SLUS539A − JUNE 2004 − REVISED AUGUST 2004 FEATURES D On-Chip Predictive Gate Drivet for D D D D D D D D D D APPLICATIONS D Secondary-Side Post Regulation (SSPR) for High-Efficiency Synchronous Buck Operation Dual ±3-A TrueDrivet Outputs 1-MHz High Frequency Operation with 70-ns Delay from SYNCIN to G1 Output Leading Edge Modulation Overcurrent Protection using a Parallel Average Current Mode Control Loop 3 Modes to Support 2.7-V to 35-V Bias Operation Reverse Current Protection for Output Stage User Programmable Shutdown ±1.0% Initial Tolerance Bandgap Reference High Bandwidth Error Amplifiers Thermally Enhanced HTSSOP 20-Pin PowerPADt Package D D DESCRIPTION The UCC2540 is a secondary-side synchronous buck PWM controller for high current and low output voltage applications. It can be used either as the local secondary-side controller for isolated dc-to-dc converters using two-stage cascaded topologies or as a secondary-side post regulator (SSPR) for multiple output power supplies. The UCC2540 runs with the synchronization signal from either the primary side or the high duty cycle quasi-dc output of bus converters or dc transformers. For higher efficiency, it also incorporates the Predictive Gate Drivet technology that virtually eliminates body diode conduction losses in synchronous rectifiers. SIMPLIFIED APPLICATION DIAGRAM Input Multiple Output Power Supplies Cascaded Buck Converters Post Processing Converters for Bus Converter and DC Transformer Architectures Main Output UCC25701 C1 FB OUT QP RSNS V FB AUX Output UCC2540 VEA− 7 13 VDRV 14 G2 REF 2 12 G2S TR 10 16 VDD Main Output R1 COMP 9 19 BST 18 G1 G2C 3 20 SWS 17 SW CEA− 8 15 PGND RSET 1 R2 11 SS SYNCIN RAMP 5 4 UDG−04057 Predictive Gate Drive, TrueDrive, and PowerPAD, are a trademarks of Texas Instruments Incorporated. !"#$ % &'!!($ #% )'*+&#$ ,#$(!,'&$% &!" $ %)(&&#$% )(! $.( $(!"% (/#% %$!'"($% %$#,#!, 0#!!#$1- !,'&$ )!&(%%2 ,(% $ (&(%%#!+1 &+',( $(%$2 #++ )#!#"($(!%- Copyright 2004, Texas Instruments Incorporated www.ti.com 1 SLUS539A − JUNE 2004 − REVISED AUGUST 2004 Please be aware that an important notice concerning availability, standard warranty, and use in critical applications of Texas Instruments semiconductor products and disclaimers thereto appears at the end of this data sheet. DESCRIPTION (CONT.) The UCC2540 is available in the extended temperature range of –40°C to 105°C and is offered in thermally enhanced PowerPADt 20-pin HTSSOP (PWP) package. This space saving package with standard 20-pin TSSOP footprint has a drastically lower thermal resistance of 1.4°C/W θJC to accommodate the dual high-current drivers on board. ABSOLUTE MAXIMUM RATINGS over operating free-air temperature range (unless otherwise noted)(1)(2) Supply voltage range, VDD Supply current, IVDD VDD CEA−, COMP, G2C, RAMP, SS, TR, VEA− UNIT 36 V 50 mA −0.3 to 3.6 VDRV Analog input voltages UCC2540 −0.3 to 9 G1, BST SW−0.3 to SW+9 SW, SWS −1 to 36 G2, G2S −1 to 9 SYNCIN −0.3 to 8.0 Sink current (peak), IOUT_SINK G1, G2 3.5 Source current (peak), IOUT_SOURCE G1, G2 −3.5 Operating junction temperature range, TJ −55 to 150 Storage temperature, Tstg −65 to 150 V A °C C Lead temperature 1,6 mm (1/16 inch) from case for 10 seconds 300 (1) 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 under “recommended operating conditions” is not implied. Exposure to absolute-maximum-rated conditions for extended periods may affect device reliability. (2) All voltages are with respect to GND. Currents are positive into, and negative out of the specified terminal. RECOMMENDED OPERATING CONDITIONS MIN 2 TYP MAX Supply voltage, VDD Mode 1 8.5 35 Supply voltage, VDRV Mode 2 4.75 8.00 Supply voltage, REF Mode 3 3.0 3.3 Supply voltage bypass, CVDD 1.0 2.2 Reference bypass capacitor, CREF 0.1 1.0 VDRV bypass capacitor, CVDRV 0.2 BST−SW bypass capacitor, CBST−SW 0.1 2.2 10 50 PWM ramp capacitor range, CRAMP 100 680 Turn-off capacitor range, CG2C 120 1000 6.5 Junction operating temperature, TJ −40 www.ti.com V 3.6 Timer current resistor range, RRSET COMP pin load range, RLOAD UNIT µF F kΩ pF kΩ 105 °C SLUS539A − JUNE 2004 − REVISED AUGUST 2004 ORDERING INFORMATION HTSSOP−20 (PWP)(1) TA = TJ Bulk −40°C to +105°C UCC2540PWP (1) The PWP package is also available at 70 devices per tube and taped and reeled at 2,000 devices per reel. Add an R suffix to the device type (i.e., UCC2540PWPR). See the application section of the data sheet for PowerPAD drawing and layout information. CONNECTION DIAGRAM PWP PACKAGE (TOP VIEW) RSET REF G2C SYNCIN RAMP GND VEA− CEA− COMP TR 1 2 3 4 5 6 7 8 9 10 20 19 18 17 16 15 14 13 12 11 SWS BST G1 SW VDD PGND G2 VDRV G2S SS NOTE: The PowerPADt is not directly connected to any lead of the package. It is electrically and thermally connected to the substrate of the device which acts as ground and should be connected to PGND on the PCB. The exposed dimension is 1.3 mm x 1.7 mm. However, the tolerances can be +1.05 mm / −0.05 mm (+41 mils / −2 mils) due to position and mold flow variation. THERMAL INFORMATION PACKAGE FAMILY PACKAGE DESIGNATOR θJA (°C/W) (with PowerPAD) θJC (°C/W) (without PowerPAD) θJC (°C/W) (with PowerPAD) MAXIMUM DIE TEMPERATURE PowerPAD HTSSOP−20 PWP 22.3 to 32.6 (500 to 0 LFM) 19.9 1.4 125°C www.ti.com 3 SLUS539A − JUNE 2004 − REVISED AUGUST 2004 ELECTRICAL CHARACTERISTICS VDD = 12 V, 1-µF capacitor from VDD to GND, 1-µF capacitor from BST to SW, 1-µF capacitor from REF to GND, 0.1-µF and 2.2-µF capacitors from VDRV to PGND, fSYNCIN = 200 kHz, TA = TJ = −40°C to 105°C, (unless otherwise noted). PARAMETER TEST CONDITIONS MIN TYP MAX 8 11 13 9 12 30 UNIT OVERALL DC IVDD Operating current fS = 200 kHz, CLOAD = 2.2 nF mA UNDERVOLTAGE LOCKOUT VVDD VVDD Start threshold voltage MODE 1 8.0 8.5 9.0 Stop threshold voltage MODE 1 7.5 8.0 8.5 VVDD VVDRV Hysteresis MODE 1 0.3 0.5 0.8 Start threshold voltage MODE 2, 4.30 4.65 4.85 VVDRV VVDRV Stop threshold voltage MODE 2 4.0 4.3 4.6 Hysteresis MODE 2 0.15 0.35 0.55 VREF VREF Start threshold voltage MODE 3 2.75 3.00 3.20 Stop threshold voltage MODE 3 2.25 2.50 2.70 MODE 3 0.3 0.5 0.8 3.28 3.30 3.32 3.2 3.3 3.4 10 13 20 0 1.5 15 0 30 70 VREF Hysteresis VOLTAGE REFERENCE (REF) VREF Reference output voltage ISC Short circuit current VVDD = 4 V VVDD= VVDRV= 2.7 V TA = 25°C Total variation Line regulation VREF = 0 V, TA = 25°C 5.25 V ≤ VREF ≤ 7.2 V Load regulation 0 mA ≤ IREF ≤ 5 mA V V mA mV PWM (RAMP) DMIN Minimum duty cycle VRAMP Offset voltage 0% Timeout threshold voltage tDEAD IRAMP G1 deadtime at maximum duty cycle ratio Ramp charge current fSYNC = 200 kHz RRSET = 10 kΩ 0.10 0.25 0.45 2.3 2.5 2.8 150 175 200 ns −325 −300 −275 µA 45 50 55 mV 3 4 V CURRENT ERROR AMPLIFIER VCEA+ GBW VOL Offset voltage Total variation Gain bandwidth(3) Low-level output voltage VOH High-level output voltage AVOL IBIAS Open loop ICOMP = 0 A, VVEA− = 2.0 V VCEA− = 3.3 V, ICOMP = 200 µA, VVEA− = 1 V VCEA− = 1.5 V ICOMP = 0 A, VVEA− = 1 V VCEA− = 0 V, Bias current ISINK Sink current CMR Common mode input range(3) VCOMP = 1.0 V, VVEA− = 0 V 0.1 V 0 0.60 0.83 2.2 2.5 3.0 V 60 100 140 dB −200 −80 −10 nA 0.35 0.80 1.70 mA 0 (3) Ensured by design. Not production tested. 