ISL62873 ® Data Sheet June 30, 2009 FN6930.0 PWM DC/DC Controller with VID Inputs for Portable GPU Core-Voltage Regulator Features The ISL62873 is a Single-Phase Synchronous-Buck PWM voltage regulator featuring Intersil’s Robust Ripple Regulator (R3) Technology™. The wide 3.3V to 25V input voltage range is ideal for systems that run on battery or AC-adapter power sources. The ISL62873 is a low-cost solution for applications requiring dynamically selected slew-rate controlled output voltages. The soft-start and dynamic setpoint slew-rates are capacitor programmed. Voltage identification logic-inputs select two resistor-programmed setpoint reference voltages that directly set the output voltage of the converter between 0.5V to 1.5V, and up to 3.3V using a feedback voltage divider. Optionally, an external reference such as the DAC output from a microcontroller, can be used by either IC to program the setpoint reference voltage, and still maintain the controlled slew-rate features. Robust integrated MOSFET drivers and Schottky bootstrap diode reduce the implementation area and lower component cost. • Output Voltage Range: 0.5V to 3.3V Intersil’s R3 Technology™ combines the best features of both fixed-frequency and hysteretic PWM control. The PWM frequency is 300kHz during static operation, becoming variable during changes in load, setpoint voltage, and input voltage when changing between battery and AC-adapter power. The modulators ability to change the PWM switching frequency during these events in conjunction with external loop compensation produces superior transient response. For maximum efficiency, the converter automatically enters diode-emulation mode (DEM) during light-load conditions such as system standby. • Choice of Overcurrent Detection Schemes - Lossless Inductor DCR Current Sensing - Precision Resistive Current Sensing Pinout Applications GND 1 • Output Load up to 30A • Flexible Output Voltage Programmability - 1-Bit VID Selects Two Independent Setpoint Voltages - Simple Resistor Programming of Setpoint Voltages - Accepts External Setpoint Reference such as DAC • ±0.75% System Accuracy: -10°C to +100°C • One Capacitor Programs Soft-start and Setpoint Slew-rate • Fixed 300kHz PWM Frequency in Continuous Conduction • External Compensation Affords Optimum Control Loop Tuning • Automatic Diode Emulation Mode for Highest Efficiency • Integrated High-current MOSFET Drivers and Schottky Boot-Strap Diode for Optimal Efficiency • Power-Good Monitor for Soft-Start and Fault Detection • Fault Protection - Undervoltage - Overcurrent (DCR-Sense or Resistive-Sense Capability) - Over-Temperature Protection - Fault Identification by PGOOD Pull-Down Resistance • Pb-Free (RoHS compliant) • Mobile PC Graphical Processing Unit VCC Rail • Mobile PC I/O Controller Hub (ICH) VCC Rail • Mobile PC Memory Controller Hub (GMCH) VCC Rail 13 VCC 14 PVCC 15 LGATE 16 PGND ISL62873 (16 LD 2.6X1.8 µTQFN) TOP VIEW • Input Voltage Range: 3.3V to 25V • Built-In Voltage Margin for System-Level Test Ordering Information 12 BOOT EN 2 11 UGATE VID0 3 10 PHASE SREF 4 9 OCSET PART NUMBER (Note) TEMP RANGE (°C) PACKAGE (Pb-Free) PKG. DWG. # ISL62873HRUZ-T* -10 to +100 16 Ld 2.6x1.8 µTQFN L16.2.6x1.8A Tape and Reel FB 7 1 VO 8 PGOOD 6 SET0 5 *Please refer to TB347 for details on reel specifications. NOTE: These Intersil Pb-free plastic packaged products employ special Pb-free material sets; molding compounds/die attach materials and NiPdAu plate - e4 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. 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. Block Diagram EN VCC 100kΩ POR 2 − EA + FB VW VCOMP BOOT RUN RUN FAULT PWM H L IN DRIVER UGATE PHASE SHOOT-THROUGH PROTECTION OTP PVCC PWM RUN LGATE DRIVER 100pF PGND VCC gmVIN + + − VSET ISL62873 − Cr VR SW0 + SREF SW1 SET0 gmVO − OCP + VID DECODER VID0 FB EXT VREF GND 500mV INT SW4 − UVP + − FAULT VO OCSET IOCSET 10µF PGOOD FIGURE 1. SIMPLIFIED FUNCTIONAL BLOCK DIAGRAM OF ISL62873 FN6930.0 June 30, 2009 ISL62873 Application Schematics RVCC +5V CPVCC CVCC VCC 13 9 QLS OCSET CBOOT COC COB COCSET VO SET0 VOUT 0.5V TO 3.3V LO PHASE 8 4 UGATE ROCSET PVCC 14 10 7 3 RO RCOMP ROFS RPGOOD VCC CINB QHS BOOT RSET2 CSOFT RSET1 LGATE 11 5 SREF 2 FB VID0 12 6 EN GPIO CINC 1 PGOOD GND 15 16 PGND VIN 3.3V TO 25V CCOMP RFB GPIO FIGURE 2. ISL62873 APPLICATION SCHEMATIC WITH TWO OUTPUT VOLTAGE SETPOINTS AND DCR CURRENT SENSE RVCC +5V VCC QHS BOOT UGATE LO RSNS PHASE QLS OCSET ROCSET 9 CBOOT VO FB CINB 13 4 8 10 7 3 SET0 VOUT 0.5V TO 3.3V COC COB COCSET RO RCOMP ROFS RPGOOD VCC PVCC 11 RSET2 CSOFT RSET1 14 2 5 SREF CINC 12 6 VID0 VIN 3.3V TO 25V 1 PGOOD EN GPIO 15 16 GND LGATE CVCC PGND CPVCC CCOMP RFB GPIO FIGURE 3. ISL62873 APPLICATION SCHEMATIC WITH TWO OUTPUT VOLTAGE SETPOINTS AND RESISTOR CURRENT SENSE 3 FN6930.0 June 30, 2009 ISL62873 Application Schematics (Continued) RVCC +5V VCC UGATE LO PHASE QLS OCSET ROCSET 9 CBOOT VO 8 4 QHS BOOT ROFS RPGOOD VOUT 0.5V TO 3.3V COC COB COCSET RO RCOMP GPIO CINB 13 PVCC 10 7 3 SET0 VCC 14 11 5 CSOFT SREF 2 FB EXT_REF CINC 12 6 VID0 15 16 EN GPIO VIN 3.3V TO 25V 1 PGOOD GND LGATE CVCC PGND CPVCC CCOMP RFB FIGURE 4. ISL62873 APPLICATION SCHEMATIC WITH EXTERNAL REFERENCE INPUT AND DCR CURRENT SENSE 4 FN6930.0 June 30, 2009 ISL62873 Absolute Maximum Ratings Thermal Information VCC, PVCC, PGOOD to GND . . . . . . . . . . . . . . . . . . -0.3V to +7.0V VCC, PVCC to PGND . . . . . . . . . . . . . . . . . . . . . . . . -0.3V to +7.0V GND to PGND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . -0.3V to +0.3V EN, SET0, VO, VID0, FB, OCSET, SREF. . . . . . . . . . . . -0.3V to GND, VCC + 0.3V BOOT Voltage (VBOOT-GND). . . . . . . . . . . . . . . . . . . . . -0.3V to 33V BOOT To PHASE Voltage (VBOOT-PHASE) . . . . . . -0.3V to 7V (DC) -0.3V to 9V (<10ns) PHASE Voltage . . . . . . . . . . . . . . . . . . . . . . . . . . GND - 0.3V to 28V GND -8V (<20ns Pulse Width, 10µJ) UGATE Voltage . . . . . . . . . . . . . . . . VPHASE - 0.3V (DC) to VBOOT VPHASE - 5V (<20ns Pulse Width, 10µJ) to VBOOT LGATE Voltage . . . . . . . . . . . . . . . GND - 0.3V (DC) to VCC + 0.3V GND - 2.5V (<20ns Pulse Width, 5µJ) to VCC + 0.3V Thermal Resistance (Typical, Note 1) θJA (°C/W) 16 Ld µTQFN Package . . . . . . . . . . . . . . . . . . . . . . 