SC452 Dual Phase Single Chip IMVP 6 Vcor
SC452 Dual Phase Single Chip IMVP 6 Vcor
SC452 Dual Phase Single Chip IMVP 6 Vcor
IG ED
Delayed Power Good Signal with Blanking
technology provides a loss-less current sensing scheme Programmable Soft-Start and DAC Slew Control
which is extremely robust and easy to lay out. Eliminat- Programmable OCP Threshold
ES D
ing the sense resistors reduces costs and PCB area, plus Supports all Ceramic Decoupling Solutions
N
D EN
increases system efficiency. Integrated SmartDriver™ 44 pin MLP (7x7)
technology initially turns on the high side driver with ‘soft’ Lead-free Package
drive to reduce ringing, EMI, and capacitive turn-on of the
M
low side MOSFET, while also increasing overall efficiency.
EW M
SC452 changes from dual-phase to single-phase opera- Graphics and Other Processor Cores
tion whenever PSI# asserts, providing optimal efficiency
across the entire power range.
O
N
POWER MANAGEMENT
Typical Application Circuit
+VDC
+V5S
1
+VDC C2 1 C1 2
9
8
7
6
5
9
8
7
6
5
CR1
1μF MBR0530 R1 1
0.1μF/25V Q2 1 1
2 2 Q1 C4
1 10.0 D D C3 C5 1.4V@
R2
BST1 1 2 BSTRCD1 4 4
10μF
48A
VPR1 0 25V 10μF +VCC_Core
10μF
1 25V
1 25V
3
2
1
2
1
3
C7 2 2 2
1
R3 C6
N FO
R4
1nF 1 2
100 1uF
2 100k L1 R6
2 R5 2 5 4
1 2 CS1P 1 2 1 2
9
8
7
6
5
9
8
7
6
5
DRN1 0.0005
1
BG1 Q4
VR_ON Q3 1 C9 C10
43 VPN1
ISH D D
To ICH7 DPRSLPVR CR2
T
1 4 4
C11 MBRS14OL
44
42
41
40
39
38
37
36
35
34
2
VPN1
VIN1
TG1
DRN1
BG1
V5_1
ISH1
BST1
100pF 330μF
DPRSL
DPRSTP#
EN
3
2
1
3
2
1
2
2
330μF
R NE
2
CLKEN#1 1 CLKEN#1 CS1+ 33
R7 32
100k R9 VREF 2 VREF CS1- CS1N
HYS 3 HYS CS2- 31 CS2N
10K 30
CLSET 4 CLSET CS2+
R12
EC W
U1 Note: Route CS1x, CS2x, DRPx and FBx
1 8
ERROUT 29 as differential pairs.
10K R10 54.9k R8 5 VID6 ERROUT
28 2 7
6 VCCA VCCA 1 Note: Terminate droop
VID5 27
7 AGND 3 6 feedback to sense
2 1 VID4 SC452 26
8 DAC DAC 2 C15 resistors with Kelvin
VID3 25 4 5
1nF C12 9 SS SS connections.
VID2 24
10 DRP+ R11 1k
VID1 23 1 1 0.1μF
11 DRP-
VID0 R22
C14 10k 2
O D
1 +VDC
C13
1 No_Pop
CRC
PWRGD
CLKEN# 1 1μF 1
DRN2
VPN2
BST2
VID6 C16
V5_2
GND
VID6
VIN2
C17
PSI#
C18
BG2
TG2
1nF
FB+
FB-
VID5 VID5 2
M E
VID4 2
9
8
7
6
5
9
8
7
6
5
VID4 100pF
2
1
VPN2 13
14
15
16
17
18
19
20
21
22
12
VID2 2 2 1
VID2
VID1
Q5 Q6 1
C20
VID1 2 D D C21
VID0 C19
VID0 4 4
To ICH7
M SI
10μF
or CPU 10μF 25V 10μF
25V 25V
3
2
1
2
1
3
2 2 2
IMVP PG
PSI# DRP+
EN G
DRP-
9
8
7
6
5
9
8
7
6
5
10K L2 0.0005
1
1 2 CS2P 1 2 Q8
Q7 1
C26
1
R16 16.2k C25 10nF D D
D
CR3
1 2 4 4 C27
MBRS140L
TG2 R18 100k
2 C24 2 2
C23 BG2 330μF
3
2
1
3
2
1
ED
DRN2
2
1 1 1
2
10nF 10nF 1 C29
1 C28 1μF 1
2
R19 C30
N
2
100 2 100pF
VPR2
2
United States Patent No. 6,441,597
R20
+VDC BST2 1 2 BSTRCD2
0 1
MBR0530
CR4 +V5S
2
1
C31
1
0.1μF/25V C32
2
1μF
2
SC452
NOT RECOMMENDED FOR NEW DESIGN
SC452
POWER MANAGEMENT
Absolute Maximum Ratings
Exceeding the specifications below may result in permanent damage to the device or device malfunction. Operation outside of the parameters specified in the
Electrical Characteristics section is not implied.
Static -0.3 30 V
BST1, BST2 to PGND
Transient <100ns -0.3 34 V
Transient <100ns -5 29 V
IG ED
TG1, TG2 to PGND DRN 1, 2-0.3 BST1, 2+0.3 V
ES D
BG1, BG2 to PGND -0.3 V5_1, 2+0.3 V
N
VIN1, VIN2 to PGND
-0.3
25
VIN1, 2+0.3
V
V
M
PGND to AGND -0.3 0.3 V
EW M
o
Operating Junction Temperature Range TJ -40 125 C
FO T R
o
Storage Temperature Range TSTG -65 150 C
o
Peak IR Reflow (10-40sec) TIRreflow 260 C
O
Electrical Characteristics
Unless otherwise specified, VccA = V5_1 = V5_2 =5 V. -40<TJ<+125°C.
VccA,V5_1, V5_2
4.5 5.0 5.5 V
Operating Range
Operating (Static) 5 10 15 mA
POWER MANAGEMENT
Electrical Characteristics (Cont.)
Parameter Condition Min Typ Max Units
In UVLO 10 μA
IG ED
Operating, Static, TG2 Low 0.3 0.9 1.2 mA
ES D
Vin1 and Vin2 Current Static TG when respective TG Low 500 μA
N
D EN Static TG when respective TG High
When in Powersave
900
0
μA
μA
M
(EN, VID[6:0],
Logic Inputs (EN, VID[6:0], DPRSLP,
DPRSL, PSI #)
DPRSTP#, PSI#)
EW M
Threshold
Input Impedance 40 kΩ
FO T R
85°C)
Reference (DAC, SS, VREF), (0 < TJ <125°C)
0.75V - 0.50V -7 7 mV
N
0.4875 - 0.30V
0.50 - 0.30V -14 14 mV
POWER MANAGEMENT
Electrical Characteristics (Cont.)