4 VCEA− = 1.5 V, www.ti.com MHz 2 V SLUS539A − JUNE 2004 − REVISED AUGUST 2004 ELECTRICAL CHARACTERISTICS VDD = 12 V, 1-µF capacitor from VDD to GND, 1-µF capacitor from BST to SW, 1-µF capacitor from REF to GND, 0.1-µF and 2.2-µF capacitors from VDRV to PGND, fSYNCIN = 200 kHz, TA = TJ = −40°C to 105°C, (unless otherwise noted) PARAMETER TEST CONDITIONS MIN TYP MAX UNIT 0.40 0.75 1.00 V VOLTAGE ERROR AMPLIFIER VSS_OFF VTR_OFF Offset voltage from soft-start input Offset voltage from tracking input VVEA+ Threshold voltage (from VEA− to COMP) GBW Gain bandwidth(3) VOL Low-level output voltage VOH High-level output voltage AVOL IBIAS Open loop ISINK Sink current(4) ICOMP = VVEA−, VTR = 1.0 V, VSS− = 1.5 V VCOMP = VVEA− 0°C ≤ TA ≤ 105°C Total variation ICOMP = 0 A, VVEA− = 2.0 V, VCEA− = 1.75 V, ICOMP = 200 µA, VVEA− = 1 V, VCEA− = 0 V, VTR = 0 V ICOMP = 0 A, VVEA− = 1 V VCEA− = 0 V 25 48 70 1.485 1.500 1.515 1.47 1.50 1.53 3 4 mV V MHz 0.1 0 0.60 0.83 2.2 2.5 3.0 V 60 100 140 dB −300 −150 −50 µA 0.35 0.80 1.70 mA RRSET = 10 kΩ −158 −150 −142 µA RSET voltage RRSET = 10 kΩ SYNCHRONIZATION AND SHUTDOWN TIMER (SYNCIN, G2C) 1.42 1.50 1.58 V 2.3 2.5 2.7 1.50 1.65 1.80 Bias current VCOMP = 1.0 V, VVEA− = 1.5 V VCEA− = 0 V, CURRENT SET IOUT VRSET Output current Timer threshold SYNCIN threshold V µA ICHG(G2C) Shutdown timer charge current SOFT-START (SS) RRSET = 10 kΩ −325 −300 −275 ICH(SS) IDSCH(SS) Charge current RRSET = 10 kΩ −230 −200 −170 Discharge current RRSET = 10 kΩ 50 70 100 0.35 0.45 0.55 V 6.87 7.20 7.53 V 0 50 100 0 50 100 15 30 50 VSWS = 0 V VSWS = 0 V 1.90 2.25 3.10 1.00 1.25 1.03 VG2S = 0 V −0.70 −0.50 −0.37 VG2S = 0 V VG2S = 0 V 1.90 2.25 2.90 SWS falling threshold voltage 1.0 1.2 1.3 Current VSWS = 0 V −1.8 −1.3 −0.9 mA −0.5 −0.3 −0.1 V Discharge/shutdown threshold A µA DRIVE REGULATOR (VDRV) VVDRV Output voltage Line regulation VVDD = 8.5 V 9 V ≤ VVDD ≤ 35 V Load regulation −5 mA ≤ IVDRV ≤ 0 mA ISC Short-circuit current G2S GATE DRIVE SENSE G2S rising threshold voltage G2S falling threshold voltage IG2S Current SWS GATE DRIVE SENSE SWS rising threshold voltage ISWS Negative threshold voltage mV mA V mA V (3) Ensured by design. Not production tested. www.ti.com 5 SLUS539A − JUNE 2004 − REVISED AUGUST 2004 ELECTRICAL CHARACTERISTICS VDD = 12 V, 1-µF capacitor from VDD to GND, 1-µF capacitor from BST to SW, 1-µF capacitor from REF to GND, 0.1-µF and 2.2-µF capacitors from VDRV to PGND, fSYNCIN = 200 kHz, TA = TJ = −40°C to 105°C, (unless otherwise noted) PARAMETER TEST CONDITIONS MIN TYP MAX 0.3 0.7 1.3 10 25 45 UNIT G1 MAIN OUTPUT RSINK Sink resistance RSRC Source resistance Sink current(3) ISINK ISRCE tRISE tFALL VSW = 0 V, VSW = 0 V, VBST = 6 V, VBST = 6 V, VG1 = 0.5 V VG1 = 5.7 V Source current(3) VSW = 0 V, VSW = 0 V, VBST = 6 V, VBST = 6 V, VG1 = 3.0 V VG1 = 3.0 V Rise time CLOAD = 2.2 nF, from G1 to SW 12 25 Fall time CLOAD = 2.2 nF, from G1 to SW 12 25 15 30 Ω 3 A −3 ns G2 SYNCHRONOUS RECTIFIER OUTPUT RSINK ISINK ISRC tRISE tFALL Sink resistance Sink current(3) VG2 = 0.3 V VG2 = 3.25 V Source current(3) −3 Rise time VG2 = 3.25 V CLOAD = 2.2 nF, from G2 to PGND 12 25 Fall time CLOAD = 2.2 nF, from G2 to PGND 12 25 6.2 6.7 7.5 RAMP rising to G1 rising 90 115 130 SYNCIN falling to G1 falling 50 70 90 Delay control resolution 3.5 5.0 6.5 VOH High-level output voltage, G2 DEADTIME DELAY (see Figure 1) tON(G1) tOFF(G1) tON(G2) tOFF(G2) 5 VSW = GND tON(G2) tON(G2) G2 on-time minimum wrt G1 falling −24 G2 on-time maximum wrt G1 falling 62 tOFF(G2) tOFF(G2) G2 off-time minimum wrt G1 rising −68 G2 off-time maximum wrt G1 rising 10 (3) Ensured by design. Not production tested. tOFF(G1) SYNCIN VERR RAMP G2C tON(G1) tON(G2) G1 tOFF(G2) G2 Figure 1. Predictive Gate Drive Timing Diagram 6 Ω 3 www.ti.com A ns V ns SLUS539A − JUNE 2004 − REVISED AUGUST 2004 FUNCTIONAL BLOCK DIAGRAM RSET 1 + − 20 ISET 1.5V VREF VDRV VREF 2 VREF VDD 19 BST VDD REF SWS HIGH SIDE DRIVER UVLO REFERENCE GLO 18 G1 PWR G1D 17 SW G2C 3 2 × ISET SYNCIN 4 GLO G2 UVLO G2 TIMER G2TO G1D UVLO DRIVE REGULATOR 100 ns CLK SYNC 16 VREF 15 PGND 2 × ISET RAMP 5 GND VDD PREDICTIVE LOGIC RAMP PGND PWM RAMP & PWM LOGIC 6 LOW SIDE DRIVER BIAS PWR 14 G2 VERR VEA− 7 CEA− 8 COMP 9 PGND ERROR AMPLIFIERS AND FAULT LOGIC 13 VREF HUP VDRV 12 G2S 1.33 × ISET TR 10 11 SS 1.73 × ISET UDG−04056 www.ti.com 7 SLUS539A − JUNE 2004 − REVISED AUGUST 2004 PIN ASSIGNMENTS TERMINAL NAME NO. I/O DESCRIPTION BST 19 I Floating G1 driver supply pin. VHI is fed by an external Schottky diode during the SR MOSFET on time. Bypass BST to SW with an external capacitor. CEA− 8 I Inverting input of the current error amplifier used for output current regulation. COMP 9 I Output of the voltage and current error amplifiers for compensation. G1 18 O High-side gate driver output that swings between SW and BST. G2 14 O Low-side gate driver output that swings between PGND and VDRV. G2C 3 I Timer pin to turn off synchronous rectifier. The capacitor connected to this pin programs the maximum duration that G2 is allowed to stay HIGH. G2S 12 I Used by the predictive deadtime controller for sensing the SR MOSFET gate voltage to set the appropriate deadtime. GND 6 − Ground for internal circuitry. GND and PGND should be tied to the pc-board ground plane with vias. PGND 15 − Ground return for the G2 driver. Connect PGND to the pc-board ground plane with several vias. RAMP 5 I Input pin to connect capacitor to GND to generate the PWM ramp and serve as a maximum duty ratio timer. REF(1) 2 I/O RSET 1 I Pin to program timer currents for G2C, RAMP, SS charge and SS discharge. This pin generates a current proportional to the value of the external resistor connected from RSET pin to GND. RSET range is 10 kΩ to 50 kΩ (giving a programmable nominal ISET range of 30 µA to 150 µA, respectively). SS 11 I Soft start and shutdown pin. Connect a capacitor to GND to set the soft-start time. Add switch to GND for immediate shutdown functionality. SYNCIN 4 I Input pin for timing signal. SW 17 − G1 driver return connection. SWS 20 I Used by the predictive controller to sense SR body-diode conduction. Connect to SR MOSFET drain close to the MOSFET package. TR 10 I Tracking input to the voltage error amplifier. Connect to REF when not used. VDD 16 I Power supply pin to the device and input to the internal VDRV drive regulator. Normal VDD range is from 4.5 V to 36 V. Bypass the pin with at least 1 µF of capacitance. VDRV 13 I Output of the drive regulator and power supply pin for the G2 driver. VDRV is also the supply voltage for the internal logic and control circuitry. 3.3-V reference pin. All internal circuits are powered from this 3.3-V rail. Bypass this pin with at least 0.1 µF of capacitance for REF loads that are 0 mA to −1 mA. Bypass this pin with at least 1 µF of capacitance if it is used as an input (Mode 3) or if it has large or pulsating loads. VEA− 7 I Inverting input of the voltage error amplifier used for output voltage regulation. (1) REF is an input in Mode 3 only. 8 www.ti.com SLUS539A − JUNE 2004 − REVISED AUGUST 2004 APPLICATION INFORMATION The UCC2540 is a high-efficiency synchronous buck controller that can be used in many point-of-load applications. It can be used as a local controller for cascaded techniques such as post processing converters for isolated integrated bus converters (IBC) and dc transformer architectures. It can also be used as a general purpose secondary-side post regulator for high-accuracy multiple-output power supplies. Using UCC2540 as the Secondary-Side PWM Controller in the Cascaded Push-Pull Buck Two Stage Converter The two-stage cascaded push-pull buck topology converts higher-input bus voltage such as 48-V telecom voltage to sub 2-V output voltages. Q2’ T1 NP1 L1 NS2 Q3 NS1 V3 NP2 Q4 C2 Q1’ VIN 48−V UCC2540 VEA− 7 13 VDRV V1 C1 Q1 OUT1 RA 14 G2 Q2 REF 2 13 G2S TR 10 R1 16 VDD COMP 9 19 BST 18 G1 G2C 3 RB 20 SWS OUT1 OUT2 SYNC UCC28089 17 SW CEA− 8 15 PGND RSET 1 R2 11 SS SYNCINRAMP 5 4 CLOCK RESET UDG−02140 Figure 2. Secondary-Side Controlled Cascaded Push-Pull/Buck Converter The primary-side power stage is an open loop push-pull converter that provides voltage step-down, and galvanic isolation. This takes the high bus voltage and converts it into an intermediate voltage such as 7 V. The primary-side push-pull gate drive signals can come from either off-the-shelf oscillators or a fully integrated 50% duty dual-output oscillator such as the UCC28089. The secondary-side power stage is a buck converter that is optimized for low-output voltage regulation. The clock reset pulse signal from the primary side is transmitted using