84 Junction Temperature Range. . . . . . . . . . . . . . . . . .-55°C to +150°C Operating Temperature Range . . . . . . . . . . . . . . . .-10°C to +100°C Storage Temperature . . . . . . . . . . . . . . . . . . . . . . . .-65°C to +150°C Pb-free Reflow Profile . . . . . . . . . . . . . . . . . . . . . . . . .see link below http://www.intersil.com/pbfree/Pb-FreeReflow.asp Recommended Operating Conditions Ambient Temperature Range. . . . . . . . . . . . . . . . . .-10°C to +100°C Converter Input Voltage to GND . . . . . . . . . . . . . . . . . . 3.3V to 25V VCC, PVCC to GND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5V ±5% 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. Electrical Specifications These specifications apply for TA = -10°C to +100°C, unless otherwise stated. All typical specifications TA = +25°C, VCC = 5V. 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 UNIT EN = 5V, VCC = 5V, FB = 0.55V, SREF < FB - 1.1 1.5 mA EN = GND, VCC = 5V - 0.1 1.0 µA EN = GND, PVCC = 5V - 0.1 1.0 µA VCC and PVCC VCC Input Bias Current IVCC VCC Shutdown Current IVCCoff PVCC Shutdown Current IPVCCoff VCC POR THRESHOLD Rising VCC POR Threshold Voltage VVCC_THR 4.40 4.49 4.60 V Falling VCC POR Threshold Voltage V 4.10 4.22 4.35 V - 0.50 - V -0.75 - +0.75 % 270 300 330 kHz 0 - 3.6 V VCC_THF REGULATION Reference Voltage VREF(int) System Accuracy VID0 = VID1 = GND, PWM Mode = CCM PWM Switching Frequency FSW PWM Mode = CCM VO VO Input Voltage Range VVO EN = 5V - 600 - kΩ VO Reference Offset Current IVOSS VENTHR < EN, SREF = Soft-Start Mode - 10 - µA VO Input Leakage Current IVOoff EN = GND, VO = 3.6V - 0.1 - µA EN = 5V, FB = 0.50V -20 - +50 nA Nominal SREF Setting with 1% Resistors 0.5 - 1.5 V VO Input Impedance RVO ERROR AMPLIFIER FB Input Bias Current IFB SREF SREF Operating Voltage Range VSREF Soft-Start Current ISS SREF = Soft-Start Mode 10 20 30 µA Voltage Step Current IVS SREF = Setpoint-Stepping Mode ±60 ±100 ±140 µA 5 FN6930.0 June 30, 2009 ISL62873 Electrical Specifications These specifications apply for TA = -10°C to +100°C, unless otherwise stated. All typical specifications TA = +25°C, VCC = 5V. 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 UNIT EXTERNAL REFERENCE EXTREF Operating Voltage Range VEXT SET0 = VCC VEXT_OFS SET0 = VCC, VID0 = 0V to 1.5V EXTREF Accuracy 0 - 1.5 V -0.5 - +0.5 % POWER GOOD PGOOD Pull-down Impedance PGOOD Leakage Current RPG_SS PGOOD = 5mA Sink 75 95 150 Ω RPG_UV PGOOD = 5mA Sink 75 95 150 Ω RPG_OV PGOOD = 5mA Sink 50 65 90 Ω RPG_OC PGOOD = 5mA Sink 25 35 50 Ω - 0.1 1.0 µA - 5.0 - mA IPG PGOOD Maximum Sink Current (Note 2) PGOOD = 5V IPG_max GATE DRIVER UGATE Pull-Up Resistance (Note 2) RUGPU 200mA Source Current - 1.0 1.5 Ω UGATE Source Current (Note 2) IUGSRC UGATE - PHASE = 2.5V - 2.0 - A UGATE Sink Resistance (Note 2) RUGPD 250mA Sink Current - 1.0 1.5 Ω UGATE Sink Current (Note 2) IUGSNK UGATE - PHASE = 2.5V - 2.0 - A LGATE Pull-Up Resistance (Note 2) RLGPU 250mA Source Current - 1.0 1.5 Ω LGATE Source Current (Note 2) ILGSRC LGATE - GND = 2.5V - 2.0 - A LGATE Sink Resistance (Note 2) RLGPD 250mA Sink Current - 0.5 0.9 Ω LGATE Sink Current (Note 2) ILGSNK LGATE - PGND = 2.5V - 4.0 - A UGATE to LGATE Deadtime tUGFLGR UGATE falling to LGATE rising, no load - 21 - ns LGATE to UGATE Deadtime tLGFUGR LGATE falling to UGATE rising, no load - 21 - ns - 33 - kΩ PHASE PHASE Input Impedance RPHASE BOOTSTRAP DIODE Forward Voltage VF PVCC = 5V, IF = 2mA - 0.58 - V Reverse Leakage IR VR = 25V - 0.2 - µA CONTROL INPUTS EN High Threshold Voltage VENTHR 2.0 - - V EN Low Threshold Voltage VENTHF - - 1.0 V 1.5 2.0 2.5 µA - 0.1 1.0 µA EN Input Bias Current IEN EN Leakage Current IENoff EN = 5V EN = GND VID<0,1> High Threshold Voltage VVIDTHR 0.6 - - V VID<0,1> Low Threshold Voltage VVIDTHF - - 0.5 V VID<0,1> Input Bias Current IVID VID<0,1> Leakage Current EN = 5V, VVID = 1V IVIDoff - 0.5 - µA - 0 - µA -1.15 - 1.15 mV PROTECTION OCP Threshold Voltage VOCPTH VOCSET - VO OCP Reference Current IOCP EN = 5.0V 9.3 10 10.5 µA OCSET Input Resistance ROCSET EN = 5.0V - 600 - kΩ OCSET Leakage Current IOCSET EN = GND - 0 - µA UVP Threshold Voltage VUVTH VFB = %VSREF 81 84 87 % 6 FN6930.0 June 30, 2009 ISL62873 Electrical Specifications These specifications apply for TA = -10°C to +100°C, unless otherwise stated. All typical specifications TA = +25°C, VCC = 5V. 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 UNIT OTP Rising Threshold Temperature (Note 2) TOTRTH - 150 - °C OTP Hysteresis (Note 2) TOTHYS - 25 - °C NOTE: 2. Limits established by characterization and are not production tested. Functional Pin Descriptions OCSET (Pin 9) IC ground for bias supply and signal reference. Input for the overcurrent detection circuit. The overcurrent setpoint programming resistor ROCSET connects from this pin to the sense node. See Figure 6 on page 10. EN (Pin 2) PHASE (Pin 10) Enable input for the IC. Pulling EN above the VENTHR rising threshold voltage initializes the soft-start sequence. Return current path for the UGATE high-side MOSFET driver. VIN sense input for the R3 modulator. Inductor current polarity detector input. Connect to junction of output inductor, high-side MOSFET, and low-side MOSFET. See “Application Schematics” (Figures 2 and 3) on page 3. GND (Pin 1) VID0 (Pin 3) Logic input for setpoint voltage selector. Use to select between the two setpoint reference voltages. External reference input when enabled by connecting the SET0 pin to the VCC pin. SREF (Pin 4) UGATE (Pin 11) High-side MOSFET gate driver output. Connect to the gate terminal of the high-side MOSFET of the converter. Soft-start and voltage slew-rate programming capacitor input. Setpoint reference voltage programming resistor input. Connects internally to the inverting input of the VSET voltage setpoint amplifier. See Figure 5 on page 9 for capacitor and resistor connections. BOOT (Pin 12) SET0 (Pin 5) VCC (Pin 13) Voltage set-point