LTC1960 Dual Battery Charger/ Selector with SPI Interface DESCRIPTION FEATURES Complete Dual-Battery Charger/Selector System Serial SPI Interface Allows External C Control and Monitoring Simultaneous Dual-Battery Discharge Extends Run Time by Typically 10% Simultaneous Dual-Battery Charging Reduces Charging Time by Up to 50% Automatic PowerPathTM Switching in <10s Prevents Power Interruption Circuit Breaker Protects Against Overcurrent Faults 5% Accurate Adapter Current Limit Maximizes Charging Rate 95% Efficient Synchronous Buck Charger Charger Has Low 0.5V Dropout Voltage No Audible Noise Generation, Even with Ceramic Capacitors 11-Bit VDAC Delivers 0.8% Voltage Accuracy 10-Bit IDAC Delivers 5% Current Accuracy VIN Up to 32V; VBATT Up to 28V Available in 5mm x 7mm 38-Pin QFN and 36-Pin Narrow SSOP Packages n n n n n n n n n n n n n n APPLICATIONS n A proprietary PowerPath architecture supports simultaneous charging or discharging of both batteries. Typical battery run times are extended by 10%, while charging times are reduced by up to 50%. The LTC1960 automatically switches between power sources in less than 10s to prevent power interruption upon battery or wall adapter removal. The synchronous buck battery charger delivers 95% efficiency with only 0.5V dropout voltage, and prevents audible noise in all operating modes. Patented input current limiting with 5% accuracy charges batteries in the shortest possible time without overloading the wall adapter. The LTC1960's 5mm x 7mm 38-pin QFN and 36-pin narrow SSOP packages allow implementation of a complete SBS-compliant dual battery system while consuming minimum PCB area. L, LT, LTC, LTM, Linear Technology and the Linear logo are registered trademarks and No RSENSE and PowerPath are trademarks of Linear Technology Corporation. All other trademarks are the property of their respective owners. Protected by U.S. Patents, including 5481178, 5723970, 6304066, 6580258. Portable Computers Portable Instruments n The LTC(R)1960 is a highly integrated battery charger and selector intended for portable products using dual smart batteries. A serial SPI interface allows an external microcontroller to control and monitor status of both batteries. TYPICAL APPLICATION Dual vs Sequential Charging LTC1960 Dual Battery/Selector System Architecture BATTERY CURRENT (mA) DC IN SYSTEM POWER LTC1960 MICROCONTROLLER 4 BAT2 BAT1 SMBus SPI 1960 TA01 3500 3000 2500 2000 1500 1000 500 0 3500 3000 2500 2000 1500 1000 500 0 BAT1 CURRENT BAT2 CURRENT SEQUENTIAL BAT1 CURRENT BAT2 CURRENT DUAL 100 MINUTES 0 50 100 150 200 TIME (MINUTES) 250 300 BATTERY TYPE: 10.8V Li-Ion (MOLTECH NI2020) REQUESTED CURRENT = 3A REQUESTED VOLTAGE = 12.3V MAX CHARGER CURRENT = 4.1A 1960 TA01b 1960fb 1 LTC1960 ABSOLUTE MAXIMUM RATINGS (Note 1) Voltage from DCIN, SCP, SCN, CLP, VPLUS, SW to GND.................................................. -0.3V to 32V Voltage from SCH1, SCH2 to GND.............. -0.3V to 28V Voltage from BOOST to GND.......................-0.3V to 41V PGND with Respect to GND....................................0.3V CSP, CSN, BAT1, BAT2 to GND....................... -5V to 28V LOPWR, DCDIV to GND.............................. -0.3V to 10V SSB, SCK, MOSI, MISO to GND.................... -0.3V to 7V COMP1 to GND............................................. -0.3V to 5V Operating Ambient Temperature Range (Note 7)............................................. 0C to 70C Operating Junction Temperature............. -40C to 125C Storage Temperature.............................. -65C to 150C Lead Temperature (Soldering, 10 sec) SSOP Only......................................................... 300C PIN CONFIGURATION 34 GCH1 SCN 4 33 SCH1 SCP 5 GDCO GDCO 25 GCH2 3 GDCI 2 BAT1 GB1O BAT2 TOP VIEW GB1I 36 SCH2 GB2O 1 GB2I VPLUS LOPWR TOP VIEW 38 37 36 35 34 33 32 VSET 1 31 SCP 32 TGATE ITH 2 30 SCN 6 31 BOOST ISET 3 29 BAT1 GDCI 7 30 SW GND 4 28 BAT2 GB1O 8 29 DCIN 28 VCC 25 COMP1 VSET 13 24 CLP ITH 14 23 CSP ISET 15 22 CSN GND 16 21 MOSI DCDIV 17 20 MISO SSB 18 23 GCH1 GND 10 22 SCH1 CSN 11 21 TGATE 20 BOOST CSP 12 13 14 15 16 17 18 19 19 SCK SW LOPWR 12 24 GCH2 MOSI 9 DCIN 26 PGND VCC GB2I 11 25 SCH2 MISO 8 BGATE 27 BGATE PGND GB2O 10 26 GND 39 SCK 7 CLP 9 27 VPLUS SSB 6 COMP1 GB1I DCDIV 5 UHF PACKAGE 38-LEAD (5mm x 7mm) PLASTIC QFN G PACKAGE 36-LEAD PLASTIC SSOP TJMAX = 125C, JA = 34C/W THE EXPOSED PAD (PIN 39) IS GND. MUST BE SOLDERED TO THE PCB TJMAX = 125C, JA = 70C/W ORDER INFORMATION LEAD FREE FINISH TAPE AND REEL PART MARKING* PACKAGE DESCRIPTION TEMPERATURE RANGE LTC1960CG#PBF LTC1960CG#TRPBF LTC1960CG 36-Lead Plastic SSOP 0C to 70C LTC1960CUHF#PBF LTC1960CUHF#TRPBF 1960 38-Lead (5mm x 7mm) Plastic QFN 0C to 70C Consult LTC Marketing for parts specified with wider operating temperature ranges. Consult LTC Marketing for information on non-standard lead based finish parts. For more information on lead free part marking, go to: http://www.linear.com/leadfree/ For more information on tape and reel specifications, go to: http://www.linear.com/tapeandreel/ 1960fb 2 LTC1960 ELECTRICAL CHARACTERISTICS The l denotes the specifications which apply over the full operating temperature range (Note 7), otherwise specifications are at TA = 25C. VDCIN = 20V, VBAT1 = 12V, VBAT2 = 12V, unless otherwise noted. SYMBOL PARAMETER CONDITIONS MIN TYP MAX UNITS DCIN Operating Range DCIN Selected 6 DCIN Operating Current Not Charging (DCIN Selected) Charging (DCIN Selected) Battery Operating Voltage Range Battery Selected, PowerPath Function (Note 2) Battery Drain Current Battery Selected, Not Charging, VDCIN = 0V 175 A VFDC VFB1 VFB2 VFSCN VPLUS Diodes Forward Voltage: DCIN to VPLUS BAT1 to VPLUS BAT2 to VPLUS SCN to VPLUS IVCC = 10mA IVCC = 0mA IVCC = 0mA IVCC = 0mA 0.8 0.7 0.7 0.7 V V V V UVLO Undervoltage Lockout Threshold VPLUS Ramping Down, Measured at VPLUS to GND VPLUS Rising, Measured at VPLUS to GND Supply and Reference ICH UVHYS UV Lockout Hysteresis VVCC VCC Regulator Output Voltage VLDR VCC Load Regulation 1 1.3 6 l 28 V 1.5 2 mA mA 28 V 3 3.5 3.9 5 5.2 5.4 V 0.2 1 % 60 IVCC = 0mA to 10mA V mV Switching Regulator VTOL ITOL Overall Voltage Accuracy Overall Current Accuracy 5V VOUT < 25V, (Note 3) IDAC Value = 3FFHEX VCSP , VCSN = 12V fOSC Regulator Switching Frequency fDO Regulator Switching Frequency in Low Dropout Mode DCMAX Regulator Maximum Duty Cycle IMAX Maximum Current Sense Threshold VITH = 2.2V ISNS CA1 Input Bias Current VCSP = VCSN > 5V CMSL CAI Input Common Mode Low CMSH CAI Input Common Mode High VCL1 CL1 Turn-On Threshold TG tr TG tf TGATE Transition Time: TGATE Rise Time TGATE Fall Time BG tr BG tf l -0.8 -1 0.8 1 % % l -5 -6 5 6 % % 345 kHz Duty Cycle 99% 255 300 20 25 kHz 99 99.5 % 140 155 190 150 A 0 V VDCIN-0.2 95 mV V 100 105 mV CLOAD = 3300pF, 10% to 90% CLOAD = 3300pF, 10% to 90% 50 50 90 90 ns ns BGATE Transition Time: BGATE Rise Time BGATE Fall Time CLOAD = 3300pF, 10% to 90% CLOAD = 3300pF, 10% to 90% 50 40 90 80 ns ns VTR DCDIV/LOPWR Threshold VDCDIV or VLOPWR Falling 1.19 1.215 V Trip Points l 1.166 VTHYS DCDIV/LOPWR Hysteresis Voltage VDCDIV or VLOPWR Rising 30 IBVT DCDIV/LOPWR Input Bias Current VDCDIV or VLOPWR = 1.19V 20 200 nA VTSC Short-Circuit Comparator Threshold VSCP - VSCN, VCC 5V 90 100 115 mV VFTO Fast PowerPath Turn-Off Threshold VDCDIV Rising from VCC 6 7 7.9 V VOVSD Overvoltage Shutdown Threshold as a Percent of Programmed Charger Voltage VSET Rising from 0.8V Until TGATE and BGATE Stop Switching l 107 mV % 1960fb 3 LTC1960 ELECTRICAL CHARACTERISTICS The l denotes the specifications which apply over the full operating temperature range (Note 7), otherwise specifications are at TA = 25C. VDCIN = 20V, VBAT1 = 12V, VBAT2 = 12V, unless otherwise noted. SYMBOL PARAMETER CONDITIONS MIN TYP MAX UNITS DACs IRES IDAC Resolution tIP tILOW IDAC Pulse Period: Normal Mode Low Current Mode VRES VDAC Resolution VSTEP VDAC Granularity VOFF VDAC Offset tVP VDAC Pulse Period Guaranteed Monotonic Above IMAX /16 10 6 Guaranteed Monotonic (5V < VBAT < 25V) bits 10 50 15 11 bits 16 (Note 6) mV 0.8 7 s ms V 11 16.5 s Charge MUX Switches tONC GCH1/GCH2 Turn-On Time VGCHX - VSCHX > 3V, CLOAD = 3nF 5 10 ms tONC GCH1/GCH2 Turn-Off Time VGCHX - VSCHX < 1V, from Time of VCSN < VBATX - 30mV, CLOAD = 3nF 3 7 s VCON CH Gate Clamp Voltage GCH1 GCH2 ILOAD = 1A VGCH1 - VSCH1 VGCH2 - VSCH2 5 5 5.8 5.8 7 7 V V VCOFF CH Gate Off Voltage GCH1 GCH2 ILOAD = 10A VGCH1 - VSCH1 VGCH2 - VSCH2 -0.8 -0.8 -0.4 -0.4 0 0 V V VTOC CH Switch Reverse Turn-Off Voltage VCSN - VBATX, 5V VBATX 28V l 5 20 40 mV VFC CH Switch Forward Regulation Voltage VBATX - VCSN, 5V VBATX 28V l 15 35 60 mV IOC(SRC) IOC(SNK) GCH1/GCH2 Active Regulation: Max Source Current Max Sink Current VGCHX - VSCHX = 1.5V VCHMIN BATX Voltage Below Which Charging Is Inhibited (Note 8) -2 2 3.5 A A 4.7 V PowerPath Switches tDLY Blanking Period After UVLO Trip Switches Held Off tPPB Blanking Period After LOPWR Trip Switches in 3-Diode Mode tONPO GB1O/GB2O/GDCO Turn-On Time VGS < -3V, from Time of Battery/DC Removal, or LOPWR Indication l 5 10 s tOFFPO GB1O/GB2O/GDCO Turn-Off Time VGS > -1V, from Time of Battery/DC Removal, or LOPWR Indication l 3 7 s VPONO Output Gate Clamp Voltage GB1O GB2O GDCO ILOAD = 1A Highest (VBAT1 or VSCP) - VGB1O Highest (VBAT2 or VSCP) - VGB2O Highest (VDCIN or VSCP) - VGDCO 6.25 6.25 6.25 7 7 7 V V V VPOFFO Output Gate Off Voltage GB1O GB2O GDCO ILOAD = -25A Highest (VBAT1 or VSCP) - VGB1O Highest (VBAT2 or VSCP) - VGB2O Highest (VDCIN or VSCP) - VGDCO 0.18 0.18 0.18 0.25 0.25 0.25 V V V VTOP PowerPath Switch Reverse Turn-Off Voltage VSCP - VBATX or VSCP - VDCIN 6V VSCP 28V l 5 20 60 mV VFP PowerPath Switch Forward Regulation Voltage VBATX - VSCP or VDCIN - VSCP 6V VSCP 28V l 0 25 50 mV IOP(SRC) IOP(SNK) GDCI/GB1I/GB2I Active Regulation Source Current Sink Current (Note 4) 4.75 4.75 4.75 250 ms 1 sec -4 75 A A 1960fb 4 LTC1960 ELECTRICAL CHARACTERISTICS The l denotes the specifications which apply over the full operating temperature range (Note 7), otherwise specifications are at TA = 25C. VDCIN = 20V, VBAT1 = 12V, VBAT2 = 12V, unless otherwise noted. SYMBOL PARAMETER tONPI Gate B1I/B2I/DCI Turn-On Time CONDITIONS MIN TYP MAX UNITS VGS < -3V, CLOAD = 3nF (Note 5) 300 s 10 s tOFFPI Gate B1I/B2I/DCI Turn-Off Time VGS > -1V, CLOAD = 3nF (Note 5) VPONI Input Gate Clamp Voltage GB1I GB2I GDCI ILOAD = 1A Highest (VBAT1 or VSCP) - VGB1I Highest (VBAT2 or VSCP) - VGB2I Highest (VDCIN or VSCP) - VGDCI VPOFFI Input Gate Off Voltage GB1I GB2I GDCI ILOAD = 25A Highest (VBAT1 or VSCP) - VGB1I Highest (VBAT2 or VSCP) - VGB2I Highest (VDCIN or VSCP) - VGDCI 4.75 4.75 4.75 6.7 6.7 6.7 7.5 7.5 7.5 V V V 0.18 0.18 0.18 0.25 0.25 0.25 V V V 1 A 0.8 V Logic I/O IIH /IIL SSB/SCK/MOSI Input High/Low Current l VIL SSB/MOSI/SCK Input Low Voltage l VIH SSB/MOSI/SCK Input High Voltage l VOL MISO Output Low Voltage IOL = 1.3mA l 0.4 V IOFF MISO Output Off-State Leakage Current VMISO = 5V l 2 A 4.5 sec -1 2 V SPI Timing (See Timing Diagram) TWD Watch Dog Timer tSSH SSB High Time tCYC SCK Period tSH tSL l 1.2 2.5 680 ns 2 s SCK High Time 680 ns SCK Low Time 680 ns tLD Enable Lead Time 200 ns tLG Enable Lag Time 200 ns tsu Input Data Set-Up Time l 100 ns tH Input Data Hold Time l 100 tA Access Time (From Hi-Z to Data Active on MISO) l 125 ns tdis Disable Time (Hold Time to Hi-Z State on MISO) l 125 ns tV Output Data Valid l 580 ns tHO Output Data Hold tIr SCK/MOSI/SSB Rise Time 0.8V to 2V 250 ns tIf SCK/MOSI/SSB Fall Time 2V to 0.8V 250 ns tOf MISO Fall Time 2V to 0.4V, CL = 200pF 400 ns CLOAD = 200pF RPULLUP = 4.7k on MISO CL = 200pF, RPULLUP = 4.7k on MISO l l Note 1: Stresses beyond those listed under Absolute Maximum Ratings may cause permanent damage to the device. Exposure to any Absolute Maximum Rating condition for extended periods may affect device reliability and lifetime. Note 2. Battery voltage must be adequate to drive gates of PowerPath P-channel FET switches. This does not affect charging voltage of the battery, which can be zero volts. Note 3. See Test Circuit. Note 4. DCIN, BAT1, BAT2 are held at 12V and GDCI, GB1I, GB2I are forced to 10.5V. SCP is set at 12.0V to measure source current at GDCI, l ns 0 ns GB1I and GB2I. SCP is set at 11.9V to measure sink current at GDCI, GB1I and GB2I. Note 5. Extrapolated from testing with CL = 50pF. Note 6. VDAC offset is equal to the reference voltage, since VOUT = VREF(16mV * VDAC(VALUE)/2047 + 1) Note 7. The LTC1960C is guaranteed to meet specified performance from 0C to 70C and is designed, characterized and expected to meet specified performance at -40C and 85C, but is not tested at these extended temperature limits. Note 8. Does not apply to low current mode. Refer to "The Current DAC Block" in the Operation section. 1960fb 5 LTC1960 TYPICAL PERFORMANCE CHARACTERISTICS Battery Drain Current (BAT1 Selected) 230 CLOAD = 20F 15 ILOAD = 0.8A T = 25C 14 A 220 13 LOAD VOLTAGE (V) BAT1 CURRENT (A) 210 200 190 180 10 9 160 7 12 24 18 BAT1 VOLTAGE (V) 14 11 8 6 15 12 170 150 16 16 TA = 25C 240 LOPWR THRESHOLD LOAD VOLTAGE (V) 250 PowerPath Autonomous Switching PowerPath Switching 1960 G01 10 12 60 50 40 30 20 10 0.025 0.50 0.10 IOUT (A) 2.5 4.0 1960 G04 8 6 4 OUTPUT CURRENT ERROR (mA) BAT1 VOLTAGE (V) 12.2 12.1 12.0 11.9 VIN = 20V VDAC = 12.288V IDAC = 4000mA TA = 25C 11.6 0 1000 2000 3000 CHARGE CURRENT (mA) 4000 1960 G07 LOAD CONNECTED 4 0 -0.05 0 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 TIME (SEC) 0 -4 -2 LOAD DISCONNECTED 0 2 500 VDCIN = 20V 100 VBAT1 = 12V RSNS = 0.025 T = 25C 80 A 10 12 14 16 1960 G06 VIN = 20V VBAT1 = 12V RSNS = 0.025 TA = 25C 450 60 40 20 0 -20 -40 4 6 8 TIME (ms) IDAC Low Current Mode vs Normal Mode 120 11.7 6 1960 G05 12.4 11.8 8 Charging Current Accuracy 12.3 VIN = 20V VDAC = 12.29V IDAC = 3000mA LOAD CURRENT = 1A TA = 25C 10 2 Charger Load Regulation 5 1960 G03 BAT1 OUTPUT 2 CHARGING CURRENT (mA) EFFICIENCY (%) 70 4 Charger Load Dump 14 BAT1 VOLTAGE (V) CHARGER OUTPUT (V) 80 BAT1 REMOVED 9 12 90 0 10 Charger Start-Up 100 0 11 7 NOTE: LIGHT LOAD TO EXAGGERATE SWITCHING EVENT 6 0 2 3 -1 1 TIME (SEC) 1960 G02 Charger Efficiency 12 8 6 -50 -40 -30 -20 -10 0 10 20 30 40 50 TIME (s) 30 13 400 350 300 LOW CURRENT MODE 250 200 NORMAL MODE 150 100 50 0 200 400 600 800 IDAC VALUE 1000 1200 1960 G08 0 0 80 160 240 320 400 480 PROGRAMMED CURRENT (mA) 560 1960 G09 1960fb 6 LTC1960 TYPICAL PERFORMANCE CHARACTERISTICS BATTERY CURRENT (mA) 50 25 0 -25 -50 -75 450 650 850 1050 VDAC VALUE 1250 1450 BAT1 CURRENT BAT1 CURRENT BAT2 CURRENT DUAL 0 50 100 150 200 TIME (MINUTES) 250 2500 15.5 2000 15.0 BAT1 CURRENT 14.5 13.5 300 3000 BAT1 VOLTAGE 16.0 BAT2 CURRENT 14.0 100 MINUTES 3500 BAT2 VOLTAGE 16.5 SEQUENTIAL Dual vs Sequential Discharge 20 0 40 1500 1000 500 0 100 120 140 160 TIME (MINUTES) 60 80 BAT1 INITIAL CAPACITY = 0% BAT2 INITIAL CAPACITY = 90% PROGRAMMED CHARGER CURRENT = 3A PROGRAMMED CHARGER VOLTAGE = 16.8V 1960 G12 Dual vs Sequential Discharge 15 12.0 BAT1 VOLTAGE 11.0 BAT2 VOLTAGE 12.0 SEQUENTIAL 11.0 10.0 BAT1 VOLTAGE 9.0 0 20 40 BAT2 VOLTAGE 13 9.0 8.0 14 DUAL BAT2 VOLTAGE 10.0 8.0 17.0 BAT2 CURRENT BATTERY TYPE: 10.8V Li-Ion (MOLTECH NI2020) REQUESTED CURRENT = 3A REQUESTED VOLTAGE = 12.3V MAX CHARGER CURRENT = 4.1A 1960 G11 1960 G10 BATTERY VOLTAGE (V) -100 250 3500 3000 2500 2000 1500 1000 500 0 3500 3000 2500 2000 1500 1000 500 0 BATTERY VOLTAGE (V) 75 DCIN = 24V TA = 25C ILOAD = 100mA BATTERY CURRENT (mA) OUTPUT VOLTAGE ERROR (mV) 100 Dual Charging Batteries with Different Charge State Dual vs Sequential Charging BATTERY VOLTAGE (V) Voltage Accuracy 12 BAT1 VOLTAGE 11 10 BAT2 VOLTAGE 15 14 60 80 100 120 140 160 180 TIME (MINUTES) BATTERY TYPE: 10.8V Li-Ion(MOLTECH NI2020) LOAD CURRENT = 3A 1960 G13 16 MINUTES 11 10 SEQUENTIAL BAT1 VOLTAGE 13 12 11 MINUTES DUAL 0 20 40 60 80 100 TIME (MINUTES) 120 140 BATTERY TYPE: 12V NIMH (MOLTECH NJ1020) 1960 G14 LOAD: 33W 1960fb 7 LTC1960 PIN FUNCTIONS (G/UHF) Input Power Related Battery Charging Related SCN (Pin 4/Pin 30): PowerPath Current Sensing Negative Input. This pin should be connected directly to the "bottom" (output side) of the sense resistor, RSC , in series with the three PowerPath switch pairs, for detecting short-circuit current events. Also powers LTC1960 internal circuitry when all other sources are absent. VSET (Pin 13/Pin 1): The Tap Point of a Programmable Resistor Divider Which Provides Battery Voltage Feedback to the Charger. A capacitor from CSN to VSET and from VSET to GND provide necessary compensation and filtering for the voltage loop. SCP (Pin 5/Pin 31): PowerPath Current Sensing Positive Input. This pin should be connected directly to the "top" (switch side) of the sense resistor, RSC , in series with the three PowerPath switch pairs, for detecting short-circuit current events. ITH (Pin 14/Pin 2): The Control Signal of the Inner Loop of the Current Mode PWM. Higher ITH voltage corresponds to higher charging current in normal operation. A capacitor of at least 0.1F to GND filters out PWM ripple. Typical full-scale