APPLICATION NOTE | AN:018 Providing a Constant Current for Powering LEDs Using the PRM® and VTM® By Joe Aguilar Product Line Applications Engineer Contents Page Introduction 1 Background: Adaptive Loop Regulation 2 Current Control Circuit 4 Overview 4 VI Chip Selection (PRM, VTM) 5 Current Sensing Sub-Circuit 6 Differential and Error Amplifier Selection (U2) 6 Shunt (Current Sense) 6 Resistor (R1) Selection Differential Amplifier 7 Gain (R2 Through R5) Voltage Reference (U1) 7 Voltage Limiting Sub-Circuit 9 Compensation 10 Components (R6 ,C2) Voltage Supply (VH) 12 Startup Sequencing of 13 the Current Regulation Circuit Current Regulation Accuracy 15 Layout Considerations 16 Conclusion 16 Appendix A Design Example 17 Introduction Light Emitting Diodes (LEDs) require a constant current for proper operation. The VI Chip® PRM® Regulator and VTM® Current Multiplier are designed to provide a regulated voltage using the Adaptive Loop Method of regulation (for further information please see www.vicorpower.com/documents/whitepapers/fpa101.pdf. In order to use the PRM and VTM to power an LED, it is therefore necessary to modify the operation of the PRM to provide a regulated current. This application note provides guidelines for implementing a constant current source using the PRM and VTM. Using the PRM and VTM to provide a constant current provides several advantages over conventional approaches. The implementation of a VTM in a system provides point of load current multiplication. The output current of a VTM is proportional to its input current by the following equation: I OUT = I IN K Equation 1 Thus in a controlled current application, the input current to the VTM can be sensed and regulated to control the output current. Sensing a lower current requires a smaller sensor which dissipates lower power and improves overall efficiency. Also the VI Chips themselves provide high efficiency and high power density, making the overall LED system small and cool and maximizing the output in Lumens per watt of dissipation. The overall system architecture is illustrated in Figure 1. A complete design example is covered in Appendix A, using the techniques outlined in this application note. Some LEDs require a pulsed current in order to operate properly. Pulsed-current operation will be covered in a future application note, due to bandwidth limitations of the circuit configuration proposed here. Also most known LED types can be driven with a single PRM + VTM pair. Parallel operation of PRMs and VTMs to provide a regulated current is not addressed in this application note. Regulated Current IIN PRM IN+ VIN VTM OUT+ IN+ + – IOUT = Current Sense IN- Regulated Current IOUT OUT- IN- LED OUT+ IIN K OUT- SC Feedback Current Amplifier Figure 1 Regulated Current Source Basic Architecture Control Signal + Compensator AN:018 vicorpower.com Applications Engineering: 800 927.9474 Page 1 Background: Adaptive Loop Regulation This application note requires a basic understanding of VI Chips, and Factorized Power Architecture (FPA), including Adaptive Loop regulation. Please refer to the following link (www.vicorpower.com/documents/whitepapers/fpa101.pdf) for more information. Before starting, the user should have a defined set of system design requirements. These requirements should include: output current set point, output voltage range, and current regulation accuracy. In most cases the specific data sheet of the LED or LED array will define many of the requirements for properly designing this circuit. It is important that the V-I characteristics of the end device (LED) are well understood to ensure that the circuit can provide the desired current within the voltage limitations of the PRM® and VTM®. The PRM is pre-configured with an internal voltage loop that regulates the output voltage of the PRM to a set value. The internal workings of the PRM should be well understood, as the external constant current circuit has been designed to work in conjunction with the internal voltage control loop, changing the PRM voltage reference in order to regulate the VTM output current. A simplified block diagram of the PRM internal voltage control loop is shown in Figure 2. Figure 2 Functional diagram of PRM Internal Error Amplifier PRM_VOUT C R R68 OS PRM_Controller – 0.961 x VSC + PRM Error Amplifier 1.24 V SC 10 k + _ 0.22 µF SG An internal reference is generated and connected to the SC port of the PRM through a 10 k resistor and a 0.22 µF capacitor, which provides a soft-start. The SC voltage can be adjusted