Parameter Condition Min Typ Max Units
Input Impedance 14 kΩ
Bandwidth(1) 2 MHz
IG ED
0 to 85°C
-85°C -0.5 0 0.5 mV
ES D
Maximum Input Signal(1) 20 mV
N
Bandwidth(1)
Bandwidth(1) 2 MHz
N O
R EC
CS Bandwidth(1) 2 MHz
N
POWER MANAGEMENT
Electrical Characteristics (Cont.)
Parameter Condition Min Typ Max Units
IG ED
HYS, CLSET Input Bias Current ±1
I500I μA
nA
ES D
HYS Error (internal HYS difference from Dual Phase -18
-24 -20 -16
-22 %
N
D EN
TG Hi to TG lo as a percentage of HYS = 1V
voltage applied at HYS pin) Single Phase ±36 ±40 ±44 %
CLKEN#, PWRGD
Leakage 1 μA
FO T R
High Impedance
Power Good Window Upper Threshold FB Rising Relative DAC +160 +200 +240 mV
Power Good Window Lower Threshold FB Falling Relative DAC -360 -300 -240 mV
POWER MANAGEMENT
Electrical Characteristics (Cont.)
Parameter Condition Min Typ Max Unit
Ω
RTG_UP, DRN < 0.5V,
3.48 5.8 9.24
-40 to 125°C
Ω
RTG_UP, DRN > 0.5V,
0.76 1.3 2.1
-40 to 125°C
IG ED
RTG_DN , 25°C 0.42 0.6 0.78 Ω
ES D
RTG_DN , -40 to 125°C 0.34 0.6 1.01
N
Rise Time (1,2)
CTG = 3nF
17
9
22
12
27
15
ns
ns
M
From Hysteretic Comparator
EW M
Ω
O
Ω
N
POWER MANAGEMENT
Pin Configuration Ordering Information
DPRSTP#
Device(2) Package(1) Temp Range (TJ)(3)
DPRSL
BST1
DRN1
VPNI
V5_1
VIN1
BG1
TG1
ISH
EN
SC452IMLTRT MLP-44 -40°C to +125°C
44 43 42 41 40 39 38 37 36 35 34
CLKEN# 1
33 CS1+ SC452EVB Evaluation Board
VREF 2
32 CS1 -
Notes:
31 CS2 - 1) Only available in tape and reel packaging. A reel contains 3000
HYS 3
CLSET 4
30 CS2+ devices.
VID6 5 29 ERROUT 2) This device is ESD sensitive. Use of standard ESD handling precau-
PGND tions is required.
VID5 6
28 VCCA
PAD
VID4 7
27 AGND
3) Lead-free package compliant with J-STD-020B. Qualified to support
IG ED
maximum IR Reflow temperature of 260°C for 30 seconds. This prod-
VID3 8
26 DAC uct is fully WEEE and RoHS compliant.
VID2 9
ES D
25 SS
VID1 10
N
24 DRP+
VID0 11
12 13 14 15 16 17 D EN
18 19 20 21 22
23 DRP -
M
DRN2
VPN2
BST2
V5_2
PWRGD
VIN2
BG2
TG2
PSI#
FB+
FB-
EW M
N O
R EC
f id
Pin Descriptions
FO T R
Pin #
Pin# Pin
PinName
Name Pin Description
Pin Description
2 VREF Internal reference voltage (2V). Bypass to AGND with a TBD pF capacitor.
1nF capacitor.
3 HYS Core comparator hysteresis. A resistor divider on this pin sets the hysteresis voltage.
4 CLSET Current Limit Set. A resistor divider on this pin sets the OCP threshold.
6 VID5
7 VID4
8 VID3
9 VID2
10 VID1
POWER MANAGEMENT
Pin Descriptions (Cont.)
PinPin#
# Pin
PinName
Name PinDescription
Pin Description
Virtual Phase Node for Phase 2. Connect an RC between this pin and the output sense
13 VPN2
point to enable Combi-Sense operation.
Input power to the DC-DC converter. Used as supply reference for internal Phase 2
14 VIN2
Combi-Sense circuitry.
Phase 2 Bootstrap pin. A capacitor is connected between BST and DRN to develop the
15 BST2
floating voltage for the high-side MOSFET.
This pin connects to the junction of the Phase 2 switching and synchronous MOSFETs .
IG ED
17 DRN2
This pin can be subjected to a -2V minimum relative to PGND without affecting operation.
18 BG2 Phase 2 output drive signal for the bottom (synchronous) MOSFET.
ES D
N
D EN
19 V5_2 Input supply for Phase 2 low-side gate drive. Connect to 5V.
26 DAC DAC output. An external cap at this pin defines VID transition timing.
O
35 DPRSTP# DPRSTP# control pin for fast/slow C4E event. Active high.
37 V5_1 Input supply for Phase 1 low-side gate drive. Connect to 5V.
POWER MANAGEMENT
Pin Descriptions (Cont.)
Pin##
Pin PinName
Pin Name Pin Description
Pin Description
38 BG1 Phase 1 output drive signal for the bottom (synchronous) MOSFET.
This pin connects to the junction of the Phase 1 switching and synchronous
39 DRN1 MOSFETs . This pin can be subjected to a -2V minimum relative to PGND without
affecting operation.
Input power to the DC-DC converter. Used as supply reference for internal Phase 1
IG ED
42 VIN1
Combi-Sense circuitry.
Virtual Phase Node for Phase 1. Connect an RC between this pin and the output sense
ES D
43 VPN1
point to enable Combi-Sense operation.
N
44 DPRSL
D EN
Deeper Sleep control pin. Active high.
M
PAD PGND Power Ground for Drivers 1 and 2. Pad must be soldered to Power Ground plane.
EW M
N O
R EC
FO T R
O
N
POWER MANAGEMENT
Block Diagram
DAC
VID +
[6:0] DAC ERROUT
-
SS ERROR AMP
FB- -
HYSTERETIC COMPARATOR
FB+ + CO
+ CO1
DRP+ + -
-
DRP- CO2
I-limit
IG ED
SELECT
Hys hi
SELECT
ES D
CURRENT SENSE AMP Hys_lo
CS1
CS1+ +
N
+
D EN
CO
CS1- - - CO
LP
FILTER DPRSLP
SELECT PHASE
PHASE
CONTROL
PSI#
M
SELECT CONTROL
LP
FILTER
CS1 DPRSTP
-
CS2+
EW M
+
CS2
- +
CS2- CS2
N O
CS1 + CL lo I-limit
CO
CS2 -
ISH
FO T R
REFERENCE +
EN & VREF
UVLO -
CL_hi
O
3ms
START UP PWRGD
HYS_hi TIME
EN EN PGNDN
BST 1,2
VccA TG 1,2
CROSS
VccA CONDUCTION DRN 1,2
PROTECTION
CO 1 or
CO 2
AGND VS 1,2
CS 1 or ZERO BG 1,2
CS 2 CROSSING
PGND 1,2
POWER MANAGEMENT
Applications Information
INTRODUCTION: Thus, customers can choose the amount of cost and per-
formance they need for any given design.