output current is 30A. Nominal voltage range for this pin is 0V to 2.4V. GDCO (Pin 6/Pin 32): DCIN Output Switch Gate Drive. Together with GDCI, this pin drives the gate of the P-channel switch in series with the DCIN input switch. ISET (Pin 15/Pin 3): A capacitor from ISET to ground is required to filter higher frequency components from the delta-sigma IDAC. GDCI (Pin 7/Pin 33): DCIN Input Switch Gate Drive. Together with GDCO, this pin drives the gate of the P-channel switch connected to the DCIN input. CSN (Pin 22/Pin 11): Current Amplifier CA1 Input. Connect this to the common output of the charger MUX switches. GB1O (Pin 8/Pin 34): BAT1 Output Switch Gate Drive. Together with GB1I, this pin drives the gate of the P-channel switch in series with the BAT1 input switch. CSP (Pin 23/Pin 12): Current Amplifier CA1 Input. This pin and the CSN pin measure the voltage across the sense resistor, RSNS, to provide the instantaneous current signals required for both peak and average current mode operation. GB1I (Pin 9/Pin 35): BAT1 Input Switch Gate Drive. Together with GB1O, this pin drives the gate of the P-channel switch connected to the BAT1 input. GB2O (Pin 10/Pin 36): BAT2 Output Switch Gate Drive. Together with GB2I, this pin drives the gate of the P-channel switch in series with the BAT2 input switch. GB2I (Pin 11/Pin 37): BAT2 Input Switch Gate Drive. Together with GB2O, this pin drives the gate of the P-channel switch connected to the BAT2 input. CLP (Pin 24/Pin 13): The Positive Input to the Supply Current Limiting Amplifier CL1. The threshold is set at 100mV above the voltage at the DCIN pin. When used to limit supply current, a filter is needed to filter out the switching noise. COMP1 (Pin 25/Pin 14): The Compensation Node for the Amplifier CL1. A capacitor is required from this pin to GND if input current amplifier CL1 is used. At input adapter current limit, this node rises to 1V. By forcing COMP1 low, amplifier CL1 will be defeated (no adapter current limit). COMP1 can source 10A. BGATE (Pin 27/Pin 16): Drives the bottom external MOSFET of the battery charger buck converter. SW (Pin 30/Pin 19): PWM switch node connected to source of the top external MOSFET switch. Used as reference for top gate driver. BOOST (Pin 31/Pin 20): Supply to Topside Floating Driver. The bootstrap capacitor is returned to this pin. Voltage swing at this pin is from a diode drop below VCC to (DCIN + VCC). 1960fb 8 LTC1960 PIN FUNCTIONS (G/UHF) TGATE (Pin 32/Pin 21): Drives the top external MOSFET of the battery charger buck converter. SCH1 (Pin 33/Pin 22), SCH2 (Pin 36/Pin 25): Charger MUX N-Channel Switch Source Returns. These two pins are connected to the sources of the back-to-back switch pairs, Q3/Q4 and Q9/Q10 (see Typical Application on back page of data sheet), respectively. A small pull-down current source returns these nodes to 0V when the switches are turned off. GCH1 (Pin 34/Pin 23), GCH2 (Pin 35/Pin 24): Charger MUX N-Channel Switch Gate Drives. These two pins drive the gates of the back-to-back switch pairs, Q3/Q4 and Q9/ Q10, between the charger output and the two batteries. External Power Supply Pins VPLUS (Pin 1/Pin 27): Supply. The VPLUS pin is connected via four internal diodes to the DCIN, SCN, BAT1, and BAT2 pins. Bypass this pin with a 1F to 2F capacitor. BAT1 (Pin 3/Pin 29), BAT2 (Pin 2/Pin 28): These two pins are the inputs from the two batteries for power to the LTC1960 and to provide voltage feedback to the battery charger. LOPWR (Pin 12/Pin 38): LOPWR Comparator Input from SCN External Resistor Divider to GND. If the voltage at LOPWR is lower than the LOPWR comparator threshold, then system power has failed and power is autonomously switched to a higher voltage source, if available. See PowerPath section of LTC1960 operation. DCDIV (Pin 17/Pin 5): External DC Source Comparator Input from DCIN External Resistor Divider to GND. If the voltage at DCDIV is above the DCDIV comparator threshold, then the DC bit is set and the wall adapter power is considered to be adequate to charge the batteries. If DCDIV rises more than 1.8V above VCC, then all of the PowerPath switches are latched off until all power is removed. DCIN (Pin 29/Pin 18): Supply. External DC power source. A 1F bypass capacitor should be connected to this pin as close as possible. No series resistance is allowed, since the adapter current limit comparator input is also this pin. Internal Power Supply Pins GND (Pin 16/Pin 4, Pin 10, Pin 26, Pin 39): Ground for Low Power Circuitry. PGND (Pin 26/Pin 15): High Current Ground Return for BGATE Driver. VCC (Pin 28/Pin 17): Internal Regulator Output. Bypass this output with at least a 2F to 4.7F capacitor. Do not use this regulator output to supply more than 1mA to external circuitry. Digital Interface Pins SSB (Pin 18/Pin 6): SPI Slave Select Input. Active low. TTL levels. This signal is low when clocking data to/from the LTC1960. SCK (Pin 19/Pin 7): Serial SPI Clock. TTL levels. MISO (Pin 20/Pin 8): SPI Master-In-Slave-Out Output, Open Drain. Serial data is transmitted from the LTC1960, when SSB is low, on the falling edge of SCK. TTL levels. A 4.7k pull-up resistor is recommended. MOSI (Pin 21/Pin 9): SPI Master-Out-Slave-In Input. Serial data is transmitted to the LTC1960, when SSB is low, on the rising edge of SCK. TTL levels. GND (Exposed Pad Pin 39, UHF Package Only): Ground. Must be soldered to the PCB ground for rated thermal performance. A capacitor from DCDIV to GND is recommended to prevent noise-induced false emergency turn-off conditions from being detected. Refer to "Fast PowerPath Turn-Off" in the Operation section and the Typical Application on the back page of this data sheet. 1960fb 9 LTC1960 (LTC1960CG Pin Numbers Shown) CHARGE PUMP SWB2 DRIVER 6 SWDC DRIVER 100mV SHORT CIRCUIT AC_PRESENT ON SELECTOR CONTROLLER ON + - SCH2 36 7 5 SCP 100 4 SCN 17 DCDIV 12 LOPWR - GCH2 35 GDCI GDCO + SCH1 33 SWB1 DRIVER 10 + - GCH1 34 11 - DCIN GB2I GB2O 8 + 9 - GB1I GB1O + BLOCK DIAGRAM 1.19V CSN 21 MOSI BAT1 3 BAT2 2 CHARGE 400k 20 MISO SPI INTERFACE CHGMON 19 SCK 18 SSB 11 1 11-BIT VOLTAGE DAC VCC REGULATOR 0.86V DCIN 29 CSP-CSN 3k 0V + - OSCILLATOR 0.8V 3k EA CA1 TON gm = 1.4m PWM LOGIC SW 30 R /15 ICMP CHARGE VCC 0.8V -+ - IREV PGND 26 - BGATE 27 40mV + gm = 0.4m + + CL1 - 100mV - CLP 24 22 CSN BUFFERED ITH + TGATE 32 CA2 S 23 CSP - Q + BOOST 31 DCIN 3k + LOW DROP DETECT + gm = 1.4m VSET 13 BGATE 15 ISET 10-BIT CURRENT DAC - VCC 28 GND 16 SCN - VPLUS 25 COMP1 14 ITH 0.75V CHGMON CLAMP 1960 BD 1960fb 10 LTC1960 TEST CIRCUIT VREF + EA CHGMON BAT1 BAT2 - VSW ITH VSET + - 0.5V 1960 TC01 TIMING DIAGRAM SPI Timing Diagram SSB tLD tCYC tLG tSSH SCK tH tsu MOSI tSH BIT 7 tA MISO tSL BIT 0 tV SLAVE BIT 7 OUT tHO tdis SLAVE BIT 0 OUT 1960 TD01 1960fb 11 LTC1960 OPERATION (Refer to Block Diagram and Typical Application) OVERVIEW The LTC1960 is composed of a battery charger controller, charge MUX controller, PowerPath controller, SPI interface, a 10-bit current DAC (IDAC) and 11-bit voltage DAC (VDAC). When coupled with a low cost microprocessor, it forms a complete battery charger/selector system for two batteries. The battery charger is programmed for voltage and current, and the charging battery is selected via the SPI interface. Charging can be accomplished only if the voltage at DCDIV indicates that sufficient voltage is available from the input power source, usually an AC adapter. The charge MUX, which selects the battery to be charged, is capable of charging both batteries simultaneously by selecting both batteries for charging. The charge MUX switch drivers are configured to allow charger current to share between the two batteries and to prevent current from flowing in a reverse direction in the switch. The amount of current that each battery receives will depend upon the relative capacity of each battery and the battery voltage. This can result in significantly shorter charging times (up to 50% for Li-Ion batteries) than sequential charging of each battery. In order to continue charging, the CHARGE_BAT information must be updated more frequently than the internal watchdog timer. The PowerPath controller selects which of the pairs of PFET switches, input and output, will provide power to the system load. The selection is accomplished over the SPI interface. If the system voltage drops below the threshold set by the LOPWR resistor divider, then all of the output side PFETs are turned on quickly and power is taken from the highest voltage source