by adding an external resistor, or by applying an external voltage. The applied voltage at the SC port should not exceed 6 Vdc. AN:018 vicorpower.com Applications Engineering: 800 927.9474 Page 2 The SC voltage is buffered and fed to the error amplifier through a resistive divider represented by the gain block of 0.961. R68 forms the top half of the voltage-sensing resistive divider. This resistor is fixed for each PRM®. Please refer to table 1 for R68 values for each PRM. The bottom half of the divider is formed by adding a resistor from the OS pin to SG (ROS). Equation 2 defines the PRM output as a function of VSC and ROS. From Equation 2, it is seen that for a given ROS resistor, adjusting the SC voltage will determine the PRM output voltage. This is the method by which the external current control circuit will control the output. Equation 2 PRM _ VOUT = 0.961 × VSC × ( R68 + ROS ) ROS Where: VSC is the voltage at the SC pin of the PRM. ROS is the resistance from OS to SG of the PRM. R68 is the PRM internal resistor specified in Table 1. Table 1 PRM Internal R68 Values PRM P048-048-24AL VIN 36 – 75 V P048-048-12AL P048-048-12AL POUT 240 W 120 W 38 – 55 V P045-048-17AL 320 W 93.1 k 170 W P024-048-12AL 18 – 36 V 120 W P036-048-12AL 18 – 60 V 120 W MP028F036M12AL 16 – 50 V 120 W AN:018 R68 vicorpower.com 69.8 k Applications Engineering: 800 927.9474 Page 3 Current Control Circuit Overview The recommended current control circuit is shown below in Figure 3. Figure 3 Constant Current Circuit NC VH PC SC TM SG SG C4 PRM IL R13 OS PR VC +IN +OUT VH R7 +OUT VEAO PC TM VTM R9 VC L1 +IN PRM_VOUT -OUT +IN CIN -IN +OUT +OUT R8 SG -OUT -OUT -IN R1 -OUT -IN Shunt VH R3 R10 VREF R2 R11 C3 SG Voltage Reference C1 SG 3+ C2 VSENSE U _A 2_ 2 U1 R12 VH SG R6 6 VREF R4 SG R5 Differential Amplifier 5 _ U2_B 7 + VEAO Error Amplifier As the VTM® is a current multiplier, the output current of the VTM can be regulated by its input current. The advantage of this approach is that the current can be sensed prior to the VTM current multiplication stage (at the higher voltage), reducing the I2R power dissipation in the external shunt. In addition, the control circuitry remains on the primary (PRM®) side, eliminating the need for isolating the feedback signal. The circuit consists of a voltage reference, shunt resistor, differential amplifier, and error amplifier. Low-side sensing is implemented at the output of the PRM using an op amp configured as a differential amplifier. The voltage across the shunt resistor (R1) is sensed and amplified with a gain determined by resistors R2 through R5. The reference voltage is generated using a precision adjustable shunt reference, and is tied to the non-inverting terminal of the error amplifier. This is the voltage to which the error amplifier will compare the differential amplifier output (VSENSE). The output of the error amplifier (VEAO) is tied to SC through resistors R7 and R8, allowing for the adjustment of the PRM output set point. The error amplifier will adjust the PRM output voltage until VSENSE is equal to the reference voltage VREF. The recommended circuit components are shown in Table 2. AN:018 vicorpower.com Applications Engineering: 800 927.9474 Page 4 Table 2 Recommended Values Ref Des Value / Part Number Manufacturer R1 CSM25120R010BXX Vishay R2 R4 R3 R5 R6 R7 R8 R9* R10 R11 R12 R13 C1 C2 C3 C4 U1 TLV431B TI U2 AD8667 Analog Devices Description Current Sense Resistor, 10 mΩ,1 W, 0.1%, 2512 1k Resistor, 0.1% 1k 100 k Resistor, 0.1%, 100 k 16.2 k 2.15 k 1.24 k 4.99 k* User Defined User Defined User Defined 10 k 0.01 µF 0.1 µF User Defined 0.01 µF Resistor, 1%, 16.2 k Resistor, 1%, 2.15 k Resistor, 1%, 1.24 k Resistor, 1%, 4.99 k* Dependent on Reference Voltage Dependent on Reference Voltage Dependent on Reference Voltage Resistor, 1%, 10 k Capacitor, Ceramic, 0.01 µF Capacitor, 0.1 µF Dependant on Startup Capacitor,Ceramic, 0.01 µF 3 terminal, Adjustable, Precision Shunt Regulator Low Noise, Precision, 16 V Dual Op amp Link CSM2512 TLV431B *If using MP028F036M12AL use 4.12 k for R9 The following are general guidelines to select the appropriate components for a straightforward, cost-effective solution with minimal component count. As there are many ways in which the circuit may be implemented, the recommended configuration may not be ideal for every application. This application note should contain enough detail for the end user to modify the circuit to fit their end application. Some aspects of the circuit, such as startup timing, are