19-2966; Rev 0; 10/03 Dual 180 Out-of-Phase Buck Controllers with Sequencing/Prebias Startup and POR Features The MAX1858A/MAX1875A/MAX1876A dual, synchronized, step-down controllers generate two outputs from input supplies ranging from 4.5V to 23V. Each output is adjustable from sub-1V to 18V and supports loads of 10A or higher. Input voltage ripple and total RMS input ripple current are reduced by synchronized 180 out-of-phase operation. 4.5V to 23V Input Supply Range The switching frequency is adjustable from 100kHz to 600kHz with an external resistor. Alternatively, the controller can be synchronized to an external clock generated by another MAX1858A/MAX1875A/MAX1876A or a system clock. One MAX1858A/MAX1875A/MAX1876A can be set to generate an in-phase, or 90 out-ofphase, clock signal for synchronization with additional controllers. This allows two controllers to operate either as an interleaved two- or four-phase system with each output shifted by 90. The MAX1858A/MAX1875A/ MAX1876A feature soft-start. The MAX1858A also features first-on/last-off power sequencing and soft-stop. The MAX1858A/MAX1875A/MAX1876A eliminate the need for current-sense resistors by utilizing the low-side MOSFET's on-resistance as a current-sense element. This protects the DC-DC components from damage during output-overload conditions or output short-circuit faults without requiring a current-sense resistor. Adjustable foldback current limit reduces power dissipation during short-circuit conditions. The MAX1858A/ MAX1876A include a power-on reset (POR) output to signal the system when both outputs reach regulation. Optional Synchronization The MAX1858A/MAX1875A/MAX1876A ensure that the output voltage does not swing negative when the input power is removed or when EN is driven low. The MAX1875A/MAX1876A also allow prebias startup without discharging the output. The MAX1858A/MAX1875A/MAX1876A are available in a 24-pin QSOP package. Use the MAX1875 evaluation kit or the MAX1858 evaluation kit to evaluate the MAX1858A/MAX1875A/MAX1876A. Applications Network Power Supplies Telecom Power Supplies DSP, ASIC, and FPGA Power Supplies Set-Top Boxes Broadband Routers Servers 0 to 18V Output Voltage Range (Up to 10A) Adjustable Lossless Foldback Current Limit Adjustable 100kHz to 600kHz Switching Frequency Clock Output for Master/Slave Synchronization 4 x 90 Out-of-Phase Step-Down Converters (Using Two Controllers, Figure 7) Prebias Startup (MAX1875A/MAX1876A) Power Sequencing (MAX1858A) RST Output with 140ms Minimum Delay (MAX1858A/MAX1876A) Fixed-Frequency Pulse-Width Modulation (PWM) Operation Ordering Information PART TEMP RANGE PIN-PACKAGE MAX1858AEEG -40C to +85C 24 QSOP MAX1875AEEG -40C to +85C 24 QSOP MAX1876AEEG -40C to +85C 24 QSOP Pin Configuration TOP VIEW COMP2 1 24 EN FB2 2 23 DH2 ILIM2 3 22 LX2 OSC 4 21 BST2 V+ 5 REF 6 GND 7 20 DL2 MAX1858A MAX1875A MAX1876A 19 VL 18 PGND CKO 8 17 DL1 SYNC 9 16 BST1 ILIM1 10 15 LX1 FB1 11 14 DH1 COMP1 12 13 RST (N.C.) Desknote Computers QSOP () ARE FOR THE MAX1875A ONLY ________________________________________________________________ Maxim Integrated Products For pricing, delivery, and ordering information, please contact Maxim/Dallas Direct! at 1-888-629-4642, or visit Maxim's website at www.maxim-ic.com. 1 MAX1858A/MAX1875A/MAX1876A General Description MAX1858A/MAX1875A/MAX1876A Dual 180 Out-of-Phase Buck Controllers with Sequencing/Prebias Startup and POR ABSOLUTE MAXIMUM RATINGS V+ to GND ..............................................................-0.3V to +25V PGND to GND .......................................................-0.3V to +0.3V VL to GND ..................-0.3V to the lower of +6V and (V+ + 0.3V) BST1, BST2 to GND ...............................................-0.3V to +30V LX1 to BST1..............................................................-6V to +0.3V LX2 to BST2..............................................................-6V to +0.3V DH1 to LX1 ..............................................-0.3V to (VBST1 + 0.3V) DH2 to LX2 ..............................................-0.3V to (VBST2 + 0.3V) DL1, DL2 to PGND........................................-0.3V to (VL + 0.3V) CKO, REF, OSC, ILIM1, ILIM2, COMP1, COMP2 to GND ..........................-0.3V to (VL + 0.3V) FB1, FB2, RST, SYNC, EN to GND...........................-0.3V to +6V VL to GND Short Circuit .............................................Continuous REF to GND Short Circuit ...........................................Continuous Continuous Power Dissipation (TA = +70C) 24-Pin QSOP (derate 9.4mW/C above +70C)...........762mW Operating Temperature Range ...........................-40C to +85C Junction Temperature ......................................................+150C Storage Temperature Range .............................-65C to +150C Lead Temperature (soldering, 10s) .................................+300C Stresses beyond those listed under "Absolute Maximum Ratings" may cause permanent damage to the device. These are stress ratings only, and functional operation of the device at these or any other conditions beyond those indicated in the operational sections of the specifications is not implied. Exposure to absolute maximum rating conditions for extended periods may affect device reliability. ELECTRICAL CHARACTERISTICS (V+ = 12V, EN = ILIM_ = VL, SYNC = GND, IVL = 0mA, PGND = GND, CREF = 0.22F, CVL = 4.7F (ceramic), ROSC = 60k, compensation components for COMP_ are from Figure 1, TA = -40C to +85C (Note 1), unless otherwise noted.) PARAMETER CONDITIONS MIN TYP MAX UNITS GENERAL V+ Operating Range (Note 2) 4.5 23.0 VL = V+ (Note 2) 4.5 5.5 V V+ Operating Supply Current VL unloaded, no MOSFETs connected 3.5 6 mA V+ Standby Supply Current EN = LX_ = FB_ = 0V 0.3 0.6 mA Thermal Shutdown Rising temperature, typical hysteresis = 10C Current-Limit Threshold PGND - LX_ VL REGULATOR Output Voltage ILIM_ = VL 75 100 RILIM_ = 100k 32 50 62 RILIM_ = 600k 225 300 375 5.5V < V+ < 23V, 1mA < ILOAD < 50mA 4.75 5 5.25 V 4.1 4.2 4.3 V VL Undervoltage Lockout Rising Trip Level VL Undervoltage Lockout Hysteresis C +160 (Note 3) 125 100 mV mV REFERENCE Output Voltage IREF = 0A Reference Load Regulation 0A < IREF < 50A 1.98 2.00 2.02 V 0 4 10 mV SOFT-START Digital Ramp Period Internal 6-bit DAC for one converter to ramp from 0V to full scale (Note 4) Soft-Start Steps 1024 DC-DC clocks 64 Steps FREQUENCY Low End of Range ROSC = 60k High End of Range ROSC = 10k DH_ Minimum Off-Time ROSC = 10k 2 0C to +85C 84 100 115 -40C to +85C 80 100 120 540 600 660 kHz 250 303 ns _______________________________________________________________________________________ kHz Dual 180 Out-of-Phase Buck Controllers with Sequencing/Prebias Startup and POR (V+ = 12V, EN = ILIM_ = VL, SYNC = GND, IVL = 0mA, PGND = GND, CREF = 0.22F, CVL = 4.7F (ceramic), ROSC = 60k, compensation components for COMP_ are from Figure 1, TA = -40C to +85C (Note 1), unless otherwise noted.) PARAMETER CONDITIONS MIN SYNC Range Switching frequency must be set to half of the SYNC frequency 200 SYNC Input Pulse Width (Note 4) High 100 Low 100 SYNC Rise/Fall Time (Note 4) TYP MAX UNITS 1200 kHz ns 100 ns 250 nA ERROR AMPLIFIER FB_ Input Bias Current FB_ Input Voltage Set Point FB_ to COMP_ Transconductance 0C to +85C 0.985 1.00 1.015 -40C to +85C 0.98 1.00 1.02 0C to +85C 1.25 1.8 2.70 -40C to +85C 1.2 1.8 2.9 V mS DRIVERS DL_, DH_ Break-Before-Make Time CLOAD = 5nF DH_ On-Resistance DL_ On-Resistance 30 ns Low 1.5 2.5 High 3 5 Low 0.6 1.5 High 3 5 LOGIC INPUTS (EN, SYNC) Input Low Level Typical 15% hysteresis, VL = 4.5V Input High Level VL = 5.5V 2.4 0.8 Input High/Low Bias Current VEN = 0 or 