The SC452 is a new generation of hysteretic converter
which combines the best features of Semtech’s hyster- The SC452 also provides a full range of features.
etic converter technology with the benefits of Semtech’s
patented Combi-Sense technology. The SC452 provides a All IMVP-6/6+ functions are implemented:
complete solution to Intel’s IMVP-6 requirements.
¨ EN
In the SC452, the ripple for the hysteretic switching con- ¨ CLKEN#
trol is provided by Combi-Sense current feedback. This ¨ IMVP-6 PWRGD
provides several advantages over plain voltage-mode hys- ¨ DPRSLPVR
teretic converters, and other topologies such as constant ¨ DPRSTP#
IG ED
on-time which switch on voltage ripple. ¨ PSI#
¨ Geyserville-3 VID changes
• No minimum amount of output ripple is required so ¨ Fast-C4 Exit
ES D
there are no controller-induced limits on capacitor value ¨ All ‘1s’ soft-OFF state
N
or ESR.
D EN
• No current sense resistors are required, resulting in
• A 2.00V voltage reference is provided
M
higher converter efficiency. • Separate hysteresis and current limit settings
EW M
• The large signal magnitude afforded by Combi-Sense • A full suite of protection features is provided:
N O
vides the fastest possible transient response without All protection features are latching, and are either reset by
N
switching at very high frequencies. This results in higher recycling power or toggling the EN signal.
efficiency with less expensive parts because switching
losses are reduced. Only the Combi-Sense ripple is used THEORY OF OPERATION
for the regulation loop, so the load-line accuracy is not af-
fected by tolerances in RDS(ON) or inductor DCR. Howev- Voltage Regulation:
er, because of the large signal magnitude, the DC is kept
for the current limiting and current sharing functions. Referring to the block diagram on the preceding page, the
hysteretic comparator is the heart of the converter. The
Load-line control is provided by a dedicated droop ampli- “+” input corresponds roughly to the CMPREF node of our
fier with uncommitted inputs. This provides users with older generations of IC; the “-” input is similar to CMP. In
maximum flexibility, as the droop source can be any of the order to regulate, the hysteretic comparator needs the fol-
following: lowing information:
POWER MANAGEMENT
Applications Information (Cont.)
CMPREF receives the reference, voltage feedback and Current Limit Regulation:
droop information. The reference is produced by the in-
tegrated seven-bit DAC. The feedback voltage is received In Current Limit, the voltage hysteretic converter is over-
by the full differential amplifier from the CPU socket. The ridden by the current limit hysteretic comparator, and the
droop amplifier reduces the voltage at the “+” node of the TG pulse is terminated when the output of the current
differential amplifier as the output current increases to sense amplifier reaches the CL_hi threshold and BG is
produce the required linear load line. A third amplifier, terminated at the CL_lo threshold. These thresholds are
labeled the “Error Amplifier”, multiplies the difference set from the CLSET resistor divider:
between the “ideal” voltage (DAC minus droop) and the
actual voltage (FB+ minus FB-) for faster response. This CL_hi = 0.33 * V(clset)
signal is the reference for the hysteretic comparator. CL_lo = 0.20 * V(clset)
IG ED
CMP has the ripple signal derived from the Combi-Sense Current limit pulses continue until 32 pulses after the
inputs plus the hysteresis signal. The DC is stripped from voltage droops to the PWRGD low threshold; then the con-
the ripple signals by the combination of low-pass filter and troller latches off. This current limit algorithm has been
ES D
summing amplifiers. Phase 2 has a current sharing in- used in several generations of IMVP controllers and have
N
D EN
put derived from an averaged difference between the two
phases. The ripple inputs are fed to a summing block via
a multiplexer which is synchronized to the active phase
been proven to be extremely robust.
M
with the select output. The hysteresis signal is added at Start-Up and Shut-Down Sequences:
EW M
The figure on Page 14 illustrates the basic switching con- typ.) then the EN signal goes high. The DAC drives 12μA
trol. Starting with the Select line (top plot, green trace) on (typ.) into the soft-start capacitor on the SS pin. The SS
Phase 2, and both CO signals low. Accordingly, both bot- and DAC pins rise slowly until the BOOT voltage (1.2V, fixed
FO T R
tom gate (BG) signals are on and the inductor currents in internally) is reached. The controller remains at BOOT
both phases are discharging as shown by the Phase 1 (or- voltage for ~30μs. At the end of the BOOT interval, the
ange) and Phase 2 (blue) ripple signals in the lower plot. VID(6:0) lines are considered valid and CLKEN# is driven
O
When CMP discharges to CMPREF, the select line toggles, fined value. Approximately 6ms after the voltage hits the
CO1 turns on, and subtracts V(hys_hi) from CMP. CO1 PWRGD threshold, IMVP-6_PWRGD goes high, and start-
remains high until CMP again charges to CMPREF. Then, up is complete.
CO1 switches low, adding V(hys_lo) to CMP. This state is
held until CMP again discharges to CMPREF. Then, the In a normal shutdown, the EN signal is driven low, the
select line toggles, CO2 turns on, and the cycle repeats. TG and BG signals are driven low, tri-stating the power
chains. An approximately 100Ω resistor on the FB+ signal
In Dual Phase mode, V(hys_lo) is zero; C0 of the initial discharges Vcore slowly and prevents normal amounts of
phase remains low while the alternate phase is in control, leakage from pulling Vcore high. The DAC and other in-
so BG of the initial phase remains on through the alter- ternal circuitry is shut down, entering a very low power
nate cycle and, as a result, the second phase will termi- (<10μA typical) state. A UVLO will also result in this type
nate at approximately -V(hys). During single phase opera- of shutdown.
tion, V(hys_lo) = -V(hys).
POWER MANAGEMENT
Applications Information (Cont.)
* D:\Products\SC452\Model_04Jun04\SC452_CBM_Top-level.sch
Date/Time run: 06/04/04 16:18:36 Temperature: 27.0
(A) SC452_CBM_Top-level.dat (active)
6.0V
IC SIGNALS
4.0V CO2
CO1
2.0V Phase 1
IG ED
elect Phase 2
0V
ES D
V(HS27.HS19:select)/2 V(HS27.CO1)/4 +3 V(HS27.CO2)/4+4
N
D EN
LOG SIGNALS
1.10V
CMPREF
M
CMP
EW M
1.05V
N O
1.00V
Vdac
SEL>> Vcore
FO T R
0.96V
220us 221us 222us 223us 224us
V(DAC) V(Vcore) V(CS2+)-V(CS2-)+1V V(CS1+)-V(CS1-) +1V V(HS27.CMPREF)-0.55V
V(HS27.CMP)-0.55V
Time
O
POWER MANAGEMENT
Applications Information (Cont.)
