19-3047; Rev 0; 10/03 KIT ATION EVALU E L B A AVAIL TFT-LCD Power-Supply Controllers Features The MAX1513/MAX1514 provide complete power-supply solutions for active-matrix thin-film transistor (TFT) liquid-crystal displays (LCDs). Both devices include a high-performance step-up regulator controller, three linear-regulator controllers, and an adjustable delay block for startup sequencing. The MAX1513 includes an additional linear-regulator controller and a high-performance buffer amplifier. The MAX1513/MAX1514 can operate from 2.7V to 5.5V input supplies and provide overload protection with timer delay latch on all the regulated outputs. The step-up regulator controller drives an external Nchannel MOSFET to generate the regulated supply voltage for the panel source-driver ICs. Its current-mode control architecture provides fast transient response to pulsed loads. The high switching frequency (up to 1.5MHz) allows the use of ultra-small inductors and ceramic capacitors while achieving efficiencies over 85% using lossless current sensing. The internal soft-start limits the input surge current during startup. ♦ 2.7V to 5.5V Input Supply Range ♦ Input-Supply Undervoltage Lockout ♦ Current-Mode Step-Up Controller Fast Transient Response to Pulsed Load High Efficiency Lossless Current Sensing 430kHz/750kHz/1.5MHz Switching Frequency ♦ Linear-Regulator Controllers for VGON, VGOFF ♦ Linear-Regulator Controller for Logic Supply ♦ High-Performance Buffer Amplifier (MAX1513 Only) ♦ Additional Linear-Regulator Controller (MAX1513 Only) ♦ Power-Up Sequence and VGON Delay Control ♦ VMAIN, VGON, VGOFF, VGAMMA Shutdown Control ♦ Timer-Delay Fault Latch for All Outputs ♦ Thermal-Overload Protection The gate-on and gate-off linear-regulator controllers of the MAX1513/MAX1514 provide regulated TFT gate-on and gate-off supplies. The gate-on supply is activated after an adjustable delay following the step-up regulator. The logic linear-regulator controller can be used to create a low-voltage logic supply. The gamma linear-regulator controller of the MAX1513 can be used to generate a gamma-correction reference supply or another generalpurpose supply rail. The MAX1513’s high-performance buffer amplifier can drive the LCD backplane (VCOM) or the gamma-correction divider string. The MAX1513/MAX1514 are available in 4mm ✕ 4mm 20-pin thin QFN packages with a maximum thickness of 0.8mm, suitable for ultra-thin LCD panel design. Ordering Information PART TEMP RANGE PIN-PACKAGE MAX1513ETP -40°C to +85°C 20 Thin QFN 4mm x 4mm MAX1514ETP -40°C to +85°C 20 Thin QFN 4mm x 4mm Minimal Operating Circuit VIN CS+ IN VMAIN CSSDFR GATE GND DEL FB Applications MAX1513 MAX1514 Notebook Computer Displays DRVL VLOGIC LCD Monitors and TVs FBL DRVP Automotive Displays FBP VMAIN VGON DRVG VGAMMA FBG SUPB DRVN Pin Configuration appears at end of data sheet. TO VCOM FBPB FBN OUTB REF VGOFF ________________________________________________________________ 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 MAX1513/MAX1514 General Description MAX1513/MAX1514 TFT-LCD Power-Supply Controllers ABSOLUTE MAXIMUM RATINGS FB, FBP, FBN, FBG, FBL, IN, CS+, CS-, SDFR to GND ...............................................-0.3V to +6V DEL, GATE, REF to GND .............................-0.3V to (VIN + 0.3V) SUPB to GND .........................................................-0.3V to +14V OUTB, FBPB to GND ..............................-0.3V to (VSUPB + 0.3V) DRVP, DRVG, DRVL to GND ..................................-0.3V to +30V DRVN to GND .....................................(VIN - 28V) to (VIN + 0.3V) OUTB Continuous Output Current ....................................±75mA Continuous Power Dissipation (TA = +70°C) 20-Pin TQFN (derate 16.9mW/°C above +70°C) .......1349mW Operating Temperature Range ...........................-40°C to +85°C Junction Temperature ......................................................+150°C Storage Temperature Range .............................-65°C to +150°C Lead Temperature (soldering, 10s) .................................+300°C 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 (Circuit of Figure 1, VIN = 3V, VSUPB = 10V, SDFR = IN, CREF = 0.22µF, TA = 0°C to +85°C. Typical values are at TA = +25°C, unless otherwise noted.) PARAMETER IN Supply Range IN Undervoltage-Lockout Threshold IN Quiescent Current SYMBOL CONDITIONS VIN VUVLO IIN MIN TYP 2.7 350mV typical hysteresis MAX UNITS 5.5 V VIN rising 2.5 2.7 2.9 VIN falling 2.2 2.35 2.5 V VFB = VFBP = VFBL = VFBG = 1.5V, VFBN = 0 1.25 mA IN Shutdown Current VSDFR = 0, VFBL = 1.5V 150 µA REF Output Voltage -2µA < IREF < 100µA, 2.7V < VIN < 5.5V 1.269 V 1.231 Temperature rising Thermal Shutdown 1.250 +160 Hysteresis °C 15 Duration to Trigger Fault Latch 43.6 ms MAIN STEP-UP CONTROLLER Operating Frequency fOSC SDFR = IN 1.275 1.500 1.725 SDFR = REF 0.60 0.75 0.90 MHz SDFR = unconnected Oscillator Maximum Duty Cycle FB Regulation Voltage VFB 0.43 80 85 90 % V VCS+ - VCS- = 0 1.237 1.25 1.263 FB Fault Trip Level VFB falling 0.96 1.00 1.04 FB Load Regulation 0 < (VCS+ - VCS-) < 50mV FB Line Regulation VIN = 2.7V to 5.5V FB Input Bias Current VFB = 1.5V CS+ Input Current 2.2V < VCS+ < 6V CS- Input Current 2.2V < VCS- < 6V Current-Limit Threshold VCS+ - VCS-, 2.2V < VCS+ < 6V Gate-Drive Output High or low Soft-Start Period -1 0.1 -100 -1 100 tSS Soft-Start Step Size V % 0.2 %/V +100 nA 90 µA +1 µA 125 150 mV 3 5 Ω 2.7 ms VREF / 128 V GATE-ON LINEAR-REGULATOR CONTROLLER (REG P) FBP Regulation Voltage IDRVP = 50µA 1.225 1.250 1.275 FBP Fault Trip Level VFBP falling 0.96 1.00 1.04 V FBP Input Bias Current VFBP = 1.5V -250 +250 nA 2 VFBP _______________________________________________________________________________________ V TFT-LCD Power-Supply Controllers (Circuit of Figure 1, VIN = 3V, VSUPB = 10V, SDFR = IN, CREF = 0.22µF, TA = 0°C to +85°C. Typical values are at TA = +25°C, unless otherwise noted.) PARAMETER SYMBOL FBP Effective Load-Regulation Error (Transconductance) DRVP Off-Leakage Current VFBP = 1.1V, VDRVP = 10V MAX UNITS -1.5 -2 % 8 1 VFBP = 1.5V, VDRVP = 28V DEL Charge Current During startup, VDEL = 1.0V VTH(DEL) DEL Discharge Switch On-Resistance Soft-Start Period TYP IDRVP = 50µA, 2.7V < VIN < 5.5V IDRVP DEL Turn-On Threshold MIN VDRVP = 10V, IDRVP = 25µA to 500µA FBP Line (IN)-Regulation Error DRVP Sink Current CONDITIONS mV mA 0.15 10 µA 4 5 6 µA 1.19 1.25 1.31 V VIN = 3.0V, VFB = 0.8V tSS Soft-Start Step Size 15 Ω 2.7 ms VREF / 128 V GAMMA LINEAR-REGULATOR CONTROLLER (REG G, MAX1513 ONLY) FBG Regulation Voltage VFBG IDRVG = 0.35mA 1.235 FBG to FB Regulation Voltage Matching IDRVG = 0.5mA, VCS+ - VCS- = 0 FBG Fault Trip Level VFBG falling 0.96 FBG Input Bias Current VFBG = 1.5V -250 FBG Effective Load-Regulation Error (Transconductance) VDRVG = 10V, IDRVG = 0.175mA to 3.5mA FBG Line (IN)-Regulation Error IDRVG = 0.5mA, 2.7V < VIN < 5.5V DRVG Sink Current IDRVG DRVG Off-Leakage Current Soft-Start Period VFBG = 1.1V, VDRVG = 10V 1.250 -1.2 1.00 -1.5 1.265 V +1.2 % 1.04 V +250 nA -2 % 5 mV 5 VFBG = 1.5V, VDRVG = 28V mA 0.15 tSS Soft-Start Step Size 10 µA 2.7 ms VREF / 128 V LOGIC LINEAR-REGULATOR CONTROLLER (REG L) FBL Regulation Voltage IDRVL = 0.8mA 1.225 1.250 1.275 FBL Fault Trip Level VFBL falling 0.96 1.00 1.04 V FBL Input Bias Current VFBL = 1.5V -250 +250 nA FBL Effective Load-Regulation Error (Transconductance) VDRVL = 3V, IDRVL = 0.4mA to 8mA -2 % FBL Line (IN)-Regulation Error IDRVL = 1mA, 2.7V < VIN < 5.5V 8 mV DRVL Sink Current VFBL IFBL DRVL Off-Leakage Current Soft-Start Period Soft-Start Step Size VFBL = 1.1V, VDRVL = VIN VFBL = 1.5V, VDRVL = 28V tSS -1.5 15 20 0.15 V mA 10 µA 2.7 ms VREF / 128 V _______________________________________________________________________________________ 3 MAX1513/MAX1514 ELECTRICAL CHARACTERISTICS (continued) MAX1513/MAX1514 TFT-LCD Power-Supply Controllers ELECTRICAL CHARACTERISTICS (continued) (Circuit of Figure 1, VIN = 3V, VSUPB = 10V, SDFR = IN, CREF = 0.22µF, TA = 0°C to +85°C. Typical values are at TA = +25°C, unless otherwise noted.) PARAMETER SYMBOL CONDITIONS MIN TYP MAX UNITS GATE-OFF LINEAR-REGULATOR CONTROLLER (REG N) FBN Regulation Voltage IDRVN = 0.2mA 220 250 280 mV FBN Fault Trip Level VFBN VFBN rising 380 420 460 mV FBN Input Bias Current VFBN = 0V -250 +250 nA FBN Effective Load-Regulation Error (Transconductance) VDRVN = -10V, IDRVN = 0.1mA to 2mA 25 mV FBN Line (IN)-Regulation Error IDRVN = 0.2mA, 2.7V < VIN < 5.5V 5 mV DRVN Source Current IFBN DRVN Off-Leakage Current VFBN = 0.3V, VDRVN = -10V 18 5 mA VFBN = -0.1V, VDRVN = -20V Soft-Start Period 0.1 tSS Soft-Start Step Size 10 µA 2.7 ms VREF / 128 V BUFFER AMPLIFIER SUPB Supply Range VSUPB SUPB Supply Current ISUPB 4.5 No load, VFBPB = 4V FBPB Input Offset Voltage VOS VFBPB = VSUPB / 2 FBPB Input Bias Current IBIAS VFBPB = VSUPB / 2 FBPB Input Common-Mode Range VCM Common-Mode Rejection Ratio CMRR 0 < VFBPB < VSUPB VOH IOUTB = 5mA Output-Voltage-Swing Low VOL IOUTB = -5mA Short-Circuit Current PSRR 0.75 1.1 mA 0 12 mV 50 nA VSUPB V DC, 6V ≤ VSUPB ≤ 13V, VFBPB = 4V 50 dB VSUPB 150 VSUPB 80 ±50 ±150 mA 60 80 dB mV 80 Slew Rate -3dB Bandwidth V 0 Output-Voltage-Swing High Power-Supply Rejection Ratio 13.0 RL = 10kΩ, CL = 10pF 150 mV 10 V/µs 12 MHz CONTROL INPUTS AND OUTPUTS