a signal transformer. www.ti.com 9 SLUS539A − JUNE 2004 − REVISED AUGUST 2004 APPLICATION INFORMATION There are many advantages to this secondary-side control circuit. The simple isolated power stage does not require any feedback across the isolation boundary. Since the primary-side oscillator is free running, there is no need for an isolated start-up power supply. This high-frequency circuit provides soft-switching operation (for all six MOSFET switches), optimum transformer core utilization, and minimizes filter requirements because there are no additional high-current inductors. The push-pull primary side permits simple direct drive control of the input stage MOSFETs. In exchange, it requires that the input MOSFETs are rated to at least twice the peak input line voltage. This configuration works well for 36-V to 72-V input line applications, because there are many suitable power MOSFETs available in the range of 150 V. For applications with larger input voltages, a half bridge or full bridge with alternating modulation might be more suitable for an input stage. Thus, the cascaded topology has a large degree of flexibility with input power stages. The cascaded topology also has flexibility in the output stages, as well. For additional information on this topology refer to Power Supply Seminar SEM−1300 Topic 1: Unique Cascaded Power Converter Topology for High Current Low Output Voltage Applications [1]. The topic discusses the operating principles, design trade-offs, and critical design procedure steps. UCC2540 in Multiple Output Power Supplies One such flexibility is an ability to easily add independently regulated auxiliary outputs. A multiple output implementation of the cascaded push-pull/buck power converter is shown in Figure 3. Q2’ L2 Q3 NP1 INPUT NP2 Q4 C1 Q1’ Q1 C2 OUT1 UCC2540 Q2 PWM PWM 14 G2 18 G1 SYNCIN 4 OUT1 OUT2 SYNC UCC28089 CLOCK RESET SYNCIN L3 Q5 Q6 C3 OUT2 UCC2540 PWM2 PWM2 14 G2 18 G1 SYNCIN 4 SYNCIN Figure 3. Multiple Output Implementation of Push-Pull/Buck Cascaded Converter 10 www.ti.com UDG−02142 SLUS539A − JUNE 2004 − REVISED AUGUST 2004 APPLICATION INFORMATION Using UCC2540 as the Secondary-Side Post Regulator UCC2540 can also be used as a secondary-side post regulator (SSPR) for precision regulation of the auxiliary voltages of multiple output power supplies, as shown in Figure 4. The UCC2540 uses leading-edge modulation so that it is compatible with either voltage-mode or current-mode primary-side control converters using any topology such as forward, half-bridge or push-pull. Q2’ L2 NP1 INPUT C2 MAIN OUTPUT C3 AUX OUT1 NP2 Q1’ C1 Q1 VFB Q2 L3 Q5 Q6 OUT1 UCC3808x OUT2 FB UCC2540 1 PWM2 14 G2 PWM2 18 G1 SYNCIN 4 UDG−02145 Figure 4. Multiple Output Converter with Primary Side Push−Pull Converter www.ti.com 11 SLUS539A − JUNE 2004 − REVISED AUGUST 2004 APPLICATION INFORMATION CEA− and VEA− pins: Current Limit and Hiccup Mode Typical power supply load voltage versus load current is shown in Figure 5. This figure shows steady state operation for no-load to overcurrent shutdown (soft-start retry is not depicted in the diagram). During the voltage regulation conditions, the voltage error amplifier output is lower than the current error amplifier, allowing the voltage error amplifier to control operation. During the current limit conditions, the current error amplifier output is lower than the voltage error amplifier, allowing the current error amplifier to control operation. The boundary between voltage and current control occurs when the difference between CEA− and VEA− tries to exceed 50 mV. VLOAD − Load Voltage − V Current limiting begins to occur when the difference between CEA− and VEA− exceeds 50 mV. For currents that exceed this operating condition, the UCC2540 controls the converter to operate as a pure current source until the output voltage falls to half of its rated steady state level. Then the UCC2540 sets both G1 and G2 outputs to LOW and it latches a fault that discharges the soft-start voltage at 30% of its charging rate. The UCC2540 inhibits a retry until the soft-start voltage falls below 0.5 V. A functional diagram of the voltage and current error amplifiers is shown in Figure 6. VREG Limited Current Shutdown ILOAD − Load Current − A Figure 5. Typical Power Supply Load Voltage vs Current 12 www.ti.com UDG−04053 SLUS539A − JUNE 2004 − REVISED AUGUST 2004 APPLICATION INFORMATION From Power MOSFET Switch Node RS C RI1 RLOAD 0.7 V + + + + SS 1.25 R Voltage Error Amplifier TR RV1 R 1.5 V 1.5 V VEA− 7 + VERR to Modulator Inverting Amplifier 50 mV + CEA− COMP + 9 8 Current Error Amplifier ZFV RFV UCC2540 CFV ZFV ZIV RI2 RFI CFI CFIR RV2 ZIV UDG−04052 Figure 6. Error Amplifier Configuration Component selection includes setting the voltage regulation threshold, then the current limit threshold, as described below. Voltage vs. Current Programming (refer to Figure 6): 1. Determine the ratio V LOAD(reg) V LOAD(reg) R V1 + *1V+ *1V R V2 V VEA* ) Threshold Voltage 1.5 V (typ) ǒ 2. Sense resistor R S + 1 ) Ǔ R V1 R V2 V CEA)offset voltage , where IS(max) is the current limit level, I S(max) VCEA+offset = 50 mV (typ). 3. Arbitrarily select either RV1 or RV2 so that the smallest of the two resistors is between 6.5 kΩ and 20 kΩ. Then calculate the value of the other resistor using the equation in the first step. If the converter is in a current-limit condition and the output voltage falls below half of the regulated output voltage, the UCC2540 enters into a hiccup (restart-retry) mode. Figure 7 shows typical signals during hiccup mode. www.ti.com 13 SLUS539A − JUNE 2004 − REVISED AUGUST 2004 APPLICATION INFORMATION SYNCIN 3.3 V SS 0.5 V ILOAD VLOAD RAMP G2C G1 G2 UDG−04046 Figure 7. Typical Hiccup Mode waveforms COMP, VEA− and CEA− pins: Voltage and Current Error Amplifiers From no-load to full rated load operating conditions, the UCC2540 operates as a voltage mode controller. Above the programmed rated current, there are two levels of over current protection; constant current limit and overcurrent reset/retry. This section gives suggestions on how to design the voltage controller and current controller so that they interact with one another in a stable fashion. Refer to the functional diagram of the voltage and current error amplifiers in Figure 6. The voltage error amplifier in the figure shows three non-inverting inputs. The lowest of the three non-inverting inputs (1.5 V, SS and TR) is summed with the non-inverting input to achieve the voltage error signal. The lowest of the two outputs drives the inverting stage which in turn, drives the modulator. During steady state voltage control operation, the feedback elements in the current loop have no effect on the loop stability. When current limit occurs, the voltage error amplifier effectively shuts OFF and the current error amplifier takes control. During steady state current limit operation, the negative feedback elements in the voltage error amplifier loop become positive feedback elements in the current error amplifier loop. In order for the current error amplifier to be stable, the impedances in the feedback path of the current error amplifier must be lower than the impedances in the feedback path of the voltage error amplifier. This means that resistors in the current error amplifier negative feedback path must be less than the resistors in the voltage error amplifier negative feedback path. Also capacitors in the current error amplifier negative feedback path must be larger than capacitors in the negative feedback path of the voltage error amplifier negative feedback path. (Capacitance is really an admittance value rather than an impedance value). This concept is illustrated in Figure 6. In order for the current loop to be stable in Figure 6, ||ZIV|| must be less than ||ZFV|| over all frequencies. This can be achieved if RFI < RFV and CFI > CFV. 14 www.ti.com SLUS539A − JUNE 2004 − REVISED AUGUST 2004 APPLICATION INFORMATION Another issue that can occur during current limit operation is modulator stability. In order for the modulator to be stable, the rising slope of the current ripple measured at the COMP pin must be smaller than the rising slope that is measured at the RAMP pin. This can be met either in the selection of the ratio of ||ZIV|| to ||ZFV||, or by the addition of a capacitor in parallel to RFI and CFI, such as CFIR, in Figure 6. Stable Dynamic Current Loop Design (refer to Figure 6): 1. Using any favorite approach, design the voltage error amplifier for stable voltage mode design. Use at least 15 kΩ for any resistors in the negative feedback path of the voltage error amplifier (between pins 9 and 7). This does not apply to resistance values between the power supply output voltage and pin 7; it also does not apply to resistance values between ground and pin 7. 