LTC3862 Multi-Phase Current Mode Step-Up DC/DC Controller FEATURES DESCRIPTION n The LTC(R)3862 is a two phase constant frequency, current mode boost and SEPIC controller that drives N-channel power MOSFETs. Two phase operation reduces system filtering capacitance and inductance requirements. The 5V gate drive is optimized for most automotive and industrial grade power MOSFETs. n n n n n n n n n n n n n Wide VIN Range: 4V to 36V Operation 2-Phase Operation Reduces Input and Output Capacitance Fixed Frequency, Peak Current Mode Control 5V Gate Drive for Logic-Level MOSFETs Adjustable Slope Compensation Gain Adjustable Max Duty Cycle (Up to 96%) Adjustable Leading Edge Blanking 1% Internal Voltage Reference Programmable Operating Frequency with One External Resistor (75kHz to 500kHz) Phase-Lockable Fixed Frequency 50kHz to 650kHz SYNC Input and CLKOUT for 2-, 3-, 4-, 6- or 12-Phase Operation (PHASEMODE Programmable) Internal 5V LDO Regulator 24-Lead Narrow SSOP Package 5mm x 5mm QFN with 0.65mm Lead Pitch and 24-Lead Thermally Enhanced TSSOP Packages APPLICATIONS n Automotive, Telecom and Industrial Power Supplies Adjustable slope compensation gain allows the user to finetune the current loop gain, improving noise immunity. The operating frequency can be set with an external resistor over a 75kHz to 500kHz range and can be synchronized to an external clock using the internal PLL. Multi-phase operation is possible using the SYNC input, the CLKOUT output and the PHASEMODE control pin allowing 2-, 3-, 4-, 6- or 12-phase operation. Other features include an internal 5V LDO with undervoltage lockout protection for the gate drivers, a precision RUN pin threshold with programmable hysteresis, soft-start and programmable leading edge blanking and maximum duty cycle L, LT, LTC and LTM are registered trademarks of Linear Technology Corporation. All other trademarks are the property of their respective owners. Protected by U.S. Patents, including 6144194, 6498466, 6611131. TYPICAL APPLICATION 84.5k 19.4H RUN GATE1 INTVCC 4.7F 10nF FB 10nF ITH 475k 12.4k 68.1k 100pF VIN = 12V 94 0.006 3V8 VOUT = 48V 96 220F 0.006 50V BLANK SENSE1- FREQ GATE2 SYNC SENSE2+ PLLFLTR SS 1nF 98 SENSE1+ LTC3862 10k Efficiency vs Output Current EFFICIENCY (%) 10nF VIN 5V TO 36V VOUT 48V 5A (MAX) VIN 24.9k 66.5k 19.4H 22F 50V SENSE2- PGND CLKOUT SLOPE DMAX PHASEMODE SGND 92 90 88 VIN = 24V VIN = 5V 86 84 82 80 100 1000 LOAD CURRENT (mA) 10000 3862 TA01b 3862 TA01 3862fb 1 LTC3862 ABSOLUTE MAXIMUM RATINGS (Notes 1, 2) Input Supply Voltage (VIN) ......................... -0.3V to 40V INTVCC Voltage ............................................ -0.3V to 6V INTVCC LDO RMS Output Current .........................50mA RUN Voltage ................................................ -0.3V to 8V SYNC Voltage ............................................... -0.3V to 6V SLOPE, PHASEMODE, DMAX, BLANK Voltage ........................................... -0.3V to 3V8 SENSE1+, SENSE1-, SENSE2+, SENSE2- Voltage ....................................... -0.3V to 3V8 SS, PLLFLTR Voltage ................................. -0.3V to 3V8 ITH Voltage ............................................... -0.3V to 2.7V FB Voltage .................................................. -0.3V to 3V8 FREQ Voltage ............................................ -0.3V to 1.5V Operating Junction Temperature Range (Notes 3, 4) LTC3862E............................................. -40C to 85C LTC3862I............................................ -40C to 125C LTC3862H .......................................... -40C to 150C Storage Temperature Range................... -65C to 150C Reflow Peak Body Temperature ........................... 260C PIN CONFIGURATION 5 20 VIN SS 6 19 INTVCC ITH 7 FB 8 17 PGND SGND 9 16 GATE2 CLKOUT 10 18 GATE1 15 NC SYNC 11 14 SENSE2- PLLFLTR 12 13 SENSE2+ FE PACKAGE 24-LEAD PLASTIC TSSOP TJMAX = 150C, JA = 38C/W EXPOSED PAD (PIN 25) IS PGND, MUST BE SOLDERED TO PCB 3 22 SENSE1- PHASEMODE FREQ 4 5 21 RUN 20 VIN SS 6 19 INTVCC ITH 7 18 GATE1 FB 8 SGND 9 CLKOUT 10 SYNC 11 PLLFLTR 12 17 PGND 16 GATE2 15 NC 14 SENSE2- 13 SENSE2+ GN PACKAGE 24-LEAD NARROW PLASTIC SSOP TJMAX = 150C, JA = 85C/W RUN FREQ BLANK SENSE1- 21 RUN 24 23 22 21 20 19 BLANK 1 18 VIN 17 INTVCC PHASEMODE 2 FREQ 3 16 GATE1 25 SS 4 15 PGND 14 GATE2 ITH 5 FB 6 13 NC 7 8 9 10 11 12 SENSE2- 4 23 SENSE1+ SENSE2 PHASEMODE SLOPE SENSE1+ 22 SENSE1- 24 3V8 2 + 3 1 3V8 BLANK DMAX DMAX 23 SENSE1+ SYNC 2 SLOPE SLOPE SGND 24 3V8 CLKOUT 1 25 TOP VIEW TOP VIEW DMAX PLLFLTR TOP VIEW UH PACKAGE 24-LEAD (5mm s 5mm) PLASTIC QFN TJMAX = 150C, JA = 34C/W EXPOSED PAD (PIN 25) IS PGND, MUST BE SOLDERED TO PCB ORDER INFORMATION LEAD FREE FINISH TAPE AND REEL PART MARKING* PACKAGE DESCRIPTION TEMPERATURE RANGE LTC3862EFE#PBF LTC3862EFE#TRPBF 3862FE 24-Lead Plastic TSSOP -40C to 85C LTC3862IFE#PBF LTC3862IFE#TRPBF 3862FE 24-Lead Plastic TSSOP -40C to 125C LTC3862HFE#PBF LTC3862HFE#TRPBF 3862FE 24-Lead Plastic TSSOP -40C to 150C LTC3862EGN#PBF LTC3862EGN#TRPBF LTC3862GN 24-Lead Plastic SSOP -40C to 85C LTC3862IGN#PBF LTC3862IGN#TRPBF LTC3862GN 24-Lead Plastic SSOP -40C to 125C LTC3862HGN#PBF LTC3862HGN#TRPBF LTC3862GN 24-Lead Plastic SSOP -40C to 150C LTC3862EUH#PBF LTC3862EUH#TRPBF 3862 -40C to 85C 24-Lead (5mm x 5mm) Plastic QFN LTC3862IUH#PBF LTC3862IUH#TRPBF 3862 -40C to 125C 24-Lead (5mm x 5mm) Plastic QFN LTC3862HUH#PBF LTC3862HUH#TRPBF 3862 -40C to 150C 24-Lead (5mm x 5mm) Plastic QFN Consult LTC Marketing for parts specified with wider operating temperature ranges. *The temperature grade is identified by a label on the shipping container. Consult LTC Marketing for information on non-standard lead based finish parts. For more information on lead free part marking, go to: http://www.linear.com/leadfree/ For more information on tape and reel specifications, go to: http://www.linear.com/tapeandreel/ 2 3862fb LTC3862 ELECTRICAL CHARACTERISTICS (Notes 2, 3) The l denotes the specifications which apply over the full operating junction temperature range, otherwise specifications are at TA = 25C. VIN = 12V, RUN = 2V and SS = open, unless otherwise noted. SYMBOL PARAMETER CONDITIONS MIN TYP MAX UNITS Supply Input and INTVCC Linear Regulator VIN VIN Supply Voltage Range IVIN VIN Supply Current Normal Mode, No Switching Shutdown INTVCC LDO Regulator Output Voltage dVINTVCC(LINE) Line Regulation 4 (Note 5) VRUN = 0V l l 1.8 30 4.8 6V < VIN < 36V dVINTVCC(LOAD) Load Regulation Load = 0mA to 20mA VUVLO INTVCC UVLO Voltage Rising INTVCC Falling INTVCC 3V8 LDO Regulator Output Voltage 36 V 3.0 80 mA A 5.0 5.2 V 0.002 0.02 %/V -2 % 3.3 2.9 V V 3.8 V Switcher Control Loop VFB Reference Voltage VITH = 0.8V (Note 6) E-Grade (Note 3) I-Grade and H-Grade (Note 3) dVFB/dVIN Feedback Voltage VIN Line Regulation VIN = 4V to 36V (Note 6) l l 1.210 1.199 1.223 1.223 1.235 1.248 V V 0.002 0.01 %/V 0.1 dVFB/dVITH Feedback Voltage Load Regulation VITH = 0.5V to 1.2V (Note 6) 0.01 gm Transconductance Amplifier Gain VITH = 0.8V (Note 6), ITH Pin Load = 5A 660 Mho f0dB Error Amplifier Unity-Gain Crossover Frequency (Note 7) 1.8 MHz VITH Error Amplifier Maximum Output Voltage (Internally Clamped) VFB = 1V, No Load 2.7 V Error Amplifier Minimum Output Voltage VFB = 1.5V, No Load 50 mV IITH % Error Amplifier Output Source Current -30 A Error Amplifier Output Sink Current 30 A IFB Error Amplifier Input Bias Currents (Note 6) VITH(PSKIP) Pulse Skip Mode Operation ITH Pin Voltage Rising ITH Voltage (Note 6) Hysteresis ISENSE(ON) SENSE Pin Current VSENSE(MAX) Maximum Current Sense Input Threshold VSLOPE = Float, Low Duty Cycle (Note 3) VSENSE(MATCH) CH1 to CH2 Maximum Current Sense Threshold Matching VSLOPE = Float, Low Duty Cycle (Note 3) (VSENSE1 - VSENSE2) IRUN RUN Source Current VRUN = 0V VRUN = 1.5V VRUN High Level RUN Channel Enable Threshold VRUNHYS RUN Threshold Hysteresis ISS SS Pull-Up Current VSS = 0V -5 A RSS SS Pull-Down Resistance VRUN = 0V 10 k Oscillator Frequency RFREQ = 45.6k RFREQ = 45.6k -50 -200 0.275 25 l 65 60 l -10 nA V mV 0.01 2 A 75 75 85 90 mV mV 10 mV RUN/Soft-Start -0.5 -5 A A 1.22 V 80 mV Oscillator fOSC Oscillator Frequency Range VFREQ Nominal FREQ Pin Voltage RFREQ = 45.6k l 280 260 l 75 300 300 1.223 320 340 kHz kHz 500 kHz V 3862fb 3 LTC3862 ELECTRICAL CHARACTERISTICS (Notes 2, 3) The l denotes the specifications which apply over the full operating junction temperature range, otherwise specifications are at TA = 25C. VIN = 12V, RUN = 2V and SS = open, unless otherwise noted. SYMBOL PARAMETER CONDITIONS MIN TYP MAX UNITS SYNC Minimum Input Frequency VSYNC = External Clock l SYNC Maximum Input Frequency VSYNC = External Clock l VSYNC SYNC Input Threshold Rising Threshold 1.5 V IPLLFLTR Phase Detector Sourcing Output Current fSYNC > fOSC -15 A fSYNC 50 650 kHz kHz Phase Detector Sinking Output Current fSYNC < fOSC 15 A CH1-CH2 Channel 1 to Channel 2 Phase Relationship VPHASEMODE = 0V VPHASEMODE = Float VPHASEMODE = 3V8 180 180 120 Deg Deg Deg CH1-CLKOUT Channel 1 to CLKOUT Phase Relationship VPHASEMODE = 0V VPHASEMODE = Float VPHASEMODE = 3V8 90 60 240 Deg Deg Deg DMAX Maximum Duty Cycle VDMAX = 0V VDMAX = Float VDMAX = 3V8 96 84 75 % % % tON(MIN)1 Minimum On-Time VBLANK = 0V (Note 8) 180 ns tON(MIN)2 Minimum On-Time VBLANK = Float (Note 8) 260 ns tON(MIN)3 Minimum On-Time VBLANK = 3V8 (Note 8) 340 ns Driver Pull-Up RDS(ON) 2.1 Driver Pull-Down RDS(ON) 0.7 Gate Driver RDS(ON) Overvoltage VFB(OV) VFB, Overvoltage Lockout Threshold VFB(OV) - VFB(NOM) in Percent Note 1: Stresses beyond those listed under Absolute Maximum Ratings may cause permanent damage to the device. Exposure to any Absolute Maximum Rating condition for extended periods may affect device reliability and lifetime. Note 2: All currents into device pins are positive; all currents out of device pins are negative. All voltages are referenced to ground unless otherwise specified. Note 3: The LTC3862E is guaranted to meet performance specifications from 0C to 85C. Specifications over the -40C to 85C operating junction temperature range are assured by design, characterization and correlation with statistical process controls. The LTC3862I is guaranteed over the full -40C to 125C operating junction temperature range and the LTC3862H is guaranteed over the full -40C to 150C operating junction temperature range. High junction temperatures degrade operating lifetimes. Operating lifetime is derated at junction temperatures greater than 125C. 8 10 12 % Note 4: This IC includes overtemperature protection that is intended to protect the device during momentary overload conditions. Continuous operation above the specified maximum operating junction temperature may impair device reliability. Note 5: Supply current in normal operation is dominated by the current needed to charge the external MOSFET gates. This current will vary with supply voltage and the external MOSFETs used. Note 6: The IC is tested in a feedback loop that adjusts VFB to achieve a specified error amplifier output voltage. Note 7: Guaranteed by design, not subject to test. Note 8: The minimum on-time condition is specified for an inductor peakto-peak ripple current = 30% (see Minimum On-Time Considerations in the Applications Information section). 