SDFR Input Level SDFR = IN (1.5MHz operation) 0.9 × VIN SDFR = unconnected (430kHz operation) 0.69 × VIN SDFR = REF (750kHz operation) 0.77 × VIN 1.00 SDFR = GND (LCD shutdown) 4 V 0.5 SDFR = IN SDFR Input Current 1.35 +3.0 SDFR = REF -3.0 SDFR = GND -3.0 _______________________________________________________________________________________ µA TFT-LCD Power-Supply Controllers (Circuit of Figure 1, VIN = 3V, VSUPB = 10V, SDFR = IN, CREF = 0.22µF, TA = -40°C to +85°C, unless otherwise noted.) (Note 1) PARAMETER IN Supply Range SYMBOL CONDITIONS MAX UNITS 2.7 5.5 V VIN rising 2.5 2.9 VIN falling 2.2 2.5 VIN IN Undervoltage-Lockout Threshold IN Quiescent Current VUVLO IIN REF Output Voltage 350mV typical hysteresis MIN VFB = VFBP = VFBL = VFBG = 1.5V, VFBN = 0 TYP V 1.25 mA V -2µA < IREF < 100µA, 2.7V < VIN < 5.5V 1.225 1.275 SDFR = IN 1.275 1.725 SDFR = REF 0.60 0.90 VCS+ - VCS- = 0 1.230 1.270 V 0.2 %/V -100 +100 nA 90 µA MAIN STEP-UP CONTROLLER Operating Frequency fOSC FB Regulation Voltage VFB FB Line Regulation VIN = 2.7V to 5.5V FB Input Bias Current VFB = 1.5V CS+ Input Current 2.2V < VCS+ < 6V CS- Input Current 2.2V < VCS- < 6V Current-Limit Threshold VCS+ - VCS-, 2.2V < VCS+ < 6V Gate-Drive Output High or low MHz -1 +1 µA 100 150 mV 5 Ω GATE-ON LINEAR-REGULATOR CONTROLLER (REG P) FBP Regulation Voltage IDRVP = 0.1mA 1.225 1.275 V FBP Input Bias Current VFBP = 1.5V -250 +250 nA FBP Effective Load-Regulation Error (Transconductance) VDRVP = 10V, IDRVP = 0.05mA to 1mA -2 % DRVP Sink Current VFBP IDRVP DEL Turn-On Threshold VFBP = 1.1V, VDRVP = 10V VTH(DEL) 2 mA 1.19 1.31 V 1.235 1.265 V GAMMA LINEAR-REGULATOR CONTROLLER (REG G, MAX1513 ONLY) FBG Regulation Voltage VFBG IDRVG = 0.5mA FBG to FB Regulation Voltage Matching IDRVG = 0.5mA, VCS+ - VCS- = 0 -1.2 +1.2 % FBG Input Bias Current VFBG = 1.5V -250 +250 nA FBG Effective Load-Regulation Error (Transconductance) VDRVG = 10V, IDRVG = 0.25mA to 5mA -2 % DRVG Sink Current IDRVG VFBG = 1.1V, VDRVG = 10V 10 mA _______________________________________________________________________________________ 5 MAX1513/MAX1514 ELECTRICAL CHARACTERISTICS MAX1513/MAX1514 TFT-LCD Power-Supply Controllers ELECTRICAL CHARACTERISTICS (continued) (Circuit of Figure 1, VIN = 3V, VSUPB = 10V, SDFR = IN, CREF = 0.22µF, TA = -40°C to +85°C, unless otherwise noted.) (Note 1) PARAMETER SYMBOL CONDITIONS MIN TYP MAX UNITS LOGIC LINEAR-REGULATOR CONTROLLER (REG L) FBL Regulation Voltage VFBL IDRVL = 1mA 1.225 1.275 V FBL Input Bias Current VFBL = 1.5V -250 +250 nA FBL Effective Load-Regulation Error (Transconductance) VDRVL = 3V, IDRVL = 0.5mA to 10mA -2 % DRVL Sink Current IFBL VFBL = 1.1V, VDRVL = VIN 20 mA GATE-OFF LINEAR-REGULATOR CONTROLLER (REG N) FBN Regulation Voltage VFBN IDRVN = 0.2mA 220 280 mV FBN Input Bias Current VFBN = 0V -250 +250 nA FBN Effective Load-Regulation Error (Transconductance) VDRVN = -10V, IDRVN = 0.1mA to 2mA 25 mV DRVN Source Current IFBN VFBN = 0.3V, VDRVN = -10V 5 mA BUFFER AMPLIFIER SUPB Supply Range VSUPB SUPB Supply Current ISUPB 13.0 V No load, VFBPB = 4V 4.5 1.1 mA FBPB Input Offset Voltage VOS VFBPB = VSUPB / 2 12 mV FBPB Input Bias Current IBIAS VFBPB = VSUPB / 2 50 nA FBPB Input Common-Mode Range VCM VSUPB V Output-Voltage-Swing High VOH IOUTB = 5mA Output-Voltage-Swing Low VOL IOUTB = -5mA 0 VSUPB - 150 mV 150 mV CONTROL INPUTS AND OUTPUTS SDFR Input Level SDFR = IN (1.5MHz operation) 0.9 × VIN SDFR = unconnected (430kHz operation) 0.69 × VIN SDFR = REF (750kHz operation) 0.77 × VIN 1.00 SDFR = GND (LCD shutdown) +3.0 SDFR = REF -3.0 SDFR = GND -3.0 Note 1: Specifications to -40°C are guaranteed by design, not production tested. 6 V 0.5 SDFR = IN SDFR Input Current 1.35 _______________________________________________________________________________________ µA TFT-LCD Power-Supply Controllers OUTPUT VOLTAGE (V) 80 VIN = 2.7V VIN = 3.3V 70 60 VIN = 5V VIN = 15V fOSC = 1.5MHz 15.0 50 VMAIN 100mV AC-COUPLED 14.9 14.8 IMAIN 500mA/div 0.1A 14.7 IL 1A/div 0A 14.6 40 0 100 200 300 400 500 600 700 800 100 200 300 400 500 600 700 4µs/div LOAD CURRENT (mA) LOAD CURRENT (mA) STEP-UP REGULATOR PULSED-LOAD-TRANSIENT RESPONSE POWER-UP SEQUENCE STEP-UP REGULATOR SOFT-START MAX1513/14 toc04 MAX1513/14 toc06 MAX1513/14 toc05 VMAIN 100mV/div AC-COUPLED VMAIN 5V/div 0V 0V 0V VGOFF 10V/div IL 1A/div VGAMMA 20V/div IL 2A/div 0A 0A 4µs/div 400µs/div LINEAR REGULATOR REG L LOAD-TRANSIENT RESPONSE BUFFER-AMPLIFIER SUPPLY CURRENT vs. SUPPLY VOLTAGE MAX1513/14 toc07 0V 4ms/div 1.2 1.0 ILOGIC 500mA/div 0mA VMAIN 0V 20V/div VGON 20V/div IMAIN 1A/div 0.1A VLOGIC 50mV/div AC-COUPLED VLOGIC 5V/div 0V BUFFER-AMPLIFIER SMALL-SIGNAL STEP RESPONSE MAX1513/14 toc09 MAX1513/14 toc08 0 SUPPLY CURRENT (mA) EFFICIENCY (%) 90 MAX1513/14 toc03 15.1 MAX1513/14 toc02 VIN = 5.0V MAX1513/14 toc01 100 STEP-UP REGULATOR LOAD-TRANSIENT RESPONSE STEP-UP OUTPUT VOLTAGE vs. LOAD CURRENT STEP-UP EFFICIENCY vs. LOAD CURRENT (1.5MHz) VFBPB 50mV/div AC-COUPLED 0.8 0.6 0.4 VOUTB 50mV/div AC-COUPLED 0.2 NO LOAD VFBPB = VSUPB / 2 0 20µs/div 4 6 8 10 12 14 400ns/div SUPPLY VOLTAGE (V) _______________________________________________________________________________________ 7 MAX1513/MAX1514 Typical Operating Characteristics (Circuit of Figure 1, VIN = 5V, VMAIN = 15V, VGON = 25V, VGOFF = -10V, VLOGIC = 3.3V, VGAMMA = 14.7V, TA = +25°C, unless otherwise noted.) MAX1513/MAX1514 TFT-LCD Power-Supply Controllers Typical Operating Characteristics (continued) (Circuit of Figure 1, VIN = 5V, VMAIN = 15V, VGON = 25V, VGOFF = -10V, VLOGIC = 3.3V, VGAMMA = 14.7V, TA = +25°C, unless otherwise noted.) BUFFER-AMPLIFIER LOAD-TRANSIENT RESPONSE BUFFER-AMPLIFIER LARGE-SIGNAL STEP RESPONSE MAX1513/14 toc11 MAX1513/14 toc10 VOUTB 1V/div AC-COUPLED VFBPB 5V/div AC-COUPLED VOUTB 5V/div AC-COUPLED IOUTB 50mA/div 0mA 1µs/div 1µs/div Pin Description PIN 1 MAX1513 MAX1514 REF REF FUNCTION Internal Reference. Connect a 0.22µF ceramic capacitor from REF to the analog ground plane, which is connected to GND. External load capability is at least 100µA. 3 FBPB N.C. LCD Shutdown and Frequency-Select Input. SDFR = GND, LCD shutdown, REF, buffer amplifier and the logic regulator (REG L) output stay on SDFR = IN, 1.5MHz switching frequency SDRF = REF, 750kHz switching frequency SDFR = unconnected, 430kHz switching frequency Buffer-Amplifier Noninverting Input for the MAX1513. Not internally connected for the MAX1514. 4 OUTB N.C. Buffer-Amplifier Output for the MAX1513. Not internally connected for the MAX1514. 5 SUPB N.C. Buffer-Amplifier Supply Input for the MAX1513. Bypass to GND with a 0.1µF capacitor. Not internally connected for the MAX1514. 6 FBN FBN Gate-Off Linear Regulator (REG N) Feedback Input. FBN regulates to 125mV nominal. Connect to the center tap of a resistive voltage-divider between the REG N output and the reference voltage (REF) to set the output voltage. Place the resistive-divider close to this pin. 7 DEL DEL Delay-Control Timing Capacitor. Connect a capacitor from DEL to GND to set the gate-on linearregulator startup delay. See the Power-Up Sequence and Delay Control Block section. 2 8 NAME SDFR SDFR _______________________________________________________________________________________ TFT-LCD Power-Supply Controllers PIN NAME FUNCTION MAX1513 MAX1514 8 DRVN DRVN REG N Base Drive. Open drain of an internal P-channel MOSFET. Connect to the base of an external NPN linear-regulator pass transistor. 9 DRVL DRVL Logic Linear-Regulator (REG L) Base Drive. Open drain of an internal N-channel MOSFET. Connect to the base of an external PNP linear-regulator pass transistor. 10 FBL FBL REG L Feedback Input. FBL regulates to 1.25V (typ). Connect to the center tap of a resistive voltage-divider between the REG L output and the analog ground plane to set the output voltage. Place the resistive voltage-divider close to this pin. 11 DRVG N.C. Gamma Linear-Regulator (REG G) Base Drive for the MAX1513. Open drain of an internal N-channel MOSFET. Connect to the base of an external PNP linear-regulator pass transistor. Not internally connected for the MAX1514. 12 FBG N.C. REG G Feedback Input for MAX1513. FBG regulates to 1.25V (typ). Connect to the center tap of a resistive voltage-divider between the REG G output and the analog ground plane to set the output voltage. Place the divider close to the FBG pin. Not internally connected for the MAX1514. 13 FBP FBP Gate-On Linear-Regulator (REG P) Feedback Input. FBP regulates to 1.25V (typ). Connect to the center tap of a resistive voltage-divider between the REG P output and the analog ground plane to set the output voltage. Place the resistive-divider close to this pin. 14 DRVP DRVP REG P Base Drive. Open drain of an internal N-channel MOSFET. Connect to the base of an external PNP linear-regulator pass transistor. 15 GND GND Ground 16 GATE GATE External MOSFET Gate Drive. Drives the gate of the step-up switching regulator’s MOSFET. 17 IN IN Supply Input. IN powers all the internal circuitry of the MAX1513/MAX1514. The input voltage range is from 2.7V to 5.5V. Bypass with a 0.1µF ceramic capacitor between IN and GND. Place the capacitor within 5mm of IN. 18 CS+ CS+ Current-Sense-Comparator Noninverting Input. Connect CS+ and CS- to the lossless current-sense network. See the Lossless Current Sense section. 