5.5V -1 V V +0.1 +1 A 0.4 V LOGIC OUTPUTS (CKO) Output Low Level VL = 5V, sinking 5mA Output High Level VL = 5V, sourcing 5mA 4.0 V COMP_ Pulldown Resistance During Shutdown and Current Limit 17 RST OUTPUT (MAX1858A/MAX1876A ONLY) Output-Voltage Trip Level Output Low Level Both FBs must be over this to allow the reset timer to start; there is no hysteresis 0.87 0.9 0.93 VL = 5V, sinking 3.2mA 0.4 VL = 1V, sinking 0.4mA 0.3 Output Leakage V+ = VL = 5V, V RST = 5.5V, VFB = 1V Reset Timeout Period VFB_ = 1V FB_ to Reset Delay FB_ overdrive from 1V to 0.85V 140 315 4 V V 1 A 560 ms s Note 1: Specifications to -40C are guaranteed by design and not production tested. Note 2: Operating supply range is guaranteed by VL line regulation test. Connect V+ to VL for 5V operation. Note 3: When VL falls and UVLO is tripped, the device is latched and VL must be discharged below 2.5V before normal operation can resume. Note 4: Guaranteed by design and not production tested. _______________________________________________________________________________________ 3 MAX1858A/MAX1875A/MAX1876A ELECTRICAL CHARACTERISTICS (continued) Typical Operating Characteristics (Circuit of Figure 1, VIN = 12V, TA = +25C, unless otherwise noted.) 60 50 40 30 20 10 0.8 0.6 0.4 0.2 OUT2 0 -0.2 OUT1 -0.4 0.5 MAX1858A/75A/76A toc03 OUT1 70 1.0 MAX1858A/75A/76A toc02 80 EFFICIENCY (%) MAX1858A/75A/76A toc01 OUT2 OUTPUT VOLTAGE ACCURACY (%) 100 90 VL VOLTAGE ACCURACY vs. LOAD CURRENT OUTPUT VOLTAGE ACCURACY vs. LOAD 0 VL VOLTAGE ACCURACY EFFICIENCY vs. LOAD -0.5 -1.0 -1.5 -0.6 -0.8 0 -2.0 -1.0 0.1 1 100 10 0 5 10 15 LOAD (A) SWITCHING FREQUENCY vs. ROSC LOAD TRANSIENT RESPONSE (OUTPUT 1) 400 50 100 MAX1858A/75A/76A toc06 VOUT2 50mV/div AC-COUPLED VOUT1 50mV/div AC-COUPLED VOUT1 50mV/div AC-COUPLED VOUT2 50mV/div AC-COUPLED 10A IOUT1 10A IOUT2 0A 0A 300 200 100 0 0 10 20 30 40 50 10s/div 60 10s/div ROSC (k) SOFT-START AND SOFT-STOP WAVEFORM (MAX1858A ONLY) SOFT-START AND SOFT-STOP WAVEFORM (MAX1858A ONLY) MAX1858A/75A/76A toc08 MAX1858A/75A/76A toc07 10V EN 0V EN PULLED HIGH BEFORE VOUT1 REACHES 0V. VOUT1 1V/div IOUT1 = 300mA 0V 2ms/div 5V EN 0V VOUT1 1V/div IOUT1 = 300mA 0V VOUT2 1V/div IOUT2 = 300mA 0V VOUT2 1V/div IOUT2 = 300mA 0V 4 150 LOAD TRANSIENT RESPONSE (OUTPUT 2) MAX1858A/75A/76A toc05 MAX1858A/75A/76A toc04 500 0 LOAD CURRENT (mA) LOAD (A) 600 SWITCHING FREQUENCY (kHz) MAX1858A/MAX1875A/MAX1876A Dual 180 Out-of-Phase Buck Controllers with Sequencing/Prebias Startup and POR 2ms/div _______________________________________________________________________________________ Dual 180 Out-of-Phase Buck Controllers with Sequencing/Prebias Startup and POR START AND STOP WAVEFORM (MAX1875A/MAX1876A ONLY) RESET TIMEOUT (MAX1858A/MAX1876A ONLY) INPUT POWER REMOVAL MAX1858A/75A/76A toc09 MAX1858A/75A/76A toc10 10V EN 0V MAX1858A/75A/76A toc11 VIN 10V/div 0V VOUT1 1V/div IOUT1 = 300mA 0V VOUT1 1V/div IOUT1 = 300mA 0V VOUT2 1V/div IOUT2 = 300mA VOUT2 1V/div IOUT2 = 300mA 0V 0V EN 0 VOUT2 0V VOUT1 0V VRST 0V PREBIAS STARTUP 2ms/div 5ms/div OUT-OF-PHASE WAVEFORM EXTERNALLY SYNCHRONIZED SWITCHING WAVEFORM 100ms/div, 5V/div MAX1858A/75A/76A toc13 MAX1858A/75A/76A toc12 VOUT1 20mV/div CKO OUTPUT WAVEFORM MAX1858A/75A/76A toc14 5V VSYNC 0V SYNC = GND 12V VLX1 5V VCK0 0V 5V VCK0 0V 0V 12V VLX2 10V VLX1 0V 10V VLX1 0V 0V VOUT1 10mV/div AC-COUPLED VOUT1 10mV/div AC-COUPLED VOUT2 20mV/div 400ns/div 1s/div 400ns/div SHORT-CIRCUIT CURRENT FOLDBACK AND RECOVERY CKO OUTPUT WAVEFORM MAX1858A/75A/76A toc15 MAX1858A/75A/76A toc16 SYNC = VL IOUT1 = 10A (5A/div) VOUT1 = 1.8V (1V/div) 5V VCK0 0V 10V VLX1 0V SHORT VOUT2 = 2.5V (1V/div) VOUT2 IOUT2 = 10A (5A/div) VOUT1 10mV/div 400ns/div 4ms/div _______________________________________________________________________________________ 5 MAX1858A/MAX1875A/MAX1876A Typical Operating Characteristics (continued) (Circuit of Figure 1, VIN = 12V, TA = +25C, unless otherwise noted.) Dual 180 Out-of-Phase Buck Controllers with Sequencing/Prebias Startup and POR MAX1858A/MAX1875A/MAX1876A Pin Description PIN 1 2 3 4 FUNCTION Compensation Pin for Regulator 2 (REG2). Compensate REG2's control loop by connecting a series resistor COMP2 (RCOMP2) and capacitor (CCOMP2A) to GND in parallel with a second compensation capacitor (CCOMP2B) as shown in Figure 1. FB2 Feedback Input for Regulator 2 (REG2). Connect FB2 to a resistive divider between REG2's output and GND to adjust the output voltage between 1V and 18V. To set the output voltage below 1V, connect FB2 to a resistive voltage-divider from REF to REG2's output. See the Setting the Output Voltage section. ILIM2 Current-Limit Adjustment for Regulator 2 (REG2). The PGND-LX2 current-limit threshold defaults to 100mV if ILIM2 is connected to VL. Connect a resistor (RILIM2) from ILIM2 to GND to adjust the REG2's current-limit threshold (VITH2) from 50mV (RILIM2 = 100k) to 300mV (RILIM2 = 600k). See the Setting the Valley Current Limit section. OSC Oscillator Frequency Set Input. Connect a resistor from OSC to GND (ROSC) to set the switching frequency from 100kHz (ROSC = 60k) to 600kHz (ROSC = 10k). The controller still requires ROSC when an external clock is connected to SYNC. When using an external clock, select ROSC as described above, and set the external clock frequency to twice the desired switching frequency. 5 V+ Input Supply Voltage. 4.5V to 23V. 6 REF 2V Reference Output. Bypass to GND with a 0.22F or greater ceramic capacitor. 7 GND Analog Ground CKO Clock Output. Clock output for external 2- or 4-phase synchronization (see the Clock Synchronization (SYNC, CKO) section). 9 SYNC Synchronization Input or Clock Output Selection Input. SYNC has three operating modes. Connect SYNC to a 200kHz to 1200kHz clock for external synchronization. Connect SYNC to GND for 2-phase operation as a master controller. Connect SYNC to VL for 4-phase operation as a master controller (see the Clock Synchronization (SYNC, CKO) section). 10 ILIM1 Current-Limit Adjustment for Regulator 1 (REG1). The PGND-LX1 current-limit threshold defaults to 100mV if ILIM1 is connected to VL. Connect a resistor (RILIM1) from ILIM1 to GND to adjust REG1's current-limit threshold (VITH1) from 50mV (RILIM1 = 100k) to 300mV (RILIM1 = 600k). See the Setting the Valley Current Limit section. 11 FB1 8 12 13 6 NAME Feedback Input for Regulator 1 (REG1). Connect FB1 to a resistive divider between REG1's output and GND to adjust the output voltage between 1V and 18V. To set the output voltage below 1V, connect FB1 to a resistive voltage-divider from REF and REG1's output. See the Setting the Output Voltage section. Compensation Pin for Regulator 1 (REG1). Compensate REG1's control loop by connecting a series resistor COMP1 (RCOMP1) and capacitor (CCOMP1A) to GND in parallel with a second compensation capacitor (CCOMP1B) as shown in Figure 1. RST Open-Drain Reset Output (MAX1858A/MAX1876A Only). RST is low when either output voltage is more than 10% below its regulation point. After soft-start is completed and both outputs exceed 90% of their nominal output voltage (VFB_ > 0.9V), RST becomes high impedance after a 140ms delay and remains high impedance as long as both outputs maintain regulation. Connect a resistor between RST and the logic supply for logic-level voltages. N.C. Connect to GND or leave unconnected for the MAX1875A. _______________________________________________________________________________________ Dual 180 Out-of-Phase Buck Controllers with Sequencing/Prebias Startup and POR PIN NAME FUNCTION 14 DH1 High-Side Gate-Driver Output for Regulator 1 (REG1). DH1 swings from LX1 to BST1. DH1 is low during UVLO. 15 LX1 External Inductor Connection for Regulator 1 (REG1). Connect LX1 to the switched side of the inductor. LX1 serves as the lower supply rail for the DH1 high-side gate driver. 