3862fb 4 LTC3862 TYPICAL PERFORMANCE CHARACTERISTICS Efficiency and Power Loss vs Input Voltage Efficiency vs Output Current 100 96 VOUT = 48V 95 95 90 85 75 EFFICIENCY (%) VIN = 35V 80 VIN = 24V 70 65 3500 94 3000 93 2500 POWER LOSS 92 VIN = 12V 60 2000 91 VOUT = 48V IOUT = 1A 55 90 50 10 100 1000 LOAD CURRENT (mA) 0 10000 10 1500 20 30 INPUT VOLTAGE (V) 40 3862 G01 Load Step POWER LOSS (mW) EFFICIENCY (%) 4000 EFFICIENCY 3862 G02 Quiescent Current vs Input Voltage Inductor Current at Light Load 3.00 ILOAD 5A/DIV 1A TO 5A 2.75 SW1 50V/DIV QUIESCENT CURRENT (mA) IL1 5A/DIV 2.50 SW2 50V/DIV IL2 5A/DIV IL1 2A/DIV IL2 2A/DIV VOUT 500mV/DIV VIN = 24V VOUT = 48V 500s/DIV 3862 G03 VIN = 12V VOUT = 48V ILOAD = 100mA 1s/DIV 2.25 2.00 1.75 1.50 1.25 1.00 0.75 0.50 3862 G04 0.25 0 4 8 12 16 20 24 28 INPUT VOLTAGE (V) 32 36 3862 G05 Shutdown Quiescent Current vs Input Voltage 45 1.85 40 1.80 1.75 1.70 1.65 1.60 1.55 1.50 -50 -25 Shutdown Quiescent Current vs Temperature 50 SHUTDOWN CURRENT (A) 1.90 SHUTDOWN CURRENT (A) QUIESCENT CURRENT (mA) Quiescent Current vs Temperature 35 30 25 20 15 10 VIN = 12V 40 30 20 10 5 0 25 50 75 100 125 150 TEMPERATURE (C) 3862 G06 0 4 8 12 16 20 24 28 INPUT VOLTAGE (V) 32 36 3862 G07 0 -50 -25 0 25 50 75 100 125 150 TEMPERATURE (C) 3862 G08 3862fb 5 LTC3862 TYPICAL PERFORMANCE CHARACTERISTICS INTVCC Line Regulation INTVCC Load Regulation INTVCC vs Temperature 5.00 5.00 5.25 4.99 4.98 5.00 INTVCC VOLTAGE (V) INTVCC VOLTAGE (V) INTVCC VOLTAGE (V) 4.95 4.90 4.97 4.96 4.95 4.94 4.93 4.85 4.92 4.91 4.75 5 0 10 15 4.80 25 20 10 30 40 20 INTVCC LOAD CURRENT (mA) 0 INPUT VOLTAGE (V) INTVCC LDO Dropout vs Load Current, Temperature 85C 800 25C 600 -40C 400 3.5 1.233 1.231 1.229 3.3 3.2 3.1 3.0 2.9 10 0 20 30 INTVCC LOAD (mA) 40 50 1.225 1.223 1.221 1.219 1.215 2.7 1.213 2.6 -50 -25 1.211 -50 -25 0 25 50 75 100 125 150 TEMPERATURE (C) 3862 G12 Current Sense Threshold vs Temperature 80 80 1.223 1.222 1.221 1.220 4 8 12 16 20 24 28 INPUT VOLTAGE (V) 32 36 3862 G15 70 CURRENT SENSE THRESHOLD (mV) CURRENT SENSE THRESHOLD (mV) 1.226 1.224 25 50 75 100 125 150 TEMPERATURE (C) 3862 G14 Current Sense Threshold vs ITH Voltage 1.225 0 3862 G13 Feedback Voltage Line Regulation FB VOLTAGE (V) 1.227 1.217 2.8 200 0 1.235 FB VOLTAGE (V) 1000 25 50 75 100 125 150 TEMPERATURE (C) Feedback Voltage vs Temperature 3.6 3.4 INTVCC VOLTAGE (V) DROPOUT VOLTAGE (mV) 150C 125C 0 3862 G11 INTVCC UVLO Threshold vs Temperature 1600 1200 50 3862 G10 3962 G09 1400 4.90 -50 -25 60 50 40 30 20 10 0 0 0.4 0.8 1.2 1.6 ITH VOLTAGE (V) 2.0 2.4 3862 G16 79 78 77 76 75 74 73 72 71 70 -50 -25 0 25 50 75 100 125 150 TEMPERATURE (C) 3862 G17 3862fb 6 LTC3862 TYPICAL PERFORMANCE CHARACTERISTICS Maximum Current Sense Threshold vs Duty Cycle RUN Threshold vs Temperature SLOPE = 0.625 70 SLOPE = 1 SLOPE = 1.66 65 60 1.4 ON 1.25 RUN PIN VOLTAGE (V) 75 1.20 OFF 1.15 50 1.10 -50 -25 10 20 30 40 50 60 70 80 90 100 DUTY CYCLE (%) 1.2 OFF 0 25 50 75 100 125 150 TEMPERATURE (C) RUN PIN CURRENT (A) -0.3 -0.5 -0.6 -0.7 15 20 25 30 INPUT VOLTAGE (V) -0.8 0 0 -1 -1 -2 -3 -4 -5 -6 -8 -50 -25 25 50 75 100 125 150 TEMPERATURE (C) 0 -2 -3 -4 -5 -7 25 50 75 100 125 150 TEMPERATURE (C) 4 8 12 16 20 24 0 -5.1 -1 32 36 3862 G23 Soft-Start Current vs Soft-Start Voltage -5.0 28 INPUT VOLTAGE (V) 1344 G06 3862 G21vv Soft-Start Current vs Temperature 40 -6 -7 -0.9 35 RUN Source Current vs Input Voltage RUN PIN CURRENT (A) 0 -0.1 -0.4 10 3862 G20 RUN (On) Source Current vs Temperature -0.2 5 3862 G09 RUN (Off) Source Current vs Temperature 0 ON 1.0 0 3862 G18 -1.0 -50 -25 1.3 1.1 55 0 RUN PIN CURRENT (A) RUN Threshold vs Input Voltage 1.5 1.30 RUN PIN VOLTAGE (V) MAXIMUM CURRENT SENSE THRESHOLD (mV) 80 Oscillator Frequency vs Temperature 307 -5.2 -5.3 -5.4 305 FREQUENCY (kHz) SOFT-START CURRENT (A) SOFT-START CURRENT (A) 306 -2 -3 -4 -5.5 -5 -5.6 -50 -25 -6 304 303 302 301 300 299 0 25 50 75 100 125 150 TEMPERATURE (C) 3862 G24 0 0.5 2.5 3 1 1.5 2 SOFT-START VOLTAGE (V) 3.5 4 3862 G25 298 -50 -25 0 25 50 75 100 125 150 TEMPERATURE (C) 3862 G26 3862fb 7 LTC3862 TYPICAL PERFORMANCE CHARACTERISTICS Oscillator Frequency vs Input Voltage RFREQ vs Frequency Frequency vs PLLFLTR Voltage 1000 320 1400 315 1200 300 295 FREQUENCY (kHz) 305 RFREQ (k) FREQUENCY (kHz) 310 100 1000 200 285 4 12 16 20 24 28 INPUT VOLTAGE (V) 8 32 10 36 0 0 100 200 300 400 500 600 700 800 900 1000 FREQUENCY (kHz) 1.235 1.233 400 1.231 350 MINIMUM ON-TIME (ns) 1.223 1.221 1.219 1.217 2.5 2 Minimum On-Time vs Input Voltage 400 BLANK = 3V8 300 BLANK = FLOAT 250 200 BLANK = 3V8 350 BLANK = SGND 150 1.215 1.5 3862 G29 Minimum On-Time vs Temperature 1.225 1 PLLFLTR VOLTAGE (V) 3862 G28 Frequency Voltage vs Temperature 1.229 1.227 0.5 0 MINIMUM ON-TIME (ns) 0 3862 G27 FREQ VOLTAGE (V) 600 400 290 280 800 300 BLANK = FLOAT 250 200 BLANK = SGND 150 1.213 1.211 -50 -25 0 25 50 75 100 125 150 TEMPERATURE (C) 100 -50 -25 0 25 50 75 100 125 150 TEMPERATURE (C) 100 4 8 12 16 20 24 28 INPUT VOLTAGE (V) 3862 G31 3862 G30 Gate Turn-On Waveform Driving Renesas HAT2266 32 36 3862 G32 Gate Turn-Off Waveform Driving Renesas HAT2266 GATE 1V/DIV GATE 1V/DIV VIN = 12V 20ns/DIV VOUT = 48V IOUT = 1A MOSFET RENESAS HAT2266 3862 G33 VIN = 12V 20ns/DIV VOUT = 48V IOUT = 1A MOSFET RENESAS HAT2266 3862 G34 3862fb 8 LTC3862 PIN FUNCTIONS 3V8: Output of the Internal 3.8V LDO from INTVCC. Supply pin for the low voltage analog and digital circuits. A low ESR 1nF ceramic bypass capacitor should be connected between 3V8 and SGND, as close as possible to the IC. ITH: Error Amplifier Output. The current comparator trip threshold increases with the ITH control voltage. The ITH pin is also used for compensating the control loop of the converter. BLANK: Blanking Time. Floating this pin provides a nominal minimum on-time of 260ns. Connecting this pin to 3V8 provides a minimum on-time of 340ns, while connecting it to SGND provides a minimum on-time of 180ns. PGND: Power Ground. Connect this pin close to the sources of the power MOSFETs. PGND should also be connected to the negative terminals of VIN and INTVCC bypass capacitors. PGND is electrically isolated from the SGND pin. The Exposed Pad of the FE and QFN packages is connected to PGND. CLKOUT: Digital Output Used for Daisy-Chaining Multiple LTC3862 ICs in Multi-Phase Systems. The PHASEMODE pin voltage controls the relationship between CH1 and CH2 as well as between CH1 and CLKOUT. DMAX: Maximum Duty Cycle.This pin programs the maximum duty cycle. Floating this pin provides 84% duty cycle. Connecting this pin to 3V8 provides 75% duty cycle, while connecting it to SGND provides 96% duty cycle. FB: Error Amplifier Input. The FB pin should be connected through a resistive divider network to VOUT to set the output voltage. FREQ: A resistor from FREQ to SGND sets the operating frequency. GATE1, GATE2: Gate Drive Output. The LTC3862 provides a 5V gate drive referenced to PGND to drive a logic-level threshold MOSFET. INTVCC: Output of the Internal 5V Low Dropout Regulator (LDO). A low ESR 4.7F (X5R or better) ceramic bypass capacitor should be connected between INTVCC and PGND, as close as possible to the IC. PHASEMODE: The PHASEMODE pin voltage programs the phase relationship between CH1 and CH2 rising gate signals, as well as the phase relationship between CH1 gate signal and CLKOUT. Floating this pin or connecting it to either 3V8, or SGND changes the phase relationship between CH1, CH2 and CLKOUT. PLLFLTR: PLL Lowpass Filter Input. When synchronizing to an external clock, this pin serves as the lowpass filter input for the PLL. A series resistor and capacitor connected from PLLFLTR to SGND compensate the PLL feedback loop. RUN: Run Control Input. A voltage above 1.22V on the pin turns on the IC. Forcing the pin below 1.22V causes the IC to shut down. There is a 0.5A pull-up current for this pin. Once the RUN pin raises above 1.22V, an additional 4.5A pull-up current is added to the pin for programmable hysteresis. 3862fb 9 LTC3862 PIN FUNCTIONS SENSE1+, SENSE2+: Positive Inputs to the Current Comparators. The ITH pin voltage programs the current comparator offset in order to set the peak current trip threshold. This pin is normally connected to a sense resistor in the source of the power MOSFET. SENSE1-, SENSE2-: Negative Inputs to the Current Comparators. This pin is normally connected to the bottom of the sense resistor. SGND: Signal Ground. All feedback and soft-start connections should return to SGND. For optimum load regulation, the SGND pin should be kelvin connected to the PCB location between the negative terminals of the output capacitors. SLOPE: This pin programs the gain of the internal slope compensation. Floating this pin provides a normalized slope compensation gain of 1.00. Connecting this pin to 3V8 increases the normalized slope compensation by 66%, and connecting it to SGND decreases the normalized slope compensation by 37.5%. See Applications Information for more details. SS: Soft-Start Input. For soft-start operation, connecting a capacitor from this pin to SGND will clamp the output of the error amp. An internal 5A current source will charge the capacitor and set the rate of increase of the peak switch current of the converter. SYNC: PLL Synchronization Input. Applying an external clock between 50kHz and 650kHz will cause the operating frequency to synchronize to the clock. SYNC is pulled down by a 50k internal resistor. The rising edge of the SYNC input waveform will align with the rising edge of GATE1 in closed-loop operation. VIN: Main Supply Input. A low ESR ceramic capacitor should be connected between this pin and SGND. 3862fb 10 LTC3862 FUNCTIONAL DIAGRAM CLKOUT SYNC SYNC DETECT PLLFLTR VIN VIN RP CIN 5V LDO CP INTVCC DMAX PHASEMODE UV CLK1 VCO FREQ OT CLK2 RFREQ CVCC 3.8V LDO UVLO L 3V8 C3V8 OVER TEMP BIAS SLOPE SLOPE COMPENSATION BLANK BLANK LOGIC DMAX S OT R1 UV BLOGIC D GATE Q R2 SD BLOGIC M COUT LOGIC + VOUT PGND PWM LATCH SENSE+ OV 3V8 SS ITRIP PSKIP + - ICMP RLOOP SENSE- 5A CSS V TO I OT UV SD DUPLICATE FOR SECOND CHANNEL VFB ITH PSKIP RC RS SD OV SGND R2 R1 CC PSKIP - + 0.275V EA + - 1.223V OV - + 1.345V RUN - + 4.5A 0.5A 3862 FD RUN 1.22V 3862fb 11 LTC3862 OPERATION The Control Loop drive supply (INTVCC) and one for the low voltage analog and digital control circuitry (3V8). A block diagram of this power supply arrangement is shown in Figure 1. The LTC3862 uses a constant frequency, peak current mode step-up architecture with its two channels operating 180 degrees out of phase. During normal operation, each external MOSFET is turned on when the clock for that channel sets the PWM latch, and is turned off when the main current comparator, ICMP, resets the latch. The peak inductor current at which ICMP trips and resets the latch is controlled by the voltage on the ITH pin, which is the output of the error amplifier, EA. The error amplifier compares the output feedback signal at the VFB pin to the internal 1.223V reference and generates an error signal at the ITH pin. When the load current increases it causes a slight decrease in VFB relative to the reference voltage, which causes the EA to increase the ITH voltage until the average inductor current matches the new load current. After the MOSFET is turned off, the inductor current flows through the boost diode into the output capacitor and load, until the beginning of the next clock cycle. The Gate Driver Supply LDO (INTVCC) The 5V output (INTVCC) of the first LDO is powered from VIN and supplies power to the power MOSFET gate drivers. The INTVCC pin should be bypassed to PGND with a minimum of 4.7F of ceramic capacitance (X5R or better), placed as close as possible to the IC pins. If two power MOSFETs are connected in parallel for each channel in order to increase the output power level, or if a single MOSFET with a QG greater than 50nC is used, then it is recommended that the bypass capacitance be increased to a minimum of 10F. An undervoltage lockout (UVLO) circuit senses the INTVCC regulator output in order to protect the power MOSFETs from operating with inadequate gate drive. For the LTC3862 the rising UVLO threshold is typically 3.3V and the hysteresis is typically 400mV. The LTC3862 was optimized for logic-level power MOSFETs and applications where the output voltage is less than 50V to 60V. For applications requiring standard threshold power MOSFETs, please refer to the LTC3862-1 data sheet. Cascaded LDOs Supply Power to the Gate Driver and Control Circuitry The LTC3862 contains two cascaded PMOS output stage low dropout voltage regulators (LDOs), one for the gate LTC3862 VIN 1.223V CIN - P-CH + SGND R2 R1 INTVCC 1.223V INTVCC - CVCC P-CH GATE + PGND SGND R4 R3 3V8 ANALOG CIRCUITS 3V8 LOGIC C3V8 SGND 3862 F01 NOTE: PLACE CVCC AND C3V8 CAPACITORS AS CLOSE AS POSSIBLE TO DEVICE PINS Figure 1. Cascaded LDOs Provide Gate Drive and Control Circuitry Power 3862fb 12 LTC3862 OPERATION In multi-phase applications, all of the FB pins are connected together and all of the error amplifier output pins (ITH) are connected together. The INTVCC pins, however, should not be connected together. The INTVCC regulator is capable of sourcing current but is not capable of sinking current. As a result, when two or more INTVCC regulator outputs are connected together, the highest voltage regulator supplies all of the gate drive and control circuit current, and the other regulators are off. This would place a thermal burden on the highest output voltage LDO and could cause the maximum die temperature to be exceeded. In multi-phase LTC3862 applications, each INTVCC regulator output should be independently bypassed to its respective PGND pin as close as possible to each IC. The Low Voltage Analog and Digital Supply LDO (3V8) The second LDO within the LTC3862 is powered off of INTVCC and serves as the supply to the low voltage analog and digital control circuitry, as shown in Figure 1. The output voltage of this LDO (which also has a PMOS output device) is 3.8V. Most of the analog and digital control circuitry is powered from the internal 3V8 LDO. The 3V8 pin should be bypassed to SGND with a 1nF ceramic capacitor (X5R or better), placed as close as possible to the IC pins. This LDO is not intended to be used as a supply for external circuitry. Thermal Considerations and Package Options The LTC3862 is offered in two package options. The 5mm x 5mm QFN package (UH24) has a thermal resistance RTH(JA) of 34C/W, the 24-pin TSSOP (FE24) package has a thermal resistance of 38C/W, and the 24-pin SSOP (GN24) package has a thermal resistance of 85C/W. The QFN and TSSOP package options have a lead pitch of 0.65mm, and the GN24 option has a lead pitch of 0.025in. The INTVCC regulator can supply up to 50mA of total current. As a result, care must be taken to ensure that the maximum junction temperature of the IC is never exceeded. The junction temperature can be estimated using the following equations: IQ(TOT) = IQ + QG(TOT) * f PDISS = VIN * (IQ + QG(TOT) * f) TJ = TA + PDISS * RTH(JA) The total quiescent current (IQ(TOT)) consists of the static supply current (IQ) and the current required to charge the gate capacitance of the power MOSFETs. The value of QG(TOT) should come from the plot of VGS vs QG in the Typical Performance Characteristics section of the MOSFET data sheet. The value listed in the electrical specifications may be measured at a higher VGS, such as 10V, whereas the value of interest is at the 5V INTVCC gate drive voltage. As an example of the required thermal analysis, consider a 2-phase boost converter with a 9V to 24V input voltage range and an output voltage of 48V at 2A. The switching frequency is 150kHz and the maximum ambient temperature is 70C. The power MOSFET used for this application is the Vishay Si7478DP, which has a typical RDS(ON) of 8.8m at VGS = 4.5V and 7.5m at VGS = 10V. From the plot of VGS vs QG, the total gate charge at VGS = 5V is 50nC (the temperature coefficient of the gate charge is low). One power MOSFET is used for each phase. For the QFN package option: IQ(TOT) = 3mA + 2 * 50nC * 150kHz = 18mA PDISS = 24V * 18mA = 432mW TJ = 70C + 432mW * 34C/W = 84.7C In this example, the junction temperature rise is only 14.7C. These equations demonstrate how the gate charge current typically dominates the quiescent current of the IC, and how the choice of package option and board heat sinking can have a significant effect on the thermal performance of the solution. 3862fb 13 LTC3862 OPERATION Thermal Shutdown Protection In the event of an overtemperature condition (external or internal), an internal thermal monitor will shut down the gate drivers and reset the soft-start capacitor if the die temperature exceeds 170C. This thermal sensor has a hysteresis of 10C to prevent erratic behavior at hot temperatures. The LTC3862's internal thermal sensor is intended to protect the device during momentary overtemperature conditions. Continuous operation above the specified maximum operating junction temperature, however, may result in device degradation. Operation at Low Supply Voltage The LTC3862 has a minimum input voltage of 4V, making it a good choice for applications that experience low supply conditions. The gate driver for the LTC3862 consists of PMOS pull-up and NMOS pull-down devices, allowing the full INTVCC voltage to be applied to the gates during power MOSFET switching. Nonetheless, care should be taken to determine the minimum gate drive supply voltage (INTVCC) in order to choose the optimum power MOSFETs. Important parameters that can affect the minimum gate drive voltage are the minimum input voltage (VIN(MIN)), the LDO dropout voltage, the QG of the power MOSFETs, and the operating frequency. If the input voltage VIN is low enough for the INTVCC LDO to be in dropout, then the minimum gate drive supply voltage is: VINTVCC = VIN(MIN) - VDROPOUT The LDO dropout voltage is a function of the total gate drive current and the quiescent current of the IC (typically 3mA). A curve of dropout voltage vs output current for the LDO is shown in Figure 2. The temperature coefficient of the LDO dropout voltage is approximately 6000ppm/C. The total Q-current (IQ(TOT)) flowing in the LDO is the sum of the controller quiescent current (3mA) and the total gate charge drive current. IQ(TOT) = IQ + QG(TOT) * f After the calculations have been completed, it is important to measure the gate drive waveforms and the gate driver supply voltage (INTVCC to PGND) over all operating conditions (low VIN, nominal VIN and high VIN, as well as from light load to full load) to ensure adequate power MOSFET enhancement. Consult the power MOSFET data sheet to determine the actual RDS(ON) for the measured VGS, and verify your thermal calculations by measuring the component temperatures using an infrared camera or thermal probe. 1600 1400 DROPOUT VOLTAGE (mV) To prevent the maximum junction temperature from being exceeded, the input supply current to the IC should be checked when operating in continuous mode (heavy load) at maximum VIN. A tradeoff between the operating frequency and the size of the power MOSFETs may need to be made in order to maintain a reliable junction temperature. Finally, it is important to verify the calculations by performing a thermal analysis of the final PCB using an infrared camera or thermal probe. As an option, an exernal regulator shown in Figure 3 can be used to reduce the total power dissipation on the IC. 150C 1200 125C 1000 85C 800 25C 600 -40C 400 200 0 0 10 20 30 INTVCC LOAD (mA) 40 50 3862 F02 Figure 2. INTVCC LDO Dropout Voltage vs Current 3862fb 14 LTC3862 OPERATION Operation at High Supply Voltage At high input voltages, the LTC3862's internal LDO can dissipate a significant amount of power, which could cause the maximum junction temperature to be exceeded. Conditions such as a high operating frequency, or the use of more than one power MOSFET per channel, could push the junction temperature rise to high levels. If the thermal equations above indicate too high a rise in the junction temperature, an external bias supply can always be used to reduce the power dissipation on the IC, as shown in Figure 3. For example, a 5V or 12V system rail that is available would be more suitable than the 24V main input power rail to power the LTC3862. Also, the bias power can be generated with a separate switching or LDO regulator. An example of an LDO regulator is shown in Figure 3. The output voltage of the LDO regulator can be set by selecting an appropriate zener diode to be higher than 5V but low enough to divide the power dissipation between LTC3862 and Q1 in Figure 3. The absolute maximum voltage rating of the INTVCC pin is 6V. R1 Q1 VIN (OPT) LTC3862 INTVCC CVCC If, however, the VIN supply to the IC comes up before the INTVCC supply, the external INTVCC supply will act as a load to the internal LDO in the LTC3862, and the LDO will attempt to charge the INTVCC output with its short-circuit current. This will result in excessive power dissipation and possible thermal overload of the LTC3862. If an independent 5V supply exists in the system, it may be possible to short INTVCC and VIN together to 5V in order to reduce gate drive power dissipation. With VIN and INTVCC shorted together, the LDO output PMOS transistor is biased at VDS = 0V, and the current demand of the internal analog and digital control circuitry, as well as the gate drive current, will be supplied by the external 5V supply. Programming the Output Voltage The output voltage is set by a resistor divider according to the following formula: R2 VOUT = 1.223V 1+ R1 VIN D1 flow from the external INTVCC supply, through the body diode of the LDO PMOS device, to the input capacitor and VIN pin. This high current flow could trigger a latchup condition and cause catastrophic failure of the IC. 3862 F03 Figure 3. Using the LTC3862 with an External Bias Supply The external resistor divider is connected to the output as shown in Figure 4. Resistor R1 is normally chosen so that the output voltage error caused by the current flowing out of the VFB pin during normal operation is negligible compared to the current in the divider. For an output voltage error due to the error amp input bias current of less than 0.5%, this translates to a maximum value of R1 of about 30k. Power Supply Sequencing As shown in Figure 1, there are body diodes in parallel with the PMOS output transistors in the two LDO regulators in the LTC3862. As a result, it is not possible to bias the INTVCC and VIN pins of the chip from separate power supplies. Independently biasing the INTVCC pin from a separate power supply can cause one of two possible failure modes during supply sequencing. If the INTVCC supply comes up before the VIN supply, high current will VOUT LTC3862 R2 FB SGND R1 3862 F04 Figure 4. Programming the Output Voltage with a Resistor Divider 3862fb 15 LTC3862 OPERATION Operation of the RUN Pin VIN The control circuitry in the LTC3862 is turned on and off using the RUN pin. Pulling the RUN pin below 1.22V forces shutdown mode and releasing it allows a 0.5A current source to pull this pin up, allowing a "normally on" converter to be designed. Alternatively, the RUN pin can be externally pulled up or driven directly by logic. Care must be taken not to exceed the absolute maximum rating of 8V for this pin. The comparator on the RUN pin can also be used to sense the input voltage, allowing an undervoltage detection circuit to be designed. This is helpful in boost converter applications where the input current can reach very high levels at low input voltage: LTC3862 INTERNAL 5V 0.5A 4.5A RUN + 10V BIAS AND START-UP CONTROL - 1.22V RUN COMPARATOR SGND 3862 F05a Figure 5a. Using the RUN Pin for a "Normally On" Converter VIN I *V IIN = OUT OUT VIN * LTC3862 INTERNAL 5V The 1.22V input threshold of the RUN comparator is derived from a precise bandgap reference, in order to maximize the accuracy of the undervoltage-sensing function. The RUN comparator has 80mV built-in hysteresis. When the voltage on the RUN pin exceeds 1.22V, the current sourced into the RUN pin is switched from 0.5A to 5A PTAT current. The user can therefore program both the rising threshold and the amount of hysteresis using the values of the resistors in the external divider, as shown in the following equations: EXTERNAL LOGIC CONTROL 10V + - 1.22V BIAS AND START-UP CONTROL RUN COMPARATOR SGND 3862 F05b Figure 5b. On/Off Control Using External Logic VIN LTC3862 INTERNAL 5V RA 0.5A B Several of the possible RUN pin control techniques are illustrated in Figure 5. 