2. The goal is to design the current limit control loop so that it drives the converter to maintain 50 mV between the VEA− pin and the CEA− pin during current-limit conditions. Select the current sense element and the voltage divider ratios for the VEA− pin to ground and the CEA− pin to ground to provide the desired current limit level. 3. Place the same configuration of components in the negative feedback path of the current error amplifier (between pins 9 and 8), that are in the negative feedback path of the voltage error amplifier (between pins 9 and 7). However, use resistors with values that are 67% of the corresponding resistors that are between pins 9 and 7 and use capacitors that are 150% of the corresponding capacitors that are between pin 9 and pin 7. 4. Check the COMP signal. If it is unstable, place a capacitor (or increase the capacitance) between pins 9 and 8 in order to attenuate the current ripple. Raise the value of the capacitor until the COMP pin voltage becomes stable. Compare the COMP voltage with the RAMP voltage. With stable operation, the rising slope of the COMP voltage ripple is less than the rising slope of the RAMP pin. www.ti.com 15 SLUS539A − JUNE 2004 − REVISED AUGUST 2004 APPLICATION INFORMATION RSET, RAMP, G2C, SS pins: Programming the Timer Currents Set the base current to the timers with a resistor between RSET and GND. The block diagram of the UCC2540 shows the interaction of the RSET pin and the dependent current sources for the RAMP, G2C and SS features. The RSET pin is a voltage source; the current of the RSET pin is reflected and multiplied by a gain and distributed to the RAMP (gain = 2), G2C (gain = 2) and SS (charge gain = 1.33, net discharge gain = 0.4). The resistance applied to the RSET pin and GND should be in the range of 10 kΩ < RRSET < 50 kΩ. RAMP, G2C and SS timers are programmed by the selection of capacitors tied between each of their respective pins and GND. G2C pin: G2 Timer G2C 2 y IRSET 2.5 V G2 Timeout Comparator *G1D + G2C Latch S Q RD Q 3 CG2C G2TO UVLO GLO G2 *G1 with delay, but not blanked UDG−04047 Figure 8. Functional diagram of the G2 Timer The G2C pin programs the maximum duration of the synchronous rectifier to facilitate low or zero duty ratio operation. FIgure 8 shows the functional diagram. This function is programmed by connecting a capacitor between the G2C pin and GND. The capacitor on G2C should be slightly larger than the capacitor on the RAMP pin. For best results, program the typical G2 time limit to be between 1.5 and 3 times the switching period (T). Notice that when the G2 timer reaches its limit, both G1 and G2 are forced to a LOW output. This feature prevents the current in the output inductor from excessive negative excursions during zero-duty ratio conditions. Program the G2 time-out (G2TO) duration using equation (1): C G2C + 2 V RSET R RSET G2 Timeout Duration , Farads G2C Timer Threshold where D VRSET = 1.5 V(typ) D 1.5 T < G2 Timeout Duration < 3TS D G2C Timer Threshold = 2.5 V (typ) 16 www.ti.com (1) SLUS539A − JUNE 2004 − REVISED AUGUST 2004 APPLICATION INFORMATION RAMP pin: PWM Modulator and G1 Timer The RAMP pin serves two purposes: (1) programming the gain of the PWM modulator and (2) programming the time-out duration of G1 in case the main power stage has not caused a SYNC pulse to occur. A diagram of the PWM modulator and G1 timer is shown in Figure 9. The UCC2540 has a leading edge modulator that compares the error output with the RAMP voltage. The modulator frequency is externally driven through the SYNCIN pin. The RAMP pin provides both a sawtooth wave for the PWM comparator and it functions as G1 time-out protection that is programmed by RSET and the value of the RAMP capacitor. A switching cycle begins with the falling edge of the SYNCIN signal, which must be LOW for at least 50 nanoseconds. The falling edge of SYNCIN generates a 100 ns discharge strobe (CLK), to the RAMP function and then, allows the RAMP capacitor to charge from the 2 × IRSET current source. 2 y IRSET PWM Comparator VERR RAMP + 5 + 250 mV PWM Latch S Q PWM RD Q ENA 2.5 V + G1 Timeout Comparator CLK UDG−04048 Figure 9. PWM Modulator and G1 Time-Out Comparator Low-line or brownout conditions can cause the primary side duty ratio to approach 100% where parasitic converter impedances may temporarily impair the quality of the SYNCIN pulse. The RAMP timing function terminates the G1 pulse when the RAMP voltage exceeds 2.5 V. The duration of the RAMP timing function should be set as follows: C RAMP w ǒ 2 Ǔ V RSET R RSET TS PWM RAMP timeout threshold voltage (2) where D TS = switching period D VRSET = 1.5 (typ) D PWM(RAMP) = 2.5 V (typ) R SETC RAMP w 1.2 å Gain (PWM modulator) w 0.4 fS (3) In order to use the G1 Timer feature, the peak RAMP voltage at the end of a switch cycle should be as close to 2.5 V as the CRAMP and RRSET tolerances allow. In other words, the PWM modulator gain should be programmed to be equal to, or slightly greater than 0.4 inverse-V. www.ti.com 17 SLUS539A − JUNE 2004 − REVISED AUGUST 2004 APPLICATION INFORMATION SYNCIN pin A falling edge applied to the SYNCIN pin generates a narrow pulse that is the base timer for internal UCC2540 functions. The SYNCIN pulse must be HI for at least 100 ns preceding the falling edge and LOW for at least 50 ns in order to be registered as a valid pulse. Due to the critical nature of the timing, avoid filtering the falling edge of the SYNCIN signal in order to avoid signal delay. The peak SYNCIN voltage can easily range from between 2.5 V and 6.6 V, which allows a simple resistive divider to scale the secondary transformer voltage in post regulator applications. Situations where the line voltage varies more extensively or there is extensive ringing may call for clamping and/or additional gain. Ground Clamping In applications where a ring or a spike causes SYNCIN to fall below GND, protect the pin with a Schottky diode (cathode = SYNCIN, anode = GND). Overvoltage Clamping The SYNCIN signal may require overvoltage clamping in applications where the peak SYNCIN voltage is perilously close to the absolute maximum level of 8 V, due to either ringing or voltage levels. The REF or VDRV can be used as clamp voltages, as in Figure 10. Make sure that REF or VDRV always sources current. The reason is that both REF and VDRV are used to detect the mode of operation when they are back-driven and they could latch into the wrong operating mode at start-up. Main Output + RSR UCC2540 2 REF RRF G1 18 1 µF Auxiliary Output 4 SYNCIN G2 14 RSY UDG−04049 Figure 10. REF Clamp for SYNCIN. Note the REF Load Resistor. 