16 BST1 Boost Flying-Capacitor Connection for Regulator 1 (REG1). Connect BST1 to an external ceramic capacitor and diode according to Figure 1. 17 DL1 Low-Side Gate-Driver Output for Regulator 1 (REG1). DL1 swings from PGND to VL. DL1 is low during UVLO. 18 PGND 19 VL 20 DL2 Low-Side Gate-Driver Output for Regulator 2 (REG2). DL2 swings from PGND to VL. DL2 is low during UVLO. 21 BST2 Boost Flying-Capacitor Connection for Regulator 2 (REG2). Connect BST2 to an external ceramic capacitor and diode according to Figure 1. 22 LX2 23 DH2 24 EN Power Ground Internal 5V Linear-Regulator Output. Supplies the regulators and powers the low-side gate drivers and external boost circuitry for the high-side gate drivers. External Inductor Connection for Regulator 2 (REG2). Connect LX2 to the switched side of the inductor. LX2 serves as the lower supply rail for the DH2 high-side gate driver. High-Side Gate-Driver Output for Regulator 2 (REG2). DH2 swings from LX2 to BST2. DH2 is low during UVLO. Active-High Enable Input. A logic low shuts down both controllers. Connect to VL for always-on operation. Detailed Description DC-DC PWM Controller The MAX1858A/MAX1875A/MAX1876A step-down converters use a PWM voltage-mode control scheme (Figure 2) for each out-of-phase controller. The controller generates the clock signal by dividing down the internal oscillator or SYNC input when driven by an external clock, so each controller's switching frequency equals half the oscillator frequency (fSW = fOSC/2). An internal transconductance error amplifier produces an integrated error voltage at the COMP pin, providing high DC accuracy. The voltage at COMP sets the duty cycle using a PWM comparator and a ramp generator. At each rising edge of the clock, REG1's high-side N-channel MOSFET turns on and remains on until either the appropriate duty cycle or until the maximum duty cycle is reached. REG2 operates out-of-phase, so the second high-side MOSFET turns on at each falling edge of the clock. During each high-side MOSFET's on-time, the associated inductor current ramps up. During the second-half of the switching cycle, the highside MOSFET turns off and the low-side N-channel MOSFET turns on. Now the inductor releases the stored energy as its current ramps down, providing current to the output. Under overload conditions, when the induc- tor current exceeds the selected valley current limit (see the Current-Limit Circuit (ILIM_) section), the high-side MOSFET does not turn on at the appropriate clock edge and the low-side MOSFET remains on to let the inductor current ramp down. Synchronized Out-of-Phase Operation The two independent regulators in the MAX1858A/ MAX1875A/MAX1876A operate 180 out-of-phase to reduce input filtering requirements, reduce electromagnetic interference (EMI), and improve efficiency. This effectively lowers component cost and saves board space, making the MAX1858A/MAX1875A/MAX1876A ideal for cost-sensitive applications. Dual-switching regulators typically operate both controllers in-phase, and turn on both high-side MOSFETs at the same time. The input capacitor must then support the instantaneous current requirements of both controllers simultaneously, resulting in increased ripple voltage and current when compared to a single switching regulator. The higher RMS ripple current lowers efficiency due to power loss associated with the input capacitor's effective series resistance (ESR). This typically requires more low-ESR input capacitors in parallel to minimize input voltage ripple and ESR-related losses, or to meet the necessary ripple-current rating. _______________________________________________________________________________________ 7 MAX1858A/MAX1875A/MAX1876A Pin Description (continued) MAX1858A/MAX1875A/MAX1876A Dual 180 Out-of-Phase Buck Controllers with Sequencing/Prebias Startup and POR VIN 6V - 23V CMPSH-3A RV+ 4.7 0.1F V+ CV+ 0.22F VL 4.7 CIN1 2 x 10F BST1 COUT1 4 x 220F NH1* L1 1.1H R1A 8.06k ** NL1* CIN2 2 x 10F BST2 CBST2 0.1F CBST1 0.1F OUTPUT1 VOUT = 1.8V CVL 4.7F 4.7 DH1 DH2 LX1 LX2 DL1 DL2 NH2* L2 1.1H ** NL2* R2A 15k OUTPUT2 VOUT = 2.5V COUT2 4 x 220F PGND FB1 FB2 RCOMP1 5.9k R1B 10k RCOMP2 8.2k COMP1 CCOMP1A 0.01F CCOMP1B 100pF D2 CMSSH-3 R2B 10k COMP2 CCOMP2B 100pF 10k OSC MAX1858A MAX1875A MAX1876A REF CCOMP2A 6800pF D3 CMSSH-3 CREF 0.22F GND CLOCK OUTPUT CKO RESET OUTPUT RST (MAX1858A/ MAX1876A ONLY) ILIM1 EN ILIM2 SYNC VL 118k 96.5k ON 140k OFF 84.5k *IRF7811W **OPTIONAL Figure 1. Standard 600kHz Application Circuit With dual, synchronized, out-of-phase operation, the MAX1858A/MAX1875A/MAX1876As' high-side MOSFETs turn on 180 out-of-phase. The instantaneous input current peaks of both regulators no longer overlap, resulting in reduced RMS ripple current and input voltage ripple. This reduces the required input capacitor ripple-current rating, allowing fewer or less expensive capacitors, and reduces shielding requirements for EMI. The Out-ofPhase Waveforms in the Typical Operating Characteristics demonstrate synchronized 180 out-of-phase operation. 8 Internal 5V Linear Regulator (VL) All MAX1858A/MAX1875A/MAX1876A functions are internally powered from an on-chip, low-dropout 5V regulator. The maximum regulator input voltage (V+) is 23V. Bypass the regulator's output (VL) with a 4.7F ceramic capacitor to PGND. The VL dropout voltage is typically 500mV, so when V+ is greater than 5.5V, VL is typically 5V. The MAX1858A/MAX1875A/MAX1876A also employs an undervoltage lockout circuit that disables both regulators when V L falls below 4.2V. V L should also be bypassed to GND with a 0.1F capacitor. When VL falls and UVLO is tripped, the device is latched and VL must be discharged below 2.5V before normal operation can resume. _______________________________________________________________________________________ Dual 180 Out-of-Phase Buck Controllers with Sequencing/Prebias Startup and POR MAX1858A/MAX1875A/MAX1876A REF V+ 5V LINEAR REGULATOR MAX1858A MAX1875A MAX1876A VREF 2.0V GND VL COMP1 BST1 FB1 CONVERTER 1 DH1 SOFT-START DAC (SEQUENCING-- MAX1858A ONLY) LX1 R Q S Q DL1 PGND 1VP-P OSC SYNC OSCILLATOR CK0 5A RST (MAX1858A/ MAX1876A ONLY) ILIM1 RESET EN VREF UVLO AND SHUTDOWN VL - 0.5V VL BST2 DH2 CONVERTER 2 LX2 DL2 COMP2 FB2 ILIM2 Figure 2. Functional Diagram The internal VL linear regulator can source over 50mA to supply the IC, power the low-side gate driver, charge the external boost capacitor, and supply small external loads. When driving large FETs, little or no regulator current may be available for external loads. For example, when switched at 600kHz, a single large FET with 18nC total gate charge requires 18nC 600kHz = 11mA. To drive larger MOSFETs, or deliver larger loads, connect VL to an external power supply from 4.5V to 5.5V. _______________________________________________________________________________________ 9 MAX1858A/MAX1875A/MAX1876A Dual 180 Out-of-Phase Buck Controllers with Sequencing/Prebias Startup and POR A UVLO B C D F E G H J I VL EN VOUT_ SS_ DH_ DL_ M N O MAX1875A/MAX1876A POWER-ON-OFF SEQUENCING DEFINITIONS SYMBOL UVLO VL EN VOUT_ SS_ DH_ DL_ A B C D E F G H I J K L M N O DEFINITION Undervoltage lockout trip level is provided in the Electrical Characteristics table. Internal 5V Linear-Regulator Output Active-High Enable Input Output Voltage Internal Soft-Start Input Signal into Error Amplifier High-Side Gate-Driver Output Low-Side Gate-Driver Output VL rising while below the UVLO threshold. EN is low. VL is greater than the UVLO threshold. EN is