programming resistor input. See Figure 5 on page 9 for resistor connection. Input for the IC bias voltage. Connect +5V to the VCC pin and decouple with at least a 1µF MLCC to the GND pin. See “Application Schematics” (Figures 2 and 3) on page 3. PGOOD (Pin 6) Power-good open-drain indicator output. This pin changes to high impedance when the converter is able to supply regulated voltage. The pull-down resistance between the PGOOD pin and the GND pin identifies which protective fault has shut down the regulator. See Table 2 on page 12. FB (Pin 7) Voltage feedback sense input. Connects internally to the inverting input of the control-loop error amplifier. The converter is in regulation when the voltage at the FB pin equals the voltage on the SREF pin. The control loop compensation network connects between the FB pin and the converter output. See Figure 9 on page 13. VO (Pin 8) Output voltage sense input for the R3 modulator. The VO pin also serves as the reference input for the overcurrent detection circuit. See Figure 6 on page 10. 7 Positive input supply for the UGATE high-side MOSFET gate driver. The BOOT pin is internally connected to the cathode of the Schottky boot-strap diode. Connect an MLCC between the BOOT pin and the PHASE pin. PVCC (Pin 14) Input for the LGATE and UGATE MOSFET driver circuits. The PVCC pin is internally connected to the anode of the Schottky boot-strap diode. Connect +5V to the PVCC pin and decouple with a 10µF MLCC to the PGND pin. See “Application Schematics” (Figures 2 and 3) on page 3. LGATE (Pin 15) Low-side MOSFET gate driver output. Connect to the gate terminal of the low-side MOSFET of the converter. PGND (Pin 16) Return current path for the LGATE MOSFET driver. Connect to the source of the low-side MOSFET. Setpoint Reference Voltage Programming Voltage identification (VID) pins select user-programmed setpoint reference voltages that appear at the SREF pin. The converter is in regulation when the FB pin voltage (VFB) FN6930.0 June 30, 2009 ISL62873 equals the SREF pin voltage (VSREF.) The IC measures VFB and VSREF relative to the GND pin, not the PGND pin. The setpoint reference voltages use the naming convention VSET(x) where (x) is the first, second, third, or fourth setpoint reference voltage where: - VSET1 < VSET2 < VSET3 < VSET4 - VOUT1 < VOUT2 < VOUT3 < VOUT4 The VSET1 setpoint is fixed at 500mV because it corresponds to the closure of internal switch SW0 that configures the VSET amplifier as a unity-gain voltage follower for the 500mV voltage reference VREF. A feedback voltage-divider network may be required to achieve the desired reference voltages. Using the feedback voltage-divider allows the maximum output voltage of the converter to be higher than the 1.5V maximum setpoint reference voltage that can be programmed on the SREF pin. Likewise, the feedback voltage-divider allows the minimum output voltage of the converter to be higher than the fixed 500mV setpoint reference voltage of VSET1. Scale the voltage-divider network such that the voltage VFB equals the voltage VSREF when the converter output voltage is at the desired level. The voltage-divider relation is given in Equation 1: R OFS V FB = V OUT ⋅ ---------------------------------R FB + R OFS (EQ. 1) Where: - VFB = VSREF - RFB is the loop-compensation feedback resistor that connects from the FB pin to the converter output - ROFS is the voltage-scaling programming resistor that connects from the FB pin to the GND pin The attenuation of the feedback voltage divider is written as: R OFS V SREF ( lim ) K = ------------------------------- = ---------------------------------V OUT ( lim ) R FB + R OFS (EQ. 2) Where: - K is the attenuation factor - VSREF(lim) is the VSREF voltage setpoint of either 500mV or 1.50V - VOUT(lim) is the output voltage of the converter when VSREF = VSREF(lim) Since the voltage-divider network is in the feedback path, all output voltage setpoints will be attenuated by K, so it follows that all of the setpoint reference voltages will be attenuated by K. It will be necessary then to include the attenuation factor K in all the calculations for selecting the RSET programming resistors. The value of offset resistor ROFS can be calculated only after the value of loop-compensation resistor RFB has been determined. The calculation of ROFS is written as Equation 3: V SET ( x ) ⋅ R FB R OFS = -------------------------------------------V OUT – V SET ( x ) (EQ. 3) is the first, second, third, or fourth programming resistor connected in series starting at the SREF pin and ending at the GND pin. When one of the internal switches closes, it connects the inverting input of the VSET amplifier to a specific node among the string of RSET programming resistors. All the resistors between that node and the SREF pin serve as the feedback impedance RF of the VSET amplifier. Likewise, all the resistors between that node and the GND pin serve as the input impedance RIN of the VSET amplifier. Equation 4 gives the general form of the gain equation for the VSET amplifier: RF ⎞ ⎛ V SET ( X ) = V REF ⋅ ⎜ 1 + ----------⎟ R ⎝ IN⎠ (EQ. 4) Where: - VREF is the 500mV internal reference of the IC - VSET(x) is the resulting setpoint reference voltage that appears at the SREF pin Component Selection for Setpoint Voltage Programming Resistors TABLE 1. ISL62873 VID TRUTH TABLE STATE RESULT VID0 CLOSE VSREF VOUT 1 SW0 VSET1 VOUT1 0 SW1 VSET2 VOUT2 First, determine the attenuation factor K. Next, assign an initial value to RSET2 of approximately 150kΩ then calculate RSET1 using Equation 5. The equation for the value of RSET1 is written as Equation 5: ⎛ KV SET2 ⎞ R SET1 = R SET2 ⋅ ⎜ ----------------------- – 1⎟ V ⎝ REF ⎠ (EQ. 5) The sum of RSET1 and RSET2 programming resistors should be approximately 300kΩ, as shown in Equation 6, otherwise adjust the value of RSET2 and repeat the calculations. R SET1 + R SET2 ≅ 300kΩ (EQ. 6) Equations 7 and 8 give the specific VSET gain equations for the ISL62873 setpoint reference voltages. The ISL62873 VSET1 setpoint is written as Equation 7: V SET1 = V REF (EQ. 7) The ISL62873 VSET2 