November 2015 I N T H I S I S S U E CAN bus transceivers operate from 3.3V or 5V and withstand ±60V faults 13 low power IQ modulator for digital communications Volume 25 Number 3 Control Individual LEDs in Matrix Headlights with Integrated 8-Switch Flicker-Free Driver Keith Szolusha 20 low profile supercapacitor power backup with input current limiting 25 boost 12V to 140V with a single converter IC 30 LEDs combine design flexibility with practical, robust circuitry, enabling automotive designers to produce striking headlight designs matched by exceptionally long life and performance. Automobile designers are increasingly incorporating LEDs in lighting because they can be arranged in distinctive eye-catching designs—helping distinguish new models from old, or high end from economy. There is no question that automobile LED lighting has arrived, but it has not yet reached its full potential. Future models will feature more LED lights, including new shapes and colors, and more control over the individual LEDs. Simple strings of LEDs will give way to matrices of LEDs that can be individually dimmed via computer control, enabling unlimited real-time pattern control and animation. The future has arrived: Linear Technology’s LT®3965 matrix LED driver makes it easy to take the next step in automotive lighting design. I 2C CONTROL OF EIGHT POWER SWITCHES WITH A SINGLE IC A basic LED headlight design operates with uniform LED current, and thus, uniform brightness. But this leaves much of the LEDs’ potential on the table. Matrix headlights take advantage of the innate abilities of LEDs by enabling control of the brightness of individual LEDs within LED strings. It is not difficult, in theory, to address the individual LEDs in a matrix via computer-controlled power switches, allowing individual LEDs to be turned on or off, or PWM dimmed, The LT3965: limitless control of automotive matrix headlights and lighting systems w w w. li n ea r.com (continued on page 4) Linear in the News In this issue... COVER STORY LINEAR TECHNOLOGY CO-FOUNDERS RECEIVE ACE LIFETIME ACHIEVEMENT AWARD Control Individual LEDs in Matrix Headlights with Integrated 8-Switch Flicker-Free Driver Linear Technology co-founders Bob Swanson and Bob Dobkin received UBM Canon’s EDN and EE Times ACE Lifetime Achievement Award at a ceremony at the Embedded Systems Conference in Santa Clara, California in July. The award is presented annually to individuals for their contributions to the electronics industry, selected by a panel of academic and business leaders, and editors of EDN and EE Times. Keith Szolusha 1 DESIGN FEATURES CAN Bus Transceivers Operate from 3.3V or 5V and Withstand ±60V Faults Ciaran Brennan 13 Low Power IQ Modulator for Digital Communications Bruce Hemp and Sunny Hsiao 20 Low Profile Supercapacitor Power Backup with Input Current Limiting David Salerno 25 DESIGN IDEAS What’s New with LTspice IV? Gabino Alonso 28 Boost 12V to 140V with a Single Converter IC Victor Khasiev 30 back page circuits 32 Bob Swanson and Bob Dobkin founded Linear Technology 34 years ago, with a goal of building unique high performance analog ICs. The company now employs 4,865 people in over 50 locations worldwide. Linear’s name has become synonymous with innovative, high quality analog solutions in a broad range of application areas. EDN Senior Technical Editor Steve Taranovich stated, “Bob Swanson and Bob Dobkin are unique as individuals in their decades-deep contributions to integrated electronics, as well as their longevity in Silicon Valley. They are also exceptional as a team because of their unprecedented continuity of management at a semiconductor company, as well as their refreshing style of management, amplifying Linear Technology’s engineering talent. For these reasons and many more, EE Times and EDN chose to honor Bob Swanson and Bob Dobkin with the 2015 ACE Lifetime Achievement Award.” Steve Taranovich continued, “The talents and experiences of these two leaders, in my estimation, have brought about a corporate culture unlike most of the companies in the electronics business today. This culture encourages innovation and strongly values and recognizes the company’s engineering talent so that when a good idea emerges, management recognizes it and ‘gets out of the way’ to allow the engineer to bring it to fruition.” The following comments are excerpted from the EDN interview with Bob Swanson and Bob Dobkin: www.linear.com/46876. Bob Swanson: “If you were an Analog Aficionado, Linear Technology looked like a bunch of great technical guys just doing analog and they wanted to be part of that. After 34 years, we are still the same. We can hire as many good people as we can afford to hire. So the great thing we have going for us is that we have a disproportionate share of really innovative people. We leverage that because the world needs innovative analog solutions.” 2 | November 2015 : LT Journal of Analog Innovation Linear in the news They are strong, talented and intelligent leaders, but with a touch of humility and compassion for their employees from which other companies can learn a great deal. The book entitled, The Company That No One Leaves gives wonderful insight into one of the key reasons this company has had such success over the last 34 years.” www.linear.com/46751 Linear Technology co-founders Bob Swanson and Bob Dobkin received UBM Canon’s EDN and EE Times ACE Lifetime Achievement Award at a ceremony at the Embedded Systems Conference. Bob Dobkin: “One the things we do is Bob Swanson: “Smart, happy people to hire engineers who want to innovate and build products. And then we don’t get in their way. So they build products, they like what they are doing, and that works well within Linear.” probably can be innovators. I tell this story—about five or six years ago somebody asked Steve Jobs: ‘You guys are acknowledged to be the best at R&D and innovation, and you’ve got the smallest R&D budget.’ And he said, ‘All you need is a handful of real innovators and you can be successful. And I’ve got a handful of really good innovators and that’s all I need.’ So R&D is not an arms race. R&D is about having innovators. Going from 100 engineers to 200 does not automatically double your innovation. If you have 25 innovators and you can get two or three more a year, you’re in good shape.” Bob Dobkin: “If you do some products right, they will sell for over 30 years. If we do some products as well as they can be done, you never have to do them again.” Bob Swanson: “Obviously we have great products, but competitors have great products too. In this analog-challenged world, we have been so good at transferring our knowledge to customers. I will occasionally see the big customers at social events. They always tell me how much they depend upon our design and field people. Our field people are brilliant FAEs.” Bob Swanson: “One of the things that Bob Dobkin told me was that power was analog. So take an automobile with hundreds of processors—every one of them needs a power supply. And it is typically a power supply that is way more difficult than it was 20 years ago. The challenge in this ‘explosion of electronics’ in the automobile seems to me to be as much about solving analog issues as digital issues.” Bob Dobkin: “One way we are unique is that the engineering group comes up with the product ideas. We don’t go and do a marketing survey. Our engineers visit customers and figure out new products. We’re excited to work on things that the customers want.” Steve Taranovich concluded: “During this interview with these two industry icons, I sensed that they know their place in the industry and in their company, and effectively use the talent of their employees in a way that I personally have never seen in my 42 years in electronics. AWARDS •The LTC®2983 digital temperature measurement IC received the EDN/EE Times Analog Ultimate Product ACE Award. •The LTC2000 16-bit, 2Gsps DAC received the Best Product Award from EDN China in the Analog and Mixed Signal IC category. CONFERENCES & EVENTS The Wireless Congress: Systems & Applications, Konferenzzentrum, Munich, Germany, Nov. 17–18— Presenting energy harvesting and wireless sensor networks. Joy Weiss presents “Low Power Wireless Sensor Networks for IoT.” www.wireless-congress.com Energy Harvesting & Storage Conference, Santa Clara Convention Center, Santa Clara, California, Nov. 18–19, Booth N28—Presenting energy harvesting and wireless sensor networks. Ross Yu on “Low Power Wireless Sensor Networks for IoT”; Tony Armstrong on “Energy Harvesting: Battery Life Extension & Storage.” www.idtechex.com/energy-harvesting-usa 3rd Annual Analog Gurus Conference, Tokyo Conference Center Shinagawa, Tokyo, Japan, Nov. 18—Linear’s analog gurus present. CTO Bob Dobkin on“Inside Precision Voltage References,” Steve Pietkiewicz, VP, Power Management products on “High Performance µModules®” and Bob Reay, VP, Mixed Signal products on“High Precision Temperature Measurement.” n November 2015 : LT Journal of Analog Innovation | 3 When combined with a suitable constant-current LED driver, the matrix dimmer LED driver allows the individual LEDs to be computer-controlled in headlights, daytime running lights, brake and tail lights, side-bending lights, and other trim lighting. (LT3965, continued from page 1) to create unique patterns and functions. Each LED (or segment of LEDs) requires either its own converter or its own shunt power switch. It is possible to build a matrix driver with traditional driver/converter ICs that include a serial communications feature, but once more than two or three switches are needed for a matrix of LEDs, designing a discrete component solution becomes challenging, involving a matrix of components that exceeds the size of the LED matrix. The LT3965 I2C 8-switch matrix LED dimmer makes it easy to control large or small LED matrices (up to 512 LEDs). Figure 1 shows the LT3965 in action on Linear’s demonstration circuit DC2218. Its highly integrated design (Figure 2) minimizes component count. The individually addressable channels of the LT3965 can be used to control LED matrices in many ways, including: •Each LT3965 can control eight dimming channels—eight LEDs or eight clusters—within a string of LEDs. •The eight channels can control the individual red, green, blue and white light on two RGBW LED modules for adjustable brightness or changing color of dashboard or trim lighting. •Multiple LT3965s can be individually addressed on a single communications bus to multiply the strings in a large array. 