low. EN is pulled high. Normal operation VL enters UVLO. VL exits UVLO. Resumes normal operation EN is pulled low. EN is pulled high. Resumes normal operation VL drops below UVLO threshold while EN is high. Resumes normal operation UVLO is activated and DL_ is latched low. Exiting UVLO: DL_ remains latched low until the first fall of DH_ is detected. DL_ is low after EN is pulled low. Figure 3. MAX1875A/MAX1876A Detailed Power-On-Off Sequencing 10 ______________________________________________________________________________________ K L Dual 180 Out-of-Phase Buck Controllers with Sequencing/Prebias Startup and POR A B C D E F G H J I K L MAX1858A/MAX1875A/MAX1876A UVLO M VL EN VOUT1 SS1 VOUT2 SS2 DH1 DL1 DH2 DL2 N O P MAX1858A POWER-ON-OFF SEQUENCING DEFINITIONS SYMBOL UVLO VL EN VOUT1 SS1 VOUT2 SS2 DH1 DL1 DH2 DL2 A B C DEFINITION Undervoltage threshold value is provided in the Electrical Characteristics table. Internal 5V Linear-Regulator Output Active-High Enable Input Regulator 1 Output Voltage Regulator 1: Internal Soft-Start Input Signal into Error Amplifier Regulator 2 Output Voltage Regulator 2: Internal Soft-Start Input Signal into Error Amplifier Regulator 1: High-Side Gate-Driver Output Regulator 1: Low-Side Gate-Driver Output Regulator 2: High-Side Gate-Driver Output Regulator 2: Low-Side Gate-Driver Output VL rising while below the UVLO threshold. EN is low. VL is greater than the UVLO threshold. EN is low. EN is pulled high. DH1 and DL1 start switching. DH2 and DL2 are off. SYMBOL D E F G H I J K L M N DEFINITION Normal operation VL enters UVLO. VL exits UVLO. Resumes normal operation. DH1 and DL1 start switching. DH2 and DL2 are off. EN is pulled low and then high. VOUT1 must reach 0V before restarting due to the cycling of the enable in region H (above). VOUT1 recovers. VOUT2 recovers. VL enters UVLO before VOUT2 fully recovers. VL exits UVLO. UVLO latches DL_ low. O Exiting UVLO: DL_ remains latched low until the first fall of DH_ is detected. P DL_ is high after EN is pulled low and soft-stop is complete. Figure 4. MAX1858A Detailed Power-On-Off Sequencing ______________________________________________________________________________________ 11 High-Side Gate-Drive Supply (BST_) Gate-drive voltages for the high-side N-channel switches are generated by the flying-capacitor boost circuits (Figure 5). A boost capacitor (connected from BST_ to LX_) provides power to the high-side MOSFET driver. On startup, the synchronous rectifier (low-side MOSFET) forces LX_ to ground and charges the boost capacitor to 5V. On the second half-cycle, after the low-side MOSFET turns off, the high-side MOSFET is turned on by closing an internal switch between BST_ and DH_. This provides the necessary gate-to-source voltage to turn on the highside switch, an action that boosts the 5V gate-drive signal above VIN. The current required to drive the highside MOSFET gates (fSWITCH QG) is ultimately drawn from VL. MOSFET Gate Drivers (DH_, DL_) The DH and DL drivers are optimized for driving moderate-size N-channel high-side and larger low-side power MOSFETs. This is consistent with the low duty factor seen with large VIN - VOUT differential. The DL_ low-side drive waveform is always the complement of the DH_ high-side drive waveform (with controlled dead time to prevent cross-conduction or "shoot-through"). An adaptive dead-time circuit monitors the DL_ output and prevents the high-side FET from turning on until DL_ is fully off. There must be a low-resistance, low-inductance path from the DL_ driver to the MOSFET gate in order for the adaptive dead-time circuit to work properly. Otherwise, the sense circuitry in the MAX1858A/MAX1875A/ MAX1876A interprets the MOSFET gate as "off" while there is actually charge still left on the gate. Use very short, wide traces (50mils to 100mils wide if the MOSFET is 1in from the device). The dead time at the DH-off edge is determined by a fixed 30ns internal delay. Current-Limit Circuit (ILIM_) The current-limit circuit employs a "valley" current-sensing algorithm that uses the on-resistance of the low-side MOSFET as a current-sensing element. If the currentsense signal is above the current-limit threshold, the MAX1858A/MAX1875A/MAX1876A do not initiate a new cycle (Figure 6). Since valley current sensing is employed, the actual peak current is greater than the current-limit threshold by an amount equal to the inductor ripple current. Therefore, the exact current-limit characteristic and maximum load capability are a function of the low-side MOSFET's on-resistance, current-limit threshold, inductor value, and input voltage. The reward for this uncertainty is robust, lossless overcurrent sensing that does not require costly sense resistors. INPUT (VIN) VL BST_ Synchronous rectification reduces conduction losses in the rectifier by replacing the normal low-side Schottky catch diode with a low-resistance MOSFET switch. Additionally, the MAX1858A/MAX1875A/MAX1876A use the synchronous rectifier to ensure proper startup of the boost gate-driver circuit and to provide the current-limit signal. The internal pulldown transistor that drives DL_ low is robust, with a 0.5 (typ) on-resistance. This low onresistance helps prevent DL_ from being pulled up during the fast rise time of the LX_ node, due to capacitive coupling from the drain to the gate of the low-side synchronous-rectifier MOSFET. However, for high-current applications, some combinations of high- and low-side FETs can cause excessive gate-drain coupling, leading to poor efficiency, EMI, and shoot-through currents. This can be remedied by adding a resistor (typically less than 5) in series with BST_, which increases the turn-on time of the high-side FET without degrading the turn-off time (Figure 5). -IPEAK 4.7 ILOAD DH_ LX_ MAX1875A Figure 5. Reducing the Switching-Node Rise Time 12 INDUCTOR CURRENT MAX1858A/MAX1875A/MAX1876A Dual 180 Out-of-Phase Buck Controllers with Sequencing/Prebias Startup and POR ILIMIT 0 TIME Figure 6. "Valley" Current-Limit Threshold Point ______________________________________________________________________________________ Dual 180 Out-of-Phase Buck Controllers with Sequencing/Prebias Startup and POR Undervoltage Lockout and Startup IF VL drops below 4.2V, the MAX1858A/MAX1875A/ MAX1876A assume that the input supply and reference voltages are too low to make valid decisions and activate the undervoltage lockout (UVLO) circuitry, which latches DL and DH low to inhibit switching. RST is also forced low during UVLO. To reset the latch and be ready for the next VL rise, VL must be pulled below 2.5V. In addition, to ensure proper startup, the value of the capacitor at REF to GND must meet the following condition: CREF > ((8.29 x 10-4) / V+_SLOPE) - (1.97 x 10-1 / fS_MAX) where V+_SLOPE is the actual input-voltage rise time's slew rate. For example, if the switching frequency is set at 600kHz nominal, which is 660kHz (max), and the inputvoltage rise time's slew rate is 1.6V/mS, then C REF should be greater than 0.22F. Make sure CREF is chosen large enough to cover for worst-case capacitance tolerances and temperature coefficient. Enable (EN), Soft-Start, and Soft-Stop Pull EN high to enable or low to shut down both regulators. See the timing diagrams, Figures 3 and 4, for more detail. Output-Voltage Sequencing After the startup circuitry enables the controller, the MAX1858A begins the startup sequence. Regulator 1 (OUT1) powers up with soft-start enabled. Once the first converter's soft-start sequence ends, regulator 2 (OUT2) powers up with soft-start enabled. Finally, when both converters complete soft-start and both output voltages