18 www.ti.com SLUS539A − JUNE 2004 − REVISED AUGUST 2004 APPLICATION INFORMATION Another overvoltage clamping option is to directly clamp the SYNCIN pin. Unfortunately, Zener diodes have excessive junction capacitance which causes too much delay in the signal. However, a base-emitter clamp that achieves the desired clamping action can be employed with minimal delay to the SYNCIN signal. See Figure 11. Simply select RSR and (RCB + RBE) to give the appropriate 0 V to 3.3 V signal at low-line conditions. Then, select the ratio of RCB to RBE to cause the transistor to turn-on when SYNCIN exceeds 4 V. Main Output + UCC2540 RSR G1 18 Auxiliary Output 4 SYNCIN RCB G2 14 RBE UDG−04050 Figure 11. VBE Clamp for SYNCIN SYNCIN Clamping for the Isolated Cascaded Buck Topology The UCC2540 is ideally suited as a secondary side controller for the cascaded buck topology, when it is partnered with the UCC28089 primary side start-up controller. The primary side controller transmits a pulse edge during its dead time. The UCC2540 uses the primary-side pulse in order to provide zero voltage conditions for primary- and secondary-side switches. The predictive delay feature tunes the secondary-side transition to minimize reverse recovery losses in the synchronous rectifier. The pulse-edge information can vary with the primary side bias voltage and therefore, it must be clamped. The circuit shown in Figure 12 includes the appropriate pulse-edge shaping circuit, clamping and 1500-V isolation. The recommended transformer, COEV part # MGBBT−00011−01, is smaller than many opto-isolators. Primary Ground UCC28089 Secondary Ground UCC2540 C1 R1 634 Ω 680 pF REF SYNC RCB QCL 2N3904 L1 15 µH RBE GND SYNCIN T1 1:1 UDG−04051 Figure 12. Isolation and Clamping the SYNCIN Signal for Cascaded Buck Converters www.ti.com 19 SLUS539A − JUNE 2004 − REVISED AUGUST 2004 APPLICATION INFORMATION VDD, VDRV, VREF and BST pins: Modes of Operation Depending on the available bias voltage for the UCC2540, the startup, shutdown, and restart conditions are different. There are three distinct configurations or modes of biasing the UCC2540. The mode is detected and latched into an internal register during power-up when VREF crosses 2 V. The register is cleared when VDD, VDRV and VREF are simultaneously less than 1 V. All modes are compatible with either cascaded buck or with secondary-side post regulator (SSPR) topologies. The main bias voltage of Modes 1 and 2 can be implemented with a diode and a capacitor from an ac-voltage such as the secondary winding of the transformer. A summary of the modes and their programming requirements are listed in Table 1. Table 1. Modes and Programming Requirements Mode VBIAS Range (V) Bias Pin UVLO ON (V) UVLO OFF (V) 1 8.5 to 36 VDD [16] VVDD = 8.5 VVDD = 8.0 2 4.75 to 8.5 VDRV [13] VVDRV = 4.65 3 3.0 to 3.6 VREF [2] VREF = 3.0 Mode Requirement at Power-Up and VVREF = 2 V V VDD u ǒV VDRV and V REFǓ Remarks Widest line operation VVDRV = 4.3 V VDRV u ǒV VDD and V REFǓ VREF= 2.5 V REF u ǒV VDD and V VDRVǓ Needs regulated bias and low VTH power MOSFETs D Mode 1, or normal operation requires the availability of a bias of 8.5 V or higher for the device. Here, the bias drives the VDD pin. The low-side drive bias, VVDRV = 7 V, is generated from an internal linear regulator and it directly draws current from the VDD pin. The high-side driver bias is a flying capacitor that is charged from the VDRV pin through the G2 pin, when G2 is HI, via a diode between G2 and BST. The UCC2540 operates in Mode 1 if VVDD > (VVDRV and VVREF) when VVREF rises above 2 V. Mode 1 permits the widest range of bias voltages, operational from 8.5 V < VVDD < 35 V. This mode is compatible with systems that have a 12 VDC bias supply already available. Alternatively, Mode 1 is particularly useful for applications where the input line voltage varies over a wide range and the bias is to be derived directly from the reflected line voltage, such as in Fig. 13. D Mode 2 is suitable for applications where the bias is typically 5 V (between 4.5 V and 8.0 V). The bias voltage is applied to the VDRV terminal of the UCC2540. The high-side driver bias is a flying capacitor that is charged from the VDRV pin through the G2 pin, when G2 is HI. Bias voltage to the VDD pin is obtained through an external voltage-doubler charge pump. If the system uses low threshold voltage power MOSFETs, VDD can be directly tied to the VDRV pin. The bias voltage could be either a bus converter output or an auxiliary supply, or the reflected converter input voltage that originates from a regulated source. D Mode 3 is for synchronous buck converter applications where the bias voltage is a regulated 3.3-V source. This is a common main output voltage in multiple output power converters. The bias voltage is applied to the VREF pin of the UCC2540. The UCC2540 operates in Mode 3 if it detects (VVREF > VVDRV and VDD) when VVREF rises above 2 V. Assorted combinations of modes and biasing schemes are shown in Figure 13 through Figure 18. In Mode 1 and Mode 2, the bias voltage can either be an independent auxiliary supply or it can be generated by rectifying and filtering the reflected line voltage, as shown in Figure 13 through Figure 16. A regulated auxiliary supply must be used with Mode 3 because the tolerance of the VREF voltage is the control tolerance of the UCC2540. In Mode 3, the regulated auxiliary supply can be independent of the power supply input voltage (as shown in Figure 18) or, the regulated auxiliary supply can be the same source as the power supply input voltage. 20 www.ti.com SLUS539A − JUNE 2004 − REVISED AUGUST 2004 APPLICATION INFORMATION UCC2540 Drive (7.2 V) Regulator VDD VDRV VREF (3.3 V) Regulator VREF BST G1 High−Side Driver SWS SW Predictive Logic G2 Low−Side Driver G2S PGND D2 16 Rectified Bias 8.5 V ≤ VVDD ≤ 35 V Vin 13 0V 8.5 V ≤ VIN ≤ 35 V 2 19 C2 C3 C4 Q1 18 C1 20 D1 17 Q2 14 12 UDG−04038 15 Figure 13. Mode 1 With Rectified Biasing for Input Voltages Between 8.5 V and 35 V UCC2540 Drive (7.2 V) Regulator VDD VDRV VREF (3.3 V) Regulator VREF BST G1 High−Side Driver SWS SW Predictive Logic G2 Low−Side Driver G2S PGND AUX Bias 8.5 V ≤ VVDD ≤ 35 V 16 13 0 V ≤ VIN ≤ 35 V 2 Vin 19 C2 C3 C4 0V Q1 18 C1 20 D1 17 Q2 14 12 15 UDG−04039 Figure 14. Mode 1 With Auxiliary Biasing for Bias Voltages Between 8.5 V and 35 V www.ti.com 21 SLUS539A − JUNE 2004 − REVISED AUGUST 2004 APPLICATION INFORMATION UCC2540 Drive (7.2 V) Regulator VDD VDRV VREF (3.3 V) Regulator VREF BST G1 High−Side Driver SWS SW Predictive Logic G2 Low−Side Driver G2S PGND Rectified Bias 4.75 V ≤ VVDRV ≤ 8.0 V D2 16 D4 13 D3 2 Vin 0V 19 C2 C3 C4 C5 Q1 18 C1 20 D1 17 Q2 13 12 11 UDG−04040 Figure 15. Mode 2 With Rectified Biasing for Input Voltages Between 4.75 V and 8.0 V UCC2540 Drive (7.2 V) Regulator VDD VDRV VREF (3.3 V) Regulator VREF BST G1 High−Side Driver SWS SW Predictive Logic G2 Low−Side Driver G2S PGND AUX Bias 4.75 V ≤ VVDRV ≤ 8.0 V D2 16 13 D3 2 0 V ≤ VIN ≤ 35 V 19 C2 C3 C4 C5 Q1 Vin 18 0V C1 20 17 D1 Q2 13 12 11 UDG−04041 Figure 16. Mode 2 With Auxiliary Biasing for Bias Voltages Between 4.75 V and 8.0 V 22 www.ti.com SLUS539A − JUNE 2004 − REVISED AUGUST 2004 APPLICATION INFORMATION UCC2540 Drive (7.2 V) Regulator VDD VDRV VREF VREF (3.3 V) Regulator BST G1 High−Side Driver SWS SW Predictive Logic G2 Low−Side Driver G2S PGND AUX Bias 4.75 V ≤ VVDRV ≤ 8.0 V 16 13 0 V ≤ VIN ≤ 35 V 2 Vin 19 C2 C3 C4 0V Q1 (Low VTH) 18 C1 20 D1 17 Q2 (Low VTH) 14 12 15 UDG−04042 Figure 17. Mode 2 With Auxiliary Biasing for Bias Voltages Between 4.75 V and 8.0 V and Low Threshold Power MOSFET Transistors UCC2540 Drive (7.2 V) Regulator VDD VDRV VREF (3.3 V) Regulator VREF BST G1 High−Side Driver SWS SW Predictive Logic G2 Low−Side Driver G2S PGND D2 16 D3 13 Regulated 3.3-VDC Bias 4.75 V 3 VVDRV 3 8.0 V DC or Pulse Train 1.8 V 3 VIN 3 5 V 2 19 C2 C4 C5 Q1 (Low VTH) 18 C1 20 17 D1 Q2 (Low VTH) 13 12 11 UDG−04043 Figure 18. Mode 3 With Regulated 3.3-VDC Bias www.ti.com 23 SLUS539A − JUNE 2004 − REVISED AUGUST 2004 APPLICATION INFORMATION Charge Pump Capacitor Selection Capacitors C1 through C5 are all part of a charge distribution network that allows the UCC2540 to pass charge to the MOSFET gates