difficult to predict and therefore must be tested and tuned to the individual application. It is up to the user to perform the necessary system testing and troubleshooting to successfully qualify the implementation of this circuit in their end application. VI Chip Selection (PRM®, VTM®) Select the PRM based on input voltage range and power level. Different load voltage requirements are addressed by appropriate VTM selection. To select a VTM, the following parameters must be known: • Maximum output current. • Minimum and maximum operating output voltage. Refer to the web (www.vicorpower.com/dc-dc-converters-board-mount/vtm) to determine if there is a VTM which will provide the desired current over the specified voltage range of the intended load device. Then, refer to the specific product data sheet for information on operation and performance. AN:018 vicorpower.com Applications Engineering: 800 927.9474 Page 5 Current Sensing Sub-circuit Figure 4 Current Sense Components R1 -IN Shunt VH R3 C1 SG 3+ R2 VSENSE SG U _A 2_2 1 SG R4 R5 Differential Amplifier Although there are other techniques, the recommended method of current sensing is low-side sensing using a differential amplifier. Differential and Error Amplifier Selection (U2) The use of a dual op amp for the differential amplifier and error amplifier is recommended in order to minimize component count. Since the overall bandwidth of the system will be limited, the amplifier selection should optimize the current sensing accuracy. Critical parameters, which contribute directly to accuracy, are the input offset voltage and input offset current. These parameters should be kept as low as possible to minimize current sensing error. Amplifier current draw will also need to be considered when selecting a voltage supply. The recommended amplifier is Analog Devices AD8667. The key parameters are summarized in Table 3. Refer to the manufacturer’s data sheet for further information. Table 3 AD8667 Parameters Parameter Offset Voltage Max Offset Current Max Bias Current Max Supply Current per Amplifier Max Symbol VOS IOS IB Value 450 65 105 Units µV pA pA Conditions -40<TAMB<125 -40<TAMB<125 -40<TAMB<125 ISY 325 µA -40<TAMB<125 Shunt (Current Sense) Resistor (R1) Selection As with the amplifier, the shunt resistor has a significant impact on the current sensing accuracy. If the expected resistance of the shunt varies by 5%, so too will the expected shunt voltage, resulting in an equivalent current sense error. It is, therefore, critical to select a shunt resistor within the desired tolerance of the current source accuracy. The magnitude of the shunt voltage should be large relative to the amplifier’s input offset voltage to avoid further inaccuracy. In addition, the presence of the shunt contributes to additional power loss. Its value should be kept low to minimize power dissipation. The recommended shunt is the Vishay CSM25120R010B. This is a 10 mΩ, 0.1% tolerance, 1 W, 2512 metal foil, four-terminal resistor with Kelvin test points for voltage sensing. At 5 A, this part will dissipate approximately 250 mW of power. AN:018 vicorpower.com Applications Engineering: 800 927.9474 Page 6 Differential Amplifier Gain (R2 through R5) For a given shunt value, the gain of the differential amplifier will determine the necessary reference voltage to achieve a desired output current. Assuming R2 is equal to R4, and R3 is equal to R5, the output of the differential amplifier is defined by Equation 3. VSENSE = VSHUNT ⋅ ( Equation 3 R3 ) R2 Where: VSENSE is the differential amplifier output. VSHUNT is the voltage across the shunt (R1 ). The recommended values equate to a gain of 100, resulting in a differential amplifier output of 1 V per Amp of PRM® current when using a 10 mΩ shunt. Voltage Reference (U1) The VTM® has the following input/output characteristics, illustrated in Figure 5: 1) PIN = VIN ⋅ IIN 2) POUT = VOUT ⋅ IOUT 3) VOUT = VIN ⋅ K − IOUT ⋅ ROUT 4) POUT = PIN ⋅ η Figure 5 VTM Operation IIN + VIN – IOUT VTM K η ROUT + VOUT – Based on the above relationships, Equation 4 can be solved for the necessary VTM input current when given output current, output voltage, VTM efficiency, and VTM output resistance. This is important since the PRM current control circuit will control the input current to the VTM. Equation 4 VTM _ IIN = PRM _ IOUT = VOUT ⋅ IOUT ⋅ K η ⋅ (VOUT + IOUT ⋅ ROUT ) Where: IOUT is the desired output current of the VTM. VOUT is the nominal output voltage of the VTM. h is the nominal efficiency of the VTM at the given