exceed 90% of their nominal values, the reset output (RST) goes high (see the Reset Output section). Soft-stop is initiated by pulling EN low. Soft-stop occurs in reverse order of soft-start, allowing last-on/first-off operation. Reset Output (RST) (MAX1858A/ MAX1876A Only) RST is an open-drain output. RST pulls low when either output falls below 90% of its nominal regulation voltage. Once both outputs exceed 90% of their nominal regulation voltages and both soft-start cycles are completed, RST goes high impedance. To obtain a logic-voltage output, connect a pullup resistor from RST to the logic supply voltage. A 100k resistor works well for most applications. If unused, leave RST grounded or unconnected. Clock Synchronization (SYNC, CKO) SYNC serves two functions: SYNC selects the clock output (CKO) type used to synchronize slave controllers, or it serves as a clock input so the MAX1858A/MAX1875A/ MAX1876A can be synchronized with an external clock signal. This allows the MAX1858A/MAX1875A/MAX1876A to function as either a master or slave. CKO provides a clock signal synchronized to the MAX1858A/MAX1875A/ MAX1876As' switching frequency, allowing either inphase (SYNC = GND) or 90 out-of-phase (SYNC = VL) synchronization of additional DC-DC controllers (Figure 7). The MAX1858A/MAX1875A/MAX1876A support the following three operating modes: * SYNC = GND: The CKO output frequency equals REG1's switching frequency (fCKO = fDH1) and the CKO signal is in phase with REG1's switching frequency. This provides 2-phase operation when synchronized with a second slave controller. * SYNC = VL: The CKO output frequency equals two times REG1's switching frequency (fCKO = 2fDH1) and the CKO signal is phase shifted by 90 with respect to REG1's switching frequency. This provides 4-phase operation when synchronized with a second MAX1858A/MAX1875A/MAX1876A (slave controller). ______________________________________________________________________________________ 13 MAX1858A/MAX1875A/MAX1876A The adjustable current limit accommodates MOSFETs with a wide range of on-resistance characteristics (see the Design Procedure section). The current-limit threshold is adjusted with an external resistor at ILIM_ (Figure 1). The adjustment range is from 50mV to 300mV, corresponding to resistor values of 100k to 600k. In adjustable mode, the current-limit threshold across the low-side MOSFET is precisely 1/10th the voltage seen at ILIM_. However, the current-limit threshold defaults to 100mV when ILIM is tied to VL. The logic threshold for switchover to this 100mV default value is approximately VL - 0.5V. Adjustable foldback current limit reduces power dissipation during short-circuit conditions (see the Design Procedure section). Carefully observe the PC board layout guidelines to ensure that noise and DC errors do not corrupt the current-sense signals seen by LX_ and PGND. The IC must be mounted close to the low-side MOSFET with short, direct traces making a Kelvin-sense connection so that trace resistance does not add to the intended sense resistance of the low-side MOSFET. MAX1858A/MAX1875A/MAX1876A Dual 180 Out-of-Phase Buck Controllers with Sequencing/Prebias Startup and POR * SYNC Driven by External Oscillator: The controller generates the clock signal by dividing down the SYNC input signal, so that the switching frequency equals half the synchronization frequency (fSW = fSYNC/2). REG1's conversion cycles initiate on the rising edge of the internal clock signal. The CKO output frequency and phase match REG1's switching frequency (f CKO = f DH1 ) and the CKO signal is in phase. Note that the MAX1858A/MAX1875A/ MAX1876A still require ROSC when SYNC is externally clocked and the internal oscillator frequency should be set to 50% of the synchronization frequency (fSW = 0.5 fSYNC). Thermal Overload Protection Thermal overload protection limits total power dissipation in the MAX1858A/MAX1875A/MAX1876A. When the device's die-junction temperature exceeds TJ = +160C, an on-chip thermal sensor shuts down the device, forcing DL_ and DH_ low, allowing the IC to cool. The thermal sensor turns the part on again after the junction temperature cools by 10C. During thermal shutdown, the regulators shut down, RST goes low, and soft-start is reset. If the VL linear-regulator output is short circuited, thermaloverload protection is triggered. Design Procedure Effective Input Voltage Range Although the MAX1858A/MAX1875A/MAX1876A controllers can operate from input supplies ranging from 4.5V to 23V, the input voltage range can be effectively limited by the MAX1858A/MAX1875A/MAX1876As' duty-cycle limitations. The maximum input voltage is limited by the minimum on-time (tON(MIN)): VIN(MAX) V + VDROP1 VIN(MIN) = OUT + VDROP2 - VDROP1 1- fSW t OFF(MIN) where VDROP1 is the sum of the parasitic voltage drops in the inductor discharge path, including synchronous rectifier, inductor, and PC board resistances. VDROP2 is the sum of the resistances in the charging path, including high-side switch, inductor, and PC board resistances. MAX1858A MAX1875A MAX1876A MAX1858A MAX1875A MAX1876A CK0 SYNC CK0 OSC SYNC SYNC VL VL MASTER SLAVE 180 PHASE SHIFT OSC SLAVE 4-OUTPUT APPLICATION 90 PHASE SHIFT DH1 MASTER SYNC MASTER 3-OUTPUT APPLICATION DH1 MASTER DH2 SLAVE t ON(MIN)fSW where tON(MIN) is 100ns. The minimum input voltage is limited by the switching frequency and minimum offtime, which determine the maximum duty cycle (DMAX = 1 - fSWtOFF(MIN)): MAX1858A MAX1875A MAX1876A OSC VOUT DH DH2 DH1 SLAVE DH2 Figure 7. Synchronized Controllers 14 ______________________________________________________________________________________ Dual 180 Out-of-Phase Buck Controllers with Sequencing/Prebias Startup and POR V R _ A = R _ B OUT -1 VSET where VSET = 1V (see the Electrical Characteristics) and VOUT can range from VSET to 18V. For output voltages below 1V, set the MAX1858A/ MAX1875A/MAX1876A output voltage by connecting a voltage-divider from the output to FB_ to REF (Figure 8). Select R_C (FB to REF resistor) in the 1k to 10k range. Calculate R_A with the following equation: V V R _ A = R _ C SET- OUT V V REF- SET where V SET = 1V, V REF = 2V (see the Electrical Characteristics), and VOUT can range from 0 to VSET. Setting the Switching Frequency The controller generates the clock signal by dividing down the internal oscillator or SYNC input signal when driven by an external oscillator, so the switching frequency equals half the oscillator frequency (fSW = fOSC/2). The internal oscillator frequency is set by a resistor (ROSC) connected from OSC to GND. The relationship between fSW and ROSC is: 6 x 109 ( - Hz) ROSC = fSW where fSW is in Hz and ROSC is in . For example, a 600kHz switching frequency is set with ROSC = 10k. Higher frequencies allow designs with lower inductor values and less output capacitance. Consequently, peak currents and I 2 R losses are lower at higher switching frequencies, but core losses, gate-charge currents, and switching losses increase. A rising clock edge on SYNC is interpreted as a synchronization input. If the SYNC signal is lost, the internal oscillator takes control of the switching rate, returning the switching frequency to that set by ROSC. This maintains output regulation even with intermittent SYNC signals. When an external synchronization signal is used, ROSC should set the switching frequency to one-half SYNC rate (fSYNC). OUT_ REF R_A R_C FB_ FB_ MAX1858A MAX1875A MAX1876A R_B MAX1858A MAX1875A MAX1876A R_A OUT_ VOUT_ < 1V VOUT_ > 1V Figure 8. Adjustable Output Voltage Inductor Selection Three key inductor parameters must be specified for operation with the