19 CS- CS- Current-Sense-Comparator Inverting Input. Connect CS+ and CS- to the lossless current-sense network. See the Lossless Current Sense section. 20 FB FB Main Step-Up Regulator Feedback Input. FB regulates to 1.25V (typ). Connect to the center tap of a resistive voltage-divider between the main output (VMAIN) and the analog ground plane to set the main step-up regulator output voltage. Place the resistive-divider close to this pin. _______________________________________________________________________________________ 9 MAX1513/MAX1514 Pin Description (continued) MAX1513/MAX1514 TFT-LCD Power-Supply Controllers VIN 4.5V TO 5.5V L1 2.2µH C1 22µF 6.3V D1 OPEN R1 110kΩ 17 CS+ 470pF 19 CS- IN GATE 2 SDFR GND 7 DEL 0.47µF FB 16 N1 R2 10.0kΩ 15 20 1.5kΩ 680Ω 9 Q1 VLOGIC 3.3V/500mA R7 16.5kΩ 10µF 10 DRVL DRVG 0.1µF 11 Q4 R8 10.0kΩ FBG R9 107kΩ 12 0.47µF R10 10.0kΩ LX 0.1µF SUPB 5 3.6kΩ 0.1µF 0.1µF FBPB 8 Q2 VGOFF -10V/30mA VGAMMA 14.7V/30mA FBL MAX1513 D2 1MΩ 0.1µF 18 2.2µF C2 10µF 16V 909Ω R11 10Ω C10 1µF VMAIN 15V/400mA LX R3 102kΩ 6 DRVN OUTB 3 4 TO VCOM BACKPLANE LX 0.1µF FBN 6.8kΩ D3 0.47µF R4 10.0kΩ 0.1µF 1 DRVP 14 REF 0.22µF FBP 13 Q3 VGON 25V/20mA R5 191kΩ R6 10.0kΩ 0.47µF Figure 1. Typical Operating Circuit of the MAX1513 10 ______________________________________________________________________________________ TFT-LCD Power-Supply Controllers VIN 4.5V TO 5.5V LX L1 2.2µH D1 C1 22µF 6.3V C2 10µF 16V 909Ω OPEN R11 10Ω R1 110kΩ 1MΩ 0.1µF 18 CS+ 17 470pF 19 CS- IN GATE C10 1µF 2 SDFR GND 7 DEL 0.47µF 2.2µF FB 16 N1 R2 10.0kΩ 15 20 LX 680Ω 0.1µF 9 Q1 VLOGIC 3.3V/500mA R7 16.5kΩ 10µF 10 6.8kΩ DRVL 0.1µF FBL DRVP R8 10.0kΩ MAX1514 LX 0.1µF D2 D3 FBP 14 13 Q3 R5 191kΩ R6 10.0kΩ VGON 25V/20mA 0.47µF 3.6kΩ 0.1µF 8 Q2 VGOFF -10V/30mA R3 102kΩ 0.47µF 6 DRVN FBN R4 10.0kΩ 1 REF 0.22µF Figure 2. Typical Operating Circuit of the MAX1514 ______________________________________________________________________________________ 11 MAX1513/MAX1514 VMAIN 15V/400mA MAX1513/MAX1514 TFT-LCD Power-Supply Controllers VIN CS+ MAX1513 MAX1514 SDFR CS- MAIN STEP-UP CONTROLLER WITH SOFT-START AND FAULT COMPARATOR DEL VMAIN GATE GND FB VIN DRVL REG L WITH SOFT-START AND FAULT COMPARATOR VLOGIC DRVP FBL REG P WITH SOFT-START AND FAULT COMPARATOR VGON FBP CONTROL BLOCK VMAIN DRVG VGAMMA FBG REG G WITH SOFT-START AND FAULT COMPARATOR DRVN REG N WITH SOFT-START AND FAULT COMPARATOR SUPB VGOFF FBN FBPB REFERENCE TO VCOM REF OUTB OP-AMP THERMAL SHUTDOWN MAX1513 ONLY Figure 3. MAX1513/MAX1514 Functional Diagram 12 ______________________________________________________________________________________ TFT-LCD Power-Supply Controllers Detailed Description The typical operating circuit of the MAX1513 (Figure 1) is a complete power-supply system for TFT LCDs. The circuit generates a +15V source-driver supply, +25V and -10V gate-driver supplies, a +3.3V logic supply for the timing controller, a 14.7V gamma-correction string supply and a VCOM buffer. The typical operating circuit of the MAX1514 (Figure 2) is similar to that of the MAX1513 except the gamma-correction string supply and the VCOM buffer have been eliminated. The input voltage range for the IC is from +2.7V to +5.5V. The typical operating circuits’ listed load currents are available from a +4.5V to +5.5V supply. Table 1 lists recommended component options, and Table 2 lists the component suppliers’ contact information. The MAX1513 and MAX1514 contain a high-performance, step-up switching-regulator controller and three linear-regulator controllers (two positive and one negative). The MAX1513 also includes an additional linear-regulator controller and a high-current buffer amplifier. Figure 3 shows the MAX1513/MAX1514 functional diagram. Table 1. Component List DESIGNATION DESCRIPTION C1 22µF ±20%, 6.3V X5R ceramic capacitor (1206) Taiyo Yuden JMK316BJ226ML C2 10µF ±20%, 16V POSCAP (D10) Sanyo 16AQU10M D1 1A, 30V Schottky diode (S-Flat) Toshiba CRS02 D2, D3 Main Step-Up Regulator Controller The main step-up regulator controller drives an external N-channel power MOSFET to generate the TFT-LCD source-driver supply. The controller employs a currentmode, fixed-frequency PWM architecture to maximize loop bandwidth and provide fast transient response to pulsed loads found in source-driver applications. The multilevel control input SDFR sets the switching frequency to 430kHz, 750kHz, or 1.5MHz. The high switching frequency allows the use of low-profile inductors and ceramic capacitors to minimize the thickness of LCD panel designs, while maintaining high efficiency using a lossless current-sense method. The IC’s built-in soft-start function reduces the inrush current during startup. The controller regulates the output voltage and the power delivered to the output by modulating the duty cycle (D) of the power MOSFET in each switching cycle. The duty cycle of the MOSFET is approximated by: 200mA, 100V diodes (SOT23) Fairchild MMBT4148SE L1 2.2µH, 3.3A inductor Sumida CLS7D16NP-2R2NC N1 3A, 20V N-channel MOSFET (SOT23) Fairchild FDN339AN Q1 3A, 60V PNP bipolar transistor (SOT23) Fairchild NZT660 Q2 200mA, 40V NPN bipolar transistor (SOT23) Fairchild MMBT3904 Q3, Q4 200mA, 40V PNP bipolar transistors (SOT23) Fairchild MMBT3906 D ≈ VMAIN - VIN VMAIN Figure 4 shows the functional diagram of the step-up regulator controller. The core of the controller is a multiinput summing comparator that sums three signals: the output-voltage error signal with respect to the reference voltage, the current-sense signal, and the slope-compensation ramp. On the rising edge of the internal clock, the controller sets a flip-flop, which turns on the external N-channel MOSFET, applying the input voltage across the inductor. The current through the inductor ramps up linearly, storing energy in its magnetic field. Once the sum of the feedback voltage error, slope compensation, and current-sense signals trip the multi- Table 2. Component Suppliers PHONE FAX Fairchild Semiconductor SUPPLIER 408-822-2000 408-822-2102 www.fairchildsemi.com WEBSITE Sumida 847-545-6700 847-545-6720 www.sumida.com Taiyo Yuden 800-348-2496 847-925-0899 www.t-yuden.com TDK 847-803-6100 847-390-4405 www.component.tdk.com Toshiba 949-455-2000 949-859-3963 www.toshiba.com ______________________________________________________________________________________ 13 MAX1513/MAX1514 Typical Operating Circuit MAX1513/MAX1514 TFT-LCD Power-Supply Controllers input PWM comparator, the flip-flop is reset and the MOSFET turns off. Since the inductor current is continuous, a transverse potential develops across the inductor that turns on the diode (D1). The voltage across the inductor then becomes the difference between the output voltage and the input voltage. This discharge condition forces the current through the inductor to ramp down, transferring the energy stored in the magnetic field to the output capacitor and the load. The N-channel MOSFET is kept off for the rest of the clock cycle. Current Limiting and Current-Sense Amplifier (CS+, CS-) The internal current-limit circuit resets the PWM flip-flop and turns off the external power MOSFET whenever the voltage difference between CS+ and CS- exceeds 125mV (typ). The tolerance on this current limit is ±20%. Use the minimum value of the current limit to select components of the current-sense network. Lossless Current Sense The lossless current-sense method uses the DC resistance (DCR) of the inductor as the sense element. Figure 5 shows a simplified step-up regulator using the basic lossless current-sensing method. An RC network is connected in parallel with the step-up inductor (L). The voltage across the sense capacitor (C S) is the RESET DOMINANT CLOCK S R GATE Q ILIM COMPARATOR 125mV input to the current-sense amplifier. To prevent the sense amplifier from seeing large common-mode switching voltages, the sense capacitor should always be connected to the nonswitching end of the inductor (i.e., the input of the step-up regulator). Lossless current sense can be easily understood using complex frequency domain analysis. The voltage across the inductor is given by: VL = IL (sL + RL ) where L is the inductance, RL is the DCR of the inductor, and IL is the inductor current. The voltage across the sense capacitor is given by: 1 VL 1 + sRSCS VS = where RS is the series resistor in the sense network and CS is the sense capacitor. The above equation can be rewritten as: VS = If sL + RL 1 + sL / RL IL = RLIL 1 + sRSCS 1 + sRSCS L = RSCS , then the equation becomes : RL VS = RLIL Therefore, the sense capacitor voltage is directly proportional to the inductor current if the time constant of the RC sense network matches the time constant of the inductor/DCR. The sense method is equivalent to using a current-sense resistor that has the same value as the inductor DCR. INDUCTOR RL CS+ LEVEL SHIFT L VIN VMAIN CS- RS CS Σ SLOPE_COMP + FB VS CS+ CSGATE TO FAULT LOGIC SOFT-START BLOCK REF 1.0V Figure 4. Step-Up Regulator-Controller Functional Diagram 14 MAX1513 MAX1514 GND FB Figure 5. Step-Up Regulator Using Lossless Current Sensing ______________________________________________________________________________________ TFT-LCD Power-Supply Controllers Gate-Off Linear-Regulator Controller The gate-off linear-regulator controller (REG N) is an analog gain block with an open-drain P-channel output. It drives an external NPN pass transistor with a 3.6kΩ base-to-emitter resistor (Figure 1). Its guaranteed basedrive source current is at least 2mA. The regulator, including Q2 in Figure 1, uses a 0.47µF ceramic output capacitor and is designed to deliver 30mA at -10V. Other output voltages and currents are possible by scaling the pass transistor, input capacitor, and output capacitor. See the Pass-Transistor Selection and Stability Requirements sections. REG N is typically used to provide the TFT-LCD gate drivers’ gate-off voltage. A negative voltage can be produced using a charge-pump circuit as shown in Figure 1. REG N is enabled after the logic linear-regulator REG L soft-start has completed. Each time it is enabled, the control goes through a soft-start routine that ramps down its internal reference DAC from VREF to 250mV in about 100 steps. Gate-On Linear-Regulator Controller The gate-on linear-regulator controller (REG P) is an analog gain block with an open-drain N-channel output. It drives an external PNP pass transistor with a 6.8kΩ base-to-emitter resistor (Figure 1). Its guaranteed basedrive sink current is at least 1mA. The regulator including Q3 in Figure 1 uses a 0.47µF ceramic output capacitor and is designed to deliver 20mA at 25V. Other output voltages and currents are possible by scaling the pass transistor, input capacitor, and output capacitor. See the Pass-Transistor Selection and Stability Requirements sections. REG P is typically used to provide the TFT-LCD gate drivers’ gate-on voltage. Use a charge pump with as many stages as necessary to obtain a voltage exceeding the required gate-on voltage (see the Selecting the Number of Charge-Pump Stages section). Note that the voltage rating of the DRVP output is 28V. If the chargepump output voltage can exceed 28V, an external cascode-connected NPN transistor should be added (Figure 6). Alternately, the linear regulator can control an intermediate charge-pump state while regulating the final charge-pump output (Figure 7). REG P is enabled after the step-up regulator soft-start has completed and the voltage on DEL exceeds 1.25V. Each time it is enabled, the controller goes through a soft-start routine that ramps up its internal reference DAC in 128 steps. LX VMAIN 0.1µF FROM CHARGEPUMP OUTPUT VMAIN 13V 0.1µF DRVP NPN CASCODE TRANSISTOR PNP PASS TRANSISTOR 6.8kΩ DRVP VGON MAX1513 MAX1514 MAX1513 MAX1514 FBP Q1 0.47µF VGON 35V 267kΩ 1% 0.22µF FBP 10.0kΩ 1% Figure 6. Using an NPN Cascode for Charge-Pump Output Voltages > 28V Figure 7. Linear Regulator Controls Intermediate Charge-Pump Stage ______________________________________________________________________________________ 15 MAX1513/MAX1514 Logic Linear-Regulator Controller The logic linear-regulator controller (REG L) is an analog gain block with an open-drain N-channel output. It drives an external PNP pass transistor with a 680Ω baseto-emitter resistor (Figure 1). Its guaranteed base-drive sink current is at least 10mA. The regulator, including transistor Q1 in Figure 1, uses a 10µF ceramic output capacitor and is designed to deliver 500mA at 3.3V. Other output voltages and currents are possible by scaling the pass transistor, input capacitor, and output capacitor. See the Pass-Transistor Selection and Stability Requirements sections. REG L is typically used to generate low-voltage logic supplies for the timing controller and the digital sections of the TFT-LCD source/gate-drive ICs. REG L is automatically enabled when the input voltage is above the UVLO threshold. Each time it is enabled, the controller goes through a soft-start routine that ramps up its internal reference DAC in 128 steps. MAX1513/MAX1514 TFT-LCD Power-Supply Controllers Gamma Linear-Regulator Controller (MAX1513 Only) The gamma linear-regulator controller REG G is an analog gain block with an open-drain N-channel output. It drives an external PNP pass transistor with a 1.5kΩ base-to-emitter resistor (Figure 1). Its guaranteed basedrive sink current is at least 5mA. The regulator, including Q4 in Figure 1, uses a 0.47µF ceramic output capacitor, and the controller is designed to deliver 40mA at 14.7V. Other output voltages and currents are possible by scaling the pass transistor, input capacitor, and output capacitor. See the Pass-Transistor Selection and Stability Requirements sections. REG G is typically used to provide the TFT-LCD gamma reference voltage, which is usually 0.3V below the source-drive supply voltage. REG G is enabled 2.7ms after REG P’s soft-start has completed. Each time it is enabled, the controller goes through a soft-start routine that ramps up its internal reference DAC in 128 steps. Buffer Amplifier (MAX1513 Only) The MAX1513 includes a buffer amplifier that is typically used to drive the LCD backplane (VCOM) or the gamma-correction divider string. The buffer amplifier features ±150mA output short-circuit current, 10V/µs slew rate, and 12MHz bandwidth. The Rail-to-Rail® input and output capability maximizes its flexibility. Short-Circuit Current Limit The MAX1513’s buffer amplifier limits short-circuit current to approximately ±150mA if the output is directly shorted to SUPB or to GND. If the short-circuit condition persists, the junction temperature of the IC rises until it reaches the thermal-shutdown threshold (+160°C typ). Once the junction temperature reaches the thermalshutdown threshold, an internal thermal sensor immediately sets the thermal fault latch, shutting off all the IC’s outputs. The device remains inactive until the input voltage is cycled below VUVLO. Driving Pure Capacitive Load The buffer amplifier is typically used to drive the LCD backplane (VCOM) or the gamma-correction divider string. The LCD backplane consists of a distributed series capacitance and resistance, a load that can be easily driven by the buffer amplifier. When driving a pure capacitive load, the amplifier’s gain peaking increases. A 5Ω to 50Ω resistor placed between OUTB and the capacitive load reduces peaking. Undervoltage Lockout The undervoltage-lockout (UVLO) circuit compares the voltage at the IN pin with the UVLO threshold (2.7V ris16 ing, 2.35V falling, typ) to ensure the input voltage is high enough for reliable operation. The 350mV (typ) hysteresis prevents supply transients from causing a restart. Once the input voltage exceeds the UVLO rising threshold, the IC is allowed to start. When the input voltage falls below the UVLO falling threshold, all the regulator outputs (including REF) are disabled until the input voltage exceeds the UVLO rising threshold. Reference Voltage (REF) The reference output is nominally 1.25V and can source at least 100µA without degrading its accuracy (see the Typical Operating Characteristics). Bypass REF with a 0.22µF ceramic capacitor connected between REF and the analog ground plane (which connects to GND). Shutdown and OscillatorFrequency Selection The four-level logic input SDFR controls shutdown and oscillator-frequency selection. Connecting SDFR to ground shuts off all the regulator outputs except the logic linear-regulator controller (REG L), buffer amplifier, and REF. Connecting SDFR to IN sets the oscillator frequency to 1.5MHz. Connecting SDFR to REF sets the oscillator frequency to 750kHz. Leaving SDFR unconnected sets the oscillator frequency to 430kHz. When SDFR is left unconnected, bypass the pin to ground with a 1000pF to 0.1µF capacitor to prevent switching noise from coupling into the pin’s high input impedance. Note the soft-start period and the fault-timer period do not change with the oscillator frequency. Power-Up Sequence and Delay Control Block Once the voltage on IN exceeds the UVLO rising threshold (2.7V typ), the internal reference is enabled. With a 0.22µF REF bypass capacitor, the reference reaches its regulation voltage of 1.25V in approximately 1ms. When the reference voltage is ready, the MAX1513/MAX1514 enable the logic linear regulator. The MAX1513 also enables the buffer amplifier at the same time. Once the logic linear-regulator soft-start is completed, the MAX1513/MAX1514 enable the step-up regulator and REG N simultaneously. Once the soft-start of the step-up regulator is completed, the MAX1513/MAX1514 enable the delay control block. An internal 5µA current starts charging the timing capacitor on DEL. When the voltage on DEL reaches 1.25V, the MAX1513/MAX1514 enable REG P. With a 0.1µF capacitor on DEL, the DEL voltage reaches 1.25V in about 25ms. The MAX1513 enables the gamma linear regulator 2.7ms after the soft-start of REG P is completed. Rail-to-Rail is a registered trademark of Nippon Motorola, Ltd. ______________________________________________________________________________________ TFT-LCD Power-Supply Controllers Fault Protection During steady-state operation, if the step-up regulator output or any of the linear-regulator outputs does not exceed its respective fault detection threshold, the MAX1513/MAX1514 activate an internal fault timer. If any condition or the combination of conditions indicates a continuous fault for the fault-timer duration (43.6ms typ), the MAX1513/MAX1514 set the fault latch, shutting down all the outputs except the reference. Once the fault