output. ROUT is the nominal output resistance of the VTM. K is the transformation ratio of the VTM. AN:018 vicorpower.com Applications Engineering: 800 927.9474 Page 7 Based on the current sensing component selection, the required reference voltage can be determined by Equation 5. VREF = PRM _ IOUT ⋅ R1 ⋅ ( Equation 5 R3 ) R2 Where: VREF is the voltage reference. PRM_IOUT is the necessary PRM current from Equation 4. R1 is the shunt resistor. R3, and R2 are the differential amplifier gain resistors. For the values given in Table 2, R1 ⋅ ( R3 ) = 1 and therefore, VREF = PRM_IOUT R2 There are multiple options available for generating the reference voltage. One simple approach is to use an adjustable shunt regulator such as the TLV431B. Figure 6 Reference Components VH R10 VREF R11 C3 U1 R12 SG Voltage Reference When selecting R10 and C3, refer to the manufacturer’s recommendations to ensure stability. Bear in mind that these components will also affect the startup timing, as described in a later section. Resistors R11 and R12 are used to adjust the output. The tolerance of these resistors will have a direct effect on accuracy; high precision resistors should be used. The current draw of the device should be kept below 1 mA for the recommended configuration to stay within the 5 mA limit of VH. This approach assumes that the reference voltage will be adjusted to achieve the correct output current based on a given shunt and differential amplifier gain. An alternative approach would be to fix the reference and adjust the gain to obtain the desired output current. In this case Equation 5 can be rearranged and the gain of the differential amplifier can be calculated for a given reference voltage, PRM output current and shunt. Equation 6 ( R3 VREF )= R2 PRM _ IOUT ⋅ R1 AN:018 vicorpower.com Applications Engineering: 800 927.9474 Page 8 Voltage Limiting Sub-circuit Figure 7 Voltage Limiting Components R7 SC VEAO R8 SG SG R9 OS The resistors R7 and R8 are required to limit the maximum voltage that appears on SC when the error amplifier is at its maximum. The resistor R9 is selected to limit the maximum PRM® output voltage during this condition. The internal SC capacitor (0.22 µF) will create a pole with the equivalent resistance formed by the parallel combination of R7, R8, and the internal 10 kΩ resistor. Equation 7 Equation 8 REQ = 1 1 + 1 + 1 R 7 R8 10 kΩ FPOLE = 1 2 ⋅ π ⋅ REQ ⋅ (0.22 µF) This pole will limit the bandwidth of the error amplifier, as described in the next section. FPOLE can be increased by decreasing R7 and R8; however, this will also increase the amount of current necessary to drive SC. When using VH as the supply, FPOLE should be limited to 1 kHz maximum. In order to select the appropriate components, the following should be defined: 1) The maximum output voltage of the error amplifier when saturated: VEAO(MAX). 2) The maximum SC voltage when the error amplifier is saturated: VSC(MAX). 3) The maximum PRM output voltage: PRM_VOUT(MAX). 4) The SC pole frequency: FPOLE. The recommended value for VSC(MAX) is 3 V. The absolute maximum voltage rating for the PRM SC port is 6 V, and this value should be avoided with margin to prevent damage to internal components. R7 and R8 will be selected based on VSC(MAX), VEAO(MAX) and FPOLE using Equations 9 and 10: Equation 9 R7 = 10 kΩ ⋅VEAO (MAX ) 10 kΩ ⋅VSC (MAX ) ⋅ 2π ⋅ FPOLE ⋅ 0.22 μF −1.24 V AN:018 vicorpower.com Applications Engineering: 800 927.9474 Page 9 Equation 10 R8 = 10 kΩ ⋅ R 7 ⋅ VSC (MAX ) 10 kΩ⋅ VEAO (MAX ) + 1.24 V⋅ R 7 − VSC (MAX ) ⋅ (10 kΩ + R 7) Where: VEAO(MAX) is the maximum error amplifier output voltage. VSC(MAX) is the maximum SC voltage. FPOLE is the SC pole frequency (Equation 7). Once the maximum SC voltage has been defined, R9 can be selected to limit the maximum PRM® output voltage as defined in Equation 11: Equation 11 R9 = ( R 68 ⋅ VSC (MAX ) ⋅ 0.961 ) PRM _VOUT (MAX ) − VSC (MAX ) ⋅ 0.961 The recommended components are designed to provide a maximum PRM output voltage that will not exceed its maximum rating. The parameters used for selecting these components are shown in Table 4. Table 4 Parameters Parameter VEAO(MAX) VSC(MAX) FPOLE PRM_VOUT(MAX) Value 8.6 V 3V 1 kHz *56 V *51 V maximum for MP028F036M12AL Compensation Components (R6, C2) Figure 8 Error Amplifier Components C2 VSENSE R6 6 VREF 5 – U2_B + 7 VEAO Error Amplifier AN:018 vicorpower.com Applications Engineering: 800 927.9474 Page 10 The compensation for this circuit consists of a single pole with the frequency response described