setpoint is written as Equation 8: R SET1⎞ ⎛ V SET2 = V REF ⋅ ⎜ 1 + ------------------⎟ R ⎝ SET2⎠ (EQ. 8) The setpoint reference voltages are programmed with resistors that use the naming convention RSET(x) where (x) 8 FN6930.0 June 30, 2009 ISL62873 RFB FB The voltage on the SREF pin rises as ISS charges CSOFT to the voltage reference setpoint selected by the state of the VID inputs at the time the EN pin is asserted. The regulator controls the PWM such that the voltage on the FB pin tracks the rising voltage on the SREF pin. Once CSOFT charges to the selected setpoint voltage, the ISS current source comes out of the 20µA current limit and decays to the static value set by VSREF ÷ RT. The elapsed time from when the EN pin is asserted to when VSREF has reached the voltage reference setpoint is the soft-start delay tSS which is given by Equation 10: VCOMP − EA ROFS VOUT + VREF + VSET − SW0 RSET1 V START-UP t SS = – ( R T ⋅ C SOFT ) ⋅ LN(1 – ------------------------------) I SS ⋅ R T SW1 SET0 (EQ. 10) Where: RSET2 CSOFT SREF FIGURE 5. ISL62873 VOLTAGE PROGRAMMING CIRCUIT External Setpoint Reference The IC can use an external setpoint reference voltage as an alternative to VID-selected, resistor-programmed setpoints. This is accomplished by removing all setpoint programming resistors, connecting the SET0 pin to the VCC pin, and feeding the external setpoint reference voltage to the VID0 pin. When SET0 and VCC are tied together, the following internal reconfigurations take place: - VID0 pin opens its 500nA pull-down current sink - Reference source selector switch SW4 moves from INT position (internal 500mV) to EXT position (VID0 pin) - VID1 pin is disabled The converter will now be in regulation when the voltage on the FB pin equals the voltage on the VID0 pin. As with resistor-programmed setpoints, the reference voltage range on the VID0 pin is 500mV to 1.5V. Use Equations 1, 2, and 3 beginning on page 8 should it become necessary to implement an output voltage-divider network to make the external setpoint reference voltage compatible with the 500mV to 1.5V constraint. Soft-Start and Voltage-Step Delay Circuit Description When the voltage on the VCC pin has ramped above the rising power-on reset voltage VVCC_THR, and the voltage on the EN pin has increased above the rising enable threshold voltage VENTHR, the SREF pin releases its discharge clamp and enables the reference amplifier VSET. The soft-start current ISS is limited to 20µA and is sourced out of the SREF pin into the parallel RC network of capacitor CSOFT and resistance RT. The resistance RT is the sum of all the series connected RSET programming resistors and is written as Equation 9: R T = R SET1 + R SET2 + …R SET ( n ) 9 - ISS is the soft-start current source at the 20µA limit - VSTART-UP is the setpoint reference voltage selected by the state of the VID inputs at the time EN is asserted - RT is the sum of the RSET programming resistors The end of soft-start is detected by ISS tapering off when capacitor CSOFT charges to the designated VSET voltage reference setpoint. The SSOK flag is set, the PGOOD pin goes high, and the ISS current source changes over to the voltage-step current source IVS which has a current limit of ±100µA. Whenever the VID inputs or the external setpoint reference, programs a different setpoint reference voltage, the IVS current source charges or discharges capacitor CSOFT to that new level at ±100µA. Once CSOFT charges to the selected setpoint voltage, the IVS current source comes out of the 100µA current limit and decays to the static value set by VSREF ÷ RT. The elapsed time to charge CSOFT to the new voltage is called the voltage-step delay tVS and is given by Equation 11: ( V NEW – V OLD ) t VS = ( R T ⋅ C SOFT ) ⋅ LN(1 – -------------------------------------------) I ⋅R VS (EQ. 11) T Where: - IVS is the ±100µA setpoint voltage-step current - VNEW is the new setpoint voltage selected by the VID inputs - VOLD is the setpoint voltage that VNEW is changing from - RT is the sum of the RSET programming resistors Component Selection for CSOFT Capacitor Choosing the CSOFT capacitor to meet the requirements of a particular soft-start delay tSS is calculated with Equation 12, which is written as: – t SS C SOFT = --------------------------------------------------------------------V START-UP ⎞ ⎛ ⎜ R T ⋅ LN(1 – ------------------------------)⎟ I SS ⋅ R T ⎠ ⎝ (EQ. 12) (EQ. 9) FN6930.0 June 30, 2009 ISL62873 Where: - tSS is the soft-start delay ISS is the soft-start current source at the 20µA limit VSTART-UP is the setpoint reference voltage selected by the state of the VID inputs at the time EN is asserted - RT is the sum of the RSET programming resistors Choosing the CSOFT capacitor to meet the requirements of a particular voltage-step delay tVS is calculated with Equation 13, which is written as: – t VS C SOFT = -----------------------------------------------------------------------------V NEW – V OLD ⎞ ⎛ ⎜ R T ⋅ LN(1 – ---------------------------------------)⎟ ± I VS ⋅ R T ⎠ ⎝ (EQ. 13) Where: tVS is the voltage-step delay VNEW is the new setpoint voltage VOLD is the setpoint voltage that VNEW is changing from - IVS is the ±100µA setpoint voltage-step current; positive when VNEW > VOLD, negative when VNEW < VOLD - RT is the sum of the RSET programming resistors - Fault Protection Overcurrent The overcurrent protection (OCP) setpoint is programmed with resistor ROCSET which is connected across the OCSET and PHASE pins. Resistor RO is connected between the VO pin and the actual output voltage of the converter. During normal operation, the VO pin is a high impedance path, therefore there is no voltage drop across RO. The value of resistor RO should always match the value of resistor ROCSET. L DCR PHASE IL + ROCSET 10µA OCSET + VROCSET VDCR CSEN _ VO CO RO VO FIGURE 6. OVERCURRENT PROGRAMMING CIRCUIT Figure 6 shows the overcurrent set circuit. The inductor consists of inductance L and the DC resistance DCR. The inductor DC current IL creates a voltage drop across DCR, which is given by Equation 14: (EQ. 14) 10 V ROCSET = 10μA ⋅ R OCSET (EQ. 15) The DC voltage difference between the OCSET pin and the VO pin, which is given by Equation 16: V OCSET – V VO = V DCR – V ROCSET = I L ⋅ DCR – I OCSET ⋅ R OCSET (EQ. 16) The IC monitors the voltage of the OCSET pin and the VO pin. When the voltage of the OCSET pin is higher than the voltage of the VO pin for more than 10µs, an OCP fault latches the converter off. Component Selection for ROCSET and CSEN The value of ROCSET is calculated with Equation 17, which is written as: I OC ⋅ DCR R OCSET = ---------------------------I OCSET (EQ. 17) Where: - ROCSET (Ω) is the resistor used to program the overcurrent setpoint - IOC is the output DC load current that will activate the OCP fault detection circuit - DCR is the inductor DC resistance For example, if IOC is 20A and DCR is 4.5mΩ, the choice of ROCSET is = 20A x 4.5mΩ/10µA = 9kΩ. Resistor ROCSET and capacitor CSEN form an R-C network to sense the inductor current. To sense the inductor current correctly not only in DC operation, but also during dynamic operation, the R-C network time constant ROCSET CSEN needs to match the inductor time constant L/DCR. The value of CSEN is then written as Equation 18: L C SEN = -----------------------------------------R OCSET ⋅ DCR _ V DCR = I L ⋅ DCR The IOCSET current source sinks 10µA into the OCSET pin, creating a DC voltage drop across the resistor ROCSET, which is given by Equation 15: (EQ. 18) For example, if L is 1.5µH, DCR is 4.5mΩ, and ROCSET is 9kΩ, the choice of CSEN = 1.5µH/(9kΩ x 4.5mΩ) = 0.037µF. When an OCP fault is declared, the PGOOD pin will pull-down to 35Ω and latch off the converter. The fault will remain latched until the EN pin has been pulled below the falling EN threshold voltage VENTHF or if VCC has decayed below the falling POR threshold voltage VVCC_THF. Undervoltage The UVP fault detection circuit triggers after the FB pin voltage is below the undervoltage threshold VUVTH for more than 2µs. For example, if the converter is programmed to regulate 1.0V at the FB pin, that voltage would have to fall below the typical VUVTH threshold of 84% for more than 2µs in order to trip the UVP fault latch. In numerical terms, that would be 84% x 1.0V = 0.84V. When a UVP fault is declared, the PGOOD pin will pull-down to 95Ω and latch-off the FN6930.0 June 30, 2009 ISL62873 converter. The fault will remain latched until the EN pin has been pulled below the falling EN threshold voltage VENTHF or if VCC has decayed below the falling POR threshold voltage VVCC_THF. The negative slope of VR can be written as Equation 20: Over-Temperature 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 controlled internally by the IC. 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 7 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. When the temperature of the IC increases above the rising threshold temperature TOTRTH, it will enter the OTP state that suspends the PWM, forcing the LGATE and UGATE gate-driver outputs low. The status of the PGOOD pin does not change nor does the converter latch-off. The PWM remains suspended until the IC temperature falls below the hysteresis temperature TOTHYS at which time normal PWM operation resumes. The OTP state can be reset if the EN pin is pulled below the falling EN threshold voltage VENTHF or if VCC has decayed below the falling POR threshold voltage V VCC_THF. All other protection circuits remain functional while the IC is in the OTP state. It is likely that the IC will detect an UVP fault because in the absence of PWM, the output voltage decays below the undervoltage threshold VUVTH. V RNEG = g m ⋅ V OUT ⁄ C R (EQ. 20) Where, gm is the gain of the transconductance amplifier. RIPPLE CAPACITOR VOLTAGE CR WINDOW VOLTAGE VW Theory of Operation The modulator features Intersil’s R3 Robust-RippleRegulator technology, a hybrid of fixed frequency PWM control and variable frequency hysteretic control. The PWM frequency is maintained at 300kHz under static continuous-conduction-mode operation within the entire specified envelope of input voltage, output voltage, and output load. If the application should experience a rising load transient and/or a falling line transient such that the output voltage starts to fall, the modulator will extend the on-time and/or reduce the off-time of the PWM pulse in progress. Conversely, if the application should experience a falling load transient and/or a rising line transient such that the output voltage starts to rise, the modulator will truncate the on-time and/or extend the off-time of the PWM pulse in progress. The period and duty cycle of the ensuing PWM pulses are optimized by the R3 modulator for the remainder of the transient and work in concert with the error amplifier VERR to maintain output voltage regulation. Once the transient has dissipated and the control loop has recovered, the PWM frequency returns to the nominal static 300kHz. Modulator 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 input voltage (VIN) at the PHASE pin and output voltage (VOUT) at the VO pin. The positive slope of VR can be written as Equation 19: V RPOS = ( g m ) ⋅ ( V IN – V OUT ) ⁄ C R 11 (EQ. 19) ERROR AMPLIFIER VOLTAGE VCOMP PWM FIGURE 7. MODULATOR WAVEFORMS DURING LOAD TRANSIENT Synchronous Rectification A standard DC/DC buck regulator uses a free-wheeling diode to maintain uninterrupted current conduction through the output inductor when the high-side MOSFET switches off for the balance of the PWM switching cycle. Low conversion efficiency as a result of the conduction loss of the diode makes this an unattractive option for all but the lowest current applications. Efficiency is dramatically improved when the free-wheeling diode is replaced with a MOSFET that is turned on whenever the high-side MOSFET is turned off. This modification to the standard DC/DC buck regulator is referred to as synchronous rectification, the topology implemented by the ISL62873 controller. Diode Emulation The polarity of the output inductor current is defined as positive when conducting away from the phase node, and defined as negative when conducting