condition is removed, toggle SDFR (below 0.4V) or cycle the input voltage (below 2.2V) to clear the fault latch and reactivate the device. The faultdetection circuit is disabled during the soft-start time of each regulator. Thermal-Overload Protection The thermal-overload protection prevents excessive power dissipation from overheating the MAX1513/ MAX1514. When the junction temperature exceeds +160°C, a thermal sensor immediately activates the fault-protection circuit, which shuts down all the outputs except the reference, allowing the device to cool down. Once the device cools down by approximately 15°C, cycle the input voltage (below the UVLO falling threshold) to clear the fault latch and reactivate the device. The thermal-overload protection protects the controller in the event of fault conditions. For continuous operation, do not exceed the absolute maximum junction temperature rating of TJ = +150°C. Design Procedure Main Step-Up Regulator Inductor Selection The minimum inductance value, peak current rating, and DC series resistance (DCR) are factors to consider when selecting the inductor. These factors influence the converter’s efficiency, maximum output load capability, transient-response time, and output voltage ripple. Size and cost are also important factors to consider. The maximum output current, input voltage, output voltage, and switching frequency determine the inductor value. Very high inductance values minimize the current ripple and therefore reduce the peak current, which decreases core losses in the inductor and I 2R losses in the entire power path. However, large inductor values also require more energy storage and more turns of wire, which increases size and can increase I 2 R losses in the inductor. Low inductance values decrease the size but increase the current ripple and peak current. Finding the best inductor involves choosing the best compromise between circuit efficiency, inductor size, and cost. The equations used here include a constant, LIR, which is the ratio of the inductor peak-to-peak ripple current to the average DC inductor current at the full load current. The best trade-off between inductor size and circuit efficiency for step-up regulators generally has an LIR between 0.3 and 0.5. However, depending on the AC characteristics of the inductor core material and the ratio of inductor resistance to other power-path resistances, the best LIR can shift up or down. If the inductor resistance is relatively high, more ripple can be accepted to reduce the number of turns required and increase the wire diameter. If the inductor resistance is relatively low, increasing inductance to lower the peak current can decrease losses throughout the power path. If extremely thin high-resistance inductors are used, as is common for LCD panel applications, the best LIR can increase to between 0.5 and 1.0. Once an inductor is chosen, higher and lower values for the inductor should be evaluated for efficiency improvements in typical operating regions. Determine the inductor value and peak current requirement as follows: Since the current delivered by charge pumps connected to LX adds to the inductor current, calculate the effective maximum output current, IMAIN(EFF): IMAIN(EFF) = IMAIN(MAX) + nNEG × INEG + (nPOS + 1) × IPOS where I MAIN(MAX) is the maximum output current including any gamma-regulator current, n NEG is the number of negative charge-pump stages, nPOS is the number of positive charge-pump stages, INEG is the negative charge-pump output current, and IPOS is the positive charge-pump output current, assuming the pump source for IPOS is VMAIN. Calculate the approximate inductor value using the typical input voltage (VIN), the expected efficiency (ηTYP) ______________________________________________________________________________________ 17 MAX1513/MAX1514 Soft-Start Each positive regulator (step-up regulator, REG P, REG L, and REG G) includes a 7-bit soft-start DAC whose input is the reference, and whose output is stepped in 128 steps from zero up to the reference voltage. The soft-start DAC of the negative regulator (REG N) steps from the reference down to 250mV in about 100 steps. The outputs of the soft-start DACs determine the set points of each regulator. The soft-start duration is 2.7ms (typ) for each positive regulator and about 2.2ms for the negative regulator. The soft-start is independent of the selected operating frequency. MAX1513/MAX1514 TFT-LCD Power-Supply Controllers taken from an appropriate curve in the Typical Operating Characteristics, and an estimate of LIR based on the above paragraphs: 2 ⎛ V ⎞ ⎛ VMAIN - VIN ⎞ ⎛ ηTYP ⎞ L = ⎜ IN ⎟ ⎜ ⎟ ⎟⎜ ⎝ VMAIN ⎠ ⎝ IMAIN(EFF) × fOSC ⎠ ⎝ LIR ⎠ Choose an available inductor value from an appropriate inductor family. Calculate the maximum DC input current at the minimum input voltage (VIN(MIN)) using the following equation: IIN(DC, MAX) = IMAIN(EFF) × VMAIN VIN(MIN) × ηMIN The expected efficiency at that operating point (ηMIN) can be taken from an appropriate curve in the Typical Operating Characteristics. Calculate the ripple current at that operating point and the peak current required for the inductor: IRIPPLE = ( VIN(MIN) × VMAIN - VIN(MIN) L × VMAIN × fOSC I IPEAK = IIN(DC, MAX) + RIPPLE 2 ) The inductor’s saturation current rating and the MAX1513/MAX1514s’ current limit (ILIM) should exceed I PEAK , and the inductor’s DC current rating should exceed IIN(DC, MAX). Considering the typical operating circuit, the maximum load current (IMAIN(MAX)) is 400mA for IMAIN directly and 30mA for REG G to provide V GAMMA. The onestage negative charge pump provides 30mA to REG N for VGOFF, and the one-stage positive charge pump provides 20mA to REG P for V GON . Altogether, the effective maximum output current (IMAIN(EFF)) is 500mA with a 15V output and a typical 5V input voltage. The switching frequency is set to 1.5MHz. Choosing an LIR of 0.6 and estimating efficiency of 85% at this operating point: 2 15V - 5V ⎞ ⎛ 0.85 ⎞ ⎛ 5V ⎞ ⎛ L = ⎜ ⎟ ≈ 2.2µH ⎟ ⎜ ⎟⎜ ⎝ 15V ⎠ ⎝ 0.5A × 1.5MHz ⎠ ⎝ 0.6 ⎠ Using the circuit’s minimum input voltage (4.5V) and estimating efficiency of 80% at that operating point: IIN(DC, MAX) = 18 0.5A × 15V ≈ 2.1A 4.5V × 0.8 The ripple current and the peak current are: IRIPPLE = 4.5V × (15V - 4.5V ) 2.2µH × 15V × 1.5MHz 1.0A ≈ 2.6A IPEAK = 2.1A + 2 ≈ 1.0A The inductor DCR should be low enough for reasonable efficiency. As a rule of thumb, do not allow the voltage drop across the inductor DCR to exceed a few percent of the input voltage at IPEAK. Many notebook panel designs have height constraints on the components. If a thin inductor with the required current rating is not available, use two thin inductors in series or parallel. Current-Sense Network Selection After selecting the inductor, use the following steps to design the current-sense network for lossless current sensing. 1) Calculate the RC time constant of the sense network using the typical inductance and typical DCR: τ = L RL(TYP) 2) Determine the component values of the sense network. Select CS, and then calculate RS using: RS = τ CS 3) Calculate the worst-case high sense voltage over temperature using the maximum DCR value (RL(MAX)) found in the inductor technical specifications: VSENSE = IPEAK × RL(MAX) × (1 + TC × ∆T) where IPEAK is the peak inductor current calculated in the Inductor Selection section, TC is the temperature coefficient of copper (0.5%/°C) and ∆T is the difference between the specified temperature for RL(MAX) and the maximum expected inductor temperature. 4) Compare the calculated sense voltage with the minimum value of the current-limit threshold in the Electrical Characteristics (100mV). If the sense voltage is between 80mV and 100mV, use the current-sense configuration in Figure 8 with the calculated CS and RS above. ______________________________________________________________________________________ TFT-LCD Power-Supply Controllers SF = 2.2µH = 91.7µs 24mΩ Select CS = 0.1µF and calculate RS: 100mV VSENSE 91.7µs = 917Ω 0.1µF RS = Assuming ∆T is 40°C and TC is 0.5%, the worst-case high sense voltage over temperature is: Calculate RS1 and RS2: VSENSE = 2.6A × 30mΩ (1 + 0.005 × 40°C) = 93.6mV RS SF R × SF RS2 = S1 1 - SF RS1 = 6) If VSENSE is less than 80mV, the current-feedback signal is low relative to the current-limit threshold. Use the Figure 8 configuration, or, if good current-limit accuracy is desired, use the optional current-sense configuration in Figure 10 to increase the amplitude of the sense signal. Calculate RS3 and RS4: RS3 = τ = VMAIN - VIN(MIN) VMAIN - VIN(MIN) - 100mV + VSENSE × RS Because VSENSE would be between 80mV and 100mV, the circuit in Figure 8 should be used. The closest 1% standard value for RS is 909Ω. If the 2.2µH inductor used in the typical operating circuit (Figures 1 and 2) has a typical DCR of 45mΩ and a maximum DCR of 56mΩ, the RC time constant of the sense network is: τ = Select CS = 0.1µF and calculate RS: RS4 = RS - RS3 If the 2.2µH inductor used in the typical operating circuit (Figures 1 and 2) had a typical DCR of 24mΩ and a maximum DCR of 30mΩ, the RC time constant of the sense network would be: 2.2µH = 48.9µs 45mΩ RS = 48.9µs = 489Ω 0.1µF Assuming ∆T is 40°C and TC is 0.5%, the worst-case high sense voltage