by Equation 12 and shown in Figure 9. Starting at the origin (f = 0 Hz), the gain will decrease at a slope of -20 dB/decade when plotted vs. frequency on a log/linear scale. The crossover frequency (FCROSS) of the error amplifier is determined by R6 and C2 as described in Equation 13. Equation 12 Equation 13 G(dB) = 20 ⋅ log( FCROSS = Figure 9 Error Amplifier Frequency Response 1 ) 2π ⋅ R 6 ⋅ C 2 ⋅ f 1 2π ⋅ R6 ⋅ C 2 Error Amplifier Frequency Response 60 40 20 Gain (dB) FCROSS= 1 2π •R6•C2 0 -20 -40 -60 1 10 100 1000 10000 100000 Frequency (Hz) In order to insure stability, the error amplifier crossover frequency (FCROSS) should be limited to a factor of 10 below the SC pole frequency. Equation 14 FCROSS = FPOLE 10 Rearranging the terms in Equation 13 allows for solving for the product of R6 and C2 to achieve the desired crossover frequency. The recommended values will provide a crossover frequency of approximately 100 Hz. Equation 15 R 6 ⋅ C2 = 1 2π ⋅ FCROSS AN:018 vicorpower.com Applications Engineering: 800 927.9474 Page 11 Voltage Supply (VH) The recommended configuration is to power the op amp and reference using VH. VH is an auxiliary 9 V supply generated internally by the PRM®. It is limited to 5 mA of current, and 0.1 µF of capacitance. If the recommended configuration has been changed, the maximum expected current draw should be determined to ensure that the 5 mA limit is not exceeded. If necessary, one method for increasing the capability of VH is shown in Figure 10. Figure 10 Increasing VH source capability PRM_VOUT VH R14 Q1 VH R7 VEAO SC VS R8 The transistor Q1 is added as an emitter follower between the output and the supply rail (VS). R14 is sized to limit the maximum VH current draw. Since the majority of the power is now sourced through the PRM output, the limitation in supply current is determined by the transistor thermal limitations. The STMicro STN715 transistor allows for an 18 mA capability at an 85°C ambient temperature, and a 55 V PRM output voltage. External supplies can be used if available, provided that the supply is primary referenced. Additional considerations for startup sequencing will need to be taken into account as described in the next section. AN:018 vicorpower.com Applications Engineering: 800 927.9474 Page 12 Startup Sequencing of the Current Regulation Circuit A typical PRM® startup sequence is shown in Figure 11. From the application of input power, there is a delay prior to the PRM beginning to ramp its output voltage. At this time, VH and VC are generated. VC is a pulse of approximately 10 msec, which allows the VTM® to temporarily operate below its minimum input of 26 V. With the VC pulse applied, the VTM output will track its input from 0 V, resulting in a soft start. The SC voltage directly controls the rate of rise of the PRM output. The same sequence would occur if enabling through the PC pin, the only difference being the delay time. Figure 11 PRM Startup from Application of VIN The startup timing for the constant current circuit is controlled not only by the rise time of VREF, but also the magnitude of the reference voltage, and the error amplifier compensation components. The compensation components, R6 and C2, limit the maximum rate of rise of the error amplifier output, leading to two startup timing conditions. The first is illustrated in Figure 12, where the rate of rise of the reference is below the maximum rate of the error amplifier. In this case, the output of the error amplifier is able to track the reference; and the result is an output current rise that closely matches that of the reference voltage. Figure 12 Constant Current Startup Condition 1 The second condition is where the rate of rise of VREF exceeds the maximum rate of the error amplifier. In this case the error amplifier output will change its rate of rise in order force a current through C2 and R6 temporarily equalizing the voltages at pin 5 and pin 6. This is illustrated in Figure 13. As the output current increases during the startup sequence, the necessary slope decreases until output current feedback is able to satisfy the error amplifier. AN:018 vicorpower.com Applications Engineering: 800 927.9474 Page 13 Figure 13 Constant Current Startup Condition 2 The startup timing for this condition is dependant on the magnitude of the reference voltage, and the characteristics of the load. Higher reference voltages will have a faster rise time, while lower reference voltages will have a slower rise time. To ensure a proper start-up, the