4.5A RUN R VIN(ON) = 1.22V 1+ A - 0.5 * RA RB R VIN(OFF ) = 1..22V 1+ A - 5 * RA R 0.5A 4.5A RUN RB 10V + 1.22V - BIAS AND START-UP CONTROL RUN COMPARATOR SGND Frequency Selection and the Phase Lock Loop 3862 F05c The selection of the switching frequency is a tradeoff between efficiency and component size. Low frequency operation increases efficiency by reducing MOSFET switching losses, but requires a larger inductor and output capacitor to maintain low output ripple. Figure 5c. Programming the Input Voltage Turn-On and Turn-Off Thresholds Using the RUN Pin 3862fb 16 LTC3862 OPERATION The operating frequency of the LTC3862 can be approximated using the following formula: 1000 RFREQ (k) The LTC3862 uses a constant frequency architecture that can be programmed over a 75kHz to 500kHz range using a single resistor from the FREQ pin to ground. Figure 6 illustrates the relationship between the FREQ pin resistance and the operating frequency. 100 RFREQ = 5.5096E9(fOSC)-0.9255 A phase-lock loop is available on the LTC3862 to synchronize the internal oscillator to an external clock source connected to the SYNC pin. Connect a series RC network from the PLLFLTR pin to SGND to compensate PLL's feedback loop. Typical compensation components are a 0.01F capacitor in series with a 10k resistor. The PLLFLTR pin is both the output of the phase detector and the input to the voltage controlled oscillator (VCO). The LTC3862 phase detector adjusts the voltage on the PLLFLTR pin to align the rising edge of GATE1 to the leading edge of the external clock signal, as shown in Figure 7. The rising edge of GATE2 will depend upon the voltage on the PHASEMODE pin. The capture range of the LTC3862's PLL is 50kHz to 650kHz. Because the operating frequency of the LTC3862 can be programmed using an external resistor, in synchronized applications, it is recommended that the free-running frequency (as defined by the external resistor) be set to the same value as the synchronized frequency. This results in a start-up of the IC at approximately the same frequency as the external clock, so that when the sync signal comes alive, no discontinuity at the output will be observed. It also ensures that the operating frequency remains essentially constant in the event the sync signal is lost. The SYNC pin has an internal 50k resistor to ground. Using the CLKOUT and PHASEMODE Pins in Multi-Phase Applications The LTC3862 features two pins (CLKOUT and PHASEMODE) that allow multiple ICs to be daisy-chained together for higher current multi-phase applications. For a 3- or 4-phase 10 0 100 200 300 400 500 600 700 800 900 1000 FREQUENCY (kHz) 3862 F06 Figure 6. FREQ Pin Resistor Value vs Frequency SYNC 10V/DIV GATE1 10V/DIV GATE2 10V/DIV CLKOUT 10V/DIV VIN = 12V 2s/DIV VOUT = 48V 1A PHASEMODE = SGND 3862 F07 Figure 7. Synchronization of the LTC3862 to an External Clock Using the PLL design, the CLKOUT signal of the master controller is connected to the SYNC input of the slave controller in order to synchronize additional power stages for a single high current output. The PHASEMODE pin is used to adjust the phase relationship between channel 1 and channel 2, as well as the phase relationship between channel 1 and CLKOUT, as summarized in Table 1. The phases are calculated relative to the zero degrees, defined as the rising edge of the GATE1 output. In a 6-phase application the CLKOUT pin of the master controller connects to the SYNC input of the 2nd controller and the CLKOUT pin of the 2nd controller connects to the SYNC pin of the 3rd controller. 3862fb 17 LTC3862 OPERATION MASTER Table 1 FREQ PHASEMODE CH-1 to CH-2 PHASE CH-1 to CLKOUT PHASE APPLICATION SGND 180 90 2-Phase, 4-Phase Float 3V8 180 120 60 240 ON/OFF CONTROL RUN LTC3862 FB 6-Phase 3-Phase INTVCC ITH VOUT SS CLKOUT INDIVIDUAL INTVCC PINS LOCALLY DECOUPLED SYNC PLLFLTR SGND Using the LTC3862 Transconductance (gm) Error Amplifier in Multi-Phase Applications ALL RUN PINS CONNNECTED TOGETHER SLAVE The LTC3862 error amplifier is a transconductance, or gm amplifier, meaning that it has high DC gain but high output impedance (the output of the error amplifier is a current proportional to the differential input voltage). This style of error amplifier greatly eases the task of implementing a multi-phase solution, because the amplifiers from two or more chips can be connected in parallel. In this case the FB pins of multiple LTC3862s can be connected together, as well as the ITH pins, as shown in Figure 8. The gm of the composite error amplifier is simply n times the transconductance of one amplifier, or gm(TOT) = n * 660S, where n is the number of amplifiers connected in parallel. The transfer function from the ITH pin to the current comparator inputs was carefully designed to be accurate, both from channel-to-channel and chip-to-chip. This way the peak inductor current matching is kept accurate. A buffered version of the output of the error amplifier determines the threshold at the input of the current comparator. The ITH voltage that represents zero peak current is 0.4V and the voltage that represents current limit is 1.2V (at low duty cycle). During an overload condition, the output of the error amplifier is clamped to 2.6V at low duty cycle, in order to reduce the latency when the overload condition terminates. A patented circuit in the LTC3862 is used to recover the slope compensation signal, so that the maximum peak inductor current is not a strong function of the duty cycle. Soft-Start The start-up of the LTC3862 is controlled by the voltage on the SS pin. An internal PNP transistor clamps the current comparator sense threshold during soft-start, thereby limiting the peak switch current. The base of the PNP is connected to the SS pin and the emitter to an internal, INTVCC FREQ ALL ITH PINS CONNECTED TOGETHER ITH RUN LTC3862 FB SS CLKOUT ALL SS PINS CONNNECTED TOGETHER SYNC PLLFLTR SGND SLAVE FREQ INTVCC RUN ITH ALL FB PINS CONNECTED TOGETHER LTC3862 FB SS CLKOUT SYNC PLLFLTR SGND 3862 F08 Figure 8. LTC3862 Error Amplifier Configuration for Multi-Phase Operation buffered ITH node (please note that the ITH pin voltage may not track the soft-start voltage during this time period). An internal 5A current source charges the SS capacitor, and clamps the peak sense threshold until the voltage on the soft-start capacitor reaches approximately 0.6V. The required amount of soft-start capacitance can be estimated using the following equation: t CSS = 5A SS 0.6 V The SS pin has an internal open-drain NMOS pull-down transistor that turns on when the RUN pin is pulled low, when the voltage on the INTVCC pin is below its undervoltage lockout threshold, or during an overtemperature condition. In multi-phase applications that use more than one 3862fb 18 LTC3862 OPERATION LTC3862 chip, connect all of the SS pins together and use one external capacitor to program the soft-start time. In this case, the current into the soft-start capacitor will be ISS = n * 5A, where n is the number of SS pins connected together. Figure 9 illustrates the start-up waveforms for a 2-phase LTC3862 application. SW1 10V/DIV SW2 10V/DIV IL1 1A/DIV IL2 1A/DIV VIN = 17V VOUT = 24V LIGHT LOAD (10mA) RUN 5V/DIV IL1 5A/DIV 1s/DIV 3862 F10 Figure 10. Light Load Switching Waveforms for the LTC3862 at the Onset of Pulse Skipping IL2 5A/DIV VOUT 50V/DIV VIN = 12V VOUT = 48V 100 LOAD 1ms/DIV 3862 F09a Figure 9. Typical Start-Up Waveforms for a Boost Converter Using the LTC3862 Pulse Skip Operation at Light Load As the load current is decreased, the controller enters discontinuous mode (DCM). The peak inductor current can be reduced until the minimum on-time of the controller is reached. Any further decrease in the load current will cause pulse skipping to occur, in order to maintain output regulation, which is normal. The minimum on-time of the controller in this mode is approximately 180ns (with the blanking time set to its minimum value), the majority of which is leading edge blanking. Figure 10 illustrates the LTC3862 switching waveforms at the onset of pulse skipping. Programmable Slope Compensation For a current mode boost regulator operating in CCM, slope compensation must be added for duty cycles above 50%, in order to avoid sub-harmonic oscillation. For the LTC3862, this ramp compensation is internal and user adjustable. Having an internally fixed ramp compensation waveform normally places some constraints on the value of the inductor and the operating frequency. For example, with a fixed amount of internal slope compensation, using an excessively large inductor would result in too much effective slope compensation, and the converter could become unstable. Likewise, if too small an inductor were used, the internal ramp compensation could be inadequate to prevent sub-harmonic oscillation. The LTC3862 contains a pin that allows the user to program the slope compensation gain in order to optimize performance for a wider range of inductance. With the SLOPE pin left floating, the normalized slope gain is 1.00. Connecting the SLOPE pin to ground reduces the normalized gain to 0.625 and connecting this pin to the 3V8 supply increases the normalized slope gain to 1.66. With the normalized slope compensation gain set to 1.00, the design equations assume an inductor ripple current of 20% to 40%, as with previous designs. Depending upon the application circuit, however, a normalized gain of 1.00 may not be optimum for the inductor chosen. If the ripple current in the inductor is greater than 40%, the normalized slope gain can be increased to 1.66 (an increase of 66%) by connecting the SLOPE pin to the 3V8 supply. If the inductor ripple current is less than 20%, the normalized slope gain can be reduced to 0.625 (a decrease of 37.5%) by connecting the SLOPE pin to SGND. To check the effectiveness of the slope compensation, apply a load step to the output and monitor the cycle-by-cycle behavior of the inductor current during the leading and trailing edges of the load current. Vary the input voltage over its full range and check for signs of cycle-by-cycle SW node instability or sub-harmonic oscillation. When 3862fb 19 LTC3862 OPERATION the slope compensation is too low the converter can suffer from excessive jitter or, worst case, sub-harmonic oscillation. When excess slope compensation is applied to the internal current sense signal, the phase margin of the control loop suffers. Figure 11 illustrates inductor current waveforms for a properly compensated loop. The LTC3862 contains a patented circuit whereby most of the applied slope compensation is recovered, in order to provide a SENSE+ to SENSE- threshold which is not a strong function of the duty cycle. This sense threshold is, however, a function of the programmed slope gain, as shown in Figure 12. The data sheet typical specification of 75mV for SENSE+ minus SENSE- is measured at a normalized slope gain of 1.00 at low duty cycle. For applications where the normalized slope gain is not 1.00, use Figure 12 to determine the correct value of the sense resistor. ILOAD 2A/DIV 200mA-3A IL1 2A/DIV IL2 2A/DIV VOUT 2V/DIV VIN = 24V VOUT = 48V 3862 F11 10s/DIV MAXIMUM CURRENT SENSE THRESHOLD (mV) Figure 11. Inductor Current Waveforms for a Properly Compensated Control Loop 80 SLOPE = 0.625 75 70 SLOPE = 1 SLOPE = 1.66 65 60 55 50 0 10 20 30 40 50 60 70 80 90 100 DUTY CYCLE (%) 3862 F12 Programmable Blanking and the Minimum On-Time The BLANK pin on the LTC3862 allows the user to program the amount of leading edge blanking at the SENSE pins. Connecting the BLANK pin to SGND results in a minimum on-time of 180ns, floating the pin increases this time to 260ns, and connecting the BLANK pin to the 3V8 supply results in a minimum on-time of 340ns. The majority of the minimum on-time consists of this leading edge blanking, due to the inherently low propagation delay of the current comparator (25ns typ) and logic circuitry (10ns to 15ns). The purpose of leading edge