over temperature is: L L VMAIN VIN VMAIN VIN RS1 RS RS2 CS CS CS+ CSGATE MAX1513 MAX1514 CS+ CS- GND FB GATE MAX1513 MAX1514 GND FB Figure 8. Lossless Current Sensing with 80mV < VSENSE < 100mV Figure 9. Lossless Current Sensing with VSENSE > 100mV ______________________________________________________________________________________ 19 MAX1513/MAX1514 5) If VSENSE is greater than 100mV, the current-feedback signal is too high and can trip the current limit before the full load current is delivered. Use the current-sense configuration in Figure 9 to attenuate the sense signal. Define the scale factor (SF) as: MAX1513/MAX1514 TFT-LCD Power-Supply Controllers Select CS = 0.1µF, RS is: L VMAIN VIN RS = RS3 CS 220µs = 2200Ω 0.1µF Assuming ∆T is 40°C and TC is 0.5%, the worst-case high sense voltage over temperature is: RS4 CS+ VSENSE = 2.6A × 14mΩ (1 + 0.005 × 40°C) = 44mV CSGATE MAX1513 MAX1514 Because VSENSE would be much less than 80mV, the circuit in Figure 10 can be used to improve the currentlimit accuracy. Calculate RS3 and RS4: GND FB RS3 = Figure 10. Lossless Current Sensing with VSENSE < 80mV 15V - 4.5V 15V - 4.5V - 0.1V + 0.044 V RS4 = 2614 Ω - 2600Ω = 14 Ω × 2600Ω = 2614 Ω The closest 1% standard values for RS3 and RS4 are 2.61kΩ and 14.0Ω, respectively. ( VSENSE = 2.6A × 56mΩ 1 + 0.005 × 40°C ) = 175mV Because VSENSE would be greater than 100mV, the circuit in Figure 9 should be used and the scale factor is: SF = 100mV = 0.571 175mV Calculate RS1 and RS2: 489Ω = 856Ω 0.571 856 × 0.571 RS2 = = 1139Ω 1 - 0.571 RS1 = The closest 1% standard values for RS1 and RS2 are 866Ω and 1.13kΩ, respectively. If the 2.2µH inductor used in the typical operating circuit (Figures 1 and 2) has a typical DCR of 10mΩ and a maximum DCR of 14mΩ, the RC time constant of the sense network is: τ = 20 2.2µH = 220µs 10mΩ Output-Capacitor Selection The output capacitor and its equivalent series resistance (ESR) affect the circuit’s stability, output voltage ripple, and transient response. The Output-Capacitor Stability Requirement section discusses the output capacitance requirement based on the loop stability. This section deals with how to determine the output capacitance according to the ripple voltage and load-transient requirements. The total output voltage ripple has two components: the ohmic ripple due to the capacitor’s equivalent series resistance (ESR), and the capacitive ripple caused by the charging and discharging of the output capacitance: VRIPPLE = VRIPPLE(ESR) + VRIPPLE(C) VRIPPLE(ESR) ≈ IPEAK × RESR VRIPPLE(C) ≈ ⎛ V IMAIN - VIN ⎞ × ⎜ MAIN ⎟ COUT ⎝ VMAIN × fOSC ⎠ where VMAIN is the output voltage of the step-up regulator, IMAIN is the output current, COUT is the output capacitance, RESR is the ESR of the output capacitor, fOSC is the switching frequency, and IPEAK is the peak inductor current (see the Inductor Selection section). ______________________________________________________________________________________ TFT-LCD Power-Supply Controllers RESR(MAX) ≤ VRIPPLE(MAX) 2 × IPEAK ⎛ V - VIN 2 × IMAIN COUT(MIN) ≥ × ⎜ MAIN VRIPPLE(MAX) ⎝ VMAIN × fOSC where VRIPPLE(MAX) is the total peak-to-peak output ripple. Since the peak inductor current calculated in the Inductor Selection section is 2.6A, the maximum ESR of the output capacitor should be less than 29mΩ. On the other hand, only 3.1µF capacitance is needed to meet the capacitive ripple requirement based on the calculation. A 10µF AQU-series POSCAP with maximum ESR of 20mΩ is selected for the typical operating circuits in Figures 1 and 2, which meets both the voltage-ripple and minimum capacitance requirements. The typical load on the step-up regulator for sourcedriver applications is a large pulsed load, with a peak current of approximately 1A and a pulse width of approximately 2µs. The shape of the pulse is close to triangular, so it is equivalent to a square pulse with 1A height and 1µs pulse width. The total voltage dip during the pulsed load transient also has two components: the ohmic dip due to the output capacitor’s ESR and the capacitive dip caused by discharging the output capacitance: VDIP = VDIP(ESR) + VDIP(C) VDIP(ESR) = IPULSE × RESR I × tPULSE VDIP(C) ≈ PULSE COUT where IPULSE is the height of the pulse load and tPULSE is the pulse width. Higher capacitance and lower ESR result in less voltage dip. Again, assume the ESR dip and the capacitive dip each should not exceed 50% of the total maximum allowed output-voltage dip caused by a load pulse (VDIP(MAX)). RESR(MAX) ≤ COUT(MIN) ≥ VDIP(MAX) 2 × IPULSE 2 × IPULSE × t PULSE VDIP(MAX) For the typical load pulse described above, assuming the voltage dip must be limited to 200mV, the minimum output capacitor is 10µF, and the maximum ESR is 100mΩ. The voltage rating and temperature characteristics of the output capacitor must also be considered. Input-Capacitor Selection The input capacitor (CIN) reduces the current peaks drawn from the input supply and reduces noise injection into the device. A 22µF ceramic capacitor is used in the typical operating circuit (Figure 1) because of the high source impedance seen in typical lab setups. Actual applications usually have much lower source impedance since the step-up regulator often runs directly from the output of another regulated supply. Typically, CIN can be reduced below the values used in the typical operating circuit. Ensure a low noise supply at IN by using adequate CIN. Alternately, greater voltage variation can be tolerated on CIN if IN is decoupled from CIN using an RC lowpass filter (see R11 and C10 in Figure 1). Rectifier Diode The MAX1513/MAX1514s’ high switching frequency demands a high-speed rectifier. Schottky diodes are recommended for most applications because of their fast recovery time and low forward voltage. In general, use a Schottky diode with a current rating exceeding the peak inductor current calculated in the Inductor Selection section. Output-Voltage Selection The output voltage of the main step-up regulator is adjustable by connecting a resistive voltage-divider from the output (VMAIN) to the analog ground plane with the center tap connected to FB (see Figure 1). Select R2 in the 10kΩ to 50kΩ range. Calculate R1 with the following equations: ⎛V ⎞ R1 = R2 × ⎜ MAIN - 1⎟ ⎝ VFB ⎠ where VFB, the step-up regulator’s feedback set point, is 1.25V. Connect the divider close to the IC. Output-Capacitor Stability Requirement The step-up regulator controller of the MAX1513/ MAX1514 uses a peak current-mode control method. The loop stability of a current-mode step-up regulator can be analyzed using a small-signal model. In continuous-conduction mode, the loop-gain transfer function consists of a DC loop gain, a dominant pole, a righthalf-plane (RHP) zero and an ESR zero. ______________________________________________________________________________________ 21 MAX1513/MAX1514 In the circuits of Figures 1 and 2, the maximum total voltage ripple is 1% (peak-to-peak) of the 15V output, which corresponds to 150mV peak-to-peak ripple. A conservative way to calculate the maximum ESR and minimum capacitance is to assume the ESR ripple and the capacitive ripple each should not exceed 50% of the total ripple budget. MAX1513/MAX1514 TFT-LCD Power-Supply Controllers The DC loop gain (ADC) is approximately: R2 1- D VMAIN ADC = × × R1 + R2 0.554 × RCS IMAIN(EFF) where R1 and R2 are the feedback-divider resistors (Figure 1), D is the duty cycle, IMAIN(EFF) is the effective maximum output current as described in the Inductor Selection section, 0.554 is the gain of the current-sense amplifier, and RCS is the equivalent senseresistor value given by: RCS = SF × RL(TYP) where RL(TYP) is the typical value of the inductor DCR, and SF is either 1 or the scale factor in step 5 of the Current-Sense Network Selection section. The frequency of the dominant pole is: fZ(RHP) = (1 - D) 2 × VMAIN 2π × L × IMAIN(EFF) The frequency of the ESR zero is: 1 fZ(ESR) = 2π × RESR × COUT The unity-gain crossover frequency is: fCROSSOVER = ADC × fP(DOMINANT) For stable operation, select an output capacitor with enough capacitance and a low enough ESR to ensure that the dominant pole is low enough so the loop gain reaches unity well before either the ESR zero or the RHP zero, the lower of which should preferably occur at or above 5 times the unity-gain frequency as long as the two zeros are well separated. Calculate the minimum output capacitance for stable operation using: COUT(MIN) = 5 × ADC × IMAIN(EFF) [ ] 2π × min fZ(RHP) , fZ(ESR) × VMAIN If the RHP zero and the ESR zero occur simultaneously, place the dominant pole so that the unity-gain frequency is less than 1/10th the frequency of the zeros. Calculate the minimum output capacitance for stable operation using: 22 2π × fZ × VMAIN Using the typical operating circuit in Figure 1 as an example: the duty cycle is 0.67, the effective maximum output current is 500mA, the inductor is 2.2µH with a typical DCR of 24mΩ, and the output capacitor is 10µF with a maximum ESR of 20mΩ. The scale factor for the current-sense network is 1, so RCS is 24mΩ. The DC loop gain ADC is 62, the RHP zero is at 236kHz, and the ESR zero is at 796kHz. Since the frequency of the ESR zero is higher than that of the RHP zero, the unity-gain crossover frequency should be