VTM® input voltage must reach 26 V within the 10 ms VC pulse duration. If the voltage is too low, the VTM will be unable to sustain its internal VCC when the VC voltage drops, and will subsiquently shut down. This puts a limitation on maximum rise time, and thus the minimum allowable reference voltage for a given C2 and R6. Low reference voltage set points may result in a condition where the output rise time is slower than the minimum 10msec to ensure a proper startup. In this case, the user should adjust the gain of the current sense amplifier to ensure that the reference voltage is high enough at the desired output current to ensure a proper start. Once running, the current can be trimmed down to a lower level without issue. An alternative solution would be to start at a higher output current and then trim down once the unit is up and running. When powering the circuit from VH, the amplifier supply voltage and reference will not be generated until VH is present. This is the instant at which the PRM® is able to respond to a control signal. If the supply and reference were present prior to this instant, the circuit would not have a controlled start. The error amplifier would rail in an unsuccessful attempt to bring the current up to the appropriate value. This is an undesirable situation. Once the PRM is enabled, the control signal would be at a maximum, forcing the PRM to its maximum voltage with no control over the rate of rise. The rise of the voltage reference must be synchronized to the VH signal to ensure that the error amplifier voltage remains low until the PRM is ready to respond to a control signal. The circuit shown in Figure 14 is one example of a modification of the circuit shown in Figure 6 which enables operation with an external supply. Figure 14 Reference Sequencing Circuit VS R 10 VH VREF R 11 Q2 Q3 C5 U1 R 12 C *VS is an external supply AN:018 vicorpower.com Applications Engineering: 800 927.9474 Page 14 Current Regulation Accuracy The fact that the current control is done at the VTM® input adds additional complexity as variations in the VTM parameters and load voltage will lead to errors in the current set point. The contributing factors to the overall accuracy are the current sensing accuracy, reference accuracy, and variation in the VTM efficiency, VTM ROUT, and VOUT. The current sensing accuracy is mostly determined by the magnitude of the input offset voltage of the AD8667 with respect to the shunt voltage. The offset error can be approximated by Equation 16. Equation 16 Offset _ %Error = ( VOFFSET ) ⋅ 100 VSHUNT Where: VOFFSET is the specified offset voltage of the op amp. VSHUNT is the shunt voltage at the operating current. Since the shunt voltage is a function of load the offset error will vary with load current and will be worse at lighter loads. If the maximum current is consistently low, consider increasing the shunt value for improved accuracy. The other contributing factors to the accuracy are the expected variation in the load voltage, VTM® Rout and VTM efficiency. When using the values for efficiency given in the data sheet, the expected variation is ±1%. This percentage error caries over to the overall accuracy. The effect of ROUT and VOUT variation is dependent on nominal operating conditions and can be predicted by Equation 17 and Equation 18: Equation 17 Voltage _ % Error = Equation 18 ROUT _ % Error = V% ⋅ 100 K ⋅ VOUT ⋅ (1 + V %) (1 − ) I IN ⋅ ROUT ⋅ η R% K ⋅ VOUT ( − (1 + R %) I IN ⋅ ROUT ⋅ η ⋅ 100 Where: IIN is the set VTM input current VOUT is the nominal output voltage of the load device V% is the % variation of the load voltage ROUT is the nominal output resistance of the VTM R% is the percent variation in ROUT (from data sheet) K is the VTM input to output ratio h is the nominal efficiency (from data sheet) AN:018 vicorpower.com Applications Engineering: 800 927.9474 Page 15 Table 5 summarizes the contributing factors to the overall error. Table 5 Current Source Error Error Source Shunt Tolerance Differential Amplifier Offset Differential Amplifier Gain TLV431B Reference TLV431 Divider VTM Efficiency VTM ROUT Error (%) ±0.1 Offset _ %Error = ( ROUT _ % Error = ( VOFFSET ) ⋅ 100 VSHUNT Load Dependent ±0.2 0.1% Tolerance Resistors ±0.5 ±0.2 ±1 0.1% Tolerance Resistors R% ⋅ 100 K ⋅ VOUT − (1 + R %) I ⋅R ⋅η Equation 18 V% ⋅ 100 K ⋅ VOUT ⋅ (1 + V %) ) I IN ⋅ ROUT ⋅ η Equation 17 Voltage _ % Error = VOUT Comments 0.1% Tolerance Shunt (1 − If the overall accuracy is not acceptable, the current sensing stage can be moved to the