4 | November 2015 : LT Journal of Analog Innovation •The LT3965 can control multiple LEDs per channel, or channels can be combined to efficiently control a single LED at higher current. When combined with a suitable constantcurrent LED driver, the matrix dimmer LED driver allows the individual LEDs to be computer-controlled in headlights, daytime running lights, brake and tail lights, side-bending lights, dashboard display and other trim lighting. The LT3965’s built-in automatic fault detection protects individual LEDs in case of a failure and reports failures to the microcontroller. The 60V LT3965 includes eight integrated 330mΩ power switches, which can be connected to one or more LEDs. The power switches act as shunt devices by turning off or PWM dimming the LEDs on a particular channel. The switches create eight individually controlled brightness channels (up to 256:1 dimming ratio) and eight faultproof segments of an LED string. The LT3965 can handle a string current of 500m A when all eight power switches are on at the same time (all LEDs off). The switches can be connected in parallel and run at 1A through four channels of LEDs as shown later in this article. Regardless of the number of LEDs or current, the LED string must be driven by a properly designed converter that has the bandwidth to handle the fast transients of the matrix dimmer. Some reference designs are included in this article. LT3797 BOOST-THEN-DUAL-BUCK MODE DRIVES TWO STRINGS, 16 LEDs AT 500mA WITH TWO LT3965s The eight shunt power switches of the LT3965 control the brightness of eight channels of LEDs at 500m A. The string voltage of the 8-LED matrix dimmer system can be between 0V and 26V, depending on how many LEDs are on or off at a given time. The recommended converter topology to drive these LEDs is a 30V step-down converter with high bandwidth and little or no output capacitor. This step-down topology requires that 9V–16V automotive input is “preboosted” to a 30V rail from which the step-down regulators can operate. The triple output LT3797 LED controller conveniently serves as a single-IC solution for both the “pre-boost” and step‑down functions—it can be configured as a step-up voltage regulator on one channel, followed by step-down LED drivers on the other two channels. Each of two stepdown LED drivers can drive a string of matrix-dimmed LEDs. This topology has a number of advantages, most notably, regardless whether the LED string voltages are above or below the battery voltage, the circuit continues to function optimally. Figure 3 shows the schematic of the demonstration board shown in Figure 1, a boost-then-dual-buck mode LT3797 and LT3965 matrix dimming headlight system with 16 LEDs at 500m A. Each LED can be individually controlled to be on, off or PWM dimmed down to 1/256 brightness. The 350kHz switching frequency of the LT3797 is outside the AM band (good design features Demonstration circuit DC2218 features a complete matrix LED dimmer system with LT3797 boost-then-dual-buck mode LED drivers and two LT3965 matrix dimmers that drive 16 LEDs at 500mA from a car battery. The board operates a matrix headlight with an attached I2C microcontroller via DC2026, the Linduino One demo circuit. 12V 3A DC POWER INPUT BOARDS CAN BE EASILY CONNECTED IN SERIES, ALL CONTROLLED BY A SINGLE µCONTROLLER TWO LT3965 MATRIX DIMMERS PER BOARD TWO 26V, 500mA SUPPLIES PRODUCED BY A SINGLE LT3797 BOOST-THEN-DUAL-BUCK MODE CONVERTER SYNC AT 350kHz CONTROLLED BY LTC6900. PWM AT 170Hz POT ALLOWS DIRECT PATTERN INTERACTION 16 HIGH POWER LEDs (EIGHT LEDs FOR EACH LT3965) MOUNTED ON BOARD WITH PROTECTIVE FILTER. CONNECTIONS FOR MORE LEDs PROVIDED. PREPROGRAMMED LINDUINO DC2026 INCLUDED. USB FOR µCONTROLLER REPROGRAMMING AND CONTOL VIA GUI SELECT FROM SEVEN PRELOADED DIMMING PATTERNS OR GUI CONTROL Photo: Steven Tanghe Figure 1. LT3965 LED matrix dimmer demonstration circuit DC2218 run as a Linduino™ shield (DC2026). This demonstration circuit runs headlight, turning light, tail light and trim patterns and can be evaluated with Linear’s graphical user interface via a USB cable. for EMI) and the resulting 170Hz PWM dimming frequency of the LT3965, generated from the same 350kHz clock, is above the visible range. With the system properly synchronized, the LT3797 and LT3965 matrix headlight operates flicker-free. The LT3797 buck mode converters are optimized for extremely fast transients with little or no output capacitor and properly compensated control loops. These >30kHz bandwidth converters tolerate fast LED transients as the LEDs are turned on and off and PWM dimmed at will. A filter capacitor placed on the LED sense resistor replaces a pole in the control system that is lost when the output capacitor is reduced or removed for the fast transient performance of the matrix dimmer. November 2015 : LT Journal of Analog Innovation | 5 LED+ VIN VIN CH8 DRN8 LED FAULT LED SW FAULT 8 LED+ FAULT DETECTOR DRIVER SRC8 EN/UVLO + – IS1 2.7µA CH7 1.24V LED SW FAULT BANDGAP REFERENCE INTERNAL BIAS SRC7 CH6 LED SW FAULT DRN6 FAULT DETECTOR DRIVER SDA SDA FAULT DETECTOR DRIVER VDD VDD 5V DRN7 SRC6 CH5 DRN5 SCL SCL ADDR1 ADDR2 ADDR3 ADDR4 LED SW FAULT I2C SERIAL INTERFACE ADDR1 ADDR2 FAULT DETECTOR DRIVER ADDR3 REGISTERS AND CONTROL LOGIC ADDR4 SRC5 CH4 LED SW FAULT FAULT DETECTOR DRIVER SRC4 TSD SET OVERHEAT FAULT –170°C ALERT ALERT CH3 LED SW FAULT SET LED FAULT 1.2V 0.6V SRC3 CH2 + – LED SW FAULT 0 PWMCLK DRN3 FAULT DETECTOR DRIVER + – DRN4 DRN2 FAULT DETECTOR DRIVER SRC2 1 Figure 2. LT3965 60V 8-switch LED matrix dimmer block diagram reveals eight power NMOS shunt switches for brightness control, a fault flag and I2C serial communications interface. CH1 INTERNAL OSCILLATOR 0.9V RTCLK RT A charge pump from the switch node is used to power the LT3965 VIN pin more than 7V above the LED+ voltage to enable the top channel NMOS to be fully enhanced when driven. The low RDS(ON) NMOS switches in the LT3965 enable high power operation without the IC getting 6 | November 2015 : LT Journal of Analog Innovation + – LED SW FAULT Q1 FAULT DETECTOR DRIVER GND hot, even when all eight shunt switches are on, turning the entire LED string off. In this case, the LT3797 LED driver survives the virtual output short created by all eight shunt switches without any issues, and is ready to quickly regulate 500m A through the next LED that is turned on. DRN1 SRC1 LEDREF LED– Demonstration circuit DC2218 (Figure 1) features the system shown in Figure 3 and operates a matrix headlight with an attached I2C microcontroller via DC2026, the Linduino™ One demo circuit. DC2218, operated as a large Linduino shield, has up to 400kHz serial code that design features The LT3965 I2C 8-switch matrix dimmer, LED driver eases the control of large or small LED matrices (up to 512 LEDs). Its highly integrated design minimizes component count, and built-in fault detection protects individual LEDs in case of a failure and reports failures to the microcontroller. Figure 3. LT3965 matrix LED dimmer system with LT3797 boost-then-dual-buck mode LED drivers and two LT3965 matrix dimmers that drive 16 LEDs at 500mA from a car battery. I2C serial communications control the brightness of individual LEDs and check for LED and channel faults. SKYHOOK VOUT ISP1 L1, L2: WURTH ELECTRONICS 7447789133 L3: WURTH ELECTRONICS 7443551151 L4: COOPER BUSSMANN SD25-470-R D1, D2: DIODES DFLS260 D3: DIODES PDS340 D4, D5, D10, D11: NXP SEMI PMEG6010CEJ D6, D7: NXP SEMI PMEG6010CEH D8, D9: NXP SEMI PMEG4010CEH M1, M2: VISHAY Si7308DN M3: VISHAY Si7850DP M4, M5: VISHAY Si7611DN Q1, Q2: ZETEX FMMT593TA 5V ALERT SDA 0.50Ω 0.50A 22µF 6.3V ISP2 SCL 0.50Ω 0.50A 10k ISN1 ISN2 10k TG1 10k D6 M4 1µF 50V 5V TO ATTACHED LINDUINO ONE DC2026C D10 D11 LED1+ 1µF D4 D5 0.50A INTVCC 0.1µF 0.1µF 0.1µF 0.1µF VLED 26V SYNC VIN 9V TO 36V L3 15µH + 33µF 50V D3 10µF 50V 10µF 50V ×6 OPTIONAL – COUT LED1 0.1µF VDD VIN SCL SCL SDA SDA ALERT 49.9k LED1+ LT3965 U1 49.9k LED2+ 1k EN/UVLO Q2 9.09k ADDR1 M3 43.2k D1 ADDR2 ADDR2 ADDR3 ADDR4 ADDR4 LEDREF LEDREF RTCLK RTCLK 5V 10Ω VIN 0.1µF 10V 57.6k 0.02Ω 1µF 50V GATE3 SENSEP3 SENSEN3 VIN 1M TG3 ISP3 ISN3 10Ω FBH3 1M 10k RT SENSEN1 SENSEP1 GATE1 SS3 47.5k D2 0.05Ω SYNC GATE2 SENSEP2 SENSEN2 0.22µF ISN1-2 FLT1-3 SW1 SW2 SS1-2 L4 47µH 1µF ON/OFF INTVCC TG1-2 ISP1-2 LT3797 VREF CTRL3 CTRL1-2 PWM1-2 OPTIONAL COUT 0.1µF 44.2k FBH2 M2 DIV 0.05Ω 499k EN/UVLO PWM3 69.8k LED2– L2 33µH OUT SET VLED 26V D9 LTC6900 GND 0.50A SRC1 GND 350kHz SYNC L1 33µH M1 ADDR1 ADDR3 D8 10Ω EN/UVLO 9.09k LED2+ DRN8 SRC8 DRN7 SRC7 DRN6 SRC6 DRN5 SRC5 DRN4 SRC4 DRN3 SRC3 DRN2 SRC2 DRN1 LT3965 U2 1k Q1 GND M5 1µF 50V SKYHOOK SRC1 44.2k 1M FBH1 D7 1µF ALERT DRN8 SRC8 DRN7 SRC7 DRN6 SRC6 DRN5 SRC5 DRN4 SRC4 DRN3 SRC3 DRN2 SRC2 DRN1 TG2 5V VIN VDD LT3470 5V REGULATOR 22µF 6.3V BOOST INTVCC 10µF 0.1µF GND VC1-2 FBH1-2 VC3 5.7k 2.2nF 15k 1nF 22nF November 2015 : LT Journal of Analog Innovation | 7 generates different headlight patterns and interfaces with Linear Technology’s graphical user interface (Figure 4). Within the GUI shown in Figure 4, LED brightness and fault protection functions can be examined with ALL CHANNEL MODE and SINGLE CHANNEL MODE commands, as well as FAULT CHECK read and write commands to check for open and short LEDs. Flicker-free operation, fault protection and transient operation can be examined with this demonstration circuit system. DC2218 can be plugged directly into a 12V DC source and it can be controlled by a personal computer running the GUI or reprogrammed from a simple USB connection. 1A MATRIX LED DRIVER USING PARALLEL CHANNELS The LT3965 can be used to drive matrices of 1A LED channels. It is easy to connect the power switches of the LT3965 in parallel so that two power switches split 1A of LED current and each LT3965 controls four 1A channels. One way to use parallel power switches for higher current is to run each of the anti-phase parallel switches for only 50% of the PWM period. By alternating and running 1A through a single NMOS power switch for half the time, the effective heating is about equal to running 500m A through the same NMOS all of the time. Figure 5 shows a 1A matrix headlight system using eight LEDs driven by two LT3965s and another boost-then-dual-buck mode LT3797. When PWM dimming, the LT3797 uses a unique 1/8-cycle phasing of the eight switches, as shown in Figure 6. In this 1A matrix system, LT3797 channels are combined in parallel pairs, so Figure 4. The PC-based interface allows designers to access control and monitoring of the LEDs driven by the LT3965. 8 | November 2015 : LT Journal of Analog Innovation design features The LT3965 can be used to drive matrices of 1A LED channels. It is easy to connect the power switches of the LT3965 in parallel so that two power switches split 1A of LED current and each LT3965 controls four 1A channels. Figure 5. 1A matrix LED driver combines anti-phase parallel channels for higher current applications in high power LED headlight systems. SKYHOOK VOUT ISP1 5V D13 TG1 0.1µF SYNC VIN 9V TO 18V + 33µF 50V M4 D10 OPTIONAL 10k 1µF 50V D4 1µF DRN4 DRN8 SRC4 SRC8 DRN3 DRN7 SRC3 SRC7 DRN2 DRN6 SRC2 SRC6 DRN1 DRN5 SRC1 SRC5 LED1+ D7 1A 0.1µF L3 10µH VLED 15V 0.1µF D3 4.7µF 25V 20V OPTIONAL 0.1µF LED1– 10µF 25V ×6 63.4k 1M FBH1 10Ω M3 66.5k D1 SCL SDA SDA LT3965 U1 0.015Ω GATE3 SENSEP3 SENSEN3 VIN 100k 249k 17.8k 38.3k TG3 1M ISP3 ISN3 DRN4 DRN8 SRC4 SRC8 DRN3 DRN7 SRC3 SRC7 DRN2 DRN6 SRC2 SRC6 DRN1 DRN5 SRC1 SRC5 49.9k LED2+ 1k LT3965 U2 1k Q1 Q2 9.09k EN/UVLO 9.09k ADDR1 ADDR1 ADDR2 ADDR2 ADDR3 ADDR3 ADDR4 ADDR4 LEDREF LEDREF RTCLK RTCLK D8 GND 5V VIN 0.1µF 10V OUT SET 1A VLED 15V LED2– 10Ω M2 DIV 0.033Ω FBH3 1M L2 15µH LTC6900 GND LED2+ D9 350kHz SYNC 10Ω D11 OPTIONAL ALERT SKYHOOK EN/UVLO L1 15µH M1 D5 VDD VIN SCL 57.6k 1µF 25V 1µF 50V 1µF 49.9k LED1+ GND TG2 M5 5V ALERT D6 22µF 6.3V ISN2 VIN VDD INTVCC 0.1µF 0.25Ω 1A 10k 5V TO ATTACHED LINDUINO ONE DC2026C D12 ISP2 SCL 10k ISN1 L1, L2: WURTH ELECTRONICS 74437349150 L3: WURTH ELECTRONICS 74432510000 L4: COOPER BUSSMANN SD25-470-R D1, D2: DIODES SBR2M30P1 D3: DIODES PDS1040 D4–D7, D10–D13: NXP SEMI PMEG4010CEJ D8, D9: NXP SEMI PMEG4010CEH M1, M2: VISHAY SiS402DN M3: INFINEON BSC0901NS M4, M5: VISHAY Si7611DN Q1, Q2: ZETEX FMMT593TA LT3470 5V REGULATOR ALERT SDA 0.25Ω 1A 22µF 6.3V OPTIONAL 0.1µF 63.4k FBH2 D2 0.033Ω SENSEN1 SENSEP1 GATE1 SYNC GATE2 SENSEP2 SENSEN2 TG1-2 EN/UVLO ISP1-2 LT3797 ISN1-2 FBH1-2 OVLO PWM3 VREF CTRL3 CTRL1-2 100k ANALOG DIM PWM1-2 RT SS3 52.3k SS1-2 0.1µF FLT1-3 SW1 SW2 L4 47µH 0.47µF ON/OFF INTVCC BOOST INTVCC 10µF 0.1µF GND VC1-2 VC3 4.7k 2.2nF 8.2k 330pF 10nF November 2015 : LT Journal of Analog Innovation | 9 Turning a high number of LEDs on or off presents a significant current load step to the DC/DC converter. The converters presented here handle these transients with grace, with a small output capacitor and high bandwidth. An ACM write transitioning a high number of LEDs produces no visible flicker or significant transient on the LED current. Figure 6. 