SC452
VID DPRSL PSI# Status Load
Mode
tSFT_START_VCC
tBOOT-VID-TR
1 1 Deeper Sleep Icc > 3A 1-phase
CPU_UP tCPU_UP
-12%
CLK_ENABLE#
IG ED
Active; Full
0 1 Icc > 15A 2-phase
Power Potential
tCPU_PWRGD
IMVP6_PWRGD
ES D
N
D EN
Response to Power Control Inputs:
Figure A - Power On Sequencing Timing Diagram Besides the EN signal, described on Page 13, the SC452
M
reacts to the other control signals in the following man-
EW M
VR_ON
ner:
N O
IMVP6_PWRGD
tPWRDOWN1 power saving enabled, always saving power at light load
regardless of the status of DPRSL or PSI#. PSI# signal is
MCH_PWRGD used to indicate the expected max level of currents de-
FO T R
VCCP
PSI# transitions no longer occurs during Deeper Sleep
mode. While in active mode, it is expected that PSI# sig-
VCC_MCH nal toggle occurs at the same Vcc-core voltage level VID.
The reason to have same VID voltage requirement is that
the superimposed charge current required to charge the
output decoupling to a new level of voltage can overcome
Figure B - Power Off Sequencing Timing Diagram single phase mode of operation (if used), during positive
dv/dt events (such as Enhanced Intel SpeedStep® or
Deeper Sleep exit).
POWER MANAGEMENT
Applications Information (Cont.)
In active mode, enhanced Intel SpeedStep transitions can A +/-0.85% 7-bit digital-to-analog converter (DAC) serves
occur. SC452 can recognize a step-up in voltage transi- as the programmable reference source of the Core Com-
tion and revert operation to full power mode to supply the parator. Programming is accomplished by logic voltage
bulk capacitor charge currents superimposed on the pro- levels applied to the DAC inputs. The VID code vs. The
cessor active mode current. During Deeper Sleep, PSI# DAC output is shown in the tables below. There are 7 volt-
indicates a very low current state. Regulator can enter age identification pins on mobile processor. These signals
asynchronous mode of operation. In rare occasions, if the can be used to support automatic selection of Vcc_core
PSI# is deasserted during Deeper Sleep, this is an indica- voltages.
tion of a high leakage component that may not benefit
They are needed to cleanly support voltage specification
from asynchronous operation.
variations on current and future processors. VID [6:0]
are defined in the table below. The VID [6:0] signals are
In the Deeper Sleep state, SC452 recognizes Deeper
0V to Vccp CMOS level inputs. These signals are not to
IG ED
Sleep exit state and its associated voltage transitions and
be pulled up externally as this will damage the proces-
reverts operation to full power mode to allow for the bulk
sor.
capacitor charge currents to superimpose with the pro-
ES D
cessor active mode current.
N
D EN
M
EW M
N O
R EC
FO T R
O
N
POWER MANAGEMENT
Applications Information (Cont.)
Table 1. Table 2.
VID vs. VCC_CORE Voltage (Active Mode) VID vs. VCC_CORE Voltage
(Active Mode/Deeper Sleep Dual Mode Region)
6 5 4 3 2 1 0 V 6 5 4 3 2 1 0 V 6 5 4 3 2 1 0 V 6 5 4 3 2 1 0 V
IG ED
0 0 0 0 0 1 1 1.4625 0 0 1 0 1 1 1 1.2125 0 1 0 1 0 1 1 0.9625 0 1 1 1 1 1 1 0.7125
ES D
N
0 0 0 0 1 0 1 1.4375 0 0 1 1 0 0 1 1.1875 0 1 0 1 1 0 1 0.9375 1 0 0 0 0 0 1 0.6875
D EN
0 0 0 0 1 1 0 1.4250 0 0 1 1 0 1 0 1.1750 0 1 0 1 1 1 0 0.9250 1 0 0 0 0 1 0 0.6750
M
0 0 0 0 1 1 1 1.4125 0 0 1 1 0 1 1 1.1625 0 1 0 1 1 1 1 0.9125 1 0 0 0 0 1 1 0.6625
EW M
1 0 1 0 0 0 0 0.5000
Table 1 - reflects VID codes to be used in Active state. The voltages represent- Table 2 - reflects VID codes to be used for both Active and Deeper Sleep
ed cover HFM through LFM. states.
POWER MANAGEMENT
Applications Information (Cont.)
Table 3. DAC Slew Rate Control:
VID vs. VCC_CORE Voltage (Deeper Sleep/
Extended Deeper Sleep Dual Mode Region) The DAC has integrated slew-rate control with multiple
current settings to charge and discharge the soft-start ca-
pacitor. The slowest setting is used for soft-start, a medi-
VID VDAC VID VDAC
um setting for slow C4-exit (DPRSLVR=’1’, DPRSTP#=’1’)
6 5 4 3 2 1 0 V 6 5 4 3 2 1 0 V and a fast setting for all other VID transitions.
1 0 1 0 0 0 1 0.4875 1 0 1 1 0 1 0 0.3750
Power Supply Protection:
1 0 1 0 0 1 0 0.4750 1 0 1 1 0 1 1 0.3625
A UVLO circuit consists of a comparator that monitors the
1 0 1 0 0 1 1 0.4625 1 0 1 1 1 0 0 0.3500 input supply voltage level, 5V. The SC452 is in UVLO mode
IG ED
1 0 1 0 1 0 0 0.4500 1 0 1 1 1 0 1 0.3375
when its supply voltage has not ramped above the upper
threshold or has dropped below the lower threshold. The
1 0 1 0 1 0 1 0.4375 1 0 1 1 1 1 0 0.3250 output of the UVLO comparator, gated with the ENABLE
ES D
signal, turns on or off the internal bias, enables or dis-
N
1 0 1 0 1 1 0 0.4250 1 0 1 1 1 1 1 0.3125
D EN
able the SC452 output, and initiates or resets the soft-
1 0 1 0 1 1 1 0.4125 1 1 0 0 0 0 0 0.3000 start timer. If an UVLO occurs, a fault is set and SC452 is
disabled until the system has shut down (and reapplied
M
1 0 1 1 0 0 0 0.4000 1 1 1 1 1 1 1 OFF power), or the enable input signal to the SC452 has tog-
EW M
gled states.
1 0 1 1 0 0 1 0.3875
N O
R EC
Table 3 - reflects VID codes likely to represent Deeper Sleep and extended ver- The OVP circuit of SC452 monitors the processor core
sions of Deeper Sleep State.
VCC_CORE voltage for an over-voltage condition. If the FB
voltage is 200mV greater than the DAC-Droop (i.e., out of
FO T R
DAC Operation Below 0.3000V: the power good window), the SC452 will latch off and hold
the low-side driver on permanently. Either the Power or
The SC452 responds to DAC codes corresponding to EN must be recycled to clear the latch. The latch is dis-
O
voltage values below 0.3V by producing voltages less abled during soft-start and VID/DeeperSleep transitions.
N
than 0.3V; however, the tolerance of these signals is not For safety, the latch is enabled if the FB voltage exceeds
specified or guaranteed. In the case of the ‘111 1111’ 1.8V even during VID/DeeperSleep transitions.