towards the phase node. The DC component of the inductor current is positive, but the AC component known as the ripple current, can be either positive or negative. Should the sum of the AC and DC components of the inductor current remain positive for the entire switching period, the converter is in FN6930.0 June 30, 2009 ISL62873 continuous-conduction-mode (CCM.) However, if the inductor current becomes negative or zero, the converter is in discontinuous-conduction-mode (DCM.) Unlike the standard DC/DC buck regulator, the synchronous rectifier can sink current from the output filter inductor during DCM, reducing the light-load efficiency with unnecessary conduction loss as the low-side MOSFET sinks the inductor current. The ISL62873 controller avoids the DCM conduction loss by making the low-side MOSFET emulate the currentblocking behavior of a diode. This smart-diode operation called diode-emulation-mode (DEM) is triggered when the negative inductor current produces a positive voltage drop across the rDS(ON) of the low-side MOSFET for eight consecutive PWM cycles while the LGATE pin is high. The converter will exit DEM on the next PWM pulse after detecting a negative voltage across the rDS(ON) of the lowside MOSFET. It is characteristic of the R3 architecture for the PWM switching frequency to decrease while in DCM, increasing efficiency by reducing unnecessary gate-driver switching losses. The extent of the frequency reduction is proportional to the reduction of load current. Upon entering DEM, the PWM frequency is forced to fall approximately 30% by forcing a similar increase of the window voltage V W. This measure is taken to prevent oscillating between modes at the boundary between CCM and DCM. The 30% increase of VW is removed upon exit of DEM, forcing the PWM switching frequency to jump back to the nominal CCM value. Power-On Reset The IC is disabled until the voltage at the VCC pin has increased above the rising power-on reset (POR) threshold voltage VVCC_THR. The controller will become disabled when the voltage at the VCC pin decreases below the falling POR threshold voltage VVCC_THF. The POR detector has a noise filter of approximately 1µs. PGOOD Monitor The PGOOD pin indicates when the converter is capable of supplying regulated voltage. The PGOOD pin is an undefined impedance if the VCC pin has not reached the rising POR threshold VVCC_THR, or if the VCC pin is below the falling POR threshold VVCC_THF. The PGOOD pull-down resistance corresponds to a specific protective fault, thereby reducing troubleshooting time and effort. Table 2 maps the pull-down resistance of the PGOOD pin to the corresponding fault status of the controller. TABLE 2. PGOOD PULL-DOWN RESISTANCE CONDITION PGOOD RESISTANCE VCC Below POR Undefined Soft-Start or Undervoltage 95Ω Overcurrent 35Ω LGATE and UGATE MOSFET Gate-Drivers The LGATE pin and UGATE pins are MOSFET driver outputs. The LGATE pin drives the low-side MOSFET of the converter while the UGATE pin drives the high-side MOSFET of the converter. The LGATE driver is optimized for low duty-cycle applications where the low-side MOSFET experiences long conduction times. In this environment, the low-side MOSFETs require exceptionally low rDS(ON) and tend to have large parasitic charges that conduct transient currents within the devices in response to high dv/dt switching present at the phase node. The drain-gate charge in particular can conduct sufficient current through the driver pull-down resistance that the VGS(th) of the device can be exceeded and turned on. For this reason the LGATE driver has been designed with low pull-down resistance and high sink current capability to ensure clamping the MOSFETs gate voltage below VGS(th). VIN and PVCC Voltage Sequence Prior to pulling EN above the VENTHR rising threshold voltage, the following criteria must be met: UGATE - VPVCC is at least equivalent to the VCC rising power-on reset voltage VVCC_THR - VVIN must be 3.3V or the minimum required by the application 1V 1V 1V 1V Start-Up Timing Once VCC has ramped above VVCC_THR, the controller can be enabled by pulling the EN pin voltage above the input-high threshold VENTHR. Approximately 20µs later, the voltage at the SREF pin begins slewing to the designated VID set-point. The converter output voltage at the FB feedback pin follows the voltage at the SREF pin. During soft-start, The regulator always operates in CCM until the soft-start sequence is complete. 12 LGATE FIGURE 8. GATE DRIVER ADAPTIVE SHOOT-THROUGH FN6930.0 June 30, 2009 ISL62873 Adaptive Shoot-Through Protection Adaptive shoot-through protection prevents a gate-driver output from turning on until the opposite gate-driver output has fallen below approximately 1V. The dead-time shown in Figure 8 is extended by the additional period that the falling gate voltage remains above the 1V threshold. 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 supplied by a boot-strap capacitor connected across the BOOT and PHASE pins. The capacitor is charged each time the phase node voltage falls a diode drop below PVCC such as when the low-side MOSFET is turned on. Compensation Design Figure 9 shows the recommended Type-II compensation circuit. The FB pin is the inverting input of the error amplifier. The COMP signal, the output of the error amplifier, is inside the chip and unavailable to users. CINT is a 100pF capacitor integrated inside the IC, connecting across the FB pin and the COMP signal. RFB, RCOMP, CCOMP and CINT form the Type-II compensator. The frequency domain transfer function is given by Equation 21: 1 + s ⋅ ( R FB + R COMP ) ⋅ C COMP G COMP ( s ) = --------------------------------------------------------------------------------------------------------------- (EQ. 21) s ⋅ R FB ⋅ C INT ⋅ ( 1 + s ⋅ R COMP ⋅ C ) COMP CINT = 100pF - CCOMP RCOMP RFB VOUT FB EA COMP + ROFS SREF reference designs that include schematics, bills of materials, and example board layouts. Selecting the LC Output Filter The duty cycle of an ideal buck converter is a function of the input and the output voltage. This relationship is