1/8 PWM flickerfree phasing of the eight LT3965 power switches limits transients during PWM dimming brightness control. POR 1 DIMMING CYCLE = 2048 RTCLK CLOCK CYCLES 1/256 DIMMING (8 CLOCK CYCLES) CH1 PHASE SHIFT OF 1/8 DIMMING CYCLE = 256 CLOCK CYCLES CH2 CH3 CH4 CH5 CH6 CH7 CH8 that paired channels are anti-phase, 180° from each other; specifically pairing channels 8 and 4, 7 and 3, 6 and 2, and 5 and 1. Parallel channels alternate shunting, effectively doubling the PWM frequency, with the advantage of spreading out the shunted current and heat. For this to work properly, the maximum duty cycle for any single shunt power switch is 50%, because two anti-phase switches that are on 50% of the time (each shunting an LED 50% of the time) turns the LED off 100% of the time. Each LT3965 controls the brightness of four 1A LEDs that are driven by two 1A buck mode LT3797 channels (from the LT3797-boosted 20V channel). This high power, robust system can be expanded to power more LEDs with more LT3965s or higher current LEDs with more channels in parallel. It is possible to drive two LEDs per channel at 1A and drive up the power of this flexible headlight system. 10 | November 2015 : LT Journal of Analog Innovation MORE THAN ONE LED PER CHANNEL The LT3965 can support one to four LEDs per channel. Although it can be advantageous to individually control every single LED for fault protection or high resolution patterns, it is not always necessary. Using more than one LED per channel reduces the number of matrix dimmers in a system and is enough to accomplish the patterns or dimming required for some designs. Segments of headlights, signal lights and tail lights can have up to four LEDs with the same brightness. Emergency LED lights can have sets of three and four LEDs that blink and wave with the same pattern. The circuit in Figure 7 demonstrates a two-LED-per-channel system—it has the same number of LEDs as the circuit in Figure 3, but uses only a single LT3965 matrix dimmer instead of two. When an I2C command tells the LT3965 to turn on, off, or dim a channel, it affects the two LEDs that are controlled by that channel’s shunt power switch. To stay within the voltage limitations of the LT3965, the 16 LEDs at 500m A still need to be split into two series LED strings as they are in Figure 2. The same LT3797 circuit in Figure 2 can be used, but only a single LT3965 controls the brightness of the two strings. This demonstrates how each NMOS shunt power switch inside the LT3965 can be configured independently of the others, allowing an endless variety of matrix designs. ALL CHANNEL MODE AND SINGLE CHANNEL MODE I 2C COMMANDS WITH FLICKER-FREE PWM AND FADE The I2C instruction set of the LT3965 includes 1-, 2- and 3-word commands. These commands are sent over the serial data line (SDA) alongside the mastergenerated clock line (SCL) at up to 400kHz speed. The master microcontroller sends all channel mode (ACM) or single channel mode (SCM) write commands to control the brightness, fade, open-circuit threshold and short-circuit threshold of the LED channels and LT3965 addresses. Broadcast mode (BCM), ACM and SCM read commands request that the LT3965s report the content of their registers, design features The LT3965 can support one to four LEDs per channel. Segments of headlights, signal lights and tail lights can have two to four LEDs with the same brightness. Emergency LED lights can have sets of three and four LEDs that blink and wave with the same pattern. Figure 7. The flexible LT3965 can drive LED channels on independent LED strings and can drive between one and four LEDs per channel. (Complete driver circuit is similar to Figure 3, but with only one LT3956, as shown here.) LED1+ LED2+ SKYHOOK 5V 1µF SKYHOOK 1µF 10k VIN VDD EN/UVLO 10k 10k SCL SCL SDA SDA ALERT DRN8 0.50A VLED 26V LT3965 LED1– including open and short registers for fault diagnostics. The LT3965 asserts an ALERT flag when there is a new fault. The micro can respond to the fault by determining which LT3965 reported the fault, as well as the type and channel of fault. In the case that multiple LT3965 ICs are reporting faults, the LT3965s can sequence fault reporting to the master to prevent overlap errors. This makes the alert response system reliable and conclusive. A complete list of the registers and command set is given in the LT3965 data sheet. ACM write commands instantly turn all of the eight channels of a single LT3965 address on or off with just two I2C words—the channels transition on DRN4 SRC8 DRN7 SRC4 DRN3 SRC7 DRN6 SRC3 DRN2 SRC6 DRN5 SRC2 DRN1 SRC5 SRC1 0.50A VLED 26V SHORT AND OPEN LED FAULT PROTECTION FOR EACH CHANNEL ADDR1–4 RTCLK 350kHz SYNC ALERT GND LEDREF LED2– or off at the same time. Turning a high number of LEDs on or off presents a significant current voltage load step to the DC/DC converter. The converters presented here handle these transients with grace, with little or no output capacitor and high bandwidth. As shown in Figure 8, an ACM write transitioning a high number of LEDs produces no visible flicker or significant transient on the LED current of other channels. The high bandwidth buck mode converter built around the LT3797 is the reason for such a small and controlled transient. Single channel mode writes produce relatively small and fast single-LED transients. SCM writes are used to set the brightness of only one channel at a time to ON, OFF, or PWM dimming with or without fade. PWM dimming values between 1/256 and 255/256 are communicated in 3-word writes while ON and OFF can be communicated in shorter, 2-word commands. A fade bit on a single SCM write command enables the LT3965 to move between two PWM dimming levels with internally determined logarithmic fade and no additional I2C traffic. The open and short thresholds of each channel can be set between one and four LEDs with SCM write commands. Short- and open-circuit protection is an inherent benefit of the matrix dimmer. Each channel’s NMOS power switch can shunt out between one and four series LEDs. Traditional LED strings have protection against the entire string being open or shorted and only some ICs have output diagnostic flags to indicate these fault conditions. In contrast, the LT3965 protects against, and rides through, individual channel shorts and opens, keeping operational channels alive and running while recording and reporting the fault conditions. When a fault occurs within a string, the LT3965 detects the fault and asserts its ALERT flag, indicating to the microcontroller that there is an issue to be addressed. If the fault is an open-circuit, the LT3965 automatically turns on its November 2015 : LT Journal of Analog Innovation | 11 Short- and open-circuit protection is an inherent benefit of the matrix dimmer. Each channel’s NMOS power switch can shunt out between one and four series LEDs. Traditional LED strings have protection against the entire string being open or shorted and only some ICs have output diagnostic flags to indicate these fault conditions. In contrast, the LT3965 protects and rides through individual channel shorts, keeping operational channels alive and running while recording and reporting the fault conditions. corresponding NMOS power switch, bypassing the faulty LED until a full diagnosis occurs or until the fault is removed. The LT3965 maintains registers of open and short faults for each channel and returns the data to the microcontroller during I2C fault read commands. The command set includes reads that leave the status register unchanged and those that clear the fault registers, allowing user-programmable fault diagnostics. Registers can be read in the various modes allowed for writes, SCM, ACM, BCM: •Single channel mode (SCM) reads return the open and short register bits for a single channel. SCM reads also check the open and short threshold register, the mode control, and the 8-bit PWM dimming value for that channel. •All channel mode (ACM) reads return the open and short register bits for all channels of a given address without clearing the bits, as well as the ACM ON and OFF bits for all eight channels. •The ACM and SCM reads can be used to check and clear faults and to read all of the registers for a robust I2C communications system. UP TO 16 ADDRESSABLE LT3965s ON THE SAME BUS Every LT3965 features four user-selectable address bits, enabling 16 unique bus addresses. Every ACM and SCM I2C command is sent to the shared communications bus, but action is only taken by the addressed LT3965. BCM commands are followed by all ICs on the bus. The 4-bit address architecture allows a single microcontroller and a single I2C 2-line communications bus to support up to 8 × 16 = 128 individually controllable channels. With the LT3965, for all but the most ambitious lighting displays, all individual LEDs in an automobile’s Figure 8. The LED matrix driver designs shown in this article feature minimal to no cross-channel transient effects. For instance, transitioning half the channels—here, simultaneously turning on two and turning off two—has little to no transient effect on the other four, untouched channels. The nontransitioned channels remain flicker free. ILED1–4 500mA/DIV (100% ON) ILED5,6 500mA/DIV (ON TO OFF) ILED7,8 500mA/DIV (OFF TO ON) ALL CHANNEL MODE (ACM) COMMAND •In more complex systems with many LT3965 matrix dimmers sharing the same bus, a broadcast mode (BCM) read first requests which, if any, LT3965 address has asserted the fault flag. 20µs/DIV 12 | November 2015 : LT Journal of Analog Innovation headlight, tail light and trim lights can be controlled by a single I2C communications bus and a single microcontroller. Given that each channel can be connected to up to four LEDs, one relatively easy-to-implement system can support matrix dimming for up to 512 LEDs. CONCLUSION The LT3965 matrix LED dimmer controls eight LED-brightness channels on a single LED string, giving lighting designers unlimited access to sophisticated and striking automotive lighting designs. The I2C communications interface allows a microprocessor to control the brightness of individual LEDs in the string. Fault protection in the I2C interface ensures LED lighting system robustness. The channels of the matrix dimmer are versatile: each channel can control multiple LEDs; channels can be combined to support higher current LEDs; or high LED-count systems can be produced with up to 16 matrix dimmer ICs on the same communications bus. Take the next step in designing automotive headlights, tail lights, front, side, dash and trim lights—the future is now. n design features CAN Bus Transceivers Operate from 3.3V or 5V and Withstand ±60V Faults Ciaran Brennan The LTC2875 is a robust CAN bus transceiver that features ±60V overvoltage and ±25kV ESD tolerance to reduce failures caused by electrical overstress. These transceivers introduce several new capabilities for high voltage tolerant CAN bus transceivers: operation from 3.3V or 5V supply voltages, up to 4Mbps data rate, ±36V common mode voltage range, continuously variable slew rate and availability in 3mm × 3mm DFN packages. The CAN bus forms the backbone of many automotive, commercial and industrial data communications systems. CAN bus networks are used in a wide variety of applications, including automotive and transportation electronics, industrial control systems, supervisory control and data acquisition systems, building automation and security, HVAC control, and other custom networked systems. Robustness to electrical overstress is an important attribute for CAN bus transceivers used in these applications, which risk