determined based on the RHP zero. The minimum output capacitance for stable operation is: 2π × VMAIN × COUT The frequency of the RHP zero is: 10 × ADC × IMAIN(EFF) where f Z is the frequency of the RHP zero and the ESR zero. COUT(MIN) = IMAIN(EFF) fP(DOMINANT) = COUT(MIN) = 5 × 62 × 500mA ≈ 6.97µF 2π × 236kHz × 15V Lead or lag compensation can be useful to compensate for particular component choices or to optimize the transient response for various output capacitor or inductor values. Adding lead compensation (the R3/C1 network from VMAIN to FB in Figure 11) increases the loop bandwidth, which can increase the speed of response to transients. Too much speed can destabilize the loop and is not needed or recommended for Figure 1’s components. Lead compensation adds a zero-pole pair, providing gain at higher frequencies and increasing loop bandwidth. The frequencies of the zero and pole for lead compensation depend on the feedback-divider resistors and the RC network between VMAIN and FB. The frequencies of the zero and pole for the lead compensation are: fZ _ LEAD = 1 2π × (R1 + R3) × C1 fP _ LEAD = 1 R1 × R2 ⎞ ⎛ 2π × ⎜ R3 + ⎟ × C1 ⎝ R1 + R2 ⎠ At high frequencies, R3 is effectively in parallel with R1, determining the amount of added high-frequency gain. If R3 is very large, there is no added gain and as R3 approaches zero, the added gain approaches the inverse of the feedback-divider’s attenuation. A typical value for R3 is greater than 1/2 of R1. The value of C1 ______________________________________________________________________________________ TFT-LCD Power-Supply Controllers 1 R1 × R2 ⎞ ⎛ 2π × ⎜ R4 + ⎟ × C2 ⎝ R1 + R2 ⎠ 1 fZ _ LAG = 2π × R4 × C2 fP _ LAG = At high frequencies, R4 is effectively in parallel with R2, increasing the divider attenuation ratio. If R4 is very large, the attenuation ratio remains unchanged and as R4 approaches zero, the attenuation ratio approaches infinity. A typical value for R4 is greater than 0.1 times R2. If high-value divider-resistors are used, choose R4 < 1.5kΩ for FB noise immunity. The value of C2 deter- L mines the frequency placement of the pole and zero. A typical value of C2 is between 100pF and 1000pF. When adding lag compensation, always check the loop stability by monitoring the transient response to a pulsed output load. Using Lead Compensation to Reduce Startup Inrush Current The digital soft-start of the main step-up regulator limits the average input current during startup. If even smoother startup is needed, add a low-frequency leadcompensation network (Figure 12). The improved softstart is active only during soft-start when the output voltage rises. Positive changes in the output are instantaneously coupled to the FB pin through D1 and the feed-forward capacitor C1. This arrangement generates a smoothly rising output voltage. When the output voltage reaches regulation, capacitor C1 charges up through R3 and diode D1 turns off. If desired, C1 and R3 can be chosen to also provide some lead compensation in normal operation. In most applications, lead compensation in normal operation is not needed and can be avoided by making R3 large. With R3 much greater than R1, the pole and the zero in the compensation network are very close to one another after startup and cancel out, eliminating the effect of the lead compensation. With R2 at 10kΩ, an effective value for C1 is approximately 1000pF. Charge Pumps L D VIN VMAIN D VIN VMAIN R3 R3 D1 R1 R1 LX LX C1 C1 COUT FB COUT FB RLOAD R4 R2 MAX1513 MAX1514 C2 GND Figure 11. Feedback Compensation R2 MAX1513 MAX1514 GND Figure 12. Using Lead Compensation for Improved Soft-Start ______________________________________________________________________________________ 23 MAX1513/MAX1514 determines the frequency placement of the zero and pole. A typical value of C1 is between 100pF and 10nF. When adding lead compensation, always check the loop stability by monitoring the transient response to a pulsed output load. Adding lag compensation (the R4/C2 network from FB to ground in Figure 11) decreases the loop bandwidth and improves FB noise immunity. Lag compensation slows the transient response but can increase the stability margin, which may be needed for particular component choice or high values of FB-divider resistors. Lag compensation adds a pole-zero pair, attenuating gain at higher frequencies and lowering loop bandwidth. The frequencies of the pole and zero for lag compensation depend on the feedback-divider resistors and the RC network between FB and GND. The frequencies of the pole and zero for the lag compensation are: POSITIVE CHARGE-PUMP OUTPUT VOLTAGE vs. VMAIN NEGATIVE CHARGE-PUMP OUTPUT VOLTAGE vs. VMAIN -0 60 VD = 0.3V TO 1V 3-STAGE CHARGE-PUMP 1-STAGE CHARGE-PUMP -5 50 -10 -15 G_OFF (V) 40 G_ON (V) MAX1513/MAX1514 TFT-LCD Power-Supply Controllers 2-STAGE CHARGE-PUMP 30 -20 2-STAGE CHARGE-PUMP -25 -30 20 3-STAGE CHARGE-PUMP -35 10 -40 1-STAGE CHARGE-PUMP VD = 0.3V TO 1V -45 0 2 4 6 8 10 12 14 2 4 6 8 10 12 14 VMAIN (V) VMAIN (V) Figure 13. Positive Charge-Pump Output Voltage vs. VMAIN Figure 14. Negative Charge-Pump Output Voltage vs. VMAIN Selecting the Number of Charge-Pump Stages For highest efficiency, always choose the lowest number of charge-pump stages that meet the output voltage requirement. Figures 13 and 14 show the positive and negative charge-pump output voltages for a given VMAIN for one-, two-, and three-stage charge pumps. The number of positive charge-pump stages is given by: Sometimes fractional stages are more desirable for better efficiency. This can be done by connecting the first stage to VIN or another available supply. If the first charge-pump stage is powered from V IN , then the above equations become: nPOS = VGON + VDROPOUT - VMAIN VMAIN - 2 × VD where nPOS is the number of positive charge-pump stages, VGON is the gate-on linear-regulator REG P output, VMAIN is the main step-up regulator output, VD is the forward-voltage drop of the charge-pump diode, and VDROPOUT is the dropout margin for the linear regulator. Use VDROPOUT = 0.3V. The number of negative charge-pump stages is given by: nNEG = -VGOFF + VDROPOUT VMAIN - 2 × VD where nNEG is the number of negative charge-pump stages, VGOFF is the gate-off linear-regulator REG N output, VMAIN is the main step-up regulator output, VD is the forward-voltage drop of the charge-pump diode, and VDROPOUT is the dropout margin for the linear regulator. Use VDROPOUT = 0.3V. The above equations are derived based on the assumption that the first stage of the positive charge pump is connected to VMAIN and the first stage of the negative charge pump is connected to ground. 24 VGON + VDROPOUT - VIN VMAIN - 2 × VD -VGOFF + VDROPOUT + VIN nNEG = VMAIN - 2 × VD nPOS = Flying Capacitors Increasing the flying capacitor (CX) value lowers the effective source impedance and increases the outputcurrent capability. Increasing the capacitance indefinitely has a negligible effect on output-current capability because the switch resistance and the diode impedance place a lower limit on the source impedance. A 0.1µF ceramic capacitor works well in most low-current applications. The flying capacitor’s voltage rating must exceed the following: VCX > n × VMAIN where n is the stage number in which the flying capacitor appears, and VMAIN is the output voltage of the main step-up regulator. Charge-Pump Output Capacitor Increasing the output capacitance or decreasing the ESR reduces the output ripple voltage and the peak-topeak voltage during load transients. With ceramic ______________________________________________________________________________________ TFT-LCD Power-Supply Controllers COUT _ CP ≥ ILOAD _ CP 2fOSCVRIPPLE _ CP where COUT_CP is the output capacitor of the charge pump, I LOAD_CP is the load current of the charge pump, and VRIPPLE_CP is the peak-to-peak value of the output ripple. The charge-pump output capacitor is typically also the input capacitor for a linear regulator. Often, its value must be increased to maintain the linear regulator’s stability. Charge-Pump Rectifier Diodes Use low-cost silicon switching diodes with a current rating equal to or greater than twice the average charge-pump input current. If their low forward voltage helps to avoid an extra stage, some or all of the diodes can be replaced with Schottky diodes with equivalent current ratings. Linear-Regulator Controllers Output-Voltage Selection Adjust the positive linear-regulator (REG P, REG L, and REG G) output voltages by connecting a resistive voltage-divider from their respective outputs to the analog ground plane (which connects to GND) with the center tap connected to FB_ (Figure 1). Select the lower resistor of the divider in the range of 10kΩ to 30kΩ. Calculate the upper resistor with the following equation: ⎛ VOUT _ ⎞ RUPPER = RLOWER × ⎜ - 1⎟ ⎝ VFB _ ⎠ where VOUT_ is the output voltage of the respective linear regulator, and VFB_ = 1.25V (typ). Adjust the gate-off linear-regulator REG N output voltage by connecting a resistive voltage-divider from VGOFF to REF with the center tap connected to FBN (Figure 1). Select R4 between 20kΩ and 50kΩ. Calculate R3 with the following equation: ⎛V - VGOFF ⎞ R3 = R4 × ⎜ FBN ⎟ ⎝ VREF - VFBN ⎠ where VFBN = 250mV, VREF = 1.25V. Note that REF can only source