available at the DCIN, BAT1 or BAT2 inputs. The input side PFETs act as diodes in this mode and power is taken from the source with the highest voltage. The input side PowerPath switch driver that is delivering power then closes its input switch to reduce the power dissipation in the PFET bulk diode. In effect, this system provides diode -like behavior from the FET switches, without the attendant high power dissipation from diodes. The microprocessor is informed of this 3-diode mode status when it polls the PowerPath status register via the SPI interface. The microprocessor can then assess which power source is capable of providing power, and program the PowerPath switches accordingly. Since high speed PowerPath switching at LOPWR trip points is handled autonomously, there is no need for real-time microprocessor resources to accomplish this task. Simultaneous discharge of both batteries is accomplished by simply programming both batteries for discharge into the system load. The switch drivers prevent reverse current flow in the switches and automatically discharge both batteries into the load, sharing current according to the relative capacity of the batteries. Simultaneous dual discharge can increase battery operating time by approximately 10% by reducing losses in the switches and reducing internal losses associated with high discharge rates. SPI Interface The SPI interface is used to write to the internal PowerPath registers, the charger control registers, the current DAC, and the voltage DAC. The SPI is also able to read internal status registers. There are two types of SPI write commands. The first write command is a 1-byte command used to load PowerPath and charger control bits. The second write command is a 2-byte command used to load the DACs. The SPI read command is a 2-byte command. In order to ensure the integrity of the SPI communication, the last bit received by the SPI is echoed back over the MISO output after the next falling SCK. The data format is set up so that the master has the option of aborting a write if the returned MISO bit is not as expected. 1960fb 12 LTC1960 OPERATION 1-Byte SPI Write Format: bit 7........byte 1..........bit 0 MOSI D0 D1 D2 X A0 A1 A2 0 MISO X D0 D1 D2 X A0 A1 A2 Charger Write Address: A[2:0] = b111 Charger Write Data: D2 = X D1 = CHARGE_BAT2 D0 = CHARGE_BAT1 PowerPath Write Address: A[2:0] = b110 PowerPath Write Data: D2 = POWER_BY_DC D1 = POWER_BY_BAT2 D0 = POWER_BY_BAT1 2-Byte SPI Write Format: bit 7........byte 1..........bit 0 bit 7..........byte 2............bit 0 MOSI D0 D1 D2 D3 D4 D5 D6 1 D7 D8 D9 D10 A0 A1 A2 0 MISO X D0 D1 D2 D3 D4 D5 D6 1 D7 D8 D9 D10 A0 A1 A2 IDAC Write Address: A[2:0] = b000 IDAC Data Bits D9-D0: IDAC value data (MSB-LSB) IDAC Data Bit D10 : Normal mode = 0, low current mode = 1 (Dual battery charging is disabled) VDAC Write Address: A[2:0] = b001 VDAC Data Bits D10-D0: VDAC value (MSB-LSB) Subsequent SPI communication is inhibited until after the addressed DAC is finished loading. It is recommended that the master transmit all zeros until MISO goes low. This handshaking procedure is illustrated in Figure 1. SSB BYTE 1 BYTE 2 SCK MOSI MISO 1960 F01 Figure 1. SPI Write to VDAC of Data = b101_0101_0101 1960fb 13 LTC1960 OPERATION 2-Byte SPI Read Format: bit 7........byte 1.......bit 0 bit 7........byte 2............bit 0 MOSI 0 0 0 0 A0 A1 A2 0 0 0 0 0 A0 A1 A2 1 MISO X 0 0 0 0 A0 A1 A2 X FA LP DC PF CH X Status Address: A[2:0] = b010 Status Read Data: LP = LOW_POWER (Low power comparator output) X DC = DCDIV (DCDIV comparator output) PF = POWER_FAIL (Set if selected power supply failed to hold up system power after three tries) CH = CHARGING (One or more batteries are being charged) FA = FAULT. This bit is set for any of the following conditions: 1) The LTC1960 is still in power-on reset. 2) The LTC1960 has detected a short circuit and has shut down power and charging. 3) The system has asserted a fast off using DCDIV. Note: All other values of A[2:0] are reserved and must not be used. A status read is illustrated in Figure 2. SSB BYTE 1 BYTE 2 SCK MOSI MISO 1960 F02 Figure 2. SPI Read of FA = 0, LP = 0, DC = 1, PF = 0, and CH = 1 1960fb 14 LTC1960 OPERATION Battery Charger Controller The LTC1960 charger controller uses a constant off-time, current mode step-down architecture. During normal operation, the top MOSFET is turned on each cycle when the oscillator sets the SR latch and turned off when the main current comparator ICMP resets the SR latch. While the top MOSFET is off, the bottom MOSFET is turned on until either the inductor current reverses, as indicated by current comparator IREV, or the beginning of the next cycle. The oscillator uses the equation: 1 (VDCIN - VCSN ) t OFF = * VDCIN fOSC to set the bottom MOSFET on time. The peak inductor current at which ICMP resets the SR latch is controlled by the voltage on ITH. ITH is in turn controlled by several loops, depending upon the situation at hand. The average current control loop converts the voltage between CSP and CSN to a representative current. Error amp CA2 compares this current against the desired current requested by the IDAC at the ISET pin and adjusts ITH until the IDAC value is satisfied. The BAT1/BAT2 MUX provides the selected battery voltage at CHGMON, which is divided down to the VSET pin by the VDAC resistor divider and is used by error amp EA to decrease ITH if the VSET voltage is above the 0.8V reference. The amplifier CL1 monitors and limits the input current, normally from the AC adapter, to a preset level (100mV/RCL). At input current limit, CL1 will decrease the ITH voltage and thus reduce battery charging current. An overvoltage comparator, 0V, guards against transient overshoots (>7%). In this case, the top MOSFET is turned off until the overvoltage condition is cleared. This feature is useful for batteries which "load dump" themselves by opening their protection switch to perform functions such as calibration or pulse mode charging. Charging is inhibited for battery voltages below the minimum charging threshold, VCHMIN. Charging is not inhibited when the low current mode of the IDAC is selected. The top MOSFET driver is powered from a floating bootstrap capacitor CB. This capacitor is normally recharged from VCC through an external diode when the top MOSFET is turned off. A 2F to 4.7F capacitor across VCC to GND is required to provide a low dynamic impedance to charge the boost capacitor. It is also required for stability and power-on reset purposes. As VIN decreases towards the selected battery voltage, the converter will attempt to turn on the top MOSFET continuously ("dropout''). A dropout timer detects this condition and forces the top MOSFET to turn off, and the bottom MOSFET on, for about 200ns at 40s intervals to recharge the bootstrap capacitor. Charge MUX Switches The equivalent circuit of a charge MUX switch driver is shown in Figure 3. If the charger controller is not enabled, the charge MUX drivers will drive the gate and source of the series-connected MOSFETs to a low voltage and the switch is off. When the charger controller is on, the charge MUX driver will keep the MOSFETs off until the voltage at CSN rises at least 35mV above the battery voltage. GCH1 is then driven with an error amplifier EAC until the voltage between BAT1 and CSN satisfies the error amplifier or until GCH1 is clamped by the internal Zener diode. The time required to close the switch could be quite long (many ms) due to the small currents output by the error amp and depending upon the size of the MOSFET switch. If the voltage at CSN decreases below VBAT1 - 20mV, a comparator CC quickly turns off the MOSFETs to prevent reverse current from flowing in the switches. In essence, this system performs as a low forward voltage diode. Operation is identical for BAT2. DCIN + 10V (CHARGE PUMPED) TO BATTERY 1 FROM CHARGER BAT1 CSN - 35mV GCH1 EAC + SCH1 + 20mV Q3 CC 10k - OFF Q4 1960 F03 Figure 3. Charge MUX Switch Driver Equivalent Circuit 1960fb 15 LTC1960 OPERATION Dual Charging Note that the charge MUX switch drivers will operate together to allow both batteries to be charged simultaneously. If both charge MUX switch drivers are enabled, only the battery with the lowest voltage will be charged until its voltage rises to equal the higher voltage battery. The charge current will then share between the batteries according to the capacity of each battery. If both batteries are selected for charging, only batteries with voltages