blanking is to filter out noise on the SENSE pins at the leading edge of the power MOSFET turn-on. During the turn-on of the power MOSFET the gate drive current, the discharge of any parasitic capacitance on the SW node, the recovery of the boost diode charge, and parasitic series inductance in the high di/dt path all contribute to overshoot and high frequency noise that could cause false-tripping of the current comparator. Due to the wide range of applications the LTC3862 is well-suited to, fixing one value of the internal leading edge blanking time would have required the longest delay time to have been used. Providing a means to program the blank time allows users to optimize the SENSE pin filtering for each application. Figure 13 illustrates the effect of the programmable leading edge blank time on the minimum on-time of a boost converter. Programmable Maximum Duty Cycle In order to maintain constant frequency and a low output ripple voltage, a single-ended boost (or flyback or SEPIC) converter is required to turn off the switch every cycle for some minimum amount of time. This off-time allows the transfer of energy from the inductor to the output capacitor and load, and prevents excessive ripple current and voltage. For inductor-based topologies like boost and SEPIC converters, having a maximum duty cycle as close as possible to 100% may be desirable, especially in low VIN to high VOUT applications. However, for transformerbased solutions, having a maximum duty cycle near 100% is undesirable, due to the need for V * sec reset during the primary switch off-time. Figure 12. Effect of Slope Gain on the Peak SENSE Threshold 3862fb 20 LTC3862 OPERATION 96% MAXIMUM DUTY CYCLE WITH DMAX = SGND MINIMUM ON-TIME AT LIGHT LOAD WITH BLANK = SGND INDUCTOR CURRENT 1A/DIV SW NODE 10V/DIV GATE 2V/DIV INDUCTOR CURRENT 2A/DIV SW NODE 20V/DIV 200ns/DIV VIN = 30V VOUT = 48V MEASURED ON-TIME = 180ns 1s/DIV 84% MAXIMUM DUTY CYCLE WITH DMAX = FLOAT MINIMUM ON-TIME AT LIGHT LOAD WITH BLANK = FLOAT INDUCTOR CURRENT 1A/DIV SW NODE 10V/DIV GATE 2V/DIV INDUCTOR CURRENT 2A/DIV SW NODE 20V/DIV VIN = 30V 200ns/DIV VOUT = 48V MEASURED ON-TIME = 260ns 1s/DIV 75% MAXIMUM DUTY CYCLE WITH DMAX = 3V8 MINIMUM ON-TIME AT LIGHT LOAD WITH BLANK = 3V8 INDUCTOR CURRENT 1A/DIV SW NODE 10V/DIV GATE 2V/DIV INDUCTOR CURRENT 2A/DIV SW NODE 20V/DIV VIN = 30V 200ns/DIV VOUT = 48V MEASURED ON-TIME = 340ns 3862 F13 Figure 13. Leading Edge Blanking Effects on the Minimum On-Time In order to satisfy these different applications requirements, the LTC3862 has a simple way to program the maximum duty cycle. Connecting the DMAX pin to SGND limits the maximum duty cycle to 96%. Floating this pin limits the duty cycle to 84% and connecting the DMAX pin to the 3V8 supply limits it to 75%. Figure 14 illustrates the effect of limiting the maximum duty cycle on the SW node waveform of a boost converter. 1s/DIV 3862 F14 Figure 14. SW Node Waveforms with Different Duty Cycle Limits The LTC3862 contains an oscillator that runs at a multiple of the switching frequency, in order to provide for 2-, 3-, 4-, 6- and 12-phase operation. A digital counter is used to divide down the fundamental oscillator frequency in order to obtain the operating frequency of the gate drivers. Since the maximum duty cycle limit is obtained from this digital counter, the percentage maximum duty cycle does not vary with process tolerances or temperature. 3862fb 21 LTC3862 OPERATION The SENSE+ and SENSE- Pins The SENSE+ and SENSE- pins are high impedance inputs to the CMOS current comparators for each channel. Nominally, there is no DC current into or out of these pins. There are ESD protection diodes connected from these pins to SGND, although even at hot temperature the leakage current into the SENSE+ and SENSE- pins should be less than 1A. Since the LTC3862 contains leading edge blanking, an external RC filter is not required for proper operation. However, if an external filter is used, the filter components should be placed close to the SENSE+ and SENSE- pins on the IC, as shown in Figure 15. The positive and negative sense node traces should then run parallel to each other to a Kelvin connection underneath the sense resistor, as shown in Figure 16. Sensing current elsewhere on the board can add parasitic inductance and capacitance to the current sense element, degrading the information at the sense pins and making the programmed current limit unpredictable. Avoid the temptation to connect the SENSE- line to the ground plane using a PCB via; this could result in unpredictable behavior. The sense resistor should be connected to the source of the power MOSFET and the ground node using short, wide PCB traces, as shown in Figure 16. Ideally, the bottom terminal of the sense resistors will be immediately VIN adjacent to the negative terminal of the output capacitor, since this path is a part of the high di/dt loop formed by the switch, boost diode, output capacitor and sense resistor. Placement of the inductors is less critical, since the current in the inductors is a triangle waveform. Checking the Load Transient Response The regulator loop response can be checked by looking at the load current transient response. Switching regulators take several cycles to respond to a step in DC (resistive) load current. When a load step occurs, VOUT shifts by an amount equal to ILOAD (ESR), where ESR is the effective series resistance of COUT . ILOAD also begins to charge or discharge COUT , generating the feedback error signal that forces the regulator to adapt to the current change and return VOUT to its steady-state value. During this recovery time VOUT can be monitored for excessive overshoot or ringing, which would indicate a stability problem. The availability of the ITH pin not only allows optimization of control loop behavior but also provides a DC-coupled and AC-filtered closed-loop response test point. The DC step, rise time and settling at this test point truly reflects the closed-loop response. Assuming a predominantly second order system, phase margin and/or damping factor can be estimated using the percentage of overshoot seen at this pin. The bandwidth can also be estimated by examining the rise time at the pin. VIN MOSFET SOURCE INTVCC LTC3862 VOUT GATE SENSE+ RSENSE RSENSE SENSE- PGND 3862 F15 FILTER COMPONENTS PLACED NEAR SENSE PINS Figure 15. Proper Current Sense Filter Component Placement TO SENSE FILTER NEXT TO CONTROLLER 3862 F16 GND Figure 16. Connecting the SENSE+ and SENSE- Traces to the Sense Resistor Using a Kelvin Connection 3862fb 22 LTC3862 OPERATION The ITH series RC * CC filter sets the dominant pole-zero loop compensation. The transfer function for boost and flyback converters contains a right half plane zero that normally requires the loop crossover frequency to be reduced significantly in order to maintain good phase margin. The RC * CC filter values can typically be modified slightly (from 0.5 to 2 times their suggested values) to optimize transient response once the final PC layout is done and the particular output capacitor type(s) and value(s) have been determined. The output capacitor configuration needs to be selected in advance because the effective ESR and bulk capacitance have a significant effect on the loop gain and phase. An output current pulse of 20% to 80% of full-load current having a rise time of 1s to 10s will produce output voltage and ITH pin waveforms that will give a sense of the overall loop stability without breaking the feedback loop. Placing a power MOSFET and load resistor directly across the output capacitor and driving the gate with an appropriate signal generator is a practical way to produce a fast load step condition. The initial output voltage step resulting from the step change in the output current may not be within the bandwidth of the feedback loop, so this signal cannot be used to determine phase margin. This is why it is better to look at the ITH pin signal which is in the feedback loop and is the filtered and compensated control loop response. The gain of the loop will be increased by increasing RC and the bandwidth of the loop will be increased by decreasing CC. If RC is increased by the same factor that CC is decreased, the zero frequency will be kept the same, thereby keeping the phase shift the same in the most critical frequency range of the feedback loop. The output voltage settling behavior is related to the stability of the closed-loop system and will demonstrate the actual overall supply performance. Figure 17 illustrates the load step response of a properly compensated boost converter. ILOAD 5A/DIV 1A TO 5A IL1 5A/DIV IL2 5A/DIV VOUT 500mV/DIV VIN = 24V VOUT = 48V 500s/DIV 3862 F17 Figure 17. Load Step Response of a Properly Compensated Boost Converter 3862fb 23 LTC3862 APPLICATIONS INFORMATION Typical Boost Applications Circuit Minimum On-Time Limitations A basic 2-phase, single output LTC3862 application circuit is shown in Figure 18. External component selection is driven by the characteristics of the load and the input supply. In a single-ended boost converter, two steady-state conditions can result in operation at the minimum on-time of the controller. The first condition is when the input voltage is close to the output voltage. When VIN approaches VOUT the voltage across the inductor approaches zero during the switch off-time. Under this operating condition the converter can become unstable and the output can experience high ripple voltage oscillation at audible frequencies. For applications where the input voltage can approach or exceed the output voltage, consider using a SEPIC or buck-boost topology instead of a boost converter. Duty Cycle Considerations For a boost converter operating in a continuous conduction mode (CCM), the duty cycle of the main switch is: V +V -V D = O F IN = tON * f VO + VF where VF is the forward voltage of the boost diode. The minimum on-time for a given application operating in CCM is: 1 VO + VF - VIN(MAX ) tON(MIN) = f VO + VF For a given input voltage range and output voltage, it is important to know how close the minimum on-time of the application comes to the minimum on-time of the control IC. The LTC3862 minimum on-time can be programmed from 180ns to 340ns using the BLANK pin. The second condition that can result in operation at the minimum on-time of the controller is at light load, in deep discontinuous mode. As the load current is decreased, the on-time of the switch decreases, until the minimum on-time limit of the controller is reached. Any further decrease in the output current will result in pulse skipping, a typically benign condition where cycles are skipped in order to maintain output regulation. VIN 5V TO 36V L1 19.4H PA2020-193 D1 PDS760 1nF 3V8 SENSE1+ SLOPE BLANK 10nF LTC3862 68.1k ITH FB 100pF 24.9k 84.5k 1F SS 12.4k SGND + RUN FREQ 10nF 0.006 100F 1W 63V 10nF PHASEMODE SENSE1- 66.5k Q1 HAT2266H 10 6.8F 50V 6.8F 50V 6.8F 50V VIN INTVCC GATE1 PGND 10 VOUT 48V 2A TO 5A 6.8F 50V 6.8F 50V 6.8F 50V 10nF SENSE2+ 0.006 1W Q2 HAT2266H GATE2 SENSE2- CLKOUT SYNC PLLFLTR 100F 6.8F 50V 63V 4.7F 475k VOUT + DMAX L2 19.4H PA2020-193 D2 PDS760 3862 F18 Figure 18. A Typical 2-Phase, Single Output Boost Converter Application Circuit 3862fb 24 LTC3862 APPLICATIONS INFORMATION Maximum Duty Cycle Limitations Another operating extreme occurs at high duty cycle, when the input voltage is low and the output voltage is high. In this case: VO + VF - VIN(MIN) DMAX = VO + VF A single-ended boost converter needs a minimum off-time every cycle in order to allow energy transfer from the input inductor to the output capacitor. This minimum off-time translates to a maximum duty cycle for the converter. The equation above can be rearranged to obtain the maximum output voltage for a given minimum input or maximum duty cycle. VO(MAX ) = VIN - VF 1 - DMAX The equation for DMAX above can be used as an initial guideline for determining