MAX1858A/MAX1875A/MAX1876A: inductance value (L), peak-inductor current (IPEAK), and DC resistance (RDC). The following equation assumes a constant ratio of inductor peak-to-peak AC current to DC average current (LIR). For LIR values too high, the RMS currents are high, and therefore I2R losses are high. Large inductances must be used to achieve very low LIR values. Typically, inductance is proportional to resistance (for a given package type), which again makes I2R losses high for very low LIR values. A good compromise between size and loss is a 30% peak-to-peak ripple current to average-current ratio (LIR = 0.3). The switching frequency, input voltage, output voltage, and selected LIR determine the inductor value as follows: V (V - V ) L = OUT IN OUT VINfSWIOUTLIR where VIN, VOUT, and IOUT are typical values (so that efficiency is optimum for typical conditions). The switching frequency is set by R OSC (see the Setting the Switching Frequency section). The exact inductor value is not critical and can be adjusted in order to make trade-offs among size, cost, and efficiency. Lower inductor values minimize size and cost, but also improve transient response and reduce efficiency due to higher peak currents. On the other hand, higher inductance increases efficiency by reducing the RMS current. However, resistive losses due to extra wire turns can exceed the benefit gained from lower AC current levels, especially when the inductance is increased without also allowing larger inductor dimensions. Find a low-loss inductor having the lowest possible DC resistance that fits in the allotted dimensions. The ______________________________________________________________________________________ 15 MAX1858A/MAX1875A/MAX1876A Setting the Output Voltage For 1V or greater output voltages, set the MAX1858A/ MAX1875A/MAX1876A output voltage by connecting a voltage-divider from the output to FB_ to GND (Figure 8). Select R_B (FB_ to GND resistor) to between 1k and 10k. Calculate R_A (OUT_ to FB_ resistor) with the following equation: MAX1858A/MAX1875A/MAX1876A Dual 180 Out-of-Phase Buck Controllers with Sequencing/Prebias Startup and POR inductor's saturation rating must exceed the peakinductor current at the maximum defined load current (ILOAD(MAX)): LIR IPEAK = ILOAD(MAX ) + I 2 LOAD(MAX ) RFBI = PFB x VOUT 5 x 10-6 (1- PFB ) and RILIM = 10 x VITH (1- PFB ) x RFBI V [ OUT -10 x VITH(1- PFB )] Setting the Valley Current Limit The minimum current-limit threshold must be high enough to support the maximum expected load current with the worst-case low-side MOSFET on-resistance value since the low-side MOSFET's on-resistance is used as the current-sense element. The inductor's valley current occurs at ILOAD(MAX) minus half of the ripple current. The current-sense threshold voltage (V ITH) should be greater than voltage on the low-side MOSFET during the ripple-current valley: LIR VITH > RDS(ONMAX , ) x ILOAD(MAX) x 1 2 where R DS(ON) is the on-resistance of the low-side MOSFET (N L). Use the maximum value for R DS(ON) from the low-side MOSFET's data sheet, and additional margin to account for RDS(ON) rise with temperature is also recommended. A good general rule is to allow 0.5% additional resistance for each C of the MOSFET junction temperature rise. Connect ILIM_ to VL for the default 100mV (typ) current-limit threshold. For an adjustable threshold, connect a resistor (R ILIM _) from ILIM_ to GND. The relationship between the current-limit threshold (VITH_) and RILIM_ is: RILIM _ = VITH _ 0.5A where RILIM_ is in and VITH_ is in V. An RILIM resistance range of 100k to 600k corresponds to a current-limit threshold of 50mV to 300mV. When adjusting the current limit, 1% tolerance resistors minimize error in the current-limit threshold. For foldback current limit, a resistor (RFBI) is added from ILIM pin to output. The value of RILIM and RFBI can then be calculated as follows: First select the percentage of foldback, PFB, from 15% to 30%, then: 16 If RILIM_ results in a negative number, select a low-side MOSFET with lower RDS(ON) or increase PFB_ or a combination of both for the best compromise of cost, efficiency, and lower power dissipation during short circuit. Input Capacitor The input filter capacitor reduces peak currents drawn from the power source and reduces noise and voltage ripple on the input caused by the circuit's switching. The input capacitor must meet the ripple current requirement (IRMS) imposed by the switching currents as defined by the following equation: IRMS = ILOAD VOUT (VIN - VOUT ) VIN IRMS has a maximum value when the input voltage equals twice the output voltage (VIN = 2VOUT), so IRMS(MAX) = ILOAD / 2. For most applications, nontantalum capacitors (ceramic, aluminum, polymer, or OS-CON) are preferred at the input due to their robustness with high inrush currents typical of systems that can be powered from very low impedance sources. Additionally, two (or more) smaller-value low-ESR capacitors can be connected in parallel for lower cost. Choose an input capacitor that exhibits less than +10C temperature rise at the RMS input current for optimal long-term reliability. Output Capacitor The key selection parameters for the output capacitor are capacitance value, ESR, and voltage rating. These parameters affect the overall stability, output ripple voltage, and transient response. The output ripple has two components: variations in the charge stored in the output capacitor, and the voltage drop across the capacitor's ESR caused by the current flowing into and out of the capacitor: VRIPPLE VRIPPLE(ESR) + VRIPPLE(C) ______________________________________________________________________________________ Dual 180 Out-of-Phase Buck Controllers with Sequencing/Prebias Startup and POR VRIPPLE(ESR) = IP-PRESR VRIPPLE(C) = To determine the loop gain (AL), consider the gain from FB to COMP (ACOMP/FB), from COMP to LX (ALX/COMP), and from LX to FB (AFB/LX). The total loop gain is: AL = A COMP / FB x ALX / COMP x AFB / LX IP-P 8COUT fSW where: V -V V IP-P = IN OUT OUT fSWL VIN where IP-P is the peak-to-peak inductor current (see the Inductor Selection section). These equations are suitable for initial capacitor selection, but final values should be verified by testing in a prototype or evaluation circuit. As a general rule, a smaller inductor ripple current results in less output ripple voltage. Since inductor ripple current depends on the inductor value and input voltage, the output ripple voltage decreases with larger inductance and increases with higher input voltages. However, the inductor ripple current also impacts transient-response performance, especially at low VIN - VOUT differentials. Low inductor values allow the inductor current to slew faster, replenishing charge removed from the output filter capacitors by a sudden load step. The amount of output-voltage sag is also a function of the maximum duty factor, which can be calculated from the minimum off-time and switching frequency: gM _ COMP V x ACOMP / FB = COMP VFB SCCOMP 1 + sRCOMPCCOMP _ A 1 + sRCOMPCCOMP _ B assuming an ideal integrator, and assuming that CCOMP_B is much less than CCOMP_A: ALX / COMP = VLX VIN = VCOMP VRAMP where VRAMP = 1VP-P: 1+ sRESRCOUT V V AFB / LX = FB = SET 2 VLX VOUT S LCOUT + SRESRCOUT + 1 1+ SRESRCOUT V SET VOUT VOUT S2LCOUT + 1 Therefore: V L(ILOAD1 - ILOAD2 )2 OUT + t OFF(MIN) VINfSW VSAG = VIN - VOUT 2COUT VOUT - t OFF(MIN) VINfSW where t OFF(MIN) is the minimum off-time (see the Electrical Characteristics), and fSW is set by ROSC (see the Setting the Switching Frequency section). Compensation Each voltage-mode controller section employs a transconductance error