above VCHMIN are allowed to charge. Dual charging is not allowed when the low current mode of the IDAC is selected. If dual charging is enabled when the IDAC enters low current mode, then only BAT1 will be charged. voltages present at the input/output. When the output PFET turns on, the voltage at SCP will be pulled up to a diode drop below the source voltage by the bulk diode of the input PFET. If the source voltage is more than 25mV above SCP, EAP will drive the gate of the input PFET low until the input PFET turns on and reduces the voltage across the input/output to the EAP set point, or until the Zener clamp engages to limit the voltage applied to the input PFET. If the source voltage drops more than 20mV below SCP, then comparator CP turns on SWP to quickly prevent large reverse current in the switch. This operation mimics a diode with a low forward voltage drop. 20mV CP + Charger Start-Up When the charger controller is enabled by the SPI Interface block, the charger output CSN will ramp from 0V until it exceeds the selected battery voltage. The clamp error amp is used to prevent the charger output from exceeding the selected battery voltage by more than 0.7V during the start-up transient while the charge MUX switches, have yet to close. Once the charge MUX switches have closed, the clamp releases ITH to allow control by another loop. OFF - FROM BATTERY 1 BAT1 SWP - GB1I Q7 EAP SCP + 25mV GB1O Q8 OFF PowerPath Controller The PowerPath switches are turned on and off via the SPI interface, in any combination. The external P-MOSFETs are usually connected as an input switch and an output switch. The output switch PFET is connected in series with the input PFET and the positive side of the short-circuit sensing resistor, RSC. The input switch is connected in series between the power source and the output PFET. The PowerPath switch driver equivalent circuit is shown in Figure 4. The output PFET is driven high and low by the output side driver controlling pin GXXO, the PFET is either on or off. The gate of the input PFET is driven by an error amplifier which monitors the voltage between the input power source (BAT1 in this case) and SCP. If the switch is turned off, the two outputs are driven to the higher of the two voltages present across the input/output terminals of the switch. When the switch is instructed to turn on, the output side driver immediately drives the gate of the output PFET approximately 6V below the highest of the 16 RSC CL TO LOAD 1960 F04 Figure 4. PowerPath Driver Equivalent Circuit Autonomous PowerPath Switching The LOPWR comparator monitors the voltage at the load through the resistor divider from pin SCN. If any POWER_BY bit is set and the LOPWR comparator trips, then all of the switches are turned on (3-diode mode) by the PowerPath controller to ensure that the system is powered from the source with the highest voltage. The PowerPath controller waits approximately 1 second, to allow power to stabilize, and then reverts to the previous PowerPath switch configuration. A power-fail counter is incremented to indicate that a failure has occurred. If the power-fail counter equals a value of 3, then the PowerPath controller sets the switches to 3-diode mode and the PF 1960fb LTC1960 OPERATION bit is set in the status register. This is a three-strikes-andyou're-out process which is intended to debounce the PowerPath PF indicator. The power-fail counter is reset by a PowerPath SPI write. Short-Circuit Protection Short-circuit protection operates in both a current mode and a voltage mode. If the voltage between SCP and SCN exceeds the short-circuit comparator threshold VTSC for more than 15ms, then all of the PowerPath switches are turned off and the FAULT bit (FA) is set. Similarly, if the voltage at SCN falls below 3V for more than 15ms, then all of the PowerPath switches are turned off and the FA bit is set. The FA bit is reset by removing all power sources and allowing the voltage at VPLUS to fall below the UVLO threshold. If the FA bit is set, charging is disabled until VPLUS exceeds the UVLO threshold and charging is requested via the SPI interface. When a hard short-circuit occurs, it might pull all of the power sources down to near 0V potentials. The capacitors on VCC and VPLUS must be large enough to keep the circuit operating correctly during the 15ms short-circuit event. The charger will stop within a few microseconds leaving a small current which must be provided by the capacitor on VPLUS. The recommended minimum values (1F on VPLUS and 2F on VCC, including tolerances) should keep the LTC1960 operating above the UVLO trip voltage long enough to perform the short-circuit function when the input voltages are greater than 8V. Increasing the capacitor across VCC to 4.7F will allow operation down to the recommended 6V minimum. Fast PowerPath Turn-Off All of the PowerPath switches can be forced off by setting the DCDIV pin to a voltage between 8V and 10V. This will have the same effect as a short-circuit event. The PF status bit will also be set. DCDIV must be less than 5V and VPLUS must decrease below the UVLO threshold to re-enable the PowerPath switches. Power-Up Strategy All three PowerPath switches are turned on after VPLUS exceeds the UVLO threshold for more than 250ms. This delay is to prevent oscillation from a turn-on transient near the UVLO threshold. The Voltage DAC Block The voltage DAC (VDAC) is a delta-sigma modulator which controls the effective value of an internal resistor, RVSET = 7.2k, used to program the maximum charger voltage. Figure 5 is a simplified diagram of the VDAC operation. The charger monitor MUX is connected to the appropriate battery indicated by the CHARGE_BATx bit. The delta-sigma modulator and switch SWV convert the VDAC value, received via SPI communication, to a variable resistance equal to (11/8)RVSET/(VDAC(VALUE)/2047). In regulation, VSET is servo driven to the 0.8V reference voltage, VREF . Therefore, programmed voltage is: VBATx = (8/11) VREF 405.3k/7.2k * (VDAC(VALUE)/2047) + VREF = 32,752mV * (VDAC(VALUE)/2047) + 0.8V Note that the reference voltage must be subtracted from the VDAC value in order to obtain the correct output voltage. This value is VREF /16mV = 50 (32HEX). Capacitors CB1 and CB2 are used to average the voltage present at the VSET pin as well as provide a zero in the voltage loop to help stability and transient response time to voltage variations. See the Applications Information section. CHGMON BAT1 BAT2 CB2 CSN RVF 405.3k VSET - TO ITH EA CB1 VREF RVSET 7.2k SWV + MODULATOR 11 DAC VALUE (11 BITS) 1960 F05 Figure 5. Voltage DAC Operation 1960fb 17 LTC1960 OPERATION The Current DAC Block When the low current mode bit (D10) is set to 1, the current DAC enters a different mode of operation. The current DAC output is pulse-width modulated with a high frequency clock having a duty cycle value of 1/8. Therefore, the maximum output current provided by the charger is IMAX/8. The delta-sigma output gates this low duty cycle signal on and off. The delta-sigma shift registers are then clocked at a slower rate, about 40ms/bit, so that the charger has time to settle to the IMAX/8 value. The resulting average charging current is equal to 1/8 of the current programmed in normal mode. Dual battery charging is disabled in low current mode. If both batteries are selected for charging, then only BAT1 will charge. The current DAC is a delta-sigma modulator which controls the effective value of an internal resistor, RSET = 18.77k, used to program the maximum charger current. Figure 6 is a simplified diagram of the DAC operation. The delta-sigma modulator and switch convert the IDAC value, received via SPI communication, to a variable resistance equal to 1.25RSET /(IDAC(VALUE)/1023). In regulation, ISET is servo driven to the 0.8V reference voltage, VREF , and the current from RSET is matched against a current derived from the voltage between pins CSP and CSN. This current is (VCSP - VCSN)/3k. Therefore, programmed current is: I AVG = IDAC(VALUE) VREF * 3k * (1.25R SNS R SET ) 1023 (VCSP - VCSN) 3k (FROM CA1 AMPLIFIER) ISET + CSET RSET 18.77k VREF TO ITH - MODULATOR 10 DAC VALUE (10 BITS) 1960 F06 Figure 6. Current DAC Operation AVERAGE CHARGER CURRENT IMAX/8 0 1960 F07 ~40ms Figure 7. Charging Current Waveform in Low Current Mode 1960fb 18 LTC1960 APPLICATIONS INFORMATION Automatic Current Sharing In a dual parallel charge configuration, the LTC1960 does not actually control the current flowing into