exposure to wiring faults, ground voltage faults and lightning induced surge voltages. by contemporary network applications. The LTC2875 transceiver is Linear Technology’s response to these requests. However, few CAN transceivers capable of operating from 3.3V supplies are available, and until now, none offer the high voltage tolerance and wide common mode operating range of the LTC2875. Many customers have requested a robust CAN bus transceiver with the performance and the expanded capabilities demanded 3.3V OR 5V OPERATION Figure 1. LTC2875 demo circuit with DFN and SO packages in foreground Most high voltage tolerant CAN bus transceivers can operate only from a 5V supply, but 5V is rarely used by most modern digital circuits. The CAN bus transceiver may be the only 5V component in the system. A high voltage tolerant CAN bus transceiver that operates from a 3.3V supply reduces design time and cost by eliminating the need for a dedicated 5V supply. The LTC2875 maintains compatibility with the ISO 11898-2 CAN bus standards when operating from a 3.3V supply, driving the full specified differential bus voltage VOD and maintaining the same receiver input threshold voltages. The only difference between 3.3V and 5V operation is that the common mode bus voltage is reduced to 1.95V while operating at 3.3V, which falls below the range of 2V to 3V specified by ISO 11898-2. This minor shift in common mode voltage falls within the minimum common mode voltage range of −2V to 7V specified in the standard (and is truly inconsequential to the ±25V common mode voltage range of the LTC2875 when operating at 3.3V), allowing the LTC2875 Ciaran Brennan November 2015 : LT Journal of Analog Innovation | 13 The LTC2875 provides a continuously variable slew rate over an approximate 20-to-1 range. The lowest slew rate is appropriate for data rates of 200kbps or less. The slew rate is programmed by a single resistor in series with the chip enable pin RS. RSL RS CANH LTC2875 TXD VCC RXD 15pF 47µF TXD RS GND CANH VCC CANL RXD SPLIT RL/2 1% VOD CM RL/2 1% 0.1µF CANL VOC GND RSL = 0Ω EXCEPT AS NOTED (a) Figure 3. Single resistor termination (a) and split termination (b) RSL RS CANH LTC2875 TXD VCC RXD 15pF 47µF TXD RS GND CANH VCC CANL RXD SPLIT RL/2 1% VOD RL/2 1% 4.7nF 0.1µF CANL GND RSL = 0Ω EXCEPT AS NOTED (b) to communicate seamlessly with any other ISO11898-2 compliant transceivers. The LTC2875 is fully interoperable with other transceivers on the same bus that are powered by 5V when operating from either a 3.3V or 5V supply. Not all systems require a high data rate. In applications where lower data rates suffice, the system designer may prefer a CAN bus driver with low electromagnetic 60 4 Mbps DATA RATE WITH HIGH SYMMETRY DRIVER AND CONTINUOUS SLEW RATE CONTROL 14 | November 2015 : LT Journal of Analog Innovation SLEW RATE (V/µs) Modern CAN bus systems may operate at data rates that exceed the capabilities of existing high voltage tolerant transceivers. For example, Linear Technology’s LT1796 CAN transceiver operates at a maximum of 125kbps. The LTC2875 offers similar high voltage tolerance to this predecessor, but can communicate 32 times faster, up to 4Mbps. 50 VCC = 5V 40 VCC = 3.3V 30 20 10 0 1 10 RSL (kΩ) 100 Figure 2. Slew rate vs slew control resistor RSL emissions (EME) slew controlled transitions. The LTC2875 provides a continuously variable slew rate over an approximate 20-to-1 range. The lowest slew rate is appropriate for data rates of 200kbps or less. The slew rate is programmed by a single resistor in series with the chip enable pin RS, as plotted in Figure 2. Considerable design effort was made to keep the switching symmetry of the CAN transmitter highly symmetrical (or more accurately, anti-symmetrical) between the CANH and CANL outputs, because any asymmetry between the switching waveforms of the two outputs produces a change in the common mode voltage. While the electromagnetic fields produced by the differential voltage along a twisted design features The LTC2875 transceivers operating from a 3.3V supply can interoperate with other CAN transceivers operating from a 5V supply on the same bus. The only major difference between operating at 3.3V and 5V is that the common mode voltages are ~1.95V and ~2.5V, respectively. pair largely cancel and produce little EME, the electromagnetic fields of the common mode voltage on the pair add together and may produce significant EME, particularly if the twisted pair is unshielded. Therefore, good CAN transmitter switching symmetry results in lower EME. The LTC2875 provides two features to reduce the EME produced by fluctuations of the common mode voltage during switching: variable slew rate control and split termination. The transmitter slew rate can be programmed by a single resistor in series with the enable pin RS. Reducing the slew rate reduces the high frequency content of the switching waveforms. Split termination entails dividing the terminator resistor at each end of the bus into two equal, series resistors of half the termination resistance value, with the center point of the resistors biased at the DC common mode voltage supplied by the SPLIT pin and a decoupling capacitor (Figure 3). Split termination provides a low impedance load for the common mode signal while maintaining the proper termination for the differential signals. The low impedance common mode loading helps suppress common mode voltage fluctuations. The effectiveness of the split termination in reducing EME from the common mode voltage fluctuations is illustrated in Figure 4. In this figure, the voltages at the CANH and CANL terminals and the common mode voltage are recorded for an LTC2875 transmitting at 1Mbps over a 10-meter unshielded twisted pair, with VCC = 3.3V, the slew rate set to maximum, and 120Ω termination resistors placed on each end of the cable. The FFT power spectra of the common mode voltage waveforms are also shown in Figure 4. The results using split termination and those using single resistor termination are both shown. The waveforms with the single resistor termination show a larger magnitude of common mode transients during the switching transition, as well as a damped oscillation after the dominant to recessive 4V transition. This damped oscillation is the result of the inductance of the line interacting with line and transceiver capacitance after the transceiver switches to its high impedance recessive state. In this example, the common mode voltage in the recessive state with the single resistor termination is loaded only by the four 40kΩ input resistors of two LTC2875 devices, one on each end of the cable, for a parallel resistance of 10k. By contrast, the common mode voltage 4V TRANSMITTER SPLIT TERMINATION MAXIMUM SLEW RATE CANH 0.5V/DIV CANH 0.5V/DIV COMMON MODE CANL 0V 0µs 0dB COMMON MODE CANL 200ns/DIV 2µs FFT POWER SPECTRUM SPLIT TERMINATION MAXIMUM SLEW RATE 0V 0µs 0dB 200ns/DIV 2µs FFT POWER SPECTRUM SINGLE TERMINATION MAXIMUM SLEW RATE 20dB/DIV 20dB/DIV −160dB 0MHz TRANSMITTER SINGLE TERMINATION MAXIMUM SLEW RATE 5MHz/DIV 50MHz −160dB 0MHz 5MHz/DIV 50MHz Figure 4. Transmitter waveforms and FFT power spectrum plots of the common mode voltage for split and single bus terminations on 10m unshielded twisted pair cable; VCC = 3.3V, 1Mbps November 2015 : LT Journal of Analog Innovation | 15 The use of the split termination results in a significant reduction in common mode noise across the frequency spectrum when transmitting at 100kbps. At these lower data rates, further reductions in the common mode noise spectrum can be obtained by setting the LTC2875 to its minimum slew rate. in the split termination case is also loaded by the four 60Ω split termination resistors, for a parallel resistance of 15Ω, in series with two parallel 4.7nF capacitors. The common mode voltage FFT power spectrum is lower in amplitude across a wide range of frequencies for the split termination compared to the single resistor termination. For transmitting at a lower data rate, a slower slew rate may be used for additional reduction in common mode EME. Figure 5 illustrates four cases with minimum and maximum slew rates, combined with split or single resistor termination. These measurements were performed with the same test configuration as those shown in Figure 4, except that the data rate was reduced to 100kbps and waveforms at both the minimum and maximum slew rates were recorded. As in the 1Mbps waveforms of Figure 4, the use of the split termination results in a significant reduction in common mode noise across the frequency spectrum when transmitting at 100kbps. At these lower data rates, further reductions in the common mode noise spectrum can be obtained by setting the LTC2875 to its minimum slew rate. In this example, combining both the split termination and the minimum slew rate reduces the common mode noise power by 20dB or more over most of the recorded spectrum, compared to the single resistor termination combined with the maximum slew rate. 16 | November 2015 : LT Journal of Analog Innovation Figure 5. Transmitter waveforms and FFT power spectrum plots of the common mode voltage for four combinations of termination and slew rate; VCC = 3.3V, 100kbps, 10m unshielded twisted pair cable 4V 4V TRANSMITTER SPLIT TERMINATION MAXIMUM SLEW RATE TRANSMITTER SPLIT TERMINATION MINIMUM SLEW RATE CANH 0.5V/DIV CANH 0.5V/DIV COMMON MODE COMMON MODE CANL 0V 0µs CANL 2µs/DIV 4V 20µs 0V 0µs 4V TRANSMITTER SINGLE TERMINATION MAXIMUM SLEW RATE CANH 0.5V/DIV COMMON MODE COMMON MODE CANL 0V 0µs 0dB CANL 2µs/DIV 20µs 0V 0µs 0dB FFT POWER SPECTRUM MAXIMUM SLEW RATE 2µs/DIV 20µs FFT POWER SPECTRUM MINIMUM SLEW RATE SINGLE TERMINATION SINGLE TERMINATION 20dB/DIV 20dB/DIV SPLIT TERMINATION −160dB 0MHz 20µs TRANSMITTER SINGLE TERMINATION MINIMUM SLEW RATE CANH 0.5V/DIV 2µs/DIV 5MHz/DIV SPLIT TERMINATION 50MHz −160dB 0MHz 5MHz/DIV 50MHz design features Another technique to reduce EME from common mode voltage fluctuations is using a common mode choke. The choke increases the source impedance of the common mode signal, and in conjunction with capacitors added between CANH and CANL and GND, forms a lowpass filter that attenuates high frequency noise. REDUCING EME WITH A COMMON MODE CHOKE Another technique to reduce EME from common mode voltage fluctuations is using a common mode choke. The choke increases the source impedance of the common mode signal, and in conjunction with capacitors added between CANH and CANL and GND, forms a lowpass filter that attenuates high frequency noise. The effectiveness of a 100µ H common mode choke in conjunction with two 33pF capacitors in reducing the common mode noise is shown in Figure 6. In this example, VCC = 3.3V, split termination was employed, the twisted pair cable was 10 meters long and the data rate was 100kbps. Figure 6. Transmitter waveforms and FFT power spectrum plots of the common mode voltage, with or without common mode choke 4V 4V SPLIT TERMINATION MAXIMUM SLEW RATE NO COMMON MODE CHOKE SPLIT TERMINATION MINIMUM SLEW RATE NO COMMON MODE CHOKE CANH 0.5V/DIV CANH 0.5V/DIV COMMON MODE COMMON MODE CANL 0V 0µs 4V CANL 2µs/DIV 20µs SPLIT TERMINATION MAXIMUM SLEW RATE 100µH COMMON MODE CHOKE 0V 0µs 4V 2µs/DIV 20µs SPLIT TERMINATION MINIMUM SLEW RATE 100µH COMMON MODE CHOKE CANH CANH MIXED 3.3V AND 5V OPERATION The LTC2875 transceivers operating from a 3.3V supply can interoperate with other CAN transceivers operating from a 5V supply on the same bus. The only major difference between operating at 3.3V and 5V is that the common mode voltages are ~1.95V and ~2.5V, respectively. The common mode of the bus therefore fluctuates depending on the logical state of the bus. When all transmitters are in the recessive state, the common mode voltage settles to some intermediate voltage depending on all the resistive loads placed on the bus, including receiver input resistors, and split termination resistors (if present). When a transmitter powered by 5V is dominant, it pulls the common mode voltage toward 2.5V. When a transmitter is powered by 0.5V/DIV 0.5V/DIV COMMON MODE COMMON MODE CANL 0V 0µs 0dB CANL 2µs/DIV 20µs 0V 0µs 0dB FFT POWER SPECTRUM MAXIMUM SLEW RATE 20µs FFT POWER SPECTRUM MINIMUM SLEW RATE NO COMMON MODE CHOKE NO COMMON MODE CHOKE 20dB/DIV 20dB/DIV 100µH COMMON MODE CHOKE −160dB 0MHz 2µs/DIV 5MHz/DIV 50MHz 100µH COMMON MODE CHOKE −160dB 0MHz 5MHz/DIV 50MHz November 2015 : LT Journal of Analog Innovation | 17 An ideal CAN bus transceiver would survive large common mode voltages and continue to send and receive data without disruption. Accordingly, the receiver in the LTC2875 is designed to operate over an expanded ±36V common mode voltage range when operating from a 5V supply, and ±25V when operating from a 3.3V supply. 