the maximum duty cycle of the application circuit. However, losses in the inductor, input and output capacitors, the power MOSFETs, the sense resistors and the controller (gate drive losses) all contribute to an increasing of the duty cycle. The effect of these losses will be to decrease the maximum output voltage for a given minimum input voltage. After the initial calculations have been completed for an application circuit, it is important to build a prototype of the circuit and measure it over the entire input voltage range, from light load to full load, and over temperature, in order to verify proper operation of the circuit. Peak and Average Input Currents The control circuit in the LTC3862 measures the input current (by means of resistors in the sources of the power MOSFETs), so the output current needs to be reflected back to the input in order to dimension the power MOSFETs properly. Based on the fact that, ideally, the output power is equal to the input power, the maximum average input current is: IIN(MAX ) = IO(MAX ) 1 - DMAX The peak current in each inductor is: IIN(PK ) = 1 IO(MAX ) * 1+ * n 2 1 - DMAX where n represents the number of phases and represents the percentage peak-to-peak ripple current in the inductor. For example, if the design goal is to have 30% ripple current in the inductor, then = 0.30, and the peak current is 15% greater than the average. Inductor Selection Given an input voltage range, operating frequency and ripple current, the inductor value can be determined using the following equation: L= VIN(MIN) IL * f * DMAX where: IL = IO(MAX ) * n 1 - DMAX Choosing a larger value of IL allows the use of a lower value inductor but results in higher output voltage ripple, greater core losses, and higher ripple current ratings for the input and output capacitors. A reasonable starting point is 30% ripple current in the inductor ( = 0.3), or: IL = 0.3 IO(MAX ) * n 1 - DMAX 3862fb 25 LTC3862 APPLICATIONS INFORMATION The inductor saturation current rating needs to be higher than the worst-case peak inductor current during an overload condition. If IO(MAX) is the maximum rated load current, then the maximum current limit value (IO(CL)) would normally be chosen to be some factor (e.g., 30%) greater than IO(MAX). IO(CL) = 1.3 * IO(MAX) Reflecting this back to the input, where the current is being measured, and accounting for the ripple current, gives a minimum saturation current rating for the inductor of: IL(SAT) 1 1.3 * IO(MAX ) * 1+ * 1 - DMAX n 2 The saturation current rating for the inductor should be determined at the minimum input voltage (which results in the highest duty cycle and maximum input current), maximum output current and the maximum expected core temperature. The saturation current ratings for most commercially available inductors drop at high temperature. To verify safe operation, it is a good idea to characterize the inductor's core/winding temperature under the following conditions: 1) worst-case operating conditions, 2) maximum allowable ambient temperature and 3) with the power supply mounted in the final enclosure. Thermal characterization can be done by placing a thermocouple in intimate contact with the winding/core structure, or by burying the thermocouple within the windings themselves. Remember that a single-ended boost converter is not short-circuit protected, and that under a shorted output condition, the output current is limited only by the input supply capability. For applications requiring a step-up converter that is short-circuit protected, consider using a SEPIC or forward converter topology. and thermal resistances RTH(JA) and RTH(JC)--both junction-to-ambient and junction-to-case. The gate driver for the LTC3862 consists of PMOS pull-up and NMOS pull-down devices, allowing the full INTVCC voltage to be applied to the gates during power MOSFET switching. Nonetheless, care must be taken to ensure that the minimum gate drive voltage is still sufficient to full enhance the power MOSFET. Check the MOSFET data sheet carefully to verify that the RDS(ON) of the MOSFET is specified for a voltage less than or equal to the nominal INTVCC voltage of 5V. For applications that require a power MOSFET rated at 6V or 10V, please refer to the LTC3862-1 data sheet. Also pay close attention to the BVDSS specifications for the MOSFETs relative to the maximum actual switch voltage in the application. Check the switching waveforms of the MOSFET directly on the drain terminal using a single probe and a high bandwidth oscilloscope. Ensure that the drain voltage ringing does not approach the BVDSS of the MOSFET. Excessive ringing at high frequency is normally an indicator of too much series inductance in the high di/dt current path that includes the MOSFET, the boost diode, the output capacitor, the sense resistor and the PCB traces connecting these components. In some challenging applications it may be necessary to use a snubber in order to limit the switch node dV/dt. Finally, check the MOSFET manufacturer's data sheet for an avalanche energy rating (EAS). Some MOSFETs are not rated for body diode avalanche and will fail catastrophically if the VDS exceeds the device BVDSS, even if only by a fraction of a volt. Avalanche-rated MOSFETs are better able to sustain high frequency drain-to-source ringing near the device BVDSS during the turn-off transition. Power MOSFET Selection Calculating Power MOSFET Switching and Conduction Losses and Junction Temperatures The peak-to-peak gate drive level is set by the INTVCC voltage is 5V for the LTC3862 under normal operating conditions. Selection criteria for the power MOSFETs include the RDS(ON), gate charge QG, drain-to-source breakdown voltage BVDSS, maximum continuous drain current ID(MAX), In order to calculate the junction temperature of the power MOSFET, the power dissipated by the device must be known. This power dissipation is a function of the duty cycle, the load current and the junction temperature itself (due to the positive temperature coefficient of its RDS(ON)). As a 3862fb 26 LTC3862 APPLICATIONS INFORMATION result, some iterative calculation is normally required to determine a reasonably accurate value. The power dissipated by the MOSFET in a multi-phase boost converter with n phases is: 2 IO(MAX ) PFET = * RDS(ON) * DMAX * T n * 1 - DMAX ( ) + k * VOUT2 * IO(MAX ) ( n * 1 - DMAX ) * CRSS * f The first term in the equation above represents the I2R losses in the device, and the second term, the switching losses. The constant, k = 1.7, is an empirical factor inversely related to the gate drive current and has the dimension of 1/current. The T term accounts for the temperature coefficient of the RDS(ON) of the MOSFET, which is typically 0.4%/C. Figure 19 illustrates the variation of normalized RDS(ON) over temperature for a typical power MOSFET. From a known power dissipated in the power MOSFET, its junction temperature can be obtained using the following formula: TJ = TA + PFET * RTH(JA) 1.5 1.0 0.5 0 -50 It is tempting to choose a power MOSFET with a very low RDS(ON) in order to reduce conduction losses. In doing so, however, the gate charge QG is usually significantly higher, which increases switching and gate drive losses. Since the switching losses increase with the square of the output voltage, applications with a low output voltage generally have higher MOSFET conduction losses, and high output voltage applications generally have higher MOSFET switching losses. At high output voltages, the highest efficiency is usually obtained by using a MOSFET with a higher RDS(ON) and lower QG. The equation above can easily be split into two components (conduction and switching) and entered into a spreadsheet, in order to compare the performance of different MOSFETs. Programming the Current Limit The peak sense voltage threshold for the LTC3862 is 75mV at low duty cycle and with a normalized slope gain of 1.00, and is measured from SENSE+ to SENSE-. Figure 20 illustrates the change in the sense threshold with varying duty cycle and slope gain. MAXIMUM CURRENT SENSE THRESHOLD (mV) RT NORMALIZED ON RESISTANCE 2.0 The RTH(JA) to be used in this equation normally includes the RTH(JC) for the device plus the thermal resistance from the case to the ambient temperature (RTH(CA)). This value of TJ can then be compared to the original, assumed value used in the iterative calculation process. 50 100 0 JUNCTION TEMPERATURE (C) 150 3862 F19 Figure 19. Normalized Power MOSFET RDS(ON) vs Temperature 80 SLOPE = 0.625 75 70 SLOPE = 1 SLOPE = 1.66 65 60 55 50 0 10 20 30 40 50 60 70 80 90 100 DUTY CYCLE (%) 3862 F20 Figure 20. Maximum Sense Voltage Variation with Duty Cycle and Slope Gain 3862fb 27 LTC3862 APPLICATIONS INFORMATION For a boost converter where the current limit value is chosen to be 30% higher than the maximum load current, the peak current in the MOSFET and sense resistor is: ISW(MAX ) = IR(SENSE) = 1 1.3 * IO(MAX ) * 1+ * n 2 1 - DMAX The sense resistor value is then: RSENSE = ( VSENSE(MAX ) * n * 1 - DMAX 1.3 * 1+ * IO(MAX ) 2 ) Again, the factor n is the number of phases used, and represents the percentage ripple current in the inductor. The number 1.3 represents the factor by which the current limit exceeds the maximum load current, IO(MAX). For example, if the current limit needs to exceed the maximum load current by 50%, then the 1.3 factor should be replaced with 1.5. The average power dissipated in the sense resistor can easily be calculated as: 2 1.3 * IO(MAX ) PR(SENSE) = * RSENSE * DMAX n * 1 - DMAX ( ) This equation assumes no temperature coefficient for the sense resistor. If the resistor chosen has a significant temperature coefficient, then substitute the worst-case high resistance value into the equation. The resistor temperature can be calculated using the equation: TD = TA + PR(SENSE) * RTH(JA) Selecting the Output Diodes To maximize efficiency, a fast switching diode with low forward drop and low reverse leakage is required. The output diode in a boost converter conducts current during the switch off-time. The peak reverse voltage that the diode must withstand is equal to the regulator output voltage. The average forward current in normal operation is equal to the output current, and the peak current is equal to the peak inductor current: ID(PEAK ) = 1 IO(MAX ) * 1+ * n 2 1 - DMAX Although the average diode current is equal to the output current, in very high duty cycle applications (low VIN to high VOUT) the peak diode current can be several times higher than the average, as shown in Figure 21. In this case check the diode manufacturer's data sheet to ensure that its peak current rating exceeds the peak current in the equation above. In addition, when calculating the power dissipation in the diode, use the value of the forward voltage (VF) measured at the peak current, not the average output current. Excess power will be dissipated in the series resistance of the diode, which would not be accounted for if the average output current and forward voltage were used in the equations. Finally, this SW NODE 10V/DIV INDUCTOR CURRENT 2A/DIV DIODE CURRENT 2A/DIV VIN = 6V VOUT = 24V 1s/DIV 3862 F21 Figure 21. Diode Current Waveform for a High Duty Cycle Application 3862fb 28 LTC3862 APPLICATIONS INFORMATION additional power dissipation is important when deciding on a diode current rating, package type, and method of heat sinking. To a close approximation, the power dissipated by the diode is: PD = ID(PEAK) * VF(PEAK) * (1 - DMAX) The diode junction temperature is: TJ = TA + PD * RTH(JA) The RTH(JA) to be used in this equation normally includes the RTH(JC) for the device plus the thermal resistance from the board to the ambient temperature in the enclosure. Once the proper diode has been selected and the circuit performance has been verified, measure the temperature of the power components using a thermal probe or infrared camera over all operating