each individual battery. The capacity, or amp-hour rating, of each battery determines how the charger current is shared. This automatic steering of current is what allows both batteries to reach their full capacity points at the same time. In other words, given all other things equal, charge termination will happen simultaneously. A battery can be modeled as a huge capacitor and hence governed by the same laws. I = C * (dV/dT), where: I = The current flowing through the capacitor C = Capacity rating of battery (using amp-hour value instead of capacitance) dV = Change in voltage dt = Change in time The equivalent model of a set or parallel batteries is a set of parallel capacitors. Since they are in parallel, the change in voltage over change in time is the same for both batteries 1 and 2. dV dV = dt BAT1 dt BAT2 From here we can simplify. IBAT1/CBAT1 = dV/dt = IBAT2/CBAT2 IBAT2 = IBAT1 CBAT2/CBAT1 At this point you can see that the current divides as the ratio of the two batteries capacity ratings. The sum of the current into both batteries is the same as the current being supply by the charger. This is independent of the mode of the charger (CC or CV). it is actual physical capacity rating at the time of charge. Capacity rating will change with age and use and hence the current sharing ratios can change over time. In dual charge mode, the charger uses feedback from the BAT2 input to determine charger output voltage. When charging batteries with significantly different initial states of charge (i.e., one almost full, the other almost depleted), the full battery will get a much lower current. This will cause a voltage difference across the charge MUX switches, which may cause the BAT1 voltage to exceed the programmed voltage. Using MOSFETs in the charge MUX with lower RDS(ON) will alleviate this problem. Adapter Limiting An important feature of the LTC1960 is the ability to automatically adjust charging current to a level which avoids overloading the wall adapter. This allows the product to operate at the same time that batteries are being charged without complex load management algorithms. Additionally, batteries will automatically be charged at the maximum possible rate of which the adapter is capable. This feature is created by sensing total adapter output current and adjusting charging current downward if a preset adapter current limit is exceeded. True analog control is used, with closed loop feedback ensuring that adapter load current remains within limits. Amplifier CL1 in Figure 8 senses the voltage across RCL, connected between the CLP and DCIN pins. When this voltage exceeds 100mV, the amplifier will override programmed charging current to limit adapter current to 100mV/RCL. A lowpass filter formed by 5k and 0.1F is required to eliminate switching noise. If the current limit is not used, CLP should be connected to DCIN. 100mV - ICHRG = IBAT1 + IBAT2 + 5k + RCL* DCIN IBAT2 = ICHRG CBAT2/(CBAT1 + CBAT2) Please note that the actual observed current sharing will vary from manufactures claimed capacity ratings since 0.1F CL1 From here we solve for the actual current for each battery. IBAT1 = ICHRG CBAT1/(CBAT1 + CBAT2) CLP + *RCL = 100mV ADAPTER CURRENT LIMIT CIN AC ADAPTER INPUT VIN 11960 F08 Figure 8. Adapter Current Limiting 1960fb 19 LTC1960 APPLICATIONS INFORMATION Watchdog Timer Charging will begin when either CHARGE_BAT1 or CHARGE_BAT2 bits are set in the charger register (address: 111). Charging will stop if the charger register is not updated prior to the expiration of the watchdog timer. Simply repeating the same data transmission to the charger register at a rate higher than once per second will ensure that charging will continue uninterrupted. Extending System to More Than Two Batteries The LTC1960 can be extended to manage systems with more than three sources of power. Contact Linear Technology Applications Engineering for more information. Charging Depleted Batteries Some batteries contain internal protection switches that disconnect a load if the battery voltage falls below what is considered a reasonable minimum. In this case, the charger may not start because the voltage at the battery terminal is less than 5V. The low current mode of the IDAC must be used in this case to condition the battery. In low current mode, there is no minimum voltage requirement (but dual charging is not allowed). Usually, the battery will detect that it is being charged and then close its protection switch, which will allow the IDAC to switch to normal mode. Smart batteries require that charging current not exceed 100mA until valid charging voltage and charging current parameters are transmitted via the SMBus. The low current IDAC mode is ideal for this purpose. Starting Charge with Dissimilar Batteries in Dual Charge Mode When charging batteries of different charger termination voltages, the charger should be started using the following procedure: Step 1. Select only the lowest termination voltage battery for charging, and set the charger to its charging parameters. Step 2. When the battery current is flowing into that battery, change to dual charging mode (without stopping the charger) and set the appropriate charging parameters for this dual charger condition. If this procedure is not followed, and BAT2 is significantly higher voltage than BAT1, the charger could refuse to charge either battery. Charge Termination Issues Batteries with constant-current charging and voltage-based charger termination might experience problems with reductions of charger current caused by adapter limiting. It is recommended that input limiting feature be defeated in such cases. Consult the battery manufacturer for information on how your battery terminates charging. Setting Output Current Limit The full-scale output current setting of the IDAC will produce VMAX = 102.3mV between CSP and CSN. To set the fullscale current of the DAC simply divide VMAX by RSNS. This is expressed by the following equation: RSNS = 0.1023/IMAX Table 1. Recommended RSNS Resistor Values IMAX (A) RSNS () 1% RSNS (W) 1.023 0.100 0.25 2.046 0.050 0.25 4.092 0.025 0.5 8.184 0.012 1 Use resistors with low ESL. Inductor Selection Higher operating frequencies allow the use of smaller inductor and capacitor values. A higher frequency generally results in lower efficiency because of MOSFET gate charge losses. In addition, the effect of inductor value on ripple current and low current operation must also be considered. The inductor ripple current IL decreases with higher frequency and increases with higher VIN. IL = 1 V VOUT 1- OUT VIN ( f )(L ) Accepting larger values of IL allows the use of low inductances, but results in higher output voltage ripple and greater core losses. A reasonable starting point for setting ripple current is IL = 0.4(IMAX). In no case should 1960fb 20 LTC1960 APPLICATIONS INFORMATION IL exceed 0.6(IMAX) due to limits imposed by IREV and CA1. Remember the maximum IL occurs at the maximum input voltage. In practice, 10H is the lowest value recommended for use. Charger Switching Power MOSFET and Diode Selection Two external power MOSFETs must be selected for use with the LTC1960 charger: An N-channel MOSFET for the top (main) switch and an N-channel MOSFET for the bottom (synchronous) switch. The peak-to-peak gate drive levels are set by the VCC voltage. This voltage is typically 5.2V. Consequently, logic-level threshold MOSFETs must be used. Pay close attention to the BVDSS specification for the MOSFETs as well; many of the logic-level MOSFETs are limited to 30V or less. Selection criteria for the power MOSFETs include the onresistance RDS(ON), reverse transfer capacitance CRSS, input voltage and maximum output current. The LTC1960 charger is always operating in continuous mode so the duty cycles for the top and bottom MOSFETs are given by: Main Switch Duty Cycle = VOUT /VIN Synchronous Switch Duty Cycle = (VIN - VOUT)/VIN The MOSFET power dissipations at maximum output current are given by: PMAIN = VOUT/VIN(IMAX)2(1 + dT)RDS(ON) + k(VIN)2 (IMAX)(CRSS)(f) PSYNC = (VIN - VOUT)/VIN(IMAX)2(1 + dT) RDS(ON) Where dT is the temperature dependency of RDS(ON) and k is a constant inversely related to the gate drive current. Both MOSFETs have I2R losses while the topside N-channel equation includes an additional term for transition losses, which are highest at high input voltages. For VIN < 20V, the high current efficiency generally improves with larger MOSFETs, while for VIN > 20V the transition