VID code, the SC452 holds BGx and TGx low, preventing
switching from occurring. In addition, a ~50Ω FET con- The device will be disabled and latched off when the inter-
nected from VCORE to GND is turned on to prevent system nal junction temperature reaches approximately 160°C.
leakage from charging up the VCORE rail. Either the Power or EN must be recycled to clear the
latch.
POWER MANAGEMENT
Applications Information (Cont.)
Design Procedure
(Based on SC452 R2 calculation with DCR Droop and R-Droop
Method. In this design procedure, we are going to use the speci-
fications required by the Intel’s IMVP VI Napa Platform T&L (Yonah)
Processor):
IG ED
V INMAX
V INMIN 10 V
19 V
ES D
V INNOM
N
V HFM_NL
I LKGMAX
1.2875
1.6 A
V
D EN Figure 1 - Hysteretic Converter Response
M
to a Positive Transient
I HFM_FL 36 A
EW M
R IMVP 2.1
mV The first condition is easy to see; if the ESR is too high,
N O
POWER MANAGEMENT
Applications Information (Cont.)
For the second condition, we need to know the inductor This value of inductance is required up to maximum load.
value, which is a function of the highest desired switching Inductors with a “swinging choke” characteristic, where the
frequency. The maximum frequency occurs at the high- zero current value of inductance is much less than the full
est input voltage. As a reasonable compromise between load current inductance can be used, as long as the above
efficiency and component size, a maximum switching restriction is met. Then, the worst-case (low input voltage)
frequency of 300kHz or less per phase is desired. Since response time (the time for the current to reach the new
we are analyzing the minimum inductance for one phase, transient value) is:
the ripple voltage will actually be twice the amount of the 1
specified output ripple, since the ripple voltage from each L1 IHFM_FL ILKGMAX
2
phase will tend to cancel. Please consult the data for your dT
specific processor. Note: The desired ripple amount comes VINMIN VHFM_NL
from Intel’s IMVP-6 Rev 0.5 specs.
IG ED
VHFM_NL 9
dMIN dT 927.8623 10 s
ES D
VINMAX
N
FS 250kHz
D EN Add ~200ns for the propagation delay from a change at
the output to the MOSFET switch turning on in reaction
due to the minimum on time requirement of the IC. Since
M
VRIPPLE 10m V the shaded area is triangular, the total charge taken out
EW M
dMIN
FS 2VRIPPLE F. C IHFM_FL ILKGMAX dT 2 10 sec
MINP
2 VDP
Ω
3
O
9 6
LMIN 361.3852 10 H CMINP 256.6036 10 F
Selected L = 470nH as the next closest value. Selected COUT = 330µF x 6 to meet transient requirements.
This condition applies only to the positive transient.
Load Step:
0.47 μH
Load Release:
L1
0.47 μH
The worst-case for the transient load release to happen is
L2 when one phase has just reached the maximum hysteresis,
(i.e., it has just turned off the high-side switch). At this point,
the second phase will be declining (approximately) through
the nominal voltage, (i.e., its low-side switch will be on).
POWER MANAGEMENT
Applications Information (Cont.)
J?
L1 Rcs1 Rcu1 IRIPPLE 13A
BG1 Rds_ON1
ESR_eq 1.6A
Rcu1_rt
R_LOAD
IHFM_FL IRIPPLE
It0_1
J? L2 Rcs2 Rcu2 Cout_eq
2 2
Rds_ON2
BG2
IHFM_FL
Rcu2_rt
It0_2
2
IG ED
0
It0_1 24.5000 10 A
ES D
0
Figure 2 - Load Release Behavior of Dual It0_2 18.0000 10 A
N
Phase Buck Converter
D EN COUT 330 10
6
F
M
Load is stepping from 36A to 1.6A:
6m Ω
EW M
The control circuit quickly turns off TG and turns on both equal to VFL / L. (without the consideration of the secondary
bottom gates to discharge the inductors as quickly as order effect, such as, Rds_on drop, current sense resistor
possible. The stored energy, once losses in the ESR and
O
BG1
VHFM_FL t
IL1 IL1 ( t) It0_1
L1
OA VHFM_FL t
IL2
IL2 ( t) It0_2
L2
OA t 0 50n s 10μ s
t= 0
POWER MANAGEMENT
Applications Information (Cont.)
Since both inductors are discharging at the same time Using Panasonic SPCAP, specified as 330µF at 2.0V with
and the same rate, there are two terms contributing to the 6mΩ ESR, we see that 6 caps are sufficient for the ESR re-
increase of the voltage on the output capacitors. First is quirements, and also meet the capacitance requirement.
due to the ESR of the output capacitor. Second is due to
the added charge contributed by the inductor currents. STEP 2: Droop Calculation
IG ED
I: VTOTAL ( t N) VESR ( t N) dVCAP ( t N) use of the droop amplifier. This IMVP gain is used to meet
the IMVP load line specification.
ES D
The goal here is to use minimum amount of N x COUT to meet The IMVP load line is defined as the High Frequency Mode
N
the transient requirement, with some headroom to allow
for component tolerance. D EN at no-load voltage, minus the High Frequency Mode at
full-load voltage, divided by the maximum (High Frequency
M
Mode) load current. Since SC452 provides remote sens-
ing for the core voltage, we no longer need to adjust for
EW M
0.15
the trace loss at full load (such adjustment was needed
N O
0.1
V TRANS_MIN
6 V HFM_FL 1.195V
FO T R
V TOTAL t
0.05
V ESR t 6
0.5m Ω
dV CAP t 6
RCS1
O
0.5m Ω
N
RCS2
0.05
6 6 6 6 5
0 2 10 4 10 6 10 8 10 1 10
0.0000 10
0 t
10.0000 10
6 In order to provide the droop required by IMVP VI applica-
tion, we will use the Block Diagram on Page 11 to determine
the component values. The reference designators that
Figure 4 - Simulated (Load Release) are used in this worksheet are from the SC452 evaluation
Transient Response board schematic.
POWER MANAGEMENT
Applications Information (Cont.)
L2 R cs2
R c s = R c s1 || R cs 2 RCS1 RCS2
RCS
C ou t
L1 +V cc_co re
RCS1 RCS2
L -R
R cs1
2.5 k Ω
2 .5k 2.5k
2.5 k Ω
R 37
0Ω
R29 R30
G ain = 1 0 V /V
R 35 DRP
We can calculate the value of R35 by using the equation
TBD
as follows:
R33
0Ω
IG ED
VDP_R35
C alculated a s
show n be lo w R35 R29 R30
VDP_CS VDP_R35
ES D
Figure 5 - Current Sense Resistive Droop Method
N
D EN
Since Rsense, R29, R30, R37 and R33 are pre-determined
to the values shown on the above diagram, we will need to
R35 26.2500 10 Ω
3
M
calculate the value of R35 to give us the correct droop.