amplifier whose output is the compensation point of the control loop. The control loop is shown in Figure 9. For frequencies much lower than Nyquist, the PWM block can be simplified to a voltage amplifier. Connect RCOMP_ and CCOMP_A from COMP to GND to compensate the loop (Figure 9). The inductor, output capacitor, compensation resistor, and compensation capacitors determine the loop stability. Since the inductor and output capacitor are chosen based on performance, size, and cost, select the compensation resistor and capacitors to optimize control-loop stability. AL gM _ COMP SCCOMP _ A x 1 + SRCOMPCCOMP _ A 1 + SRCOMPCCOMP _ B x VIN VRAMP V 1 + SRESR COUT x SET x VOUT S2 LC +1 OUT For an ideal integrator, this loop gain approaches infinity at DC. In reality the gM amplifier has a finite output impedance, which imposes a finite, but large, loop gain. It is this large loop gain that provides DC load accuracy. The dominant pole occurs due to the integrator, and for this analysis, it can be approximated to occur at DC. RCOMP creates a zero at: fZ _ COMP _ A = 1 2 x RCOMP _ CCOMP _ A The inductor and capacitor form a double pole at: fLC = 1 2 x LCOUT ______________________________________________________________________________________ 17 MAX1858A/MAX1875A/MAX1876A The output voltage ripple as a consequence of the ESR and output capacitance is: MAX1858A/MAX1875A/MAX1876A Dual 180 Out-of-Phase Buck Controllers with Sequencing/Prebias Startup and POR At some higher frequency, the output capacitor's impedance becomes insignificant compared to its ESR, and the LC system becomes more like an LR system, turning a double pole into a single pole. This zero occurs at: fESR = 1 The following loop-gain equation can be found by using these previous approximations with Figure 9: AL Setting the loop gain to 1 and solving for the crossover frequency yields: 2 x RESR COUT A final pole is added using CCOMP_B to reduce the gain and attenuate noise after crossover. This pole (fCOMP_B) occurs at: fCOMP _ B = V x SET VRAMP VOUT gM _ COMP x RCOMP x RESR x 2 x L fCO = GBW = 1 2 x RCOMPCCOMP _ B Figure 10 shows a Bode plot of the poles and zeros in their relative locations. Near crossover, the following approximations can be made to simplify the loop-gain equation: * RCOMP has much higher impedance than CCOMP. This is true if, and only if, crossover occurs above fZ_COMP_A. If this is true, CCOMP_A can be ignored (as a short to ground). * RESR is much higher impedance than COUT. This is true if, and only if, crossover occurs well after the output capacitor's ESR zero. If this is true, C OUT becomes an insignificant part of the loop gain and can be ignored (as a short to ground). * CCOMP_B is much higher impedance than RCOMP and can be ignored (as an open circuit). This is true if, and only if, crossover occurs far below fCOMP_B. P DH DL To ensure stability, select RCOMP to meet the following criteria: * Unity-gain crossover must occur below 1/5th of the switching frequency. * For reasonable phase margin using type 1 compensation, fCO must be larger than 5 fESR. Choose CCOMP_A so that fZ_COMP_A equals half fLC using the following equation: CCOMP _ A = CCOMP _ B = 1 ( ) 2 x 3 x fCO x RCOMP L VOUT LX N FB LX = COUT RESR FB COMP_ gM_COMP gM_COMP VSET RCOMP_ RCOMP Choose CCOMP_B so that fCOMP_B occurs at 3 times fCO using the following equation: RESR COMP_ 2 x LCOUT GAIN = +VIN/VRAMP L M VIN N W VC gM _ COMP x RCOMP x RESR VIN V x SET x VRAMP VOUT sL CCOMP_A CCOMP_A VSET RCOMP_ CCOMP_B CCOMP_B Figure 9. Fixed-Frequency Voltage-Mode Control Loop 18 ______________________________________________________________________________________ COUT Dual 180 Out-of-Phase Buck Controllers with Sequencing/Prebias Startup and POR All four N-channel MOSFETs must be a logic-level type with guaranteed on-resistance specifications at VGS 4.5V. For maximum efficiency, choose a high-side MOSFET (NH_) that has conduction losses equal to the switching losses at the optimum input voltage. Check to ensure that the conduction losses at minimum input voltage do not exceed MOSFET package thermal limits, or violate the overall thermal budget. Also, check to ensure that the conduction losses plus switching losses at the maximum input voltage do not exceed package ratings or violate the overall thermal budget. Ensure that the MAX1858A/MAX1875A/MAX1876A DL_ gate drivers can drive NL _. In particular, check that the dv/dt caused by NH _ turning on does not pull up the NL _ gate through NL _'s drain-to-gate capacitance. This is the most frequent cause of cross-conduction problems. Gate-charge losses are dissipated by the driver and do not heat the MOSFET. All MOSFETs must be selected so that their total gate charge is low enough that VL can power all four drivers without overheating the IC: PVL = VIN x QG _ TOTAL x fSW MOSFET package power dissipation often becomes a dominant design factor. I2R power losses are the greatest heat contributor for both high-side and low-side MOSFETs. I2R losses are distributed between NH_ and NL _ according to duty factor as shown in the equations below. Switching losses affect only the high-side MOSFET, since the low-side MOSFET is a zero-voltage switched device when used in the buck topology. Calculate MOSFET temperature rise according to package thermal-resistance specifications to ensure that both MOSFETs are within their maximum junction temperature at high ambient temperature. The worst-case dissipation for the high-side MOSFET (PNH) occurs at both extremes of input voltage, and the worst-case dissipation for the low-side MOSFET (PNL) occurs at maximum input voltage. 50 fLC 40 30 fESR 20 GAIN (dB) * Total gate charge (Qg) * Reverse transfer capacitance (CRSS) * Power dissipation BODE PLOT FOR VOLTAGEMODE CONTROLLERS 10 fZ-COMP_A fSWITCH fCO 0 -10 -20 fCOMP_B -30 -40 0.001 0.01 0.1 1 FREQUENCY (MHz) Figure 10. Voltage-Mode Loop Analysis Q + QGD PNH(SWITCHING) = VINILOADfSW GS IGATE IGATE is the average DH driver-output current capability determined by: IGATE = VL ( 2 RDS(ON)DH + RGATE + RGMOSFET ) where RDS(ON)DH is the high-side MOSFET driver's onresistance (5 max), RGATE is any series resistance between DH and BST (Figure 5), and RGMOSFET is the internal gate resistance of the external MOSFET: V PNH(CONDUCTION) = ILOAD2RDS(ON)NH OUT VIN PNH(TOTAL) = PNH(SWITCHING) + PNH(CONDUCTION) V PNL = ILOAD2RDS(ON)NL 1- OUT VIN where PNH(CONDUCTION) is the conduction power loss in the high-side MOSFET, and PNL is the total low-side power loss. To reduce EMI caused by switching noise, add a 0.1F ceramic capacitor from the high-side switch drain to the low-side switch source or add resistors in series with DL_ and DH_ to increase the MOSFETs' turn-on and turn-off times. ______________________________________________________________________________________ 19 MAX1858A/MAX1875A/MAX1876A MOSFET Selection The MAX1858A/MAX1875A/MAX1876As' step-down controller drives two external logic-level N-channel MOSFETs as the circuit switch elements. The key selection parameters are: * On-resistance (RDS(ON)) * Maximum drain-to-source voltage (VDS(MAX)) * Minimum threshold voltage (VTH(MIN)) MAX1858A/MAX1875A/MAX1876A Dual 180 Out-of-Phase Buck Controllers with Sequencing/Prebias Startup and POR Applications Information Dropout Performance When working with low input voltages, the output-voltage adjustable range for continuous-conduction operation is restricted by the minimum off-time (tOFF(MIN)). For best dropout performance, use the lowest (100kHz) switching-frequency setting. Manufacturing tolerances and internal propagation delays introduce an error to the switching frequency and