up to 50µA; using a resistor less than 20kΩ for R4 results in higher bias current than REF can supply without degrading REF accuracy. Pass-Transistor Selection The pass transistor must meet specifications for current gain (hFE), input capacitance, collector-emitter saturation voltage, and power dissipation. The transistor’s current gain limits the guaranteed maximum output current to: ⎛ V ⎞ ILOAD(MAX) = ⎜ IDRV - BE ⎟ × hFE(MIN) RBE ⎠ ⎝ where IDRV is the minimum guaranteed base-drive current and RBE is the pullup resistor connected between the transistor’s base and emitter. Furthermore, the transistor’s current gain increases the linear regulator’s DC loop gain (see the Stability Requirements section), so excessive gain destabilizes the output. Therefore, transistors with current gain over 100 at the maximum output current can be difficult to stabilize and are not recommended unless needed to meet output-current requirements. The transistor’s saturation voltage at the maximum output current determines the minimum input-to-output voltage differential that the linear regulator supports. Also, the package’s power dissipation limits the useable maximum power-dissipation capability of the transistor’s package, and mounting must exceed the actual power dissipation in the device. The power dissipation equals the maximum load current (ILOAD(MAX)_LR ) times the maximum input-to-output voltage differential: ( P = ILOAD(MAX)_ LR × VIN(MAX)_ LR - VOUT_LR ) where VIN(MAX)_LR is the maximum input voltage of the linear regulator and VOUT_LR is the output voltage of the linear regulator. Stability Requirements The MAX1513/MAX1514 linear-regulator controllers use an internal transconductance amplifier to drive an external pass transistor. The transconductance amplifier, the pass transistor, the base-emitter resistor, and the output capacitor determine the loop stability. The following applies equally to all linear regulators in the MAX1513 and MAX1514. The transconductance amplifier regulates the output voltage by controlling the pass transistor’s base current. The total DC loop gain is approximately: ______________________________________________________________________________________ 25 MAX1513/MAX1514 capacitors, the output voltage ripple is dominated by the capacitance value. Use the following equation to approximate the required capacitor value: MAX1513/MAX1514 TFT-LCD Power-Supply Controllers ⎡ ⎛I ⎛ 4⎞ × hFE ⎞ ⎤ A V _ LR ≈ ⎜ ⎟ × ⎢1 + ⎜ BIAS ⎟ ⎥ × VREF ⎝ VT ⎠ ⎢⎣ ⎝ ILOAD _ LR ⎠ ⎥⎦ where CIN = gm is the transconductance of the pass transistor, and fT is the transition frequency. Both parameters can be found in the transistor’s data sheet. Because RBE is much greater than RIN, the above equation can be simplified: where VT is 26mV at room temperature and IBIAS is the current through the base-to-emitter resistor (RBE). Each of the four linear-regulator controllers is designed for a different maximum output current, so they have different output drive currents and different bias currents (IBIAS). Each controller’s bias current can be found in the Electrical Characteristics table. The current listed in the conditions column for the FB_ regulation voltage specification is the individual controller’s bias current. The base-to-emitter resistor for each controller should be chosen to set the correct IBIAS: RBE = fPOLE _ IN = 4) First, determine the dominant pole set by the linear regulator’s output capacitor and the load resistor: fPOLE _ LR = fPOLE _ FB = 5) The unity-gain crossover of the linear regulator is: 2) The pole created by the internal amplifier delay is about 1MHz: fPOLE _ AMP ≈ 1MHz 3) Next, calculate the pole set by the transistor’s input capacitance CIN, the transistor’s input resistance RIN, and the base-to-emitter pullup resistor: fPOLE _ IN = 26 1 ( 2π × CIN × RBE || RIN 6) ( 1 2π × CFB × RUPPER || RLOWER ) where CFB is the capacitance between FB_ and ground, RUPPER is the upper resistor of the linear regulator’s feedback divider, and RLOWER is the lower resistor of the divider. Next, calculate the zero caused by the output capacitor’s ESR: fPOLE _ ESR = 2π × COUT _ LR × VOUT _ LR fT hFE Next, calculate the pole set by the linear regulator’s feedback resistance and the capacitance between FB_ and GND (including stray capacitance): ILOAD(MAX)_ LR fCROSSOVER = A V _ LR × fPOLE _ LR 2π × CIN × RIN fPOLE _ IN = VBE IBIAS 1 The equation can be further simplified: The output capacitor and the load resistance create the dominant pole in the system. However, the internal amplifier delay, the pass transistor’s input capacitance, and the stray capacitance at the feedback node create additional poles in the system. The output capacitor’s ESR generates a zero. For proper operation, use the following equations to verify the linear regulator is properly compensated: 1) gm h , RIN = FE , 2πfT gm 1 2π × COUT _ LR × RESR where R ESR is the equivalent series resistance of COUT_LR. To ensure stability, choose COUT_LR large enough so the crossover occurs well before the poles and zero calculated in steps 2 to 5. The poles in steps 3 and 4 generally occur at several megahertz and using ceramic capacitors ensures the ESR zero occurs at several megahertz as well. Placing the crossover below 500kHz is sufficient to avoid the amplifier-delay pole and generally works well, unless unusual component choices or extra capacitances move the other poles or zero below 1MHz. ) ______________________________________________________________________________________ TFT-LCD Power-Supply Controllers 1) Minimize the area of high-current loops. The highcurrent input loop goes from the positive terminals of the input capacitors to the inductor, to the power MOSFET, and to the negative terminals of the input capacitors. The high-current output loop is from the positive terminals of the input capacitors to the inductor, to the output diode, and to the positive terminals of the output capacitors, reconnecting between the output-capacitor and input-capacitor ground terminals. Connect these loops with short, wide connections. Avoid using vias in the high-current paths. If vias are unavoidable, use many vias in parallel to reduce resistance and inductance. 2) Create a power ground plane consisting of the input and output-capacitor ground terminals, the source of the power MOSFET, and any ground terminals of the charge-pump components. Connect all of these together with short, wide traces or a small ground plane. Maximizing the width of the power ground traces improves efficiency and reduces output voltage ripple and noise spikes. Create an analog ground plane consisting of the IC’s backside pad, all the feedback-divider ground connections, the bufferamplifier-divider ground connection, the REF capacitor ground connection, and the DEL capacitor ground connection. The power ground plane and the analog ground plane should be connected at only one loca- CS- CS+ IN GATE TOP VIEW FB Pin Configuration 20 19 18 17 16 1 15 GND SDFR 2 14 DRVP *FBPB 3 13 FBP *OUTB 4 12 *FBG *SUPB 5 11 *DRVG 7 8 9 10 DEL DRVN DRVL FBL FBN 6 4) Place all feedback-voltage-divider resistors as close to their respective feedback pins as possible. The divider’s center trace should be kept short. Placing the resistors far away causes their FB traces to become antennas that can pick up switching noise. Care should be taken to avoid running any feedback trace near the switching nodes in the step-up regulator and charge pumps. 5) Minimize the length and maximize the width of the traces between the output capacitors and the load for best transient responses. 6) Minimize the size of the switching node while keeping it wide and short. Keep the switching node away from feedback nodes (FB, FBP, FBL, FBG, and FBN) and analog ground. Use DC traces to shield if necessary. Refer to the MAX1513 evaluation kit for an example of proper board layout. Chip Information TRANSISTOR COUNT: 4807 PROCESS: BiCMOS REF MAX1513 MAX1514 tion, which is the IC’s GND pin. All other ground connections, such as the IN pin bypass capacitor and the linear-regulator output capacitors, should be starconnected directly to the backside pad of the IC through a via with wide traces, not otherwise connecting to either the power ground plane or the analog ground plane. Connect the IC’s backside pad to the IC’s GND pin. Make no other connections between the analog and power ground planes. 3) Place IN and REF bypass capacitors as close to the device as possible. THIN QFN 4mm x 4mm *N.C. FOR MAX1514 ______________________________________________________________________________________ 27 MAX1513/MAX1514 PC Board Layout and Grounding Careful PC board layout is important for proper operation. Use the following guidelines for good PC board layout: 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.) 24L QFN THIN.EPS MAX1513/MAX1514 TFT-LCD Power-Supply Controllers PACKAGE OUTLINE 12,16,20,24L QFN THIN, 4x4x0.8 mm 21-0139 B 1 2 PACKAGE OUTLINE 12,16,20,24L QFN THIN, 4x4x0.8 mm 21-0139 B 2 2 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. 28 ____________________Maxim Integrated Products, 120 San Gabriel Drive, Sunnyvale, CA 94086 408-737-7600 © 2003 Maxim Integrated Products Printed USA is a registered trademark of Maxim Integrated Products.