EW M
m V
R IMVP 2.1 METHOD II: DCR Droop Method
A
FO T R
L1
DRN1 L-R Rsense +Vcc_ core Cout
R28
3 R31
18.2kΩ
TH3
Calculated C 60
R32 R37
from Step 10 39nF
Since the droop amplifier has a gain of 10V/V, the actual L2
47kΩ 0Ω
voltage appears across R35 is only VDP_R35: DRN2 L-R Rsense Gain = 10 V/V
0Ω R32
47kΩ
DRP
R35
V R36 TBD.
GDP_AMP 10 R34
18.2kΩ
TH4
39nF
R32
Calculated 47kΩ
R33
from Step 10
R32 0Ω
VDP 47kΩ
VDP_R35
GDP_AMP Droop Section of the SC452. R35 is the only variable
which is calculated as shown below
Since at full load, the voltage drop across the current sense Figure 6 - Inductive DCR Droop Method
resistor is VDP_CS.
POWER MANAGEMENT
Applications Information (Cont.)
Using Inductive DCR droop method will also require the The RC filter time constant is set by C60 and parallel
external compensation for the temperature coefficient combination of R31 and R28, TH3, R32 and R35.
of the DCR resistance. We will cover the entire thermal
design in STEP 10. Here we are only interested in deriv- In order to calculate R35 and C60, use the following
ing the correct value of the droop resistor R35. simplified diagram for Phase 1,
DC Voltage Between Two Terminals is:
To simplify the design process, here we will use the nomi- VDCR = ( I FULL_LOAD / 2 ) x RDCR
RDCR
IG ED
R32
47k:
ES D
Voltage
filter placed across the output inductor. This filter must Between these
two terminals is
C60 IR35 R35
TBD
Gain = 10 V/V
DRP_AMP
N
Droop
have the same time constant that the output inductor Veq
D EN
R28
18.2k: Requirement =
2.1mV/A
and its DCR have. If the DCR value of the inductor is R32
Droop
Requirement =
R31 47k:
0.21mV/A
very low, then care must be given to include any distrib-
M
36k:
Hence, V R35 = (0.21 mV/A) x Full Load Current
uted/parasitic impedances on the board. Second Phase Connects
Here With Identical
Component Values
EW M
In the SC452 evaluation design, this low pass filter is Considering One of the Two Phases of the SC452. The Second Phase Circuit will be exactly identical contributing the
N O
and R35 (for phase 1) and R34, R36, TH4, C61, R32
and R35 (for phase 2). For phase 1, resistor combina- Figure 7 - Inductive DCR Droop Method (Simplifed
tion R28, TH3 and R32, R35 are used to scale the mag- Diagram)
FO T R
mV
directly into the droop amplifier. The effects of this filter
R IMVP 2.1
O
A
3
92.4000 u 10
From Step 1 we have the inductor and DCR value as:
VDP V
9
L1 470.0000 u 10 H
Since the droop amplifier has a gain of 10V/V, the
9
actual voltage appears across R35 is only VR35
L2 470.0000 u 10 H
VR35
VDP
GDP_AMP
DCR: = 1.2 • m • Ω
10.0000 u 10
The time constant of the output inductor is given by, 0
GDP_AMP
W
L1
DCR
6
W 391.6667 u 10 s
POWER MANAGEMENT
Applications Information (Cont.)
Consider Phase 1 Consider Phase 2
Since
Sinceat fulldroop
the load, amplifi
the voltage drop
er has across
a gain ofthe DCR is
10V/V, the Since the components and current are identical for phase
VDCR,
actual voltage appears across R35 is only VR35 2 we get the same expression for voltage across R35,
IG ED
We
R31can 16 k Ω the value of R35 as follows:
calculate
Substituting equations A,B,C, D, E, F, G into H and simplify-
ES D
ing we get the following final expression for R35:
•k•Ω
N
R
We31:can
16 calculate
D EN
the value of R35 as follows:
18.2 k Ω
M
R28 TH3 R28 2R32 TH3 R28 R31 2 R32 R31
EW M
R35_DCR
2 VDCR TH3 R28
33 k Ω
TH3 R28 R31
N O
TH3 VR35
R EC
47 k Ω
31.6633 10 Ω
R32
3
FO T R
R35_DCR
Req Rcombination
TH3 R28 R35 2R32 R32 R32 R35_DCR R31 R28 TH3
N
So,
11.1125 10 Ω
3
Rcombination
Req
Veq VDCR
Req R31
3
C60
Now, Rcombination
Veq 9
IR35 C60 35.2457 10 F
2 R32 R35
POWER MANAGEMENT
Applications Information (Cont.)
We use the standard resistor value of 32K for R35 in the Cinput 66 μF
case of DCR droop method. For C60 and C61, we use
standard capacitor value of 7nF. 1
L in
2
STEP 3: Calculate Input RMS Current (for input capaci- 2 FS C input
tor selection)
IG ED
n 2
formula is more accurate:
ES D
VHFM_FL POUT PIN
IIN_DC
N
D PIN
VINMIN
D EN 85% VINMIN
M
POUT IHFM_FL VHFM_FL 0
IIN_DC 5.1328 10 A
EW M
N O
0
POUT 43.6284 10 W
R EC
2
I HFM_FL 2
I RMS2 n I IN_DC D n I IN_DC 1 D
The simplified expression is given as: n
FO T R
2
1 2 n 2 3 0
I RMS 2 9.2970 10 A
O
IRMS1 7.93 A
Selection:
POWER MANAGEMENT
Applications Information (Cont.)
0.0012 Ω
STEP 4: Combi-Sense Component Values
Calculation R IND
Ω
3
RSENSE 3.7658 10
CS1P CS1N
L1 4
IG ED
Figure 8 - The Equivalent Circuit of 1.2481 10 s
Combi-Sense During On-Time R SENSE
ES D
Lout
In Figures 8 and 9 RCOMBI is the parallel combination of
N
D EN
Vout
R4 and R7. R4 is the gain setting resistor and R7 with R4
DCR
Ccombi sets RCOMBI.
VPN
M
Cout
Rcombi
10 Ω
Rds_on 3
EW M
R7 7.5
CS1P CS1N
N O
100 10 Ω
3
R4
R EC
Figure 9 - The Equivalent Circuit of Note: We assume R7 and R4 to be 7.5k and 100k respectively.
FO T R
Combi-Sense During Off-Time One can vary R4 to have desired gain and the signal for combi-
sense. If thermal compensation is required please refer Step
The equivalent combi-sense resistance is then given by the 10.
O
following
The equations:
equivalent Comb-Sense resistance is then given by 1
N
R COMBI
the following equations: 1 1
1.25 Ω
0.008 R7 R4
R HS_FE T ---> 2 X IRF7821 per phase @
2
6.9767 10 Ω
3
75°C, per datasheet effective Rds-on at 75°C is 1.25 x RCOMBI
8mΩ
L1
1.2 Ω
0.004
R LS _FET --->2 X IRF7832 per phase @ 75°C, CCOMBI
2 RSENSE R7
per datasheet effective Rds-on at 75°C is 1.2 x 4mΩ.