of Q1 and Q2 (all reference designators in this section refer to the schematics in Figure 13 through Figure 18). This section gives guidelines on selecting the values of C1 through C5 so that the converter functions properly. Specific capacitor values may need to be larger than the recommended value due to MOSFET characteristics, diode D1 – D4 characteristics and closed-loop converter performance. All three modes of operation require a charge pump capacitor and diode, C1 and D1, in order to drive the high-side power MOSFET. Modes 2 and 3 require additional charge pump capacitors and diodes in order to supply voltage to VDD. In general, all charge pump diodes should be Schottky diodes in order to have low forward voltage and high speed. The charge pump capacitors should be ceramic capacitors with low effective series resistance (ESR), such as X5R or X7R capacitors. The value of the charge pump capacitor C1 depends on the power MOSFET gate charge and capacitance, the voltage level of the Miller plateau threshold, the forward drop of D1 and the closed-loop response time. The unloaded high-side gate driver typically draws 2 nC of charge per rising edge plus 30 µA of direct current from C1. Usually, the unloaded high−side gate driver load is miniscule compared to the gate charge requirements of the high-side power MOSFET, Q1. Typical values for C1 are approximately 50 to 100 times the input capacitance (CISS) of MOSFET Q1. This usually allows for transient operation at extremely large duty ratio, where C1 does not have sufficient time to fully recharge. If C1 is excessively large, its ESR and ESL prevents it from recharging during transients, including the start-up transient. Capacitors C2 through C5 are then selected based on the direction of charge transfer and the requirements of the UCC2540. Selection guidelines are shown in Table 2. Keep in mind that each converter design may require adjustments for larger capacitor ratios than those that are suggested in Table 2. The selection process begins at the left side of Table 2 and progresses towards the right side of the table, which is the reverse order of the charge flow during the first few cycles of start-up. If iteration is required in the design process, review the progression of the capacitors in the order from left to right that is shown in the table. Table 2. Charge Pump and Bias Capacitor Selection Guidelines Mode High-Side Drive Capacitor (≥ 0.1 µF) VDRV Filter Capacitor VREF Filter Capacitor VDD Filter Capacitor VDD Charging Capacitor 1 C1 > 50 CISS C3 > 2 × C1 C2 > 0.1 µF C4 > 1 µF n/a 2 C1 > 50 CISS C3 > 2 × C1 C2 > 0.1 µF C4 > 1 µF, 2 × C3 C5 > 2 × C4 C1 > 50 CISS C4 > 1 µF 2 × C1 C2 > 1.0 µF C4 > 1 µF, 2 × C1 C5 > 2 × C4 3 For Modes 2 and 3, the VDD filter capacitor, C4, in Table 2 must supply the IVDD idle current to the UCC2540 (approximately 11 mA) plus the charge to drive the gates G1 and G2. Capacitor C4 must be large enough to sustain adequate operating voltages during start-ups and other transients under the full operational IVDD current. Knowing the operating frequency and the MOSFET gate charges (QG), the average IVDD current can be estimated as: I VDD + I VDD(idle) ) ǒQ G1 ) Q G2Ǔ fS (4) D where fS is switching frequency In order to prevent noise problems, C4 must be at least 1 µF. Furthermore, it needs to be large enough to pass charge along to the power MOSFET gates. Thus C4 often needs to have at least twice the capacitance of the VDRV filter capacitor, as shown in Table 2. 24 www.ti.com SLUS539A − JUNE 2004 − REVISED AUGUST 2004 APPLICATION INFORMATION Output Stage The UCC2540 includes dual gate drive outputs and each is capable of ±3-A peak current. The pull-up/ pull-down circuits of the driver are bipolar and MOSFET transistors in parallel. High-side and low-side dual drivers provide a true 3-A high-current capability at the MOSFET’s Miller Plateau switching region where it is most needed. The peak output current rating is the combined current from the bipolar and MOSFET transistors. The output resistance is the RDS(on) of the MOSFET transistor when the voltage on the driver output is less than the saturation voltage of the bipolar transistor. The output drivers can switch from VDD to GND. Each output stage also provides a very low impedance to overshoot and undershoot. This means that in many cases, external-schottky-clamp diodes are not required. The outputs are also designed to withstand 500-mA reverse current without either damage to the device or logic upset. For additional information on drive current requirements at MOSFET’s Miller plateau region, refer to the Power Supply Seminar SEM−1400[2] and the UCC37323/4/5 datasheet[3]. Predictive Gate DriveTM Technology The Predictive Gate Drive technology maximizes efficiency by minimizing body diode conduction. It utilizes a digital feedback system to detect body diode conduction, and adjusts the deadtime delays to minimize the conduction time interval. This closed loop system virtually eliminates body diode conduction while adjusting for different MOSFETs, temperature, and load dependent delays. Since the power dissipation is minimized, a higher switching frequency can be utilized, allowing for a smaller component size. Precise gate timing at the nanosecond level reduces the reverse recovery time of the synchronous rectifier MOSFET body diode, which reduces reverse recovery losses seen in the main (high-side) MOSFET. Finally, the lower power dissipation results in increased reliability. 19 BST 18 G1 17 SW Predictive Logic 20 SWS 12 G2S VDRV 14 G2 15 PGND UDG−02149 Figure 19. For additional information on Predictive Gate Drive control and efficiency comparisons to earlier adaptive delay and adaptive control techniques, refer to the UCC27223 datasheet [3]. www.ti.com 25 SLUS539A − JUNE 2004 − REVISED AUGUST 2004 APPLICATION INFORMATION VDD and IDD Although quiescent VDD current is low, total supply current is higher, depending on output gate drive requirements and the programmed oscillator frequency. Total VDD current (IVDD) is the sum of quiescent VDD current and the average output currents of G1 and G2, as described in equation (3). Knowing the operating frequency and the MOSFET gate charge (QG), average driver output current, per gate, can be calculated from: IG + QG fS (5) where D fS is switching frequency To prevent noise problems, connect a 1-µF ceramic capacitor between the VDD and GND pins. Place the 1-µF ceramic capacitor as close to the UCC2540 as possible. This capacitor is in addition to any electrolytic energy storage capacitors that may be used in the bias supply design. Soft-Start and Tracking Features Separate pins are provided for the soft-start feature and the tracking feature. Soft-start or tracking (sequencing) can be easily implemented with this configuration using a minimum number of external components. During a power-up transient, the converter output tracks the lower of the SS voltage, the TR voltage or a 1.5-V internal reference, provided the system is not in current limit. In other words, the voltage control loop is closed during power-up, provided the system is not current limited. Figure 20 shows the UCC2540 configured for soft-start operation. For applications that do not use the tracking feature, connect the TR pin to either SS or REF, as shown in the figure. Remote shutdown and sequential power-up can be easily implemented as a transistor switch across CSS. TR UCC2540 10 REF (3.3 V) Voltage 1.33 y IRSET SS Error 0.7 V + 11 1.5 V + + + Amplifier COMP CSS 1.73 y IRSET UVLO VEA− HUP 50 mV + 7 To Positive Input of Current Error Amplifier UDG−04045 Figure 20. Using the Soft-Start Feature 26 www.ti.com SLUS539A − JUNE 2004 − REVISED AUGUST 2004 APPLICATION INFORMATION The soft-start interval begins when the UCC2540 recognizes that the appropriate voltage (see Mode 1, 2 or 3) is above the UVLO