operation, select bulk capacitors with voltage and current
ratings above the maximum input voltage and capable of
supplying the RMS current required by the switching circuit.
Their voltage rating should be at least 1.25x greater than the
maximum input voltage, while a voltage rating of 1.5x is a
preferred rating. Figure 11 is a graph of the input RMS ripple
current, normalized relative to output load current, as a
function of duty cycle for a single-phase regulator that is
adjusted for converter efficiency.
0.60
0.55
NORMALIZED INPUT RMS
RIPPLE CURRENT (IRMS/IO)
compared to the switching frequency, the output filter limits
the system transient response. The output capacitors must
supply or sink load current while the current in the output
inductors increases or decreases to meet the demand.
IP-P,N = 0.75
0.50
0.45
0.40
IP-P,N = 0
0.35
0.30
IP-P,N = 0.25
0.25
0.20
0.15
0.10
0.05
0
0
0.1
I PP
V C = ----------------------------8 • CO • f
(EQ. 21)
SW
If the output of the converter has to support a load with high
pulsating current, several capacitors will need to be paralleled
to reduce the total ESR until the required VP-P is achieved.
The inductance of the capacitor can cause a brief voltage dip
if the load transient has an extremely high slew rate. Capacitor
ESL can significantly impact output voltage ripple. Low
inductance capacitors should be considered. A capacitor
dissipates heat as a function of RMS current and frequency.
Be sure that IP-P is shared by a sufficient quantity of paralleled
capacitors so that they operate below the maximum rated
RMS current at FSW. Take into account that the rated value of
a capacitor can degrade as much as 50% as the DC voltage
across it increases.
20
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
DUTY CYCLE (VIN/VO)
FIGURE 11. NORMALIZED RMS INPUT CURRENT FOR
SINGLE PHASE CONVERTER
The normalized RMS current calculation is written as
Equation 22:
I IN_RMS, N =
and Equation 21:
IP-P,N = 0.50
IP-P,N = 1
2
D
D ⋅ ( 1 – D ) + ⎛ ------⎞ ⋅ I PP ,N
⎝ 12⎠
(EQ. 22)
Where:
-
IMAX is the maximum continuous ILOAD of the converter
IPP,N is the ratio of inductor peak-to-peak ripple current
to IMAX
- D is the duty cycle that is adjusted to take into account
the efficiency of the converter which is written as:
VO
D = -----------------V IN ⋅ η
(EQ. 23)
- where η is converter efficiency
Figure 12 provides the same input RMS current information
for two-phase designs.
In addition to the bulk capacitance, some low ESL ceramic
capacitance is recommended to decouple between the drain
of the high-side MOSFET and the source of the low-side
MOSFET.
FN6884.0
May 11, 2009
ISL6265A
- tON is the time required to drive the device into
saturation
- tOFF is the time required to drive the device into cut-off
NORMALIZED INPUT RMS
RIPPLE CURRENT (IRMS/IO)
0.3
Selecting The Bootstrap Capacitor
0.2
All three integrated drivers feature an internal bootstrap
schottky diode. Simply adding an external capacitor across
the BOOT and PHASE pins completes the bootstrap circuit.
The bootstrap function is also designed to prevent the
bootstrap capacitor from overcharging due to the large
negative swing at the PHASE node. This reduces voltage
stress on the BOOT and PHASE pins.
IP-P,N = 0.5
IP-P,N = 0.75
0.1
IP-P,N = 0
0
0
0.2
0.4
0.6
0.8
1.0
DUTY CYCLE (VIN/VO)
FIGURE 12. NORMALIZED RMS INPUT CURRENT FOR
2-PHASE CONVERTER
MOSFET Selection and Considerations
The choice of MOSFETs depends on the current each
MOSFET will be required to conduct, the switching
frequency, the capability of the MOSFETs to dissipate heat,
and the availability and nature of heat sinking and air flow.
Typically, a MOSFET cannot tolerate even brief excursions
beyond their maximum drain to source voltage rating. The
MOSFETs used in the power stage of the converter should
have a maximum VDS rating that exceeds the sum of the
upper voltage tolerance of the input power source and the
voltage spike that occurs when the MOSFETs switch.
There are several power MOSFETs readily available that are
optimized for DC/DC converter applications. The preferred
high-side MOSFET emphasizes low gate charge so that the
device spends the least amount of time dissipating power in the
linear region. The preferred low-side MOSFET emphasizes low
r DS(ON) when fully saturated to minimize conduction loss.