output of the VTM®. As the VTM is an isolated device, this will require the addition of an opto coupler to transfer the feedback signal to the primary side. The implementation of this additional stage is beyond the scope of this document. Please contact Vicor applications engineering for additional information if required. Layout Considerations Application Note AN:005 details board layout using VI Chip components. Additional consideration must be given to the external current control circuit components. The shunt voltage is on the millivolt level and is highly sensitive to noise. As such, current sensing circuitry should be located close to the shunt to avoid routing the sense signal over any distance. A 4 terminal Kelvin contact shunt is recommended for best results, eliminating error caused by solder resistance from the shunt to the current carrying connection on the PCB. The control signal from the sense circuit to the PRM should be shielded. Avoid routing this signal directly underneath the PRM® if possible. Components that tie directly to the PRM should be located close to their respective pins. It is also critical that all components be referenced to SG, and that SG not be tied to any other ground in the system, including –IN and –OUT of the PRM. Ensure that there is no unintentional bypass path which effectively shorts the shunt resistor. Conclusion The high power density and high efficiency of VI Chips can be used to power LEDs and other loads requiring regulated current operation by using the circuit and guidelines discussed in this document. Appendix A covers a complete design example using the Constant Current LED Driver Demonstration board. For additional assistance, circuit, schematic, or board layout review please contact Vicor Applications Engineering at: Tel: 800-735-6200 Customer Service: [email protected] Technical Support: [email protected] AN:018 vicorpower.com Applications Engineering: 800 927.9474 Page 16 Appendix A - Design Example An application requires that eight, 1 A Opto-Semiconductor LED strings be placed in parallel for appropriate luminance intensity. The current control accuracy required is ±5%. The forward voltage of the cells in question ranges from 20 V to 30 V and is nominally 25 V. The input voltage is 48V ±10%. The maximum ambient temperature is 50°C. 1) Select the appropriate VTM® from the product listing: V048F320T009 is chosen due to its operating voltage range of 17.3 V to 36.7 V, and a maximum output current of 9 A. This voltage and current range fall within the specification of the LED string. 2) Find the required PRM® output current: The efficiency plot of the V048F320T009 is located on the data sheet (Figure 3, pg. 3) and used to determine the VTM efficiency, which is approximately 96.3% at 8 A. Figure A1 V048F320T009 Efficiency vs. Load Graph 96.3% The nominal value for ROUT is found to be 79 mΩ from the output specifications table on pg. 2 of the data sheet. Figure A2 V048F320T009 Output Specifications Table AN:018 vicorpower.com Applications Engineering: 800 927.9474 Page 17 Using these numbers, and the nominal output voltage of the LED string (25 V), the necessary PRM® output current is calculated for a VTM output current of 8 A using Equation 4 from the Application Note. VTM _ IIN = PRM _ IOUT = 25 V ⋅ 8 A ⋅ 2 3 VOUT ⋅ IOUT K = = 5.4 A η⋅(VOUT + IOUT ⋅ ROUT ) 0.963 . (25 V + 8 A . 0.079 Ω) The P045F048T32AL is selected for its 6.67 A output current capability and 38 V - 55 V input voltage range. 3) Find the necessary reference voltage: The recommended values for the shunt resistor, and gain resistors are used. Equation 5 is used to determine the necessary reference voltage for a 5.4 A PRM output current. VREF = PRM _ IOUT ⋅ R1 ⋅ ( R3 100 kΩ ) = 5.4 A . 10 mΩ . ( ) = 5.4 V 1 kΩ R2 Using a TLV431B shunt regulator, R11, and R12 are selected to provide a 5.4 V output using 0.1% tolerance resistors. R10 is selected to limit the current to 1 mA. R10 = VH – VREF 9 V– 5.4 V = = 36 kΩ 1 mA 1 mA The closest standard 1% value is selected as 35.7 KΩ. 4) Determine the maximum PRM output voltage: The maximum PRM output voltage is selected to ensure that the PRM and VTM® can provide the maximum operating voltage of 30 V taking into account the maximum output resistance (ROUT(MAX)) of the VTM. For additional margin, the maximum output voltage is increased by 1 V. PRM _VOUT (MAX ) = (VOUT (MAX ) + IOUT ⋅ ROUT (MAX ) ) 31 V + 8 A ⋅(98 mΩ) = = 47.7 V 2 K 3 5) Find R7, R8 and R9: The recommended parameters are used for the maximum