losses rapidly increase to the point that the use of a higher RDS(ON) device with lower CRSS actually provides higher efficiency. The synchronous MOSFET losses are greatest at high input voltage or during a short-circuit when the duty cycle in this switch is nearly 100%. The term (1 + dT) is generally given for a MOSFET in the form of a normalized RDS(ON) vs Temperature curve, but d = 0.005/C can be used as an approximation for low voltage MOSFETs. CRSS is usually specified in the MOSFET characteristics. The constant k = 1.7 can be used to estimate the contributions of the two terms in the main switch dissipation equation. If the LTC1960 charger is to operate in low dropout mode or with a high duty cycle greater than 85%, then the topside N-channel efficiency generally improves with a larger MOSFET. Using asymmetrical MOSFETs may achieve cost savings or efficiency gains. The Schottky diode D1, shown in the Typical Application on the back page, conducts during the dead-time between the conduction of the two power MOSFETs. This prevents the body diode of the bottom MOSFET from turning on and storing charge during the dead-time, which could cost as much as 1% in efficiency. A 1A Schottky is generally a good size for 4A regulators due to the relatively small average current. Larger diodes can result in additional transition losses due to their larger junction capacitance. The diode may be omitted if the efficiency loss can be tolerated. Calculating IC Power Dissipation The power dissipation of the LTC1960 is dependent upon the gate charge of QTG and QBG (refer to Typical Application). The gate charge is determined from the manufacturer's data sheet and is dependent upon both the gate voltage swing and the drain voltage swing of the FET. PD = (VDCIN - VVCC) * [fOSC(QTG + QBG) + IVCC] + VDCIN * IDCIN Example: VVCC = 5.2V, VDCIN = 19V, fOSC = 345kHz, QG2 = QG3 = 15nC, IVCC = 0mA. PD = 165mW 1960fb 21 LTC1960 APPLICATIONS INFORMATION VSET/ISET Capacitors Capacitor C7 is used to filter the delta-sigma modulation frequency components to a level which is essentially DC. Acceptable voltage ripple at ISET is about 10mVP-P . Since the period of the delta-sigma switch closure, T , is about 10s and the internal IDAC resistor, RSET , is 18.77k, the ripple voltage can be approximated by: VISET V *T = REF R SET * C7 Then the equation to extract C7 is: C7 = VREF * T VISET * R SET = 0.8/0.01/18.77k(10s) @ 0.043F In order to prevent overshoot during start-up transients, the time constant associated with C7 must be shorter than the time constant of C5 at the ITH pin. If C7 is increased to improve ripple rejection, then C5 should be increased proportionally and charger response time to average current variation will degrade. Capacitor CB1 and CB2 are used to filter the VDAC deltasigma modulation frequency components to a level which is essentially DC. CB2 is the primary filter capacitor and CB1 is used to provide a zero in the response to cancel the pole associated with CB2. Acceptable voltage ripple at VSET is about 10mVP-P . Since the period of the deltasigma switch closure, T , is about 11s and the internal VDAC resistor, RVSET , is 7.2k, the ripple voltage can be approximated by: VREF * T VVSET = R VSET (CB1 || CB2 ) Then the equation to extract CB1 || CB2 is: CB1 || CB2 = VREF * T R VSET VVSET CB2 should be 10x to 20x CB1 to divide the ripple voltage present at the charger output. Therefore CB1 = 0.01F and CB2 = 0.1F are good starting values. In order to prevent overshoot during start-up transients the time constant associated with CB2 must be shorter than the time constant of C5 at the ITH pin. If CB2 is increased to improve ripple rejection, then C5 should be increased proportionally and charger response time to voltage variation will degrade. Input and Output Capacitors In the 4A Lithium Battery Charger (Typical Application section), the input capacitor (CIN) is assumed to absorb all input switching ripple current in the converter, so it must have adequate ripple current rating. Worst-case RMS ripple current will be equal to one-half of output charging current. Actual capacitance value is not critical. Solid tantalum, low ESR capacitors have a high ripple current rating in a relatively small surface mount package, but caution must be used when tantalum capacitors are used for input or output bypass. High input surge currents can be created when the adapter is hot-plugged to the charger or when a battery is connected to the charger. Solid tantalum capacitors have a known failure mechanism when subjected to very high turn-on surge currents. Only Kemet T495 series of "surge robust" low ESR tantalums are rated for high surge conditions such as battery to ground. The relatively high ESR of an aluminum electrolytic for C15, located at the AC adapter input terminal, is helpful in reducing ringing during the hot-plug event. Highest possible voltage rating on the capacitor will minimize problems. Consult with the manufacturer before use. Alternatives include new high capacity ceramic (at least 20F) from Tokin, United Chemi-Con/Marcon, et al. Other alternative capacitors include OSCON capacitors from Sanyo. The output capacitor (COUT) is also assumed to absorb output switching current ripple. The general formula for capacitor current is: IRMS V 0.29(VBAT ) 1- BAT VDCIN = (L1)(f) For example: VDCIN = 19V, VBAT = 12.6V, L1 = 10H, and f = 300kHz, IRMS = 0.41A. 1960fb 22 LTC1960 APPLICATIONS INFORMATION EMI considerations usually make it desirable to minimize ripple current in the battery leads, and beads or inductors may be added to increase battery impedance at the 300kHz switching frequency. Switching ripple current splits between the battery and the output capacitor depending on the ESR of the output capacitor and the battery impedance. If the ESR of COUT is 0.2 and the battery impedance is raised to 4 with a bead or inductor, only 5% of the current ripple will flow in the battery. PowerPath and Charge MUX MOSFET Selection Three pairs of P-channel MOSFETs must be used with the wall adapter and the two battery discharge paths. Two pairs of N-channel MOSFETs must be used with the battery charge path. The nominal gate drive levels are set by the clamp drive voltage of their respective control circuitry. This voltage is typically 6.25V. Consequently, logic-level threshold MOSFETs must be used. Pay close attention to the BVDSS specification for the MOSFETs as well; many of the logic-level MOSFETs are limited to 30V or less. Selection criteria for the power MOSFETs include the on-resistance RDS(ON), input voltage and maximum output current. For the N-channel charge path, the maximum current is the maximum programmed current to be used. For the P-channel discharge path maximum current typically occurs at end of life of the battery when using only one battery. The upper limit of RDS(ON) value is a function of the actual power dissipation capability of a given MOSFET package that must take into account the PCB layout. As a starting point, without knowing what the PCB dissipation capability would be, derate the package power rating by a factor of two. R DS(ON)MAX = P MOSFET 2 (IMAX ) 2 If you are using a dual MOSFET package with both MOSFETs in series, you must cut the package power rating in half again and recalculate. R DS(ON)MAX = P MOSFETDUAL 4 (IMAX ) 2 If you use identical MOSFETs for both battery paths, voltage drops will track over a wide current range. The LTC1960 linear 25mV CV drop regulation will not occur until the current has dropped below: ILINEARMAX = 25mV 2 RDS(ON)MAX However, if you try to use the above equation to determine RDS(ON) to force linear mode at full current, the MOSFET RDS(ON) value becomes unreasonably low for MOSFETs available at this time. The need for the LTC1960 voltage drop regulation only comes into play for parallel battery configurations that terminate charge or discharge using voltage. At first this seems to be a problem, but there are several factors helping out: 1. When batteries are in parallel current sharing, the current flow through any one battery is less than if it is running standalone. 2. Most batteries that charge in constant-voltage mode, such as Li-Ion, charge terminate at a current value of C/10 or less which is well within the linear operation range of the MOSFETs. 