For the low-side (LS) MOSFET, the power loss can be
assumed to be conductive only and is written as Equation 24:
2
P CON_LS ≈ I LOAD ⋅ r DS ( ON )_LS • ( 1 – D )
(EQ. 24)
For the high-side (HS) MOSFET, the its conduction loss is
written as Equation 25:
P CON_HS = I LOAD
2
•
r DS ( ON )_HS • D
(EQ. 25)
For the high-side MOSFET, the switching loss is written as
Equation 26:
V IN • I PEAK • t OFF • f
V IN • I VALLEY • t ON • f
SW
SW
P SW_HS = ----------------------------------------------------------------- + ------------------------------------------------------------2
2
(EQ. 26)
Where:
- IVALLEY is the difference of the DC component of the
inductor current minus 1/2 of the inductor ripple current
- IPEAK is the sum of the DC component of the inductor
current plus 1/2 of the inductor ripple current
21
The bootstrap capacitor must have a maximum voltage
rating above PVCC + 4V and its capacitance value is
selected per Equation 27:
Qg
C BOOT ≥ -----------------------ΔV BOOT
(EQ. 27)
Where:
- Qg is the total gate charge required to turn on the
high-side MOSFET
- ΔVBOOT, is the maximum allowed voltage decay across
the boot capacitor each time the high-side MOSFET is
switched on
As an example, suppose the high-side MOSFET has a total
gate charge Qg, of 25nC at VGS = 5V, and a ΔVBOOT of
200mV. The calculated bootstrap capacitance is 0.125µF; for
a comfortable margin, select a capacitor that is double the
calculated capacitance. In this example, 0.22µF will suffice.
Use a low temperature-coefficient ceramic capacitor.
PCB Layout Considerations
Power and Signal Layers Placement on the PCB
As a general rule, power layers should be close together,
either on the top or bottom of the board, with the weak analog
or logic signal layers on the opposite side of the board. The
ground-plane layer should be adjacent to the signal layer to
provide shielding. The ground plane layer should have an
island located under the IC, the compensation components,
and the FSET components. The island should be connected
to the rest of the ground plane layer at one point.
Component Placement
There are two sets of critical components in a DC/DC
converter; the power components and the small signal
components. The power components are the most critical
because they switch large amount of energy. The small
signal components connect to sensitive nodes or supply
critical bypassing current and signal coupling.
The power components should be placed first and these
include MOSFETs, input and output capacitors, and the
inductor. It is important to have a symmetrical layout for each
power train, preferably with the controller located equidistant
from each power train. Symmetrical layout allows heat to be
dissipated equally across all power trains. Keeping the
FN6884.0
May 11, 2009
ISL6265A
distance between the power train and the control IC short
helps keep the gate drive traces short. These drive signals
include the LGATE, UGATE, PGND, PHASE and BOOT.
VIAS TO
GROUND
PLANE
GND
VOUT
INDUCTOR
OUTPUT
CAPACITORS
SCHOTTKY
DIODE
PHASE
NODE
HIGH-SIDE
MOSFETS
VIN
VIN PIN
The VIN pin should be connected close to the drain of the
high-side MOSFET, using a low- resistance and
low-inductance path.
VCC PIN
For best performance, place the decoupling capacitor very
close to the VCC and GND pins.
LOW-SIDE
MOSFETS
PVCC PIN
INPUT
CAPACITORS
For best performance, place the decoupling capacitor very
close to the PVCC and respective PGND pins, preferably on
the same side of the PCB as the ISL6265A IC.
FIGURE 13. TYPICAL POWER COMPONENT PLACEMENT
When placing MOSFETs, try to keep the source of the upper
MOSFETs and the drain of the lower MOSFETs as close as
thermally possible (see Figure 13). Input high-frequency
capacitors should be placed close to the drain of the upper
MOSFETs and the source of the lower MOSFETs. Place the
output inductor and output capacitors between the
MOSFETs and the load. High-frequency output decoupling
capacitors (ceramic) should be placed as close as possible
to the decoupling target (microprocessor), making use of the
shortest connection paths to any internal planes. Place the
components in such a way that the area under the IC has
less noise traces with high dV/dt and di/dt, such as gate
signals and phase node signals.
ENABLE AND PGOOD PINS
Signal Ground and Power Ground
LGATE ROUTING
The bottom of the ISL6265A QFN package is the signal
ground (GND) terminal for analog and logic signals of the IC.