conditions to ensure a good thermal design. Finally, remember to keep the diode lead lengths short and to observe proper switch-node layout (see Board Layout Checklist) to avoid excessive ringing and increased dissipation. ripple waveform are illustrated in Figure 22 for a typical boost converter. The choice of component(s) begins with the maximum acceptable ripple voltage (expressed as a percentage of the output voltage), and how this ripple should be divided between the ESR step and the charging/discharging V. For the purpose of simplicity we will choose 2% for the maximum output ripple, to be divided equally between the ESR step and the charging/discharging V. This percentage ripple will change, depending on the requirements of the application, and the equations provided below can easily be modified. One of the key benefits of multi-phase operation is a reduction in the peak current supplied to the output capacitor by the boost diodes. As a result, the ESR requirement of the capacitor is relaxed. For a 1% contribution to the total ripple voltage, the ESR of the output capacitor can be determined using the following equation: ESRCOUT 0.01 * VOUT ID(PEAK ) where: Output Capacitor Selection Contributions of ESR (equivalent series resistance), ESL (equivalent series inductance) and the bulk capacitance must be considered when choosing the correct combination of output capacitors for a boost converter application. The effects of these three parameters on the output voltage ID(PEAK ) = 1 IO(MAX ) * 1+ * n 2 1 - DMAX The factor n represents the number of phases and the factor represents the percentage inductor ripple current. SW1 50V/DIV SW2 50V/DIV IL1 2A/DIV IL2 2A/DIV VOUT 50mV/DIV AC COUPLED VIN = 10V VOUT = 48V 500mA LOAD 1s/DIV 3862 F22 Figure 22. Switching Waveforms for a Boost Converter 3862fb 29 LTC3862 APPLICATIONS INFORMATION For the bulk capacitance, which we assume contributes 1% to the total output ripple, the minimum required capacitance is approximately: COUT IO(MAX ) 0.01 * n * VOUT * f For many designs it will be necessary to use one type of capacitor to obtain the required ESR, and another type to satisfy the bulk capacitance. For example, using a low ESR ceramic capacitor can minimize the ESR step, while an electrolytic capacitor can be used to supply the required bulk C. The voltage rating of the output capacitor must be greater than the maximum output voltage, with sufficient derating to account for the maximum capacitor temperature. IORIPPLE/IOUT Because the ripple current in the output capacitor is a square wave, the ripple current requirements for this capacitor depend on the duty cycle, the number of phases and the maximum output current. Figure 23 illustrates the normalized output capacitor ripple current as a function of duty cycle. In order to choose a ripple current rating for the output capacitor, first establish the duty cycle range, based on the output voltage and range of input voltage. Referring to Figure 23, choose the worst-case high normalized ripple current, as a percentage of the maximum load current. 3.25 3.00 2.75 2.50 2.25 2.00 1.75 1.50 1.25 1.00 0.75 0.50 0.25 0 0.1 The output ripple current is divided between the various capacitors connected in parallel at the output voltage. Although ceramic capacitors are generally known for low ESR (especially X5R and X7R), these capacitors suffer from a relatively high voltage coefficient. Therefore, it is not safe to assume that the entire ripple current flows in the ceramic capacitor. Aluminum electrolytic capacitors are generally chosen because of their high bulk capacitance, but they have a relatively high ESR. As a result, some amount of ripple current will flow in this capacitor. If the ripple current flowing into a capacitor exceeds its RMS rating, the capacitor will heat up, reducing its effective capacitance and adversely affecting its reliability. After the output capacitor configuration has been determined using the equations provided, measure the individual capacitor case temperatures in order to verify good thermal performance. Input Capacitor Selection The input capacitor voltage rating in a boost converter should comfortably exceed the maximum input voltage. Although ceramic capacitors can be relatively tolerant of overvoltage conditions, aluminum electrolytic capacitors are not. Be sure to characterize the input voltage for any possible overvoltage transients that could apply excess stress to the input capacitors. 1-PHASE 2-PHASE 0.2 0.3 0.4 0.5 0.6 0.7 0.8 DUTY CYCLE OR (1-VIN/VOUT) 0.9 3862 F23 Figure 23. Normalized Output Capacitor Ripple Current (RMS) for a Boost Converter 3862fb 30 LTC3862 APPLICATIONS INFORMATION The value of the input capacitor is a function of the source impedance, and in general, the higher the source impedance, the higher the required input capacitance. The required amount of input capacitance is also greatly affected by the duty cycle. High output current applications that also experience high duty cycles can place great demands on the input supply, both in terms of DC current and ripple current. The input ripple current in a multi-phase boost converter is relatively low (compared with the output ripple current), because this current is continuous and is being divided between two or more inductors. Nonetheless, significant stress can be placed on the input capacitor, especially in high duty cycle applications. Figure 24 illustrates the normalized input ripple current, where: INORM = VIN L*f 1. The duty cycle range (where 5A is available at the output) is: V +V -V DMAX = O F IN VO + VF 48 V + 0.5V - 24V = 50.5% = 48 V + 0..5V 2. The operating frequency is chosen to be 300kHz so the period is 3.33s. From Figure 6, the resistor from the FREQ pin to ground is 45.3k. 0.90 0.80 0.70 $IIN/INORM Consider the LTC3862 application circuit is shown in Figure 25a. The output voltage is 48V and the input voltage range is 5V to 36V. The maximum output current is 5A when the input voltage is 24V to 36V. Below 24V, current limit will linearly reduce the maximum load to 1A at 5V in (see Figure 25b). 48 V + 0.5V - 36 V = 25.8% DMIN = 48 V + 0.5V 1.00 0.60 3. The minimum on-time for this application operating in CCM is: 1-PHASE 0.50 0.40 1 VO + VF - VIN(MAX ) 1 * tON(MIN) = * = f VO + VF 300kHz 2-PHASE 0.30 0.20 0.10 0 A Design Example 0 0.2 0.6 0.4 DUTY CYCLE 0.8 48 V + 0.5V - 36 V 48 V + 0.5V = 859ns 1.0 3862 F24 Figure 24. Normalized Input Peak-to-Peak Ripple Current The maximum DC input current is: IIN(MAX ) = IO(MAX ) 1 - DMAX = 5A = 10.1A 1 - 0.505 3862fb 31 LTC3862 APPLICATIONS INFORMATION VIN 5V TO 36V L1 18.7H PB2020-223 D1 30BQ060 1nF DMAX 3V8 SENSE1+ SLOPE BLANK LTC3862 ITH FB 100pF 84.5k 1F SS 30.1k 24.9k 6.8F 50V 6.8F 50V 6.8F 50V VIN 100F 10F 50V 63V 4.7F INTVCC 12.4k GATE1 PGND SGND 0.008 1W 10 475k VOUT Q2 HAT2266H GATE2 SENSE2- CLKOUT SYNC PLLFLTR VOUT 48V 5A (MAX) + 10nF + RUN FREQ 4.7nF 0.008 100F 1W 63V 10nF PHASEMODE SENSE1- 45.3k Q1 HAT2266H 10 10F 50V 10F 50V 10F 50V 10nF SENSE2+ L2 18.7H PB2020-223 D2 30BQ060 3862 F25a Figure 25a. A 5V to 36V Input, 48V/5A Output 2-Phase Boost Converter Application Circuit 5. The inductor ripple current is: OUTPUT LOAD CURRENT (A) 6 5 IL = 4 IO(MAX ) 0.4 5A * = * = 2.02A n 1 - DMAX 2 1 - 0.505 6. The inductor value is therefore: 3 2 L= 1 0 0 10 20 30 INPUT VOLTAGE (V) 40 3862 F25b Figure 25b. Output Current vs Input Voltage 4. A ripple current of 40% is chosen so the peak current in each inductor is: IIN(PK ) = = 1 IO(MAX ) * 1- * n 2 1 - DMAX 1 0.4 5A * = 6.06 A * 1+ 2 2 1 - 0.505 VIN(MIN) IL * f * DMAX = 24V * 0.505 2.02A * 300kHz = 20H 7. For a current limit value 30% higher than the maximum load current: IO(CL) = 1.3 * IO(MAX) = 1.3 * 5A = 6.5A The saturation current rating of the inductors must therefore exceed: IL(SAT) = 1 1.3 * IO(MAX ) * 1+ * 1 - DMAX n 2 1 0.4 1.3 * 5A * = 7.9 A * 1+ 2 2 1 - 0.505 3862fb 32 LTC3862 APPLICATIONS INFORMATION The inductor value chosen was 18.7H and the part number is PB2020-223, manufactured by Pulse Engineering. This inductor has a saturation current rating of 20A. 8. The power MOSFET chosen for this application is a Renesas HAT2266H. This MOSFET has a typical RDS(ON) of 11m at VGS = 4.5V and 9.2m at VGS = 10V. The BVDSS is rated at a minimum of 60V and the maximum continuous drain current is 30A. The typical gate charge is 25nC for a VGS = 4.5V. Last but not least, this MOSFET has an absolute maximum avalanche energy rating EAS of 34mJ, indicating that it is capable of avalanche without catastrophic failure. 9. The total IC quiescent current, IC power dissipation and maximum junction temperature are approximately: IQ(TOT) = IQ + 2 * QG(TOT) * f = 3mA + 2 * 25nC * 300kHz = 18mA PDISS = 24V * 18mA = 432mW TJ = 70C + 432mW * 34C/W = 84.7C 10. The inductor ripple current was chosen to be 40% and the maximum load current is 5A. For a current limit set at 30% above the maximum load current, the maximum switch and sense resistor currents are: ISW(MAX ) = IR(SENSE) = = 1 1.3 * IO(MAX ) * 1+ * n 2 1 - DMAX 1 0.4 1.3 * 5A = 7.9 A * 1+ * 2 2 1 - 0.505 11. The maximum current sense threshold for the LTC3862 is 75mV at low duty cycle and a normalized slope gain of 1.0. Using Figure 20, the maximum sense voltage drops to 73mV at a duty cycle of 51% with a normalized slope gain of 1, so the sense resistor is calculated to be: RSENSE = VSENSE(MAX ) ISW(MAX ) = 73mV = 9.2m 7.9 A For this application a 8m, 1W surface mount resistor was used for each phase. 12. The power dissipated in the sense resistors in current limit is: 2 1.3 * IO(MAX ) PR(SENSE) = * RSENSE * DMAX n * 1 - DMAX ( ) 2 1.3 * 5 = * 0.009 * 0.505 2 * 1 - 0.505 ( ) = 0.20 W 13. The average current in the boost diodes is half the output current (5A/2 = 2.5A), but the peak current in each diode is: ID(PEAK ) = = 1 IO(MAX ) * 1+ * n 2 1 - DMAX 1 0.4 5A = 6.06 A * 1+ * 2 2 1 - 0.505 The diode chosen for this application is the 30BQ060, manufactured by International Rectifier. This surface mount diode has a maximum average forward current of 3A at 125C and a maximum reverse voltage of 60V. The maximum forward voltage drop at 25C is 0.65V and is 0.42V at 125C (the positive TC of the series resistance is compensated by the negative TC of the diode forward voltage). The power dissipated by the diode is approximately: PD = ID(PEAK) * VF(PEAK) * (1 - DMAX) = 6.06A * 0.42V * (1 - 0.505) = 1.26W 14. Two types of output capacitors are connected in parallel for this application; a low ESR ceramic capacitor and an aluminum electrolytic for bulk storage. For a 1% contribution to the total ripple voltage, the maximum ESR of the composite output capacitance is approximately: ESRCOUT 0.01 * VOUT 0.01 * 48 V = = 0.109 ID(PEAK ) 4.4A 3862fb 33 LTC3862 APPLICATIONS INFORMATION For the bulk capacitance, which we assume contributes 1% to the total output ripple, the minimum required capacitance is approximately: 2. In order to help dissipate the power from the MOSFETs and diodes, keep the ground plane on the layers closest to the power components. Use power planes for the MOSFETs and diodes in order to maximize the heat IO(MAX ) 5A spreading from these components into the PCB. COUT = T 0.01 * n * VOUT * f 0.01 * 2 * 48 V * 300kHz 3. Place all power components in a tight area. This will minimize the size of high current loops. The high di/dt = 17.5F loops formed by the sense resistor, power MOSFET, For this application, in order to obtain both low ESR the boost diode and the output capacitor should be and an adequate ripple current rating (see Figure 23), kept as small as possible to avoid EMI. two 100F, 63V aluminum electrolytic capacitors were 4. Orient the input and output capacitors and current connected in parallel with four 6.8F, 50V ceramic sense resistors in a way that minimizes the distance capacitors. Figure 26 illustrates the switching wavebetween the pads connected to the ground plane. forms for this application circuit. Keep the capacitors for INTVCC, 3V8 and VIN as close as possible to LTC3862. SW1 50V/DIV IL1 5A/DIV 5. Place the INTVCC decoupling capacitor as close as possible to the INTVCC and PGND pins, on the same layer as the IC. A low ESR (X5R or better) 4.7F to 10F ceramic capacitor should be used. SW2 50V/DIV IL2 5A/DIV VOUT 100mV/DIV AC COUPLED 6. Use a local via to ground plane for all pads that connect to the ground. Use multiple vias for power components. VIN = 24V VOUT = 48V, 1.5A 2.5s/DIV 3862 F26 Figure 26. LTC3862 Switching Waveforms for Boost Converter PC Board Layout Checklist When laying out the printed circuit board, the following checklist should be used to ensure proper operation of the converter: 1. For lower power applications a 2-layer PC board is sufficient. However, for higher power levels, a multilayer PC board is recommended. Using a solid ground plane and proper component placement under the circuit is the easiest way to ensure that switching noise does not affect the operation. 