level. The voltage of CSS then linearly increases until it is clamped at the REF voltage of 3.3V. Regulation should be reached when the soft-start voltage reaches about 2.2 V (1.5 V plus a diode drop). Select a CSS capacitor value using equation (5) to program a desired soft-start duration, ∆tSS. V RSET R SET C SS + 1.33 Dt SS + 1.33 DV SS Dt SS Farads 2.2 V 1.5 V R SET (6) If a UVLO fault is encountered, both outputs of the UCC2540 are disabled and the soft-start pin (SS) is discharged to GND. The UCC2540 does not retry until the UVLO fault is cleared. Using the TR pin, the UCC2540 can be programmed to track another converter output voltage. If the voltage to be tracked is between 0 V and 3.3 V, simply connect the TR pin to the voltage to be tracked with a resistor that is approximately equal to the DC impedance that is connected to the VEA− terminal (RV1 || RV2, in Figure 6). If the voltage is above that range, use a voltage divider, again with an equivalent resistance that approximately equals the DC impedance that is connected to the VEA− terminal. Other strategies can be used to achieve sequential, ratiometric or simultaneous power supply tracking[14]. An implementation of sequential sequencing of a multiple output power supply[5] is shown in Figure 21. Applications where the loads include a processor with a core voltage of 1.5 V and I/O ports that require 3.3 V can require sequential sequencing in order to resolve system level bus contention problems during start-up. In this circumstance the core must power-up first, then after an initialization period of 130 ms, the ports are allowed to power-up. 5V From Transformer Secondary UCC2540 TPS3103K33 TR G1 SS G2 0V I/O 3.3 V RESET VDD CSS GND PFO 1.6 kΩ MR PFI 1 kΩ UCC2540 10 kΩ REF TR G1 Core 1.5 V G2 SS CSS UDG−04061 Figure 21. Sequencing a Multiple Output Post Regulated Power Supply www.ti.com 27 SLUS539A − JUNE 2004 − REVISED AUGUST 2004 APPLICATION INFORMATION Regulation loss due to loss of primary line voltage V − Voltage − V 130 ms 3.3 1.5 1.43 VI/O VCORE 0 t − Time UDG−04061 Figure 22. Using the TR pin, the UCC2540 can be programmed to ratio-metrically track another converter output voltage[5]. Ratio-metric tracking is when the ratio of the output voltages is constant from zero volts to the point where one or more of the outputs lock into regulation. The TR pin is easier to use for tracking than the SS pin because the external currents that would be applied to the SS pin may interfere with SS discharge currents and fault recovery. It should be understood that the voltage that is being tracked must lag the bias voltages (VDD, VDRV and REF) on start-up and lead the bias voltages during shutdown. Furthermore, the output that is being tracked must not reach its steady state DC level before the output that is tracking reaches its steady state DC level. Figure 23 illustrates the concept of programming an output voltage VC, to ratio-metrically track another output, VM. VM (Leader) Main Power Supply (Leader) + MM VC (Tracker) VM (a) ratio−metric sequencing MC VM (Leader) MM Core Power Supply (Trader) VC UCC2540 7 + MC TR VC (Tracker) VM (Leader) Tracking Ratio AT ^ ǒ Ǔ MC MM MM VC (Tracker) MC Figure 23. Ratio−Metric Tracking 28 www.ti.com (b) simultaneous sequencing (c) ratio−metric sequencing UDG−04061 SLUS539A − JUNE 2004 − REVISED AUGUST 2004 APPLICATION INFORMATION The general circuit to program the UCC2540 to track the leader supply voltage by the tracking ratio AT is shown in Figure 24. To program the tracking profile gains GT1 and GT2, follow the ratio-metric tracking design procedure that is listed below. The special case of simultaneous sequencing for VM > 1.5V is the simplest to design; set RT1= RV1 and RT2= RV2, GT2 is not needed. In many other cases, the circuit can be simplified with the removal of the operational amplifier for GT2 and the Zener clamping diode. If an operational amplifier is necessary, it should be capable of rail to rail operation and usually low voltage bias; the TLV271 is an inexpensive solution for both of those requirements. Notice that the tracking circuit in Figure 24 also has a soft-start capacitor, CSS. The soft-start capacitor is useful for limiting the time between short-circuit retry attempts and it can prevent overshoot when recovering from a fault that is experienced in only the tracking supply but not the main supply. Ratio-Metric Tracking Design Procedure (see Figures 21 and 22) 1. Determine the tracking ratio, AT. AT + MC MM (7) where MC and MM are the soft-start slopes of VC and VM, respectively. 2. Determine GV. GV + R V2 R V1 ) R V2 (8) where RV2 and RV1are selected when designing the voltage control loop. 3. Test GT2 if necessary when VM ≤ 1.5 V or ATGV > 1. a. If GT2 is needed, set GT2 so that both equations (8) and (9) apply. G T2 + 1 ) R F1 R F2 (9) so that both of the following apply: G T2 + ǒ Ǔ 1.5 V V M G T1 and G T2 u ǒA T G VǓ (10) b. If GT2 is not needed, set GT2 = 1. 4. Set GT1 . G T1 + AT GV R T2 + R T1 ) R T2 G T2 (11) 5. Select RT1 and RT2 so that RT1 || RT2 ≈ RV1 || RV2 to minimize offset differences. www.ti.com 29 SLUS539A − JUNE 2004 − REVISED AUGUST 2004 RF2 TLV271 RF1 G T2 + 1 ) R F1 R F2 Main Power Supply (Leader) Use GT2 stage if ATGV > 1 OR if VMGT1 ≤ 1.5 V at steady-state + + VM Rectified Secondary Voltage TR RT2 G T1 + R T2 R T1 ) R T2 nVIN UCC2540 RT1 0V G1 + *DZ 3.3 V SS VC G2 CSS VEA− RV1 RV2 *DZ needed only if VMGT1GT2 > 3 V GV + R V2 R V1 ) R V2 Determined by voltage loop design UDG−04059 Figure 24. Programming the UCC2540 to Track Another Output More elaborate power supply sequencing and tracking can easily be implemented by extending the above techniques. Consult reference [5] for further information THERMAL INFORMATION The useful temperature range of a controller that contains high-current output drivers is greatly affected by the drive power requirements of the load and the thermal characteristics of the device package. In order for a power driver to be useful over a particular temperature range the package must allow for the efficient removal of the heat produced while keeping the junction temperature within rated limits. The UCC2540 is available in the 20-pin HTSSOP PowerPADt package. The PowerPADTM HTSSOP-20 (PWP) offers the most effective means of removing the heat from the semiconductor junction and therefore long term reliability improvement. As illustrated in [5], the PowerPAD packages offer a leadframe die pad that is exposed at the base of the package. This pad is soldered to the copper on the PC board directly underneath the device package, reducing the θjc down to 2°C/W. Data is presented in [5] to show that the power dissipation can be quadrupled in the PowerPAD configuration when compared to the standard packages. The PC board must be designed with thermal lands and thermal vias to complete the heat removal subsystem, as summarized in [6] to realize a significant improvement in heat−sinking over standard non-PowerPAD surface mount packages. 