expressed in Equation 22: VO D = --------V IN (EQ. 22) The output inductor peak-to-peak ripple current is expressed in Equation 23: VO ⋅ ( 1 – D ) I P-P = ------------------------------F SW ⋅ L (EQ. 23) A typical step-down DC/DC converter will have 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 using Equation 24: 2 P COPPER = I LOAD ⋅ DCR (EQ. 24) Where, ILOAD is the converter output DC current. The copper loss can be significant so attention has to 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 expressed in Equations 25 and 26: ΔV ESR = I P-P ⋅ E SR (EQ. 25) I P-P ΔΔV C = --------------------------------8 ⋅ CO ⋅ F (EQ. 26) SW FIGURE 9. COMPENSATION REFERENCE CIRCUIT The LC output filter has a double pole at its resonant frequency that causes rapid phase change. The R3 modulator used in the IC makes the LC output filter resemble a first order system in which the closed loop stability can be achieved with the recommended Type-II compensation network. Intersil provides a PC-based tool that can be used to calculate compensation network component values and help simulate the loop frequency response. 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. In addition to this guide, Intersil provides complete 13 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. 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 fade as much as 50% as the DC voltage across it increases. Selection of the Input Capacitor 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 FN6930.0 June 30, 2009 ISL62873 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 10 is a graph of the input RMS ripple current, normalized relative to output load current, as a function of duty cycle that is adjusted for converter efficiency. The ripple current calculation is written as Equation 27: 2 2 D 2 ( I MAX ⋅ ( D – D ) ) + ⎛ x ⋅ I MAX ⋅ ------ ⎞ ⎝ 12 ⎠ I IN_RMS = ----------------------------------------------------------------------------------------------------I MAX (EQ. 27) Where: Where: - QGATE is the amount of gate charge required to fully charge the gate of the upper MOSFET - ΔVBOOT is the maximum decay across the BOOT capacitor As an example, suppose an upper MOSFET has a gate charge, QGATE , of 25nC at 5V and also assume the droop in the drive voltage over a PWM cycle is 200mV. One will find that a bootstrap capacitance of at least 0.125µF is required. The next larger standard value capacitance is 0.15µF. A good quality ceramic capacitor such as X7R or X5R is recommended. 2.0 VO D = -------------------------V IN ⋅ EFF (EQ. 28) 1.8 1.6 CBOOT_CAP (µF) - IMAX is the maximum continuous ILOAD of the converter - x is a multiplier (0 to 1) corresponding to the inductor peak-to-peak ripple amplitude expressed as a percentage of IMAX (0% to 100%) - D is the duty cycle that is adjusted to take into account the efficiency of the converter Duty cycle is written as Equation 28: 1.0 0.8 QGATE = 100nC nC 50 NORMALIZED INPUT RMS RIPPLE CURRENT 1.2 0.6 0.4 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. 0.2 20nC 0.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 ΔVBOOT_CAP (V) 0.8 0.9 1.0 FIGURE 11. BOOT CAPACITANCE vs BOOT RIPPLE VOLTAGE 0.60 x=1 0.55 Driver Power Dissipation 0.50 0.45 x = 0.75 0.40 0.35 x = 0.50 x = 0.25 0.30 0.25 0.20 x=0 0.15 0.10 0.05 0 1.4 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 DUTY CYCLE FIGURE 10. NORMALIZED RMS INPUT CURRENT FOR x = 0.8 Switching power dissipation in the driver is mainly a function of the switching frequency and total gate charge of the selected MOSFETs. Calculating the power dissipation in the driver for a desired application is critical to ensuring safe operation. Exceeding the maximum allowable power dissipation level will push the IC beyond the maximum recommended operating junction temperature of +125°C. When designing the application, it is recommended that the following calculation be performed to ensure safe operation at the desired frequency for the selected MOSFETs. The power dissipated by the drivers is approximated as Equation 30: P = Fsw ( 1.5V U Q + V L Q ) + P L + P U U L (EQ. 30) Selecting The Bootstrap Capacitor Adding an external capacitor across the BOOT and PHASE pins completes the bootstrap circuit. We selected the bootstrap capacitor breakdown voltage to be at least 10V. Although the theoretical maximum voltage of the capacitor is PVCC-VDIODE (voltage drop across the boot diode), large excursions below ground by the phase node requires we select a capacitor with at least a breakdown rating of 10V. The bootstrap capacitor can be chosen from Equation 29: Q GATE C BOOT ≥ -----------------------ΔV BOOT (EQ. 29) 14 Where: - Fsw is the switching frequency of the PWM signal VU is the upper gate driver bias supply voltage VL is the lower gate driver bias supply voltage QU is the charge to be delivered by the upper driver into the gate of the MOSFET and discrete capacitors - QL is the charge to be delivered by the lower driver into the gate of the MOSFET and discrete capacitors - PL is the quiescent power consumption of the lower driver - PU is the quiescent power consumption of the upper driver FN6930.0 June 30, 2009 ISL62873 1000 QU = 100nC QL = 200nC 900 - tOFF is the time required to drive the device into cut-off QU = 50nC QL = 100nC QU = 50nC QL = 50nC Layout Considerations The IC, analog signals, and logic signals should all be on the same side of the PCB, located away from powerful emission sources. The power conversion components should be arranged in a manner similar to the example in Figure 13 where the area enclosed by the current circulating through the input capacitors, high-side MOSFETs, and low-side MOSFETs is as small as possible and all located on the same side of the PCB. The power components can be located on either side of the PCB relative to the IC. 800 POWER (mW) 700 600 QU = 20nC QL = 50nC 500 400 300 200 100 0 GND 0 200 400 600 800 1k 1.2k 1.4k 1.6k 1.8k 2k FREQUENCY (Hz) + + OUTPUT CAPACITORS FIGURE 12. POWER DISSIPATION vs FREQUENCY VOUT MOSFET Selection and Considerations 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 