3.3V, it pulls the common mode voltage toward 1.95V. The common mode voltage fluctuates between 2.5V and 1.95V, resulting in a modest increase in EME. An example of mixed voltage operation is shown in Figure 7. The experimental setup consists of two LTC2875 transceivers, each connected to the end of a 10-meter twisted pair, with split termination employed. Each transceiver alternates in driving a dominant state on the bus. The waveforms are recorded on the CANH and CANL pins of the near side transceiver. In the plots shown on the left, both the near and far transceivers are powered by a 3.3V supply. The common mode voltage remains near 1.95V with only minor perturbations. In the plots on the right, by comparison, the near transceiver remains powered by 3.3V, while the far transceiver is powered by 5V. The recessive common mode voltage settles to about 2.23V, the average of 2.5V and 1.95V. When the near side transceiver is dominant, the common mode voltage is pulled down close to 1.95V, whereas when the far side transceiver is dominant, the common mode voltage is pulled close to 2.5V. The difference in EME resulting from common mode voltage fluctuations can be seen by comparing the FFT power spectra of the common mode voltage recorded at the terminals of the near transceiver. An increase in power of approximately 8dB is observed from 0MHz to 25MHz for the mixed power supply voltage case, with the difference rolling off above that frequency. 18 | November 2015 : LT Journal of Analog Innovation 4V 4V NEAR SIDE DOMINANT 3.3V FAR SIDE DOMINANT 3.3V NEAR SIDE DOMINANT 3.3V 0.5V/DIV 0.5V/DIV SPLIT TERMINATION MAXIMUM SLEW RATE NEAR & FAR SIDES POWERED BY 3.3V 0V 0µs 0dB FAR SIDE DOMINANT 5V SPLIT TERMINATION MAXIMUM SLEW RATE NEAR SIDE POWERED BY 3.3V; FAR SIDE 5V 5µs/DIV 50µs 0V 0µs 0dB FFT POWER SPECTRUM MAXIMUM SLEW RATE 5µs/DIV 50µs FFT POWER SPECTRUM MINIMUM SLEW RATE NEAR: 3.3V; FAR: 5V 20dB/DIV 20dB/DIV NEAR & FAR: 3.3V −160dB 0MHz 5MHz/DIV 50MHz NEAR & FAR: 3.3V −160dB 0MHz 5MHz/DIV 50MHz Figure 7. Transmitter waveforms and FFT power spectrum plots of the common mode voltage for two transmitters powered by identical and mixed power supply voltages; 10m unshielded twisted pair cable, 100kbps, maximum slew rate ±36V COMMON MODE VOLTAGE RANGE Standard CAN bus transceivers operate over a limited common mode voltage range that extends from −2V to +7V. In commercial or industrial environments, ground faults, noise, and other electrical interference can induce common mode voltages that greatly exceed these limits. An ideal CAN bus transceiver would survive large common mode voltages and continue to send and receive data without disruption. Accordingly, the receiver in the LTC2875 is designed to operate over an expanded ±36V common mode voltage range when operating from a 5V supply, and ±25V when operating from a 3.3V supply. The receiver uses low offset bipolar differential inputs, combined with high precision resistor dividers, to maintain precise receiver thresholds over the wide common mode voltage range. The design features The LTC2875 is a groundbreaking ±60V overvoltage tolerant CAN bus transceiver that operates from either 3.3V or 5V supplies. Its industrial robustness is matched by superior performance, application flexibility and excellent EME characteristics. transmitters operate up to the absolute maximum voltages of ±60V, and will sink or source current up to the limits imposed by their current limit circuitry. HOT PLUGGING, HOT SWAPPING, AND GLITCH-FREE POWER-UP AND POWER-DOWN The LTC2875 features glitch-free powerup and power-down protection to meet hot plugging (or hot swap) requirements. These transceivers do not produce a differential disturbance on the bus when they are connected to the bus while unpowered, or while powered but disabled. Similarly, these transceivers do not produce a differential disturbance on the bus when they are powered up in the disabled state while already connected to the bus. In all of these cases the receiver output RXD remains high impedance (with an internal 500k pull-up resistor), while the CANH and CANL outputs remain in the high impedance recessive state. If the transceiver is powered up in the enabled state, the chip goes active shortly after the supply voltage passes through the transceiver’s internal power good detector threshold. The RXD output reflects the state of the bus data when the chip goes active, while the transmitter remains in the recessive state with its outputs high impedance until the first recessive-to-dominant transition of TXD after the chip goes active. If the transmitter is in the enabled state when the chip is powered down, the chip goes inactive shortly after the supply voltage passes through the transceiver’s internal supply undervoltage detector threshold. If the transmitter is in the dominant state at this time, the outputs smoothly switch to the recessive state. Regardless of whether it is outputting a dominant or recessive state, the receiver output RXD smoothly switches to a high impedance state (weakly pulled up through an internal 500k pull-up resistor). ±60V FAULT AND ±25KV ESD TOLERANCE CAN bus wiring connections in industrial installations are sometimes made by connecting the bare twisted wire to screw terminal blocks. The apparatus containing the CAN bus interface may house circuits powered by 24V AC/DC or other voltages that are also connected with screw terminals. The handling of exposed wires and screw terminals by service personnel introduces the risk of ESD damage, while the possibility of wiring the cables to the wrong screw terminals introduces the risk of overvoltage damage. The high fault voltage and ESD tolerance make the LTC2875 exceptionally resistant to damage from these hazards. LTC2875 is protected from ±60V faults even with GND open, or VCC open or grounded. The LTC2875 is protected from electrostatic discharge from personnel or equipment up to ±25kV (HBM) to the A, B, Y and Z pins with respect to GND. On-chip protection devices begin to conduct at voltages greater than approximately ±78V and conduct the discharge current safely to the GND pin. Furthermore, these devices withstand up to ±25kV discharges even when the part is powered up and operating without latching up. All the other pins are protected to ±8kV (HBM). CONCLUSION The LTC2875 is a groundbreaking ±60V overvoltage tolerant CAN bus transceiver that operates from either 3.3V or 5V supplies. Its industrial robustness is matched by superior performance, application flexibility and excellent EME characteristics. n The ±60V fault protection of the LTC2875 is achieved by using high voltage BiCMOS integrated circuit technology. The naturally high breakdown voltage of this technology provides protection in powered-off and high impedance conditions. The driver outputs use a progressive foldback current limit design to protect against overvoltage faults while still allowing high current output drive. The November 2015 : LT Journal of Analog Innovation | 19 Low Power IQ Modulator for Digital Communications Bruce Hemp and Sunny Hsiao IQ modulators are versatile building blocks for RF systems. The most common application is generating RF signals for digital communication systems. This article illustrates the modulation accuracy of the LTC5599 low power IQ modulator, and shows by simple example how to integrate the device into a digital communication system. MODULATOR APPLICATIONS Figure 1. Test setup to measure basic modulation accuracy LO ROHDE & SCHWARZ SMJ 100A 450MHz, 0 dBm BASEBAND IQ SIGNAL GENERATOR ROHDE & SCHWARZ SMJ 100A OR EQUIVALENT DAC (14-BIT) + LPF DAC (14-BIT) + LPF I+ I− Q+ Q− LTC5599 RF OUT −4 dBm RMS LOW POWER IQ MODULATOR Virtually any type of RF modulation can be generated with IQ modulation, within the center frequency, bandwidth and accuracy capabilities of the modulator device. Table 1 shows some of the applications of the LTC5599. MODULATION: 16-QAM, PRBS 9, ROOT COSINE ALPHA 0.35, 30k SYM./SEC VBIAS = 1.4V, IQ DRIVE = 1.1VP–P(DIFF), CREST FACTOR 5.4dB Table 1. Some possible applications for the LTC5599 low power IQ modulator. APPLICATION MOD STD MODULATION TYPE (REFERENCE 1) MAX RF BW Digital wireless microphones Proprietary QPSK, 16/32/64-DAPSK, Star-QAM 200kHz 802.11af OFDM: BPSK, QPSK, 16/64/256-QAM Up to 4× 6MHz channels DOCSIS 16-QAM 6MHz Custom Wide programmability range — — AM, FM/PM, SSB, DSB-SC — TETRA π/4-DQPSK, π/8-D8PSK, 4/16/64-QAM 25kHz to 150kHz • Commercial TETRAPOL GMSK 10kHz, 12.5kHz • Industrial P-25 C4FM, CQPSK 6.25kHz to 12.5kHz DMR 4FSK 6.25kHz, 12.5kHz Wireless networking • White-space radios • Cognitive radio CATV upstream Military radios (portable, manpack) Software defined radios (SDR) Portable test equipment Analog modulation 2-way radios • Public safety 20 | November 2015 : LT Journal of Analog Innovation design features Figure 2. LTC5599 EVM measured using lab-grade baseband and LO signal generators. Note that the MER measures over 49dB, basically “Broadcast Quality.” MEASURED IQ VECTOR MEASURED EYE DIAGRAM MODULATION ACCURACY AND EVM The error vector magnitude or EVM is a measure of modulation accuracy in digital radio communication systems. Modulation accuracy is important because any error on the modulated signal can cause reception difficulty or excessive occupied bandwidth. If left unchecked, the receiver could exhibit excessive bit errors, the effective receiver sensitivity could be degraded or the transmit adjacent channel power (ACP) can become elevated. EVM VS TIME An error vector is a vector in the I-Q plane between the actual received or transmitted symbol and the ideal reference symbol. EVM is the ratio of the average of the error vector power over the average ideal reference symbol vector power. It is frequently expressed in either dB or percentage. Figure 1 is a test setup example showing the modulation accuracy attainable with the LTC5599 low power direct quadrature modulator. Figure 2 shows the results. In this test, precision lab equipment generates a 30k symbol/second 16-QAM baseband (120kbps), and 450MHz LO input signal to the modulator. A vector signal analyzer (VSA) examines the modulator output. In Figure 2, the EVM vs time results show EVM uniformly low across all symbols, while the error summary shows EVM approximately 0.24% RMS, and 0.6% peak. This is indeed excellent performance, shown by a modulation error ratio (MER) of 49.6dB. The LTC5599 has internal trim registers that facilitate fine adjustments of I and ERROR SUMMARY TEST CIRCUIT SHOWN IN FIGURE 1 DUT: LTC5599 REGISTERS[0…8] = [0X31, 84, 80, 80, 80, 10, 50, 06, 00] FREQUENCY SET TO 450MHz; OTHERWISE ALL DEFAULTS VSA: AGILENT 89441A, ANALYSIS SPAN = 100kHz MEASURING FILTER: ROOT COSINE ALPHA 0.35 REFERENCE FILTER: RAISED COSINE What’s an IQ Modulator? An IQ modulator is a device that converts baseband information into RF signals. Internally, two doublebalanced mixers are combined as shown below. By modulating with both in-phase (I) and quadrature (Q) inputs, any arbitrary output amplitude and phase can be selected. By targeting specific points in amplitude and phase, high order modulation is created. Shown below is 16QAM. There are four possible I values, which decodes into two bits. Likewise for the Q axis. So each symbol can convey four bits of information. I-CHANNEL MODULATOR INPUT LO INPUT Q+ 90° COMBINER Q-CHANNEL MODULATOR INPUT I− I+ Q− Fundamental architecture of an IQ modulator November 2015 : LT Journal of Analog Innovation | 21 Q DC offset, amplitude imbalance, and quadrature phase imbalance to further optimize modulation accuracy—results are even better if trim registers are adjusted. In many ways, this test demonstrates the best-case capabilities of the modulator without optimization: baseband bandwidth is large, DAC accuracy and resolution are superb and digital filtering is nearly ideal.1 While these test results are useful for measuring the true performance of the modulator, practical low power wireless implementations necessitate some compromises, as discussed below. DRIVING FROM PROGRAMMABLE LOGIC OR AN FPGA Figure 3. Transmit exciter block diagram. (Full schematic is in Figure 4.) LO 0 dBm Many FPGAs and programmable devices support digital filter block (DFB) functionality, an essential building block for digital communications. Raw transmit data is readily IQ mapped and digitally filtered. Figure 3 shows an example of how a device such as the Cypress PSoC 5LP can be utilized to drive IQ modulators such as the LTC5599. MCU WITH PROGRAMMABLE LOGIC DFB DFB DAC 8-BIT DAC 8-BIT I I LT6238 I SINGLE ENDED TO DIFF Q Q VOCM = 1.4V MODULATION: 16-QAM, PRBS 9, ROOT COSINE ALPHA 0.35, 30k SYM./SEC LTC5599 LC RECONSTRUCT FILTERS LOW POWER IQ MODULATOR RF OUT −4 dBm RMS Q 5th ORDER BESSEL FULL SCHEMATIC IN FIGURE 4 OCCUPIED BW = 40.5kHz Figure 4. Driving an IQ modulator with programmable logic and DACs. The passive Bessel filter attenuates DAC images and provides lowest RF output noise floor, while imposing negligible symbol error vector. VCC MCU WITH PROGRAMMABLE LOGIC UNMODULATED LO INPUT 0 dBm VCC VREF 1.74k 0.10µF 2.49k VCC 133Ω 470µH Vcc I CH. DAC 0~1.024V FULL-SCALE 1.47k +1.4V U2 LT6238 27nF 2.49k 2.49k 13nF 10nF 2.2nF 220µH 4.7nF 1nF 1nF RL(I) 267Ω VCC VCTRL EN +1.4V U2 133Ω 470µH 27nF LT6238 GAIN SCALING AND DC SHIFT BBMI SINGLE-ENDED TO DIFFERENTIAL DRIVER VREF 1.74k 2.49k 133Ω DAC LC RECONSTRUCTION FILTER 1.47k +1.4V U2 2.49k LT6238 +1.4V VOLTAGE OUTPUT DAC VREF 2.49k U2 LT6238 SPI MASTER 2.49k 0.10µF 27nF 10nF RF 2.2nF SDI SCLK CSB GNDRF 1.5nF 220µH 13nF 4.7nF 1nF 470µH 27nF 220µH 10nF 2.2nF MODULATED RF OUTPUT RL(Q) 267Ω 133Ω BESSEL RESPONSE, −3dB @ 50kHz −0.7dB @ 20kHz −44dB @ 220kHz INDUCTORS: COILCRAFT DS1608C SERIES, OR EQUIVALENT SPI BUS SETS MODULATOR CENTER FREQUENCY 22 | November 2015 : LT Journal of Analog Innovation U3 LTC5599 BBPQ 1.82k +1.4V SDO BBMQ 470µH Q CH. DAC 0~1.024V FULL-SCALE LOL TEMP 2.2nF BBPI U1 CY8C58LP LOC TTCK 220µH 10nF 39nH 15pF VCC = 3.3V 10µF GND SPI BUS 54mA design features Digital interpolation is used to increase the DAC clock frequency, and hence the DAC image frequencies. This lowers the filter order requirement of the LC reconstruct filter, which serves to attenuate DAC images to acceptable levels, while minimizing phase error and wideband noise. Figure 4 shows the complete circuit. The differential baseband drive to the modulator, as opposed to single-ended baseband drive, offers the highest RF output power and lowest EVM. The LTC6238 low noise amplifier, U2, converts the DAC singleended I and Q outputs to differential.2 Input amplifier U2 gain is designed to scale the DAC out voltage range to the modulator input voltage range, after the 2:1 attenuation effect of filter terminating resistors RL(I) and RL(Q) is taken into account. The input amplifier U2 is also designed to supply the required input common mode voltage for the IQ modulator—important for maintaining proper modulator DC operating point and linearity. Classical LC filter synthesis methods are used for the DAC reconstruction lowpass filter (LPF) design. Some of the filter shunt capacitance is implemented as common mode capacitors to ground. This also reduces common mode noise, which can find its way to the modulator output. If active filters are used here, the final filter stage before the modulator should be a passive LC roofing filter for lowest broadband RF noise floor. Table 2, Figure 5 and Figure 6 show the performance results. In this case, EVM is limited by the digital accuracy of the baseband waveforms, here determined by MEASURED IQ VECTOR EVM VS TIME MEASURED EYE DIAGRAM ERROR SUMMARY Figure 5. EVM measurement detail. Two IC devices replace the lab signal generator. It’s not perfect, but is usually ‘good enough.’ the number of U1 FIR filter taps (63), and by the DAC resolution (eight bits). For this reason, EVM does not substantially improve when IQ modulator impairments are adjusted out, as shown in Table 2. For lower EVM, use more FIR filter taps and higher resolution DACs. When comparing the results shown in Figures 2 and 5, we see the price paid for replacing a high grade lab signal generator with a circuit composed of programmable logic and op amp filters. EVM increased from 0.24% RMS to 0.8% RMS. The increased EVM is primarily due to the fact that the waveforms generated by the programmable logic IC are not as accurate as the lab instrument. Such is the case in a real world implementation, but Figure 5 shows a fairly decent eye diagram, and a summary measurement that shows the modulation accuracy is sufficient for most applications. In Figure 6 we see the output spectrum is quite clean. The amplitude of the DAC image spurs, relative to the desired signal, is estimated by sin(x)/x, where x = πf/fCLK , plus attenuation afforded by the DAC LC reconstruct filter. For lowest adjacent channel power, a long FIR filter (many taps) is essential, as is a low phase noise LO. Table 2. EVM performance. Even with a 63-tap FIR filter design and 8-bit dual-DACs, the 0.8% RMS EVM achievement is entirely adequate for most applications. TX FIR FILTER DESIGN INTERPOLATION FACTOR SYMBOL RATE (ksps) DATA RATE (kbps) 63-tap RRC, α = 0.35 8 30 120 EVM (% RMS) EVM (% PEAK) NOTES 0.8 2.0 LTC5599 Unadjusted (MER = 39.1dB) 0.8 1.8 LTC5599 Adjusted (MER = 39.8dB) November 2015 : LT Journal of Analog Innovation | 23 Linear Technology’s LTC5599 IQ modulator is a versatile RF building block, offering low power consumption, high performance, wide frequency range and unique optimization capabilities. It simplifies radio transmitter design without sacrificing performance or efficiency. Table 3. Output noise density levels off at approximately 17dB over kTB. 0 −10 −20 AMPLITUDE (dBm) Figure 6. Output spectrum. In this design, the closest image spurs are about 70dB down, reasonably good for most systems. Modulator RMS output power measures −4dBm. Harmonic filtering is still required. −30 RBW = 5kHz DAC IMAGE SPURS, −70dBc: −25dB FROM (sin x)/x + −45dB FROM LPF −70dB TOTAL −40 −50 −60 −70 FREQUENCY OFFSET (MHz) RF OUTPUT NOISE DENSITY (dBM/Hz) +6 −156.7 +10 −156.8 +20 −156.8 −80 −90 −100 449.5 450 FREQUENCY (MHz) Higher frequency span sweeps show no visible spurious products except for the harmonics of the carrier, which must be filtered as usual. Low output noise floor is also important in many cases, such as when a transmitter and receiver are duplexed or co-located, when high PA gain is used, or when multiple transmitters run simultaneously. Table 3 shows the measured output noise density for the system of Figure 3, while transmitting at a modulated carrier frequency of 460MHz. The low U2 op amp noise, combined with the 5th order roll-off of the LC reconstruct filter, keeps the baseband noise contribution as low as possible. 450.5 Total current consumption at 3.3V measures 96m A, as summarized in Table 4. The majority of the DC power is consumed by U1, the programmable logic device, for which each DFB is specified to typically consume 21.8m A at the 67MHz clock frequency of this application.3 In summary, the DFBs account for 81% of the digital power consumption. Clearly the key to reduced current consumption for the digital section is optimization of the DFB architecture, which is beyond the scope of this article.4 Linear Technology’s LTC5599 IQ modulator is a versatile RF building block, offering low power consumption, high performance, wide frequency range and unique optimization capabilities. It simplifies radio transmitter design without sacrificing performance or efficiency. n Notes 1Test equipment FIR filters are synthesized in software, so hundreds or thousands of filter taps are feasible and preferred, since signal quality is most important, and delay is inconsequential. In contrast, a real-time wireless application typically requires trade-offs between filter delay and EVM/ACP. 2For lower symbol rate applications, the LTC1992 low power fully differential input/output amplifier/driver could also be used for this purpose, offering improved DC accuracy and lower DC power consumption in exchange for a higher transmit noise floor within the channel passband. 3In Table 4. Total power consumption STAGE DESCRIPTION ICC (mA) POWER (mW) U1 CY8C58LP Programmable System on Chip 54 178 U2 LT6238 Quad Op Amp 13 43 U3 LTC5599 Low Power IQ Modulator 29 96 Total: 96 317 24 | November 2015 : LT Journal of Analog Innovation CONCLUSION this example, the minimum DFB clock frequency = 30kHz symbol rate • 8x interpolation • 63 FIR filter taps • 2 cycles for multiply and accumulate (MAC) • 2 cycles for arithmetic logic (ALU) = 60.5MHz. 4DFBs that are faster and more highly optimized are available from Altera and Xilinx. References 1“Digital Modulation in Communications Systems – An Introduction,” Application Note 1298, Keysight Technologies design features Low Profile Supercapacitor Power Backup with Input Current Limiting David Salerno Supercapacitors are increasingly used as backup power sources, due in large part to their continually improving volumetric energy capacity and robust nature. Large output capacitors can strain the load capabilities of an input source, especially when that source is limited by protocol (USB or PCMCIA) or a high source resistance. Input source limitations can complicate designs. The LTC3128 simplifies power backup by adding a programmable accurate input current limit to a complete supercapacitor charger. Figure 1 shows that only a few components are needed to produce a supercapacitor charger with a 3.0A input current limit. The LTC3128 is a buck-boost DC/DC supercapacitor charger with programmable accurate input current limit (up to 3A) and active balancing, offered in 4mm × 5mm × 0.75mm QFN or 24-lead TSSOP packages. The 1.2MHz switching frequency, along with low resistance, low gate charge integrated switches provide an efficient, compact and low profile solution for charging large output capacitors. The high accuracy (±2%) of the programmable input current limit allows designers to limit the maximum current draw to just below the capability of the input source. Capacitor voltage monitoring and protection, combined with the integrated active charge balancer, prevents mismatched capacitors from being overvoltaged and keeps capacitors with mismatched leakages in balance. This makes the LTC3128 ideal for backup or pulsed load