9
CCOMBI 16.6408 10 F
POWER MANAGEMENT
Applications Information (Cont.)
STEP 5: Hysteresis (Frequency) Setting Using the Taylor Series we can expand the right hand
side.
The next step is to calculate the desired hysteresis voltage
level for a fixed switching frequency. Since SC452 is a If we consider only the first term of the Taylor series the
hysteretic converter with the benefits of Combi-Sense equation is greatly simplified with about a 20% error in
technology, the ripple for the hysteretic switching control is the final value. If first 3 terms are considered the equation
provided by the Combi-Sense current feedback. will give virtually zero error,
The ripple inputs are fed to a summer via a multiplexer Ton Vcombi ( Ton)
which is synchronized to the active phase with the select Vhys_low
Rcombi Ccombi Vin Vout
ouput. The hysteresis signal is added at the summer. The
hystersisis voltage is set directly by the resistor divider from
IG ED
the 2V REF output. Ton
Vcombi ( Ton) ( Vin Vout) Vhys_low
Rcombi Ccombi
ES D
From the circuit in Figure 8 on Page 27, we can write the
N
following equation for Combi-Sense operation during on
time (high side MOSFETs are on),
Vin Vout Vcombi ( t)
D EN
Rcombi Ccombi
dVcombi ( t)
For more accuracy we have to consider more than the
first term of the Taylor series for the hysteretic frequency
M
prediction.
dt
EW M
The Feedback
1 1
R EC
dt dVcombi
Rcombi Ccombi Vin Vout Vcombi ( t( t) ) The feedback voltage essentially contains two components,
the voltage across the output capacitor and the voltage
FO T R
1
Vfb( t) i esr i dt
N
Cout
Hence, the solution for the above equation is:
POWER MANAGEMENT
Applications Information (Cont.)
The Droop Example Showing Variation with VHYS_LOW Voltage:
The droop circuit is the simple inductor sense feedback for LOUT 0.47 μ H
the SC452 with a gain of 0.00021.
6
Vin Vout C OUT 330.0000 10 F
Vdrp( Ton) Ton 0.00021
2 Lout
COMBI RCOMBI CCOMBI
IG ED
0.0126 Ω
Final Expresssion
R calc 6 ESR MAX
2 Vcombi ( Ton) 10 Vdrp( Ton) Vfb( Ton) 12
ES D
N
following final result for tON, D EN
Substituting 1, 2 and 3 in the above equation we get the RCALC is a constant obtained from equation 4 above. It
enables better presentation of the solution of equation 4
M
as follows:
3 2 2 0.0252 6 e sr
EW M
To n To n
L ou t Co u t Rco mb i Cco mb i 2 L ou t L ou t d NOM
Fs V HYS_LO
N O
Ton V HYS_LO
R EC
2 Vhys_low 0 2
2L OUT COMBI R calc 2L OUT COMBI R calc 3
4 2V HYS _LO
FO T R
of tON only. Finding the expresssion for tOFF is a similar pro- 399.7341 10
3 4 10
5
5
3.5 10
Where, 2.5 10
5
(Hz)
Fs V HYS _LO 5
2 10
D
Fs 1.5 10
5
Ton
5
1 10
4
Vout Iout Rdrop 5 10
D 3
Vin 46.8798 10 0
0.005 0.01 0.015 0.02 0.025 0.03 0.035
0.005 V HYS _LO 0.04
(V )
POWER MANAGEMENT
Applications Information (Cont.)
STEP 6: Current Limit Calculation In current limit, the voltage hysteretic converter is over-rid-
den by the current limit hysteretic comparator, and the TG
Setting the threshold for current limit is a relatively straight- pulse is terminated when the output of the current sense
forward process. To do this we must calculate the peak amplifier reaches the CL_hi threshold and BG terminated
current based on the maximum DC value plus the worst- at the CL_lo threshold. These thresholds are set from the
case ripple current. Because the SC452 has a current-limit CLSET resistor divider:
comparator for each phase, the following calculations apply
for a single phase. CL_hi 0.33 V( clset )
Worst-case ripple occurs at the highest input voltage. Since
CL_lo 0.20 V( clset )
ripple is also inversely proportional to inductance, it is rec-
ommended that the minimum inductance value be used Current limit pulses continue until 32 pulses after the volt-
IG ED
based on the manufacturer’s specified tolerance: age droops to the PWRGD low threshold; then the controller
latches off.
ES D
3 L T OL 20%
N
D EN
d MIN 64. 3750 10 ICLIM 28.1233A
Ω
3
RSENSE 3.7658 10
M
V INMAX V HFM_NL d MIN
I RIPPLE_MAX
Per phase current limit is set by ICLIM, Rsense (Combi-Sense),
EW M
L MIN F S
current sense amplifier gain in current limit mode and CL-
N O
V 1 V
GaCS_CL 2 GaCLSET
FO T R
make an adjustment for the dynamic current-sharing tol- RSENSE ICLIM GaCS_CL
erance. We then add the maximum DC current and the VCLSET
N
GaCLSET
maximum ripple value to obtain peak current.
3
I_SHR TOL 5% ICC_MAX IHFM_FL VCLSET 635.4477 10 V
1 IRIPPLE_MAX
IPEAK ICC_MAX 1 I_SHR TOL We calculate R14 and R13 based on the VCLSET voltage,
2 2
100 10 Ω
3 VCLSET R14
IPEAK 25.5667A R14 R13
2V VCLSET
It is recommended that the current limit be set at 110%
of the peak value to allow for inductor current overshoot
R13 46568.2149Ω
during load transients:
ICLIM 1.10 IPEAK ICLIM 28.1233A We use 46.5K for R13 and 100K for R14.
POWER MANAGEMENT
Applications Information (Cont.)
STEP 7: OVP 1. Start-Up:
12 μ A
dV
No calculations are necessary for Over Voltage Protection. iSS_STARTUP iSS_STARTUP CSS
dt
The OVP circuit of SC452 monitors the processor core
Vcc_core voltage for an over-voltage condition. If the FB
dV VHFM_NL dt 3m s
voltage is 200mV greater than the DAC_Droop (i.e., out of
the powergood window), the SC452 will latch off and hold
iSS_STARTUP dt 9
the low-side driver on permanently. Either the Power or EN CSS CSS 27.9612 10 F
must be recycled to clear the latch. The latch is disabled dV
120 μ A
during soft-start and VID/DeeperSleep transitions. For
safety, the latch is enabled if the FB voltage exceeds 1.8V iSS_GVIII
even during VID/DeeperSleep transition.