30 www.ti.com SLUS539A − JUNE 2004 − REVISED AUGUST 2004 TYPICAL CHARACTERISTICS RAMP CURRENT vs TEMPERATURE OUTPUT REFERENCE VOLTAGE vs TEMPERATURE −275 3.40 −285 IRAMP − Ramp Current − µA VVREF − Reference Voltage − V RRSET = 10 kΩ 3.35 3.30 3.25 −295 −305 −315 3.20 −50 0 50 100 −325 −50 150 0 Figure 25 1.1 IG2C/IRAMP AND ISS/IRAMP vs TEMPERATURE 7.6 VVDRV − Regulator Output Voltage − V IG2C/IRAMP, RRSET =50kΩ 0.9 µA/µA 150 REGULATOR OUTPUT VOLTAGE vs TEMPERATURE MODE 1 1.0 0.8 ISS/IRAMP, RRSET = 10kΩ 0.7 ISS/IRAMP, RRSET = 50kΩ 0.6 0 50 100 TJ − Junction Temperature − °C 100 Figure 26 IG2C/IRAMP, RRSET =10kΩ 0.5 −50 50 TJ − Junction Temperature − °C TJ − Junction Temperature − °C 150 7.4 7.2 7.0 6.8 −50 0 50 100 150 TJ − Junction Temperature − °C Figure 27 Figure 28 www.ti.com 31 SLUS539A − JUNE 2004 − REVISED AUGUST 2004 TYPICAL CHARACTERISTICS TRACKING TO VOLTAGE ERROR AMPLIFIER OFFSET vs TEMPERATURE 50 40 30 20 10 0 50 100 TJ − Junction Temperature − °C 53 51 49 47 45 −50 150 0 Figure 29 SYNCIN THRESHOLD VOLTAGE vs TEMPERATURE 150 INVERTING AMPLIFIER GAIN AND PHASE vs FREQUENCY 1.75 1.70 5 0 0 −45 −5 1.65 1.60 Phase −90 Gain −10 −135 −15 −180 −20 −225 1.55 1.50 −50 −25 1k 0 50 100 TJ − Junction Temperature − °C 150 10 k 100 k 1M f − Frequency − Hz Figure 32 Figure 31 32 100 Figure 30 Gain − dB VSYNCHIN − Timing Signal Voltage − V 1.80 50 TJ − Junction Temperature − °C www.ti.com 10 M −270 100 M Phase − ° VTR(OS) − Offset Voltage − mV 60 0 −50 55 VCEA− − Current Error Amplifier Offset Voltage − mV 70 CURRENT ERROR AMPLIFIER OFFSET vs TEMPERATURE SLUS539A − JUNE 2004 − REVISED AUGUST 2004 TYPICAL CHARACTERISTICS VOLTAGE ERROR AMPLIFIER GAIN AND PHASE vs FREQUENCY CURRRENT ERROR AMPLIFIER GAIN AND PHASE vs FREQUENCY 120 0 120 100 100 Gain 80 −45 20 60 Gain − dB −90 40 Phase − ° 60 −90 40 Phase 20 Phase 0 −135 0 −45 Phase − ° Gain 80 Gain − dB 0 −135 −20 −20 −40 1 10 100 1k −180 10 k 100 k 1 M 10 M 100 M −40 1 10 100 1k −180 10 k 100 k 1 M 10 M 100 M f − Frequency − Hz f − Frequency − Hz Figure 33 Figure 34 OPERATING CURRENT (DC) vs BIAS VOLTAGE 12 IVDD − Bias Current − mA 10 8 6 4 2 0 0 5 10 15 20 25 30 VVDD − Bias Voltage − V 35 40 Figure 35 www.ti.com 33 SLUS539A − JUNE 2004 − REVISED AUGUST 2004 TYPICAL CHARACTERISTICS G2 Lower Gate Drive (5V/div) G2 Lower Gate Drive (5V/div) Predictive Delay Adjustment SW Node (500 mV/div) SW Node (5V/div) Synchronous FET Body Diode Conduction t − Time − 20 ns/div t − Time − 20 ns/div Figure 36. Predictive Gate Drive − G2 Falling Figure 37. Predictive Gate Drive − G2 Falling G2 Lower Gate Drive Predictive Delay Adjustment SW Node t − Time − 20 ns/div Figure 38. Predictive Gate Drive − G2 Falling 34 www.ti.com SLUS539A − JUNE 2004 − REVISED AUGUST 2004 RELATED PRODUCTS D D D D D D UCC28089 Primary Side Push-Pull Oscillator UCC27223 High Efficiency Predictive Synchronous Buck Driver with Enable UCC3583 Switch Mode Secondary Side Post Regulator UCC25701 Advanced Voltage Mode Pulse Width Modulator UCC3808A Low-Power Currrent-Mode Push-Pull PWM UCC38083/4/5/6 8-Pin Current-Mode Push-Pull PWM with Programmable Slope Compensation REFERENCES 1. Power Supply Seminar SEM−1300 Topic 1: Unique Cascaded Power Converter Topology for High Current Low Output Voltage Applications, by L. Balogh, C. Bridge, and B. Andreycak, (SLUP118) 2. Power Supply Seminar SEM−1400 Topic 2: Design And Application Guide For High Speed MOSFET Gate Drive Circuits, by L. Balogh, (SLUP133) 3. Datasheet, UCC27223 High Efficiency Predictive Synchronous Buck Driver, (SLUS558) 4. Datasheet, UCC37323/4/5 Dual 4−A Peak High Speed Low−Side Power MOSFET Drivers, (SLUS492A) 5. Power Supply Seminar SEM1600 Topic 2: Sequencing Power Supplies in Multiple Voltage Rail Environments, by D. Daniels, D. Gehrke, and M. Segal, (SLUP224) 6. Technical Brief, PowerPAD Thermally Enhanced Package, (SLMA002) 7. Application Brief, PowerPAD Made Easy, (SLMA004) 8. Datasheet, TPS3103K33 Ultra-Low Supply Current/Supply Voltage Supervisory Circuits, (SLVS363) 9. Application Note, A Revolutionary Power Management Solution for Highly Efficient, Multiple Output Applications, by Bill Andreycak, (SLUA255) 10. Application Note, Predictive Gate DriveE FAQ, by Steve Mappus (SLUA285) www.ti.com 35 SLUS539A − JUNE 2004 − REVISED AUGUST 2004 MECHANICAL DATA PWP (R-PDSO-G**) PowerPAD PLASTIC SMALL-OUTLINE 20 PINS SHOWN 0,30 0,19 0,65 20 0,10 M 11 Thermal Pad (See Notes D and F) 4,50 4,30 0,15 NOM 6,60 6,20 Gage Plane 1 10 0,25 A 0°−ā 8° 0,75 0,50 Seating Plane 0,15 0,05 1,20 MAX PINS ** 0,10 14 16 20 24 28 A MAX 5,10 5,10 6,60 7,90 9,80 A MIN 4,90 4,90 6,40 7,70 9,60 DIM 4073225/F 10/98 NOTES: A. B. C. D. All linear dimensions are in millimeters. This drawing is subject to change without notice. Body dimensions do not include mold flash or protrusions. The package thermal performance may be enhanced by bonding the thermal pad to an external thermal plane. This pad is electrically and thermally connected to the backside of the die and possibly selected leads. E. Falls within JEDEC MO-153 F. The PowerPADt is not directly connected to any lead of the package. It is electrically and thermally connected to the substrate of the device which acts as ground and should be connected to PGND on the PCB. The exposed dimension is 1.3 mm x 1.7 mm. However, the tolerances can be +1.05 mm / −0.05 mm (+41 mils / −2 mils) due to position and mold flow variation. Powe 36 www.ti.com IMPORTANT NOTICE Texas Instruments Incorporated and its subsidiaries (TI) reserve the right to make corrections, modifications, enhancements, improvements, and other changes to its products and services at any time and to discontinue any product or service without notice. Customers should obtain the latest relevant information before placing orders and should verify that such information is current and complete. All products are sold subject to TI’s terms and conditions of sale supplied at the time of order acknowledgment. TI warrants performance of its hardware products to the specifications applicable at the time of sale in accordance with TI’s standard warranty. Testing and other quality control techniques are used to the extent TI deems necessary to support this warranty. Except where mandated by government requirements, testing of all parameters of each product is not necessarily performed. TI assumes no liability for applications assistance or customer product design. Customers are responsible for their products and applications using TI components. To minimize the risks associated with customer products and applications, customers should provide adequate design and operating safeguards. TI does not warrant or represent that any license, either express or implied, is granted under any TI patent right, copyright, mask work right, or other TI intellectual property right relating to any combination, machine, or process in which TI products or services are used. Information published by TI regarding third-party products or services does not constitute a license from TI to use such products or services or a warranty or endorsement thereof. Use of such information may require a license from a third party under the patents or other intellectual property of the third party, or a license from TI under the patents or other intellectual property of TI. Reproduction of information in TI data books or data sheets is permissible only if reproduction is without alteration and is accompanied by all associated warranties, conditions, limitations, and notices. Reproduction of this information with alteration is an unfair and deceptive business practice. TI is not responsible or liable for such altered documentation. Resale of TI products or services with statements different from or beyond the parameters stated by TI for that product or service voids all express and any implied warranties for the associated TI product or service and is an unfair and deceptive business practice. TI is not responsible or liable for any such statements. Following are URLs where you can obtain information on other Texas Instruments products and application solutions: Products Applications Amplifiers amplifier.ti.com Audio www.ti.com/audio Data Converters dataconverter.ti.com Automotive www.ti.com/automotive DSP dsp.ti.com Broadband www.ti.com/broadband Interface interface.ti.com Digital Control www.ti.com/digitalcontrol Logic logic.ti.com Military www.ti.com/military Power Mgmt power.ti.com Optical Networking www.ti.com/opticalnetwork Microcontrollers microcontroller.ti.com Security www.ti.com/security Telephony www.ti.com/telephony Video & Imaging www.ti.com/video Wireless www.ti.com/wireless Mailing Address: Texas Instruments Post Office Box 655303 Dallas, Texas 75265 Copyright 2004, Texas Instruments Incorporated