applications. Supercapacitors, because of their long lifetime, large cycle capability (up to 10 years and 500,000 cycles) and relatively straightforward charging profiles, are ideal for backup solutions. SUPERCAPACITOR CHARGE TIME AND HOLDUP TIME When designing a backup system, two of the most important criteria are charge time and holdup time. The charge time determines the minimum amount of time the system needs to be in operation before it can withstand a power failure, and holdup time determines how long a system can maintain operation from its backup source. Charge time is determined by a combination of programmed input current limit, programmed output voltage, converter efficiency and output capacitance. Figure 2 shows the charge time for a 1F output capacitance at a programmed input current of 3.0A. This curve takes into account VIN , VOUT and the converter efficiency. If the output capacitance is larger or smaller than 1F, the charge time scales proportionally to the output capacitance. At the end of charging, the LTC3128 dials back the input current to top off the Figure 2. LTC3128 charge time 5.5 5.0 3.3µH SW1 VIN 2.4V TO 5.5V 3.0A 10µF 10µF 4.5 SW2 VOUT = 4.2V RSENP VOUTP RSENS VOUTS LTC3128 VIN RUN MID PFI FB PFO PGOOD MAXV PROG GND 10µF 4.0 TO LOAD 1.87M 100F VOUT (V) Figure 1. Complete supercapacitor charging circuit with input current limit 3.5 3.0 2.5 2.0 1.5 1.0 3.57k 470pF 301k COUT = 1F IIN = 3.0A 0.5 0 VIN = 2.4V VIN = 3.3V VIN = 4.2V VIN = 5.0V 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 TIME (s) November 2015 : LT Journal of Analog Innovation | 25 The active charge balancer uses the inductor of the LTC3128 to efficiently move charge from one capacitor to another to balance them, maintaining the same voltage across each capacitor. Active charge balancing eliminates the high quiescent current and continuous power dissipation of passive balancing. charge on the output capacitor stack. This is done to prevent chattering in and out of regulation due to the ESR of the output capacitors. Figure 3 shows an example of the charge current being dialed back as the capacitor approaches full charge. The current is typically dialed back at 95% of programmed output voltage, and this is the voltage that should be used for the charge time calculation. The circuit of Figure 1 charges 100F to 4.2V with a programmed input current of 3.0A and a VIN of 3.3V. Figure 2 shows that it takes 1.3 seconds to charge a 1.0F capacitor to 4.0V (4.0V ≈ 0.95 • 4.2V) from 0V. Since the capacitor in this example is one hundred times larger, it will take roughly 130 seconds to charge a 100F capacitor to 4.0V from 0V. To determine how long backup power can be provided to the system, the operational voltage range on the output needs to be determined. For this application, the operational output voltage is 4.2V down to 1.0V. The stored energy on the 100F capacitor can be determined as: Figure 4. LTC3122 boost converter powered by the LTC3128 1 2 2 1 W = COUT ( VINITIAL ) − COUT ( VFINAL ) 2 2 1 2 1 2 = 100 • 4.2 − 100 • 1.0 = 832J 2 2 where W is the work done in joules, COUT is the total output capacitance, VINITIAL is the beginning voltage on COUT, and VFINAL is the minimum voltage COUT can run down to. VIN 2.4V TO 5.5V 3.0A 10µF 10µF tBACKUP = VMID 5V/DIV 5.00s/DIV Figure 3. Input current charging profile capacitor, and PLOAD is the power draw from the secondary converter. BALANCING SUPERCAPACITORS Achieving higher output voltages with supercapacitors requires putting two or more cells in series with each other because the maximum voltage for each capacitor is typically specified between 2.3V and 2.7V, depending on the manufacturer and type of capacitor. The life of the capacitor is dependent on the voltage across the capacitor. To extend capacitor lifetime the voltage on the capacitor should be regulated below the rated WSTORED 832J = = 554.66s PLOAD 1.5W where tBACKUP is the holdup time of the system, WSTORED is the available stored energy on the output 3.3µH SW2 SW VOUT = 4.2V RSENP VOUTP RSENS VOUTS LTC3128 VIN RUN MID PFI FB PFO PGOOD MAXV PROG GND 10µF 1.87M 4.7µF 100F VIN OFF ON BURST PWM SD 470pF 301k LTC3122 PWM/SYNC 100nF CAP FB VCC VC SGND 4.7µF VOUT 12V 800mA VOUT RT 57.6k 3.57k 26 | November 2015 : LT Journal of Analog Innovation VOUT 2V/DIV If a secondary boost converter is connected to VOUT, it acts as a constant power draw from the supercapacitor. Figure 4 shows an example of a secondary boost converter being powered by VOUT of the LTC3128. The LTC3122 data sheet shows that for a 12V output with a 100m A load, the average converter efficiency across a 1V to 4.2V input is approximately 80%, resulting in a 1.5W constant power load on the holdup capacitor. The holdup time can be determined by: 3.3µH SW1 IIN 2A/DIV PGND 1.02M 22µF 113k 210k 390pF 10pF design features supercapacitors. The LTC3128 allows the secondary converter to pull its current through a current sense resistor internal to the LTC3128. This allows the secondary converter to draw the required current from the supply, up to 4A, and the LTC3128 will charge the output capacitors with the programmed input current, less the current drawn from the secondary converter. As long as the secondary converter never draws more than the programmed input current, the LTC3128 limits the total current draw from the input supply to the programmed value, while charging the backup capacitors with the remaining available current. 3.3µH SW1 VOUT = 4.2V MAXV = 2.7V SW2 RSENP VOUTP RSENS VOUTS LTC3128 VIN MID RUN PFI PFO FB PGOOD PROG MAXV GND VIN 2.4V TO 5.5V 3.0A 10µF 10µF 135k 10µF 1.87M TO LOAD 200F 10µF 470pF 3.57k maximum voltage. Capacitor vendors typically specify how to derate the voltage on their supercapacitors to extend life. The LTC3128 integrates a programmable maximum capacitor voltage comparator and an efficient active charge balancer. The maximum capacitor voltage comparators look at the voltage across each individual capacitor and ensure that the programmed voltage is not exceeded while charging. If the maximum programmed capacitor voltage is reached on either capacitor, the LTC3128 halts charging to balance the cells and then resumes charging. The active charge balancer uses the inductor of the LTC3128 to efficiently move charge from one capacitor to another to balance them, so that the capacitors maintain the same voltage across them. This is important because during a holdup event, if the capacitors are far enough out of balance, the polarity of one of the cells could become 200F Figure 5. LTC3128 with charge balancer and maximum capacitor voltage protection 301k reversed, damaging the capacitor. The LTC3128 will only balance the cells if one of the cells has violated its programmed maximum capacitor voltage, or if the output voltage is in regulation and the capacitors are out of balance but the maximum voltage has not been violated. To extend backup time, the LTC3128 draws less than 1µ A from VOUT when in shutdown, or less than 2µ A when in input UVLO. Figure 6 shows a power ride-through application using the LTC3128 and LTC3122. Active charge balancing eliminates the high quiescent current and continuous power dissipation of passive balancing. Figure 5 shows the LTC3128 configured with 100F of total output capacitance, a programmed output voltage of 4.2V, and a maximum capacitor voltage of 2.7V, each. CONCLUSION The LTC3128 3A buck-boost DC/DC supercapacitor charger is a streamlined solution for efficiently charging and protecting supercapacitors in high reliability, long-life applications. It features a ±2% accurate programmable input current limit, programmable maximum capacitor voltage comparators and active charge balancing. n POWER RIDE-THROUGH APPLICATION In a backup system, the ability to wait for the storage capacitors to charge before you begin operating is not always an option. A power ride-through application provides a means to power the secondary converter directly from the input supply while simultaneously charging the 3.3µH Figure 6. Power ride-through application using the LTC3128 and the LTC3122 boost converter 3.3µH VOUT = 4.2V MAXV = 2.7V SW2 RSENP VOUTP RSENS VOUTS LTC3128 VIN MID RUN PFI PFO FB PGOOD PROG MAXV GND SW1 VIN 2.4V TO 5.5V 3.0A 10µF 10µF 135k 10µF 1.87M 200F OFF ON BURST PWM 10µF 4.7µF 470pF 3.57k SW VIN 200F 301k SD LTC3122 PWM/SYNC 100nF CAP RT FB VCC VC SGND 57.6k 4.7µF VOUT 12V 800mA VOUT PGND 1.02M 22µF 113k 210k 390pF 10pF November 2015 : LT Journal of Analog Innovation | 27 What’s New with LTspice IV? Gabino Alonso —Follow @LTspice at www.twitter.com/LTspice —Like us at facebook.com/LTspice BLOG BY ENGINEERS, FOR ENGINEERS Check out the LTspice® blog (www.linear.com/solutions/LTspice) for tech news, insider tips and interesting points of view. New Article: “Achieving Low On-Resistance with Guaranteed SOA in High Current Hot Swap Applications” by Dan Eddleman www.linear.com/solutions/5722 The requirement for live insertion and removal in high current backplane applications demands MOSFETs that exhibit both low on-resistance during steady state operation and high safe operating area (SOA) for transient conditions. Often, modern MOSFETs optimized for low on-resistance are unsuitable for high SOA hot swap applications. This article overviews an application that provides the best of both worlds by utilizing the LTC4234 to satisfy SOA requirements and an external low on-resistance MOSFET reduces the overall power dissipation. SELECTED DEMO CIRCUITS For a complete list of simulations utilizing Linear Technology’s devices, please visit www.linear.com/democircuits. Buck Regulators • LT3697: 5V step-down converter with cable drop compensation & output current limit (8V–35V to 5V at 6A) www.linear.com/solutions/5476 • LT8613: 5V step-down converter with 6A output current limit (5.8V–42V to 5V at 6A) www.linear.com/solutions/5751 • LT8616: 5V, 3.3V, 2MHz step-down converter (5.8V–42V to 5V at 1.5A & 3.3V at 2.5A) www.linear.com/solutions/5753 • LT8640: 2MHz µPower ultralow EMI buck converter (5.7V–42V to 5V at 5A) www.linear.com/solutions/5635 What is LTspice IV? LTspice IV is a high performance SPICE simulator, schematic capture and waveform viewer designed to speed the process of power supply design. LTspice IV adds enhancements and models to SPICE, significantly reducing simulation time compared to typical SPICE simulators, allowing one to view waveforms for most switching regulators in minutes compared to hours for other SPICE simulators. LTspice IV is available free from Linear Technology at www.linear.com/LTspice. Included in the download is a complete working version of LTspice IV, macro models for Linear Technology’s power products, over 200 op amp models, as well as models for resistors, transistors and MOSFETs. 