minimum off-time specifications. This error is more significant at higher frequencies. Also, keep in mind that transient response performance of buck regulators operated close to dropout is poor, and bulk output capacitance must often be added (see the VSAG equation in the Design Procedure section). The absolute point of dropout is when the inductor current ramps down during the minimum off-time (IDOWN) as much as it ramps up during the maximum on-time (IUP). The ratio h = IUP/IDOWN is an indicator of the ability to slew the inductor current higher in response to increased load, and must always be greater than 1. As h approaches 1, the absolute minimum dropout point, the inductor current cannot increase as much during each switching cycle and V SAG greatly increases unless additional output capacitance is used. A reasonable minimum value for h is 1.5, but adjusting this up or down allows tradeoffs between VSAG, output capacitance, and minimum operating voltage. For a given value of h, the minimum operating voltage can be calculated as: V + VDROP1 VIN(MIN) = OUT + VDROP2 - VDROP1 1- hfSW t OFF(MIN) where VDROP1 is the sum of the parasitic voltage drops in the inductor discharge path, including synchronous rectifier, inductor, and PC board resistances; VDROP2 is the sum of the resistances in the charging path, including high-side switch, inductor, and PC board resistances; and t OFF(MIN) is from the Electrical Characteristics. The absolute minimum input voltage is calculated with h = 1. If the calculated V+(MIN) is greater than the required minimum input voltage, then reduce the operating frequency or add output capacitance to obtain an acceptable VSAG. If operation near dropout is anticipated, calculate V SAG to be sure of adequate transient response. 20 Dropout design example: VOUT = 5V fSW = 600kHz tOFF(MIN) = 250ns VDROP1 = VDROP2 = 100mV h = 1.5 5V + 100mV VIN(MIN) = 1- 1.5(600kHz)(250ns) +100mV - 100mV = 6.58V Calculating again with h = 1 gives the absolute limit of dropout: 5V + 100mV VIN(MIN) = ( )( ) 1 600 250 kHz ns +100mV - 100mV = 6V Therefore, VIN must be greater than 6V, even with very large output capacitance, and a practical input voltage with reasonable output capacitance would be 6.58V. Improving Noise Immunity Applications where the MAX1858A/MAX1875A/ MAX1876A must operate in noisy environments can typically adjust their controller's compensation to improve the system's noise immunity. In particular, high-frequency noise coupled into the feedback loop causes jittery duty cycles. One solution is to lower the crossover frequency (see the Compensation section). PC Board Layout Guidelines Careful PC board layout is critical to achieve low switching losses and clean, stable operation. This is especially true for dual converters where one channel can affect the other. Refer to the MAX1858 EV kit or MAX1875 EV kit data sheet for specific layout examples. If possible, mount all the power components on the top side of the board with their ground terminals flush against one another. Follow these guidelines for good PC board layout: * Isolate the power components on the top side from the analog components on the bottom side with a ground shield. Use a separate PGND plane under the OUT1 and OUT2 sides (referred to as PGND1 and PGND2). Avoid the introduction of AC currents into the PGND1 and PGND2 ground planes. Run the power plane ground currents on the top side only. ______________________________________________________________________________________ Dual 180 Out-of-Phase Buck Controllers with Sequencing/Prebias Startup and POR * Keep the high-current paths short, especially at the ground terminals. This practice is essential for stable, jitter-free operation. * Connect GND and PGND together close to the IC. Do not connect them together anywhere else. Carefully follow the grounding instructions under step 4 of the Layout Procedure section. * Keep the power traces and load connections short. This practice is essential for high efficiency. Use thick copper PC boards (2oz vs. 1oz) to enhance full-load efficiency by 1% or more. * LX_ and PGND connections to the synchronous rectifiers for current limiting must be made using Kelvinsense connections to guarantee the current-limit accuracy. With 8-pin SO MOSFETs, this is best done by routing power to the MOSFETs from outside using the top copper layer, while connecting PGND and LX_ underneath the 8-pin SO package. * When trade-offs in trace lengths must be made, allow the inductor-charging path to be made longer than the discharge path. Since the average input current is lower than the average output current in step-down converters, this minimizes the power dissipation and voltage drops caused by board resistance. For example, allow some extra distance between the input capacitors and the high-side MOSFET rather than to allow distance between the inductor and the low-side MOSFET or between the inductor and the output filter capacitor. * Ensure that the feedback connection to COUT_ is short and direct. * Route high-speed switching nodes (BST_, LX_, DH_, and DL_) away from the sensitive analog areas (REF, COMP_, ILIM_, and FB_). Use PGND1 and PGND2 as EMI shields to keep radiated noise away from the IC, feedback dividers, and analog bypass capacitors. * Make all pin-strap control input connections (ILIM_, SYNC, and EN) to analog ground (GND) rather than power ground (PGND). Layout Procedure 1) Place the power components first, with ground terminals adjacent (NL _ source, CIN_, and COUT_). Make all these connections on the top layer with wide, copper-filled areas (2oz copper recommended). 2) Mount the controller IC adjacent to the synchronousrectifier MOSFETs (NL _), preferably on the back side in order to keep LX_, PGND_, and DL_ traces short and wide. The DL_ gate trace must be short and wide, measuring 50mils to 100mils wide if the low-side MOSFET is 1in from the controller IC. 3) Group the gate-drive components (BST_ diodes and capacitors, and VL bypass capacitor) together near the controller IC. 4) Make the DC-DC controller ground connections as follows: create a small analog ground plane near the IC. Connect this plane to GND and use this plane for the ground connection for the reference (REF) V+ bypass capacitor, compensation components, feedback dividers, OSC resistor, and ILIM_ resistors (if any). Connect GND and PGND together under the IC (this is the only connection between GND and PGND). 5) On the board's top side (power planes), make a star ground to minimize crosstalk between the two sides. Chip Information TRANSISTOR COUNT: 6688 PROCESS: BiCMOS ______________________________________________________________________________________ 21 MAX1858A/MAX1875A/MAX1876A * Use a star-ground connection on the power plane to minimize the crosstalk between OUT1 and OUT2. Package Information (The package drawing(s) in this data sheet may not reflect the most current specifications. For the latest package outline information go to www.maxim-ic.com/packages.) QSOP.EPS MAX1858A/MAX1875A/MAX1876A Dual 180 Out-of-Phase Buck Controllers with Sequencing/Prebias Startup and POR PACKAGE OUTLINE, QSOP .150", .025" LEAD PITCH 21-0055 F 1 1 Maxim cannot assume responsibility for use of any circuitry other than circuitry entirely embodied in a Maxim product. No circuit patent licenses are implied. Maxim reserves the right to change the circuitry and specifications without notice at any time. 22 ____________________Maxim Integrated Products, 120 San Gabriel Drive, Sunnyvale, CA 94086 408-737-7600 (c) 2003 Maxim Integrated Products Printed USA is a registered trademark of Maxim Integrated Products.