SC voltage (VSC(MAX) = 3 V) and SC pole frequency (FPOLE = 1 kHz) defined in Table 4. The maximum error amplifier output voltage (VEAO(MAX)) is determined from the AD8667 data sheet which specifies the output dropout voltage as a function of temperature on page 8 (www.analog.com/UploadedFiles/Data_Sheets/AD8663_AD8667_AD8669.pdf). At 50°C, the dropout voltage is 250 mV, resulting in a VEAO(MAX) of 8.75 V. Using Equation 9, R7 is calculated: R7 = 10 kΩ ⋅ 8.75 V 10 kΩ ⋅VEAO (MAX ) = = 2.4 KΩ 10 kΩ ⋅VSC (MAX ) ⋅ 2π ⋅ FPOLE ⋅ 0.22 μF −1.24 V 10 kΩ ⋅ 3 V ⋅ 2 π ⋅ 1 kHz ⋅ 0.22 µF – 1.24 V AN:018 vicorpower.com Applications Engineering: 800 927.9474 Page 18 The closest standard 1% value is selected as 2.37 kΩ Using Equation 10, R8 is selected: R8 = 10 kΩ⋅ R 7 ⋅ VSC (MAX ) 10 kΩ⋅ VEAO (MAX ) + 1.24 V⋅ R 7 − VSC (MAX ) ⋅ (10 kΩ + R 7) = 10 kΩ . 2.37 kΩ . 3 V 10 kΩ . 8.75 V + 1.24 V . 2.37 kΩ − 3 V . (10 kΩ + 2.37 kΩ) = 1.33 kΩ The closest standard 1% value is selected as 1.33 kΩ. Using Equation 11, R9 is selected based on VSC(MAX), PRM_VOUT(MAX), and the R68 value from Table 1: R9 = ( kΩ ⋅ 3 V ⋅ 0.961 R 68 ⋅ VSC (MAX ) ⋅ 0.961 ) = ( 93.1 PRM _VOUT (MAX ) − VSC (MAX ) ⋅ 0.961 47.7 V – 3 V ⋅ 0.961 ) = 5.99 kΩ The closest standard 1% value is selected as 6.04 kΩ. 6) Determine the compensation components R6 and C2 The crossover frequency is selected as 100 Hz which is a factor of 10 below the SC pole frequency of 1 kHz. C2 is fixed at a standard value of 0.1 µF, and R6 is calculated using Equation 15: R6 = 1 1 = = 15.9 kΩ C 2 ⋅ 2π ⋅ FCROSS 0.1 µF ⋅ 2π ⋅ 100 Hz The closest standard 1% value is selected as 16 kΩ. 7) Determine the overall accuracy: The sources of error are specified in Table 5. These factors are added up to determine the overall % error. The shunt error is 0.1%. The offset error is calculated using Equation 16, assuming a maximum input offset voltage of 300 µV for the AD8667 at 50°C. Offset _ %Error = ( VOFFSET 300 μV ) ⋅100 = ( ) ⋅100 = .55% 5.4 A ⋅10 mΩ VSHUNT The error due to the gain resistors is .2% The error due to the voltage reference and resistors is .7% The error due to variation in efficiency is 1% AN:018 vicorpower.com Applications Engineering: 800 927.9474 Page 19 In order to calculate the error due to output voltage variation, the percent variation in the load voltage from the nominal is calculated based on the specifications. V% = VMAX − VNOM 30 V − 25 V = = 20% = 0.2 25 V VNOM The maximum percent error due to this voltage variation is predicted using Equation 17. Voltage _ %Error = V% 0.2 ⋅ 100 = 0.4% ⋅100 = 2 ⋅ 25V ⋅ (1+ 0.2) K ⋅ VOUT ⋅ (1+ V %) 3 (1− ) ) (1 − IIN ⋅ ROUT ⋅ η 5.4 A ⋅ 79 mΩ ⋅ .963 In order to calculate the error due to variation in ROUT, the percent variation in ROUT from the nominal is calculated based on the data sheet specifications. R% = ROUT (MAX ) − ROUT (NOM ) 98 mΩ − 79 mΩ = = 24% = 0.24 ROUT (NOM ) 79 mΩ The maximum percent error due to this variation is predicted from Equation 17. ROUT _ %Error = R% 0.24 x100 = 0.61% x100 = 2 K ⋅ VOUT ⋅ 25 V 3 − (1+ .24) ( − (1+ R%) 5.4 A ⋅ 79 mΩ ⋅ 0.963 IIN ⋅ ROUT ⋅ η The total error is the sum of all the errors. Total _ %Error = Shunt _ %Error + Offset _ %Error + Gain _ %Error + Reference _ %Error + Efficiency _ %Error + VOUT _ %Error + ROUT _ %Error Total _ %Error = 0.1% + 0.55% + 0.2% + 0.7% + 1.0% + 0.4% + 0.6% = 3.6% Designing an LED driver circuit can be a challenging task due to the design variabilities and unknowns which may occur during the process. A Constant Current LED Driver Demonstration Board is available to assist in the design process. The board contains the basic circuit outlined in AN:018 along with the ability to adjust the output voltage and current settings and match the PRM with any standard VTM®. For further information please consult the User’s Guide (UG:007 www.vicorpower.com/documents/user_guides/vichip/ug-ConstantCurrent.pdf). AN:018 vicorpower.com Applications Engineering: 800 927.9474 Page 20 Figure A3 Constant Current LED Driver Demonstration Board Information furnished by Vicor is believed to be accurate and reliable. However, no responsibility is assumed by Vicor for its use. Vicor components are not designed to be used in applications, such as life support systems, wherein a failure or malfunction could result in injury or death. All sales are subject to Vicor’s Terms and Conditions of Sale, which are available upon request. Specifications are subject to change without notice. 10/2013 vicorpower.com Applications Engineering: 800 927.9474 Rev 1.3 Page 21