IG ED
2. During GV-III Transition:
STEP 8: Thermal Protection
dV
ES D
iSS_GVIII CSS
SC452 will be disabled and latched off when the internal
N
dt
D EN
junction temperature reaches approximately 160°C. Either
the Power or EN must be recycled to clear this fault. Slew_Rate GVIII 3.2
mV
μ s
M
STEP 9: Soft-Start/DAC Slew Control dV
EW M
Slew_Rate GVIII
dt
N O
define the DAC slew rate during DeeperSleep and VID 3. During fast C4 exit:
transition. During VID/DeeperSleep transitions, the SS
1
FO T R
9
O
Resistance of the node is set to provide a fixed gain of C S S_GVIII 37. 5000 10 F
10; as a result, only 0.21mW of resistance is required to
N
120 μ A
produce the Intel-defined load line of 2.1mV/A.
dV
iSS_C4E iSS_C4E CSS
According to Intel’s IMVP-VI Timing Requirements, the dt
maximum t_SFT_START_CC is specified at 3ms(max). And
the slew rate for CPU_UP due to, dV mV
μ s
Slew_Rate C4E Slew_Rate C4E 10
dt
1) GV-III VID change is 3.2mV/μsec
2) Deeper Sleep exit is specified at 10mV/μsec 1
C S S_C4 i S S_C4E
Slew_Rate C4E
POWER MANAGEMENT
Applications Information (Cont.)
Component Manufacturer Series or Part Number
International Rectifier
IRF7821, IRF6602,
Fairchild Semiconductor
High Side MOSFET, HSFET SSC3002S,
Siliconix
Si4860DY,Si4410BDY
Infenion Technologies
International Rectifier
Fairchild Semiconductor
Low Side MOSFET, LSFET Depends on Application
Siliconix
Infenion Technologies
IG ED
Boost Diode, Dbst Various Schottky, 200mA or greater
ES D
N
Output Inductor, L Panasonic / NEC TOKIN 0.5μH
Decoupling Capacitors D EN
Various X5R or better
M
Current Sense Resistor IRC, Panasonic ERJ-M1WTJ
EW M
Output Bulk Capacitors Panasonic / NEC-TOKIN SPCAP 330μF, max ESR 6mΩ
N O
R EC
FO T R
Company Contact
Web: http://www.irf.com/product-info/
International Rectifier
O
Web: http://www.panasonic.com/pic/ecg/
Panasonic
Phone: (201) 348-7522
Web: http://www.irctt.com
IRC
Phone: (888) 472-4376
Web: http://www.kernet.com/
Kernet
Phone: (864) 963-6300
Web: http://www.sanyovideo.com/
Sanyo
Phone: (619) 661-6835
Web: http://www.component.tdk.com/components/components.html
TDK
Phone: (847) 390-4373
Web: http://www.vishay.com/brands/dale
Vishay/Dale
Phone: (402) 564-3131
Web: http://www.vishay.com/brands/siliconix
Vishay/Siliconix
Phone: (800) 554-5565
POWER MANAGEMENT
Typical Characteristics
High Frequency Line and Load Regulation Low Frequency Line and Load Regulation
VOUT = 1.2875V, IOUT = 0A to 36A (Spec bounds @ 25C) VOUT = 0.8375V, IOUT = 0A to 9.5A (Spec bounds @ 25C)
1.320 0.855
0.850
1.300
0.845
0.840
1.280
0.815
1.220
0.810
IG ED
1.200
0.805
0.800
1.180 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 22.0 24.0 26.0 28.0 30.0 32.0 34.0 36.0
ES D
I (A)
OUT
I (A)
OUT
N
Deeper Sleep Line and Load Regulation
VOUT = 0.7625V, IOUT = 0A to 3.5A (Spec bounds @ 25C) D EN SC452 High Frequency Mode Output Power Efficiency
VOUT=1.2875V, IOUT=2A to 36A
M
0.780 95.0%
EW M
90.0%
0.775
N O
85.0%
0.770
80.0%
R EC
10Vin (125C)
Spec Min.
0.755
Spec Max.
65.0%
0.750
60.0%
O
0.745 55.0%
N
0.740 50.0%
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 22.0 24.0 26.0 28.0 30.0 32.0 34.0 36.0
I (A) I (A)
OUT OUT
SC452 Low Frequency Mode Output Power Efficiency SC452 Deeper Sleep (C4) Mode Output Power Efficiency
VOUT=0.8375V, IOUT=1A to 9.5A VOUT=0.7625V, IOUT=0.25A to 3.5A
95.0%
100.0%
90.0%
95.0%
85.0% 90.0%
80.0% 85.0%
65.0%
65.0%
60.0%
60.0%
55.0% 55.0%
50.0%
50.0%
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5
I (A)
I (A) OUT
OUT
POWER MANAGEMENT
Typical Characteristics (Cont.)
Phase 1 vs. Phase 2 Current Share Balance
(specified at IOUT > 50% load)
50.0%
45.0%
40.0%
35.0%
15.0%
10.0%
5.0%
IG ED
0.0%
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 22.0 24.0 26.0 28.0 30.0 32.0 34.0 36.0
IOUT (A)
ES D
N
D EN
M
EW M
N O
R EC
FO T R
O
N
POWER MANAGEMENT
Typical Characteristics (Cont.)
IG ED
ES D
N
D EN
M
EW M
N O
R EC
FO T R
O
N
POWER MANAGEMENT
Outline Drawing - MLP-44
A D B
DIMENSIONS
INCHES MILLIMETERS
DIM
MIN NOM MAX MIN NOM MAX
A .031 - .040 0.80 - 1.00
A1 .000 - .002 0.00 - 0.05
A2 - (.008) - - (0.20) -
PIN 1 b
INDICATOR .007 .010 .012 0.18 0.25 0.30
E D .271 .275 .279 6.90 7.00 7.10
(LASER MARK)
D1 .197 .203 .207 5.00 5.15 5.25
E .271 .275 .279 6.90 7.00 7.10
E1 .197 .203 .207 5.00 5.15 5.25
l
e .020 BSC 0.50 BSC
IG ED
L .017 .021 .025 0.45 0.55 0.65
tia
N 44 44
aaa .003 0.08
bbb .004 0.10
A2
ES D
A SEATING
N
D EN
aaa C en PLANE
A1 C
D1
M
LxN
EW M
id
E/2
N O
R C
E1
nf O RE
2
1
F o
N
T
bxN
bbb C A B
O
e
C
NOTES: D/2
N
POWER MANAGEMENT
Land Pattern - MLP-44
DIMENSIONS
DIM INCHES MILLIMETERS
C (.268) (6.80)
G .228 5.80
(C) K Z H .207 5.25
G
K .207 5.25
P .021 0.50
IG ED
X .011 0.30
Y .039 1.00
Z .307 7.80
Y
ES D
N
D EN
X
P
M
NOTES:
1. THIS LAND PATTERN IS FOR REFERENCE PURPOSES ONLY.
EW M
Contact Information
Taiwan Branch Tel: 886-2-2748-3380 Semtech Switzerland GmbH Tel: 81-3-6408-0950
Fax: 886-2-2748-3390 Japan Branch Fax: 81-3-6408-0951
www.semtech.com