3. Voltage tracking for the discharge process does not need such precise voltage tracking values. The LTC1960 has two transient conditions that force the discharge path P-channel MOSFETs to have two additional parameters to consider. The parameters are gate charge QGATE and single pulse power capability. When the LTC1960 senses a LOW_POWER event, all the P-channel MOSFETs are turned on simultaneously to allow voltage recovery due to a loss of a given power source. However, there is a delay in the time it takes to turn on all the MOSFETs. Slow MOSFETs will require more bulk capacitance to hold up all the system's power supply function during the transition and fast MOSFET will require less bulk capacitance. The transition speed of a MOSFET to an on or off state is a direct function of the MOSFET gate charge. Q t = GATE IDRIVE 1960fb 23 LTC1960 APPLICATIONS INFORMATION IDRIVE is the fixed drive current into the gate from the LTC1960 and "t" is the time it takes to move that charge to a new state and change the MOSFET conduction mode. Hence, time is directly related to QGATE. Since QGATE goes up with MOSFETs of lower RDS(ON), choosing such MOSFETs has a counterproductive increase in gate charge making the MOSFET slower. Please note that the LTC1960 recovery time specification only refers to the time it takes for the voltage to recover to the level just prior to the LOW_POWER event as opposed to full voltage. The single pulse current rating of the MOSFET is important when a short-circuit takes place. The MOSFET must survive a 15ms overload. MOSFETs of lower RDS(ON) or MOSFETs that use more powerful thermal packages will have a high power surge rating. Using too small of a pulse rating will allow the MOSFET to blow to the open-circuit condition instantly like a fuse. Typically there is no outward sign of failure because it happens so fast. Please measure the surge current for all discharge power paths under worse case conditions and consult the MOSFET data sheet for the limitations. Voltage sources with the highest voltage and the most bulk capacitance are often the biggest risk. Specifically the MOSFETs in the wall adapter path with wall adapters of high voltage, large bulk capacitance and low resistance DC cables between the adapter and device are the most common failures. Remember to only use the real wall adapter with a production DC power cord when performing the wall adapter path test. The use of a laboratory power supply is unrealistic for this test and will force you to over specify the MOSFET ratings. A battery pack usually has enough series resistance to limit the peak current or are too low in voltage to create enough instantaneous power to damage their respective PowerPath MOSFETs. The highest frequency switching loop has the highest layout priority. For best results, avoid using vias in this loop and keep the entire high frequency loop on a single external PCB layer. If you must, use multiple vias to keep the impedance down (see Figure 9). SWITCH NODE L1 VBAT VIN CIN HIGH FREQUENCY CIRCULATING PATH 1. Keep the highest frequency loop path as small and tight as possible. This includes the bypass capacitors, with the higher frequency capacitors being closer to the noise source than the lower frequency capacitors. 24 COUT BAT 1960 F09 Figure 9. High Speed Switching Path 2. Run long power traces in parallel. Best results are achieved if you run each trace on separate PCB layer one on top of the other for maximum capacitance coupling and common mode noise rejection. 3. If possible, use a ground plane under the switcher circuitry to minimize capacitive interplane noise coupling. 4. Keep signal or analog ground separate. Tie this analog ground back to the power supply at the output ground using a single point connection. 5. For best current programming accuracy provide a Kelvin connection from RSENSE to CSP and CSN. See Figure 10 as an example. DIRECTION OF CHARGING CURRENT PCB Layout Considerations For maximum efficiency, the switch node rise and fall time is kept as short as possible. To prevent magnetic and electrical field radiation and high frequency resonant problems, proper layout of the components connected to the IC is essential. D1 RSNS 1960 F10 CSP CSN Figure 10. Kelvin Sensing of Charging Current 1960fb LTC1960 PACKAGE DESCRIPTION G Package 36-Lead Plastic SSOP (5.3mm) (Reference LTC DWG # 05-08-1640) 12.50 - 13.10* (.492 - .516) 1.25 0.12 7.8 - 8.2 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 5.3 - 5.7 0.42 0.03 7.40 - 8.20 (.291 - .323) 0.65 BSC 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 RECOMMENDED SOLDER PAD LAYOUT 2.0 (.079) MAX 5.00 - 5.60** (.197 - .221) 0 - 8 0.09 - 0.25 (.0035 - .010) 0.55 - 0.95 (.022 - .037) NOTE: 1. CONTROLLING DIMENSION: MILLIMETERS MILLIMETERS 2. DIMENSIONS ARE IN (INCHES) 0.65 (.0256) BSC 0.22 - 0.38 (.009 - .015) TYP 0.05 (.002) MIN G36 SSOP 0204 3. DRAWING NOT TO SCALE *DIMENSIONS DO NOT INCLUDE MOLD FLASH. MOLD FLASH SHALL NOT EXCEED .152mm (.006") PER SIDE **DIMENSIONS DO NOT INCLUDE INTERLEAD FLASH. INTERLEAD FLASH SHALL NOT EXCEED .254mm (.010") PER SIDE 1960fb 25 LTC1960 PACKAGE DESCRIPTION UHF Package 38-Lead Plastic QFN (5mm x 7mm) (Reference LTC DWG # 05-08-1701 Rev C) 0.70 0.05 5.50 0.05 5.15 0.05 4.10 0.05 3.00 REF 3.15 0.05 PACKAGE OUTLINE 0.25 0.05 0.50 BSC 5.5 REF 6.10 0.05 7.50 0.05 RECOMMENDED SOLDER PAD LAYOUT APPLY SOLDER MASK TO AREAS THAT ARE NOT SOLDERED 5.00 0.10 0.75 0.05 PIN 1 NOTCH R = 0.30 TYP OR 0.35 x 45 CHAMFER 3.00 REF 37 0.00 - 0.05 38 0.40 0.10 PIN 1 TOP MARK (SEE NOTE 6) 1 2 5.15 0.10 5.50 REF 7.00 0.10 3.15 0.10 (UH) QFN REF C 1107 0.200 REF 0.25 0.05 0.50 BSC R = 0.125 TYP R = 0.10 TYP BOTTOM VIEW--EXPOSED PAD NOTE: 1. DRAWING CONFORMS TO JEDEC PACKAGE OUTLINE M0-220 VARIATION WHKD 2. DRAWING NOT TO SCALE 3. ALL DIMENSIONS ARE IN MILLIMETERS 4. DIMENSIONS OF EXPOSED PAD ON BOTTOM OF PACKAGE DO NOT INCLUDE MOLD FLASH. MOLD FLASH, IF PRESENT, SHALL NOT EXCEED 0.20mm ON ANY SIDE 5. EXPOSED PAD SHALL BE SOLDER PLATED 6. SHADED AREA IS ONLY A REFERENCE FOR PIN 1 LOCATION ON THE TOP AND BOTTOM OF PACKAGE 1960fb 26 LTC1960 REVISION HISTORY (Revision history begins at Rev B) REV DATE DESCRIPTION PAGE NUMBER B 04/11 Updated Absolute Maximum Ratings section Added Note 8 2 5 Updated Pin Functions 8, 9 Updated equation in "The Current DAC Block" section 18 Updated equation in "Calculating IC Power Dissipation" section 21 Updated Typical Application 28 1960fb Information furnished by Linear Technology Corporation is believed to be accurate and reliable. However, no responsibility is assumed for its use. Linear Technology Corporation makes no representation that the interconnection of its circuits as described herein will not infringe on existing patent rights. 27 LTC1960 TYPICAL APPLICATION (LTC1960CG Pin Numbers Shown) Dual Battery Selector and 4A Charger PowerPath MUX BAT2 BAT1 VIN R1 5.1k 1% 4.7k R4 14k 1% MISO SCK MOSI SSB R5 1k 1% 100 C1 0.1F VDD C9 100pF C2 1F C6 1F RCL 0.03 R7 49.9k 1% C3 0.01F C8 0.1F 24 29 3 2 20 19 21 18 17 25 35 36 34 33 13 28 16 LTC1960 CLP DCIN BAT1 BAT2 MISO SCK MOSI SSB DCDIV COMP1 GCH2 SCH2 GCH1 SCH1 VSET VCC GND Q1, Q2, Q5, Q6, Q7, Q8: Si4925DY Q3, Q4, Q9, Q10, QTG, QBG: FDS6912A D1: MBR130T3 D2: CMDSH-3 TYPE D3, D4: BAT54A TYPE VPLUS GDCI GDCO GB1I GB1O GB2I GB2O SCP SCN LOPWR CSN CSP ITH ISET SW BOOST TGATE BGATE PGND BAT2 D3 1 7 6 9 8 11 10 5 4 12 22 23 14 15 30 31 32 27 26 Q6 Q7 Q2 Q5 Q8 RSC 0.02 R9 3.3k 1% C7 0.1F D2 C5 0.15F CIN 20F 25V RSNS 0.025 1% QBG CB1 0.01F CL 20F 25V L1 10H 4A C4 0.1F C6 2F LOAD R2 649k 1% R3 100k 1% QTG BAT1 D4 R11 1k CB2 0.1F Q1 COUT 20F 25V CHARGE MUX Q9 D1 R6 100 Q4 Q10 Q3 1960 TA02 RELATED PARTS PART NUMBER DESCRIPTION COMMENTS LT1505 Up to 97% Efficiency; AC Adapter Current Limit High Efficiency Battery Charger LTC1628-PG 2-Phase, Dual Synchronous Step-Down Controller Minimizes CIN and COUT ; Power Good Output; 3.5V VIN 36V LTC1709 2-Phase, Dual Synchronous Step-Down Controller with VID Up to 42A Output; Minimum CIN and COUT ; Uses Smallest Components for Intel and AMD Processors LTC3711 No RSENSETM Synchronous Step-Down Controller with VID 3.5V VIN 36V; 0.925V VOUT 2V; for Transmeta, AMD and Intel Mobile Processors LTC1759 SMBus Controlled Smart Battery Charger Synchronous Operation for High Efficiency; Integrated SMBus Accelerator; AC Adapter Current Limit LT1769 2A Battery Charger Constant-Current/Constant-Voltage Switching Regulator; Input Current Limiting Maximizes Charge Current 1960fb 28 Linear Technology Corporation LT 0411 REV B * PRINTED IN USA 1630 McCarthy Blvd., Milpitas, CA 95035-7417 (408) 432-1900 FAX: (408) 434-0507 www.linear.com LINEAR TECHNOLOGY CORPORATION 2001