Connect the GND pad of the ISL6265A to the island of
ground plane under the top layer using several vias, for a
robust thermal and electrical conduction path. Connect the
input capacitors, the output capacitors, and the source of the
lower MOSFETs to the power ground plane.
The LGATE trace has a signal going through it that is both
high dV/dt and di/dt, with high peak charging and
discharging current. Route this trace in parallel with the trace
from the PGND pin. These two traces should be short, wide,
and away from other traces. There should be no other weak
signal traces in proximity with these traces on any layer.
Routing and Connection Details
The signals going through these traces are both high dv/dt
and high di/dt, with high peak charging and discharging
current. Route the UGATE and PHASE pins in parallel with
short and wide traces. There should be no other weak signal
traces in proximity with these traces on any layer.
Specific pins (and the trace routing from them), require extra
attention during the layout process. The following
sub-sections outline concerns by pin name.
PGND PINS
This is the return path for the pull-down of the LGATE
low-side MOSFET gate driver. Ideally, PGND should be
connected to the source of the low-side MOSFET with a
low-resistance, low-inductance path.
These are logic signals that are referenced to the GND pin.
Treat as a typical logic signal.
FB PINS
The input impedance of the FB pin is high, so place the
voltage programming and loop compensation components
close to the COMP, FB, and GND pins keeping the high
impedance trace short.
FSET_NB PIN
This pin requires a quiet environment. The resistor RFSET
should be placed directly adjacent to this pin. Keep fast
moving nodes away from this pin.
BOOT AND PHASE ROUTING
Copper Size for the Phase Node
The parasitic capacitance and parasitic inductance of the
phase node should be kept very low to minimize ringing. It is
best to limit the size of the PHASE node copper in strict
accordance with the current and thermal management of the
application. An MLCC should be connected directly across
the drain of the upper MOSFET and the source of the lower
MOSFET to suppress the turn-off voltage.
All Intersil U.S. products are manufactured, assembled and tested utilizing ISO9000 quality systems.
Intersil Corporation’s quality certifications can be viewed at www.intersil.com/design/quality
Intersil products are sold by description only. Intersil Corporation reserves the right to make changes in circuit design, software and/or specifications at any time without
notice. Accordingly, the reader is cautioned to verify that data sheets are current before placing orders. Information furnished by Intersil is believed to be accurate and
reliable. However, no responsibility is assumed by Intersil or its subsidiaries for its use; nor for any infringements of patents or other rights of third parties which may result
from its use. No license is granted by implication or otherwise under any patent or patent rights of Intersil or its subsidiaries.
For information regarding Intersil Corporation and its products, see www.intersil.com
22
FN6884.0
May 11, 2009
ISL6265A
Package Outline Drawing
L48.6x6
48 LEAD THIN QUAD FLAT NO-LEAD PLASTIC PACKAGE
Rev 1, 4/07
4X 4.4
6.00
44X 0.40
A
B
6
PIN 1
INDEX AREA
6
PIN #1 INDEX AREA
48
37
1
6.00
36
4 .40 ± 0.15
25
12
0.15
(4X)
13
24
0.10 M C A B
0.05 M C
TOP VIEW
48X 0.45 ± 0.10
4 48X 0.20
BOTTOM VIEW
SEE DETAIL "X"
(
SEATING PLANE
0.08 C
( 44 X 0 . 40 )
( 5. 75 TYP )
C
0.10 C
BASE PLANE
MAX 0.80
SIDE VIEW
4. 40 )
C
0 . 2 REF
5
( 48X 0 . 20 )
0 . 00 MIN.
0 . 05 MAX.
( 48X 0 . 65 )
DETAIL "X"
TYPICAL RECOMMENDED LAND PATTERN
NOTES:
1. Dimensions are in millimeters.
Dimensions in ( ) for Reference Only.
2. Dimensioning and tolerancing conform to AMSE Y14.5m-1994.
3. Unless otherwise specified, tolerance : Decimal ± 0.05
4. Dimension b applies to the metallized terminal and is measured
between 0.15mm and 0.30mm from the terminal tip.
5. Tiebar shown (if present) is a non-functional feature.
6. The configuration of the pin #1 identifier is optional, but must be
located within the zone indicated. The pin #1 indentifier may be
either a mold or mark feature.
23
FN6884.0
May 11, 2009