7. Place the small-signal components away from high frequency switching nodes on the board. The pinout of the LTC3862 was carefully designed in order to make component placement easy. All of the power components can be placed on one side of the IC, away from all of the small-signal components. 8. The exposed area on the bottom of the QFN package is internally connected to PGND; however it should not be used as the main path for high current flow. 9. The MOSFETs should also be placed on the same layer of the board as the sense resistors. The MOSFET source should connect to the sense resistor using a short, wide PCB trace. 3862fb 34 LTC3862 APPLICATIONS INFORMATION 10. The output resistor divider should be located as close as possible to the IC, with the bottom resistor connected between FB and SGND. The PCB trace connecting the top resistor to the upper terminal of the output capacitor should avoid any high frequency switching nodes. 14. Keep the MOSFET drain nodes (SW1, SW2) away from sensitive small-signal nodes, especially from the opposite channel's current-sensing signals. The SW nodes can have slew rates in excess of 1V/ns relative to ground and should therefore be kept on the "output side" of the LTC3862. 11. Since the inductor acts like a current source in a peak current mode control topology, its placement on the board is less critical than the high di/dt components. 15. Check the stress on the power MOSFETs by independently measuring the drain-to-source voltages directly across the devices terminals. Beware of inductive ringing that could exceed the maximum voltage rating of the MOSFET. If this ringing cannot be avoided and exceeds the maximum rating of the device, choose a higher voltage rated MOSFET or consider using a snubber. 12. The SENSE+ and SENSE- PCB traces should be routed parallel to one another with minimum spacing in between all the way to the sense resistor. These traces should avoid any high frequency switching nodes in the layout. These PCB traces should also be Kelvinconnected to the interior of the sense resistor pads, in order to avoid sensing errors due to parasitic PCB resistance IR drops. 16. When synchronizing the LTC3862 to an external clock, use a low impedance source such as a logic gate to drive the SYNC pin and keep the lead as short as possible. 13. If an external RC filter is used between the sense resistor and the SENSE+ and SENSE- pins, these filter components should be placed as close as possible to the SENSE+ and SENSE- pins of the IC. Ensure that the SENSE- line is connected to the ground only at the point where the current sense resistor is grounded. 3862fb 35 LTC3862 TYPICAL APPLICATIONS A 12V Input, 24V/5A Output 2-Phase Boost Converter Application Circuit VIN 5V TO 24V L1 4.2H CDEP145-4R2 D1 MBRD835L 1nF DMAX 3V8 SENSE1+ SLOPE BLANK 100k 22F 25V 1F SS ITH FB 6.98k SGND 22F 25V VIN LTC3862 100pF 22F 25V + 10nF 26.7k + 15k RUN FREQ 1nF 0.007 100F 1W 35V 10nF PHASEMODE SENSE1- 45.3k Q1 Si7386DP 10 100F 10F 50V 35V 4.7F INTVCC GATE1 PGND 130k VOUT 10F 50V Q2 Si7386DP GATE2 SENSE2- CLKOUT SYNC PLLFLTR 10F 50V 0.007 1W 10 VOUT 24V 5A (MAX) 10F 50V 10nF L2 4.2H CDEP145-4R2 SENSE2+ D2 MBRD835L 3862 TA02a Start-Up Load Step ILOAD 5A/DIV RUN 5V/DIV IL1 5A/DIV IL1 5A/DIV IL2 5A/DIV IL2 5A/DIV VOUT 20V/DIV VOUT 500mV/DIV VIN = 12V VOUT = 24V ILOAD = 5A 3862 TA02b 1ms/DIV VIN = 12V VOUT = 24V ILOAD = 2A TO 5A 500s/DIV 3862 TA02c Efficiency 100 10000 VIN = 12V VOUT = 24V 95 EFFICIENCY (%) 1000 POWER LOSS 85 POWER LOSS (mW) EFFICIENCY 90 80 75 100 1000 LOAD CURRENT (mA) 100 10000 3862 TA02d 3862fb 36 LTC3862 TYPICAL APPLICATIONS A 4.5V to 5.5V Input, 12V/15A Output 4-Phase Boost Converter Application Circuit VIN 4.5V TO 5.5V L1 2.7H CDEP145-2R7 D1 MBRB2515LT41 1nF 3V8 BLANK 10nF ON/OFF CONTROL FREQ 1F SS 10nF 33F 10V 33F 10V RUN 10nF LTC3862 3.83k ITH FB 330pF 0.005 220F 1W 16V - + PHASEMODE SENSE1 45.3k Q1 HAT2165H 10 SENSE1+ SLOPE 33F 10V VIN 220F 15F 25V 16V 4.7F INTVCC 18.7k GATE1 PGND SGND 165k VOUT 15F 25V 0.005 1W 10 15F 25V Q3 HAT2165H GATE2 15F 25V SENSE2- CLKOUT SYNC PLLFLTR VOUT 12V 15A + DMAX 10nF L2 2.7H CDEP145-2R7 SENSE2+ L1 2.7H CDEP145-2R7 D2 MBRB2515LT41 D1 MBRB2515LT41 1nF 3V8 BLANK - + RUN FREQ 33F 10V 33F 10V 1F SS 330pF 0.005 220F 1W 16V 10nF PHASEMODE SENSE1 45.3k Q1 HAT2165H 10 SENSE1+ SLOPE LTC3862 ITH FB SGND + DMAX 33F 10V VIN 220F 15F 25V 16V 4.7F INTVCC GATE1 PGND 15F 25V 0.005 1W 10 15F 25V Q3 HAT2165H GATE2 15F 25V SENSE2- 10nF CLKOUT SYNC PLLFLTR 10k 10nF L2 2.7H CDEP145-2R7 SENSE2+ D2 MBRB2515LT41 3862 TA03a Start-Up Efficiency Load Step 100 VOUT 200mV/DIV VIN = 5V VOUT = 12V RLOAD = 10 1ms/DIV 3862 TA03b 100000 VIN = 5V VOUT = 12V 90 EFFICIENCY (%) VOUT 10V/DIV VIN 5V/DIV 95 EFFICIENCY 10000 85 80 75 POWER LOSS 1000 POWER LOSS (mW) IL1 MASTER 5A/DIV IL2 MASTER 5A/DIV IL1 SLAVE 5A/DIV IL2 SLAVE 5A/DIV ILOAD 2.5A-5A 5A DIV IL1 MASTER 5A/DIV IL2 MASTER 5A/DIV IL1 SLAVE 5A/DIV IL2 SLAVE 5A/DIV 70 65 VIN = 5V VOUT = 12V 250s/DIV 3862 TA03c 60 100 1000 10000 LOAD CURRENT (mA) 100 100000 3862 TA03d 3862fb 37 LTC3862 PACKAGE DESCRIPTION FE Package 24-Lead Plastic TSSOP (4.4mm) (Reference LTC DWG # 05-08-1663) Exposed Pad Variation AA 7.70 - 7.90* (.303 - .311) 3.25 (.128) 3.25 (.128) 24 23 22 21 20 19 18 17 16 15 14 13 6.60 p0.10 2.74 (.108) 4.50 p0.10 6.40 2.74 (.252) (.108) BSC SEE NOTE 4 0.45 p0.05 1.05 p0.10 0.65 BSC 1 2 3 4 5 6 7 8 9 10 11 12 RECOMMENDED SOLDER PAD LAYOUT 4.30 - 4.50* (.169 - .177) 0.09 - 0.20 (.0035 - .0079) 0.25 REF 1.20 (.047) MAX 0o - 8o 0.65 (.0256) BSC 0.50 - 0.75 (.020 - .030) NOTE: 1. CONTROLLING DIMENSION: MILLIMETERS MILLIMETERS 2. DIMENSIONS ARE IN (INCHES) 3. DRAWING NOT TO SCALE 0.05 - 0.15 (.002 - .006) 0.195 - 0.30 (.0077 - .0118) TYP FE24 (AA) TSSOP 0208 REV O 4. RECOMMENDED MINIMUM PCB METAL SIZE FOR EXPOSED PAD ATTACHMENT *DIMENSIONS DO NOT INCLUDE MOLD FLASH. MOLD FLASH SHALL NOT EXCEED 0.150mm (.006") PER SIDE GN Package 24-Lead Plastic SSOP (Narrow .150 Inch) (Reference LTC DWG # 05-08-1641) .337 - .344* (8.560 - 8.738) 24 23 22 21 20 19 18 17 16 15 1413 .033 (0.838) REF .045 p.005 .229 - .244 (5.817 - 6.198) .254 MIN .150 - .157** (3.810 - 3.988) .150 - .165 1 .0165 p.0015 2 3 4 5 6 7 8 9 10 11 12 .0250 BSC RECOMMENDED SOLDER PAD LAYOUT .015 p .004 s 45o (0.38 p 0.10) .0075 - .0098 (0.19 - 0.25) .0532 - .0688 (1.35 - 1.75) .004 - .0098 (0.102 - 0.249) 0o - 8o TYP .016 - .050 (0.406 - 1.270) NOTE: 1. CONTROLLING DIMENSION: INCHES INCHES 2. DIMENSIONS ARE IN (MILLIMETERS) 3. DRAWING NOT TO SCALE .008 - .012 (0.203 - 0.305) TYP .0250 (0.635) BSC GN24 (SSOP) 0204 *DIMENSION DOES NOT INCLUDE MOLD FLASH. MOLD FLASH SHALL NOT EXCEED 0.006" (0.152mm) PER SIDE **DIMENSION DOES NOT INCLUDE INTERLEAD FLASH. INTERLEAD FLASH SHALL NOT EXCEED 0.010" (0.254mm) PER SIDE 3862fb 38 LTC3862 PACKAGE DESCRIPTION UH Package 24-Lead Plastic QFN (5mm x 5mm) (Reference LTC DWG # 05-08-1747 Rev A) 0.75 p0.05 5.40 p0.05 3.90 p0.05 3.20 p 0.05 3.25 REF 3.20 p 0.05 PACKAGE OUTLINE 0.30 p 0.05 0.65 BSC RECOMMENDED SOLDER PAD LAYOUT APPLY SOLDER MASK TO AREAS THAT ARE NOT SOLDERED 5.00 p 0.10 R = 0.05 TYP 0.75 p 0.05 BOTTOM VIEW--EXPOSED PAD R = 0.150 TYP 23 0.00 - 0.05 PIN 1 NOTCH R = 0.30 TYP OR 0.35 s 45o CHAMFER 24 0.55 p 0.10 PIN 1 TOP MARK (NOTE 6) 1 2 3.20 p 0.10 5.00 p 0.10 3.25 REF 3.20 p 0.10 (UH24) QFN 0708 REV A 0.200 REF NOTE: 1. DRAWING IS NOT A JEDEC PACKAGE OUTLINE 2. DRAWING NOT TO SCALE 3. ALL DIMENSIONS ARE IN MILLIMETERS 4. DIMENSIONS OF EXPOSED PAD ON BOTTOM OF PACKAGE DO NOT INCLUDE MOLD FLASH. MOLD FLASH, IF PRESENT, SHALL NOT EXCEED 0.20mm ON ANY SIDE 5. EXPOSED PAD SHALL BE SOLDER PLATED 6. SHADED AREA IS ONLY A REFERENCE FOR PIN 1 LOCATION ON THE TOP AND BOTTOM OF PACKAGE 0.30 p 0.05 0.65 BSC 3862fb Information furnished by Linear Technology Corporation is believed to be accurate and reliable. However, no responsibility is assumed for its use. Linear Technology Corporation makes no representation that the interconnection of its circuits as described herein will not infringe on existing patent rights. 39 LTC3862 RELATED PARTS PART NUMBER DESCRIPTION COMMENTS LTC1624 Current Mode DC/DC Controller SO-8, 300kHz Operating Frequency, Buck, Boost, SEPIC Design, VIN Up to 36V LTC1700 No RSENSETM Synchronous Step-Up Controller Up to 95% Efficiency, Operating as Low as 0.9V Input LTC1871/LTC1871-1 Wide Input Range, No RSENSE Current Mode Boost, Flyback and SEPIC Controller LTC1871-7 Programmable Frequency from 50kHz to 1MHz in MSOP-10 Package. LT(R)1950 Wide Input Range Forward, Flyback, Boost or SEPIC Controller Suitable for 36V to 72V Inputs Single Switch PWM Controller with Auxiliary Boost Converter LT1952/LT1952-1 Single Switch Synchronous Forward Controllers Ideal for High Power 48V Input Applications LTC3704 Positive-to-Negative DC/DC Controller No RSENSE, Current Mode Control, 50kHz to 1MHz LTC3706/LTC3705 Isolated Synchronous Forward Converter Chip Set with PolyPhase(R) Capability Ideal for High Power 48V Input Applications LTC3726/LTC3725 Isolated Synchronous Forward Converter Chip Set Ideal for High Power 48V Input Applications LT3782A 2-Phase Step-Up DC/DC Controller 6V VIN 40V, 4A Gate Drive, 150kHz to 500kHz LTC3803 LTC3803-3 LTC3803-5 Constant-Frequency Current Mode Flyback DC/DC Controllers in ThinSOTTM Wide Input Range Flyback, Boost and SEPIC Controller. High Temperature Grade Available LTC3805 Adjustable Frequency Current Mode DC/DC Controller Wide Input Range Flyback, Boost and SEPIC Controller with Programmable Frequency, Run and Soft-Start LTC3806 Synchronous Flyback DC/DC Controller Current Mode Flyback Controller with Synchronous Gate Drive LTC3813 100V Current Mode Synchronous Step-Up Controller Large 1 Gate Drivers, No Current Sense Resistor Required LTC3814-5 60V Current Mode Synchronous Step-Up Controller Large 1 Gate Drivers, No Current Sense Resistor Required LT3825 Isolated No-OPTO Synchronous Flyback Controller with Wide Input Supply Range Input Voltage Limited Only by External Components, Ideal for 48V Input Applications LT3837 Isolated No-OPTO Synchronous Flyback Controller Suitable for Industrial 9V to 36V Inputs LTC3872 No RSENSE Current Mode Boost DC/DC Controller 550kHz Fixed Frequency, 2.75V VIN 9.8V, ThinSOT Package LTC3873 LTC3873-5 No RSENSE Constant Frequency Current Mode Flyback, Boost and SEPIC DC/DC Controllers Programmable Soft-Start, Adjustable Current Limit, 2mm x 3mm DFN or 8-Lead TSOT-23 Packages LTC3780 High Efficiency, Synchronous, 4-Switch Buck-Boost Controller VIN: 4V to 36V, VOUT : 0.8V to 30V, IQ = 1.5mA, ISD < 55A, SSOP-24, QFN-32 Packages PolyPhase is a registrated treademark of Linear Technology Corporation. No RSENSE and ThinSOT are trademarks of Linear Technology Corporation. 3862fb 40 Linear Technology Corporation LT 1208 REV B * PRINTED IN USA 1630 McCarthy Blvd., Milpitas, CA 95035-7417 (408) 432-1900 FAX: (408) 434-0507 www.linear.com (c) LINEAR TECHNOLOGY CORPORATION 2008