Power Factor Correction in Digital World By Nitin Agarwal, STMicroelectronics Pvt. Ltd., India Abstract There are various reasons why power factor correction circuit is used in various power supplies in large systems. The PFC (Power Factor Correction) at input stage becomes real necessary when the input power goes above 1kW typically. Recent advances in PFC include interleaving of two out-of-phase circuits to gain many system level advantages. Enough has been discussed about the advantages of interleaved PFC; in this article we will focus on a comparison of the analog approach to the digital approach. 1 Introduction We are very well aware that almost every load in modern electrical distribution systems is inductive. So, consumer is educating themselves to improve power factor by using various devices available in market. Some are having power factor correction capacitors, which improves power factor and reduces power consumption and finally saves money for the consumer. Power factor correction is usually used in industrial settings, where large inductive load exists. So, by using power factor corrector, tremendous power savings are realized, as well as greater longevity of the appliancesis widespread. 2 Future Trends There are analog power factor corrector designs are available from many IC suppliers. The trends in the market are coming up for interleaved Power factor corrector. The interleaving of PFC converters using analog controllers requires very sophisticated ICs that are dedicated for this purpose because of multiple loops within the system. A system based on analog controllers is forced to achieve a balance between complexity of the system and flexibility for adoption to various application needs. On the other hand, using an MCU (Micro Controller Unit) to manage the feedback loop and management of peripheral functions provides utmost flexibility with only a slight increase in complexity. So the trends of using microcontroller or DSP are coming up. 3 Advantages of Digital PFC over traditional Analog PFC Digital approach has two different meanings. One is the management of peripheral functions, which is commonly referred to as digital management, and the other is feedback loop control, which is called digital control