MOSFET switches off. There are several power MOSFETs readily available that are optimized for DC/DC converter applications. The preferred high-side MOSFET emphasizes low switch charge so that the device spends the least amount of time dissipating power in the linear region. Unlike the low-side MOSFET which has the drain-source voltage clamped by its body diode during turn-off, the high-side MOSFET turns off with VIN - VOUT, plus the spike, across it. The preferred low-side MOSFET emphasizes low r DS(ON) when fully saturated to minimize conduction loss. For the low-side MOSFET, (LS), the power loss can be assumed to be conductive only and is written as Equation 31: 2 P CON_LS ≈ I LOAD ⋅ r DS ( ON )_LS ⋅ ( 1 – D ) (EQ. 31) For the high-side MOSFET, (HS), its conduction loss is written as Equation 32: 2 P CON_HS = I LOAD ⋅ r DS ( ON )_HS ⋅ D (EQ. 32) For the high-side MOSFET, its switching loss is written as Equation 33: PHASE NODE HIGH-SIDE MOSFETS VIN Signal Ground The GND pin is the signal-common also known as analog ground of the IC. When laying out the PCB, it is very important that the connection of the GND pin to the bottom setpoint-reference programming-resistor, bottom feedback voltage-divider resistor (if used), and the CSOFT capacitor be made as close as possible to the GND pin on a conductor not shared by any other components. In addition to the critical single point connection discussed in the previous paragraph, the ground plane layer of the PCB should have a single-point-connected island located under the area encompassing the IC, setpoint reference programming components, feedback voltage divider components, compensation components, CSOFT capacitor, and the interconnecting traces among the components and the IC. The island should be connected using several filled vias to the rest of the ground plane layer at one point that is not in the path of either large static currents or high di/dt currents. The single connection point should also be where the VCC decoupling capacitor and the GND pin of the IC are connected. Power Ground Where: VCC and PVCC Pins 15 INPUT CAPACITORS FIGURE 13. TYPICAL POWER COMPONENT PLACEMENT V IN ⋅ I VALLEY ⋅ t ON ⋅ F V IN ⋅ I PEAK ⋅ t OFF ⋅ F SW SW P SW_HS = ---------------------------------------------------------------------- + -----------------------------------------------------------------2 2 (EQ. 33) - 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 - tON is the time required to drive the device into saturation LOW-SIDE MOSFETS Anywhere not within the analog-ground island is Power Ground. Place the decoupling capacitors as close as practical to the IC. In particular, the PVCC decoupling capacitor should have a very short and wide connection to the PGND pin. The VCC decoupling capacitor should not share any vias with the PVCC decoupling capacitor. FN6930.0 June 30, 2009 ISL62873 EN, PGOOD, VID0, and VID1 Pins LGATE, PGND, UGATE, BOOT, and PHASE Pins These are logic signals that are referenced to the GND pin. Treat as a typical logic signal. The current-sensing network consisting of ROCSET, RO, and CSEN needs to be connected to the inductor pads for accurate measurement of the DCR voltage drop. These components however, should be located physically close to the OCSET and VO pins with traces leading back to the inductor. It is critical that the traces are shielded by the ground plane layer all the way to the inductor pads. The procedure is the same for resistive current sense. The signals going through these traces are high dv/dt and high di/dt, with high peak charging and discharging current. The PGND pin can only flow current from the gate-source charge of the low-side MOSFETs when LGATE goes low. Ideally, route the trace from the LGATE pin in parallel with the trace from the PGND pin, route the trace from the UGATE pin in parallel with the trace from the PHASE pin, and route the trace from the BOOT pin in parallel with the trace from the PHASE pin. These pairs of traces should be short, wide, and away from other traces with high input impedance; weak signal traces should not be in proximity with these traces on any layer. FB, SREF, SET0, SET1, and SET2 Pins Copper Size for the Phase Node The input impedance of these pins is high, making it critical to place the loop compensation components, setpoint reference programming resistors, feedback voltage divider resistors, and CSOFT close to the IC, keeping the length of the traces short. 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 spike. OCSET and VO Pins 16 FN6930.0 June 30, 2009 ISL62873 Ultra Thin Quad Flat No-Lead Plastic Package (UTQFN) D L16.2.6x1.8A B 16 LEAD ULTRA THIN QUAD FLAT NO-LEAD PLASTIC PACKAGE MILLIMETERS 6 INDEX AREA 2X A SYMBOL E N 0.10 C 1 2 2X MIN NOMINAL MAX NOTES A 0.45 0.50 0.55 - A1 - - 0.05 - 0.10 C A3 TOP VIEW 0.10 C C A 0.05 C 0.127 REF - b 0.15 0.20 0.25 5 D 2.55 2.60 2.65 - E 1.75 1.80 1.85 - e 0.40 BSC - SEATING PLANE A1 SIDE VIEW e PIN #1 ID K 1 2 NX L L1 K 0.15 - - - L 0.35 0.40 0.45 - L1 0.45 0.50 0.55 - N 16 2 Nd 4 3 Ne 4 3 θ NX b 5 16X 0.10 M C A B 0.05 M C (DATUM B) (DATUM A) 0 - 12 4 Rev. 5 2/09 NOTES: 1. Dimensioning and tolerancing conform to ASME Y14.5-1994. 2. N is the number of terminals. BOTTOM VIEW 3. Nd and Ne refer to the number of terminals on D and E side, respectively. 4. All dimensions are in millimeters. Angles are in degrees. CL 5. Dimension b applies to the metallized terminal and is measured between 0.15mm and 0.30mm from the terminal tip. (A1) NX (b) L 5 e SECTION "C-C" 6. The configuration of the pin #1 identifier is optional, but must be located within the zone indicated. The pin #1 identifier may be either a mold or mark feature. 7. Maximum package warpage is 0.05mm. TERMINAL TIP C C 8. Maximum allowable burrs is 0.076mm in all directions. 9. JEDEC Reference MO-255. 10. For additional information, to assist with the PCB Land Pattern Design effort, see Intersil Technical Brief TB389. 3.00 1.80 1.40 1.40 2.20 0.90 0.40 0.20 0.50 0.20 0.40 10 LAND PATTERN 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 17 FN6930.0 June 30, 2009