28 | November 2015 : LT Journal of Analog Innovation Buck-Boost Controllers • LT8709: Negative buck-boost regulator with output current monitor and power good (−4.5V to −38V input to −12V output at 5A) www.linear.com/solutions/5719 SEPIC Converter • LT8494: 450k Hz , 5V output SEPIC converter (3V–60V to 5V at 1A) www.linear.com/solutions/5848 Isolated Regulator • LTM®8057: 2kV AC isolated low noise µModule regulator (3.1V–29V to 5V at 300m A) www.linear.com/solutions/5206 LED Driver • LT3952: Short-circuit robust boost LED driver (7V–42V to 50V LED string at 333m A) www.linear.com/solutions/5749 SELECT MODELS To search the LTspice library for a particular device model, choose Component from the Edit menu or press F2. Since LTspice is often updated with new features and models, it is good practice to update to the current version by choosing Sync Release from the Tools menu. The changelog.txt file (see root installation directory) list provides a revision history of changes made to the program. Linear Regulators • LT3042: 20V, 200m A, ultralow noise, ultrahigh PSRR RF linear regulator www.linear.com/LT3042 • LT3088: 800m A single resistor rugged linear regulator www.linear.com/product/LT3088 Buck Regulators • LT8602: 42V quad monolithic synchronous step-down regulator www.linear.com/LT8602 • LTC3887: Dual output PolyPhase® step-down DC/DC controller with digital power system management www.linear.com/LTC3887 • LTC7138: High efficiency, 140V 400m A step-down regulator www.linear.comt/LTC7138 • LTM4622: Dual ultrathin 2.5A step- down DC/DC µModule regulator www.linear.com/LTM4622 design ideas Be sure to check for recently added demonstration circuit simulations, such as this wide input voltage range boost/SEPIC/inverting controller: 2.5V to 36V input, 12V/2A output SEPIC converter (automotive 12V regulator). Multitopology Controller • LT8570: Boost/SEPIC/inverting DC/DC converter with 65V switch, soft-start and synchronization www.linear.com/LT8570 Surge Stopper • LTC7860: High efficiency switching surge stopper www.linear.com/LTC7860 • LTM4675: Dual 9A or single 18A µModule regulator with digital power system management www.linear.com/LTM4675 • LTM4676A: Dual 13A or single 26A µModule regulator with digital power system management www.linear.com/LTM4676A Wake-Up Timer • LTC2956: Wake-up timer with pushbutton control www.linear.com/LTC2956 n Power User Tip TIGHTEN UP YOUR SCHEMATICS: COMBINE MULTIPLE MODEL INSTANCES INTO ONE SYMBOL When you need multiple instances of a model, it is easy to copy and paste a symbol, but sometimes you can tighten up your schematics by using a single symbol to define multiple instances of same device. For instance, instead of placing four identical capacitor symbols in parallel, use one symbol times four, “x4”. This feat can be accomplished using the M (parallel units) or N (series units) parameters. A number of intrinsic devices support the M (parallel units) parameter, such as the capacitor, inductor, diode and MOSFET models. If you are not sure if the model supports the M (parallel units) parameter, try it, and if you do not get an error message, you should be good. The diode (including LED) model is the only intrinsic model that supports N (series units) parameter. To define multiple instances of a model in a device symbol: 1.Ctrl + right-click the symbol to edit the component attributes. 2.Insert “m=<number>” or “n=<number>” into the Value2 field. Note that non-integer <number> values are allowed. 3.Make the multiple instances visible in your schematic by selecting the Value2 attribute and clicking the Vis column. Parallel Capacitors To match certain electrical schematic standards you can define parallel capacitors either using “m=<number>” or “x<number>” syntax as in “x4”. Series (String) of LEDs Diodes are the only intrinsic models that support the N (series units) parameter. Happy simulations! November 2015 : LT Journal of Analog Innovation | 29 Boost 12V to 140V with a Single Converter IC Victor Khasiev Generating a high voltage from a much lower voltage presents a number of challenges for the classical single stage boost topology. For instance, the maximum duty cycle limitation of a boost controller may not allow the required step-up ratio. Even if it does, there is often a sharp decrease in efficiency at high duty cycles. The duty cycle can be shortened by choosing discontinuous mode of operation, but this leads to high peak input current, higher losses and EMI challenges. An alternative to a single boost converter is a 2-stage boost converter, where the first stage produces an intermediate voltage and the second stage boosts to the final high voltage. A 2-stage converter can be produced with a single controller IC, such as the LTC3788, a high performance 2-phase dual output synchronous boost controller, which drives all N‑channel power MOSFETs. The LTC3788 can be configured such that the first boost stage takes advantage of its synchronous rectification feature, which maximizes efficiency, reduces power losses Figure 1. Block diagram of LTC3788-based 2-stage boost converter CINT 40V ABS MAX TG1 SW1 PGOOD1 ILIM SS1 ITH1 FB1 SENSE1+ Q1 RS1 Q2 SENSE1– VIN GND BOOST1 FREQ BG1 PHASMD VBIAS CLKOUT PGND LTC3788 BG2 RUN2 BOOST2 SS2 ITH2 FB2 SENSE2 TG2 RUN1 SW2 INTVCC 5V PGOOD2 SGND + EXTVCC SENSE2– PLLIN/MODE GND L1 U1 30 | November 2015 : LT Journal of Analog Innovation VOUT COUT D1 L2 DR 7V TO 10V MOSFET DRIVER Q3 GND RS2 GND and eases thermal requirements. The maximum output voltage of this controller is 60V, when using synchronous rectification. If greater than 60V is required, the second stage can be designed to run non-synchronously, as described below. 2-STAGE BOOST PRODUCES 140V FROM 12V The block diagram in Figure 1 shows the LTC3788 in a 2-stage boost configuration. This block diagram also reveals a few caveats that must be observed in this design: •The output of the first stage (Q1, CINT) is connected to the input of second stage (RS2, L2). The output of the first stage should not exceed 40V, because the maximum absolute rating of the SENSE pins is 40V. •The gate drive voltage of 5V is suitable for logic level MOSFETs, but not for high voltage standard MOSFETs, with typical gate voltages of 7V to 12V. The external gate driver DR, controlled by the BG2 signal can be used as shown here to drive high voltage standard MOSFETs. •To generate an output voltage above maximum limit of 60V, the synchronous rectification MOSFET is replaced by a single diode D1. Figure 2 shows the complete solution. Transistors Q1, Q2 and inductor L1 compose the first stage, which generates an intermediate bus voltage of 38V. The first stage employs synchronous rectification for maximum efficiency. The output of the first stage is connected as input to the second stage, comprised design ideas 12.1k V_INT 2.21k V_INT PLLIN/MODE EXTVCC SGND INTVCC RUN1 BG2 SW2 SS2 PGOOD2 ITH2 FB2 SENSE2+ SENSE2– GND 12.1k DRIVER BIAS CIRCUIT 5.1k Q4 MMBTA42LT1G 8.2V SNS2+ SNS2– 1μF 100pF 2.7k 6.98k + Q2 BSC028N06LS3G VIN 82μF/50V 50HVH82M GND V_INT VOUT 140V AT 1.0A 2x0.47μF/450V C4532X7T2W474M + 22μF/200V EEVEB2D220SQ VOUT GND D1 SBR10U200P5130 0.1μF 15nF VIN 3V TO 36V 4x4.7μF 4.7μF BOOST2 RUN2 309k SENSE1– SENSE1+ L1, 6.8μF SER2915H682 RS1, 0.002Ω 0.1μF PGND LTC3788EUH 82μF/50V 50HVH82M Q1 BSC067N06LS3G VBIAS VBIAS CLKOUT VBIAS BAS140W BG1 PHASMD + 4x4.7μF 0.1μF BOOST1 FREQ 42.2k TG1 SW1 ILIM SS1 ITH1 SENSE1– U1 TG2 SNS1– FB1 SENSE1+ 0.1μF INTERMEDIATE BUS 15nF PGOOD1 8.66k SNS1+ 100pF 374k U2 VBIAS LTC4440 BST VCC GND TG IN TS L2, 100μH PCV210405L SENSE2– SENSE2+ RS2, 0.01Ω Q3 BSC320N20NS3 806k Figure 2. Full schematic of 2-stage 140V output, 1A boost converter of Q3, D1, L2. The output of the second stage produces 140V at 1A. Transistor Q3 is a standard level MOSFET, driven by the LTC4440. Here, an LDO, based on transistor Q4, biases the gate driver, but a switching 94 regulator can be employed instead (such as one built around the LTC3536) to further increase overall efficiency. This solution features an input voltage range from 3V to 36V, nominal 12V. To decrease components’ thermal stress, the VOUT = 140V EFFICIENCY (%) CONCLUSION VIN = 12V IOUT = 1A 92 LTC3788 is a high performance 2-phase dual output synchronous boost controller, suitable for high power, high voltage applications. Its dual outputs can be used in tandem to achieve extremely high step-up ratios to high voltages. n 90 VOUT 20V/DIV 88 86 84 output current should be reduced when the input voltages fall below 10V. Figure 3 shows measured efficiency, and Figure 4 shows the start-up waveforms. A 93% efficiency is shown with VIN = 24V and with the 140V output loaded from 0.4A to 1A. This converter can operate at full load with no airflow. VIN = 24V VIN = 12V VIN = 8V 0 0.2 0.6 0.4 LOAD CURRENT (A) 0.8 Figure 3. Efficiency of the 2-stage converter in Figure 2 1 5ms/DIV Figure 4. Start-up waveforms November 2015 : LT Journal of Analog Innovation | 31 highlights from circuits.linear.com 100V INPUT, 5V OUTPUT, LOW RIPPLE 1A STEP-DOWN CONVERTER The LT8631 is a current mode PWM step-down DC/DC converter with internal synchronous switches that provides current for output loads up to 1A. The wide input range of 3V to 100V makes the LT8631 suitable for regulating power from a wide variety of sources, including automotive and industrial systems and 36V to 72V telecom supplies. Low ripple Burst Mode® operation enables high efficiency operation down to very low output currents while keeping the output ripple below 10mVP-P. Resistor programmable 100kHz to 1MHz frequency range and synchronization capability enable optimization between efficiency and external component size. The soft-start feature controls the ramp rate of the output voltage, eliminating input current surge during start-up, while also providing output tracking. A power good flag signals when the output voltage is within ±7.5% of the regulated output. Undervoltage lockout can be programmed using the EN/UV pin. Shutdown mode reduces the total quiescent current to < 5μA. The LT8631 is available in a 20-lead TSSOP package with exposed pad for low thermal resistance and high voltage lead spacing. www.linear.com/solutions/5855 VIN 3.3V VOUT1 0.1µF 10µF 60.4k PGOOD1 PGOOD2 VIN VOUT1 RUN1 RUN2 LTM4622 VOUT2 INTVCC COMP1 SYNC/MODE COMP2 TRACK/SS1 FB1 TRACK/SS2 FREQ 60.4k 10µF 10µF VOUT1 1.5V, 2.5A 60.4k VOUT2 1.2V, 2.5A 40.2k 4622 F27 4.35V TO 5V INPUT SUPPLY (PROTECTED TO 40V) 5V BACKUP APPLICATION WITH OVERVOLTAGE PROTECTION AND NON-BACKED-UP-LOAD OPTION (CHARGE CURRENT SETTING: 2.5A, INPUT CURRENT LIMIT SETTING: 4A) The LTC4040 is a complete 3.5V to 5.5V supply rail battery backup system. It contains a high current step-up DC/DC regulator to back up the supply from a single-cell Li-Ion or LiFePO4 battery. When external power is available, the step-up regulator operates in reverse as a step-down battery charger. www.linear.com/solutions/5821 VPWR VIN 2.2µF BST LT8631 EN/UV 0.1µF SW 22µH PG IND INTVCC 2.2µF VOUT 47pF RT 25.5k 1M VOUT 5V, 1A FB SYNC/MODE 191k TR/SS GND 47µF ×4 1210, 16V 0.1µF FSW = 400kHz L: TDK SLF1O145T-22OM1R9 ULTRATHIN µModule 3.3V INPUT, 1.5V AND 1.2V OUTPUTS AT 2.5A DESIGN WITH OUTPUT COINCIDENT TRACKING The LTM4622 is a complete dual 2.5A step-down switching mode µModule® (micromodule) regulator in a tiny ultrathin 6.25mm × 6.25mm × 1.82mm LGA package. Included in the package are the switching controller, power FETs, inductor and support components. Operating over an input voltage range of 3.6V to 20V, the LTM4622 supports an output voltage range of 0.6V to 5.5V, set by a single external resistor. Its high efficiency design delivers dual 2.5A continuous, 3A peak, output current. Only a few ceramic input and output capacitors are needed. www.linear.com/solutions/5841 FB2 GND VIN 6.5V TO 100V RS 6mΩ MN1 2.2µF 6.2k 1/4W OVP OPT 178k MN2 VIN CLN OVSNS PFI 60.4k FAULT PFO RST CHRG CLPROG TO NON-BACKED-UP LOAD VSYS 4.35V TO 5V SW BAT LTC4040 100µF 1690k VSYS BSTFB RSTFB IGATE TO BACKED-UP SYSTEM LOAD 324k 2.2µH 10µF VIN RBIAS NTC CHGOFF BSTOFF GND F0 F1 F2 PROG VSYS 800Ω NTC + Li-Ion BATTERY 4.1V L1: COILCRAFT XAL-5030-222 MN1: VISHAY/SILICONIX SiS488DN MN2: VISHAY/SILICONIX SIR424DP-T1-GE3 L, LT, LTC, LTM, Linear Technology, the Linear logo, Burst Mode, PolyPhase, LTspice and µModule are registered trademarks and Linduino is a trademark of Linear Technology Corporation. All other trademarks are the property of their respective owners. © 2015 Linear Technology Corporation/Printed in U.S.A./71.5K Linear Technology Corporation 1630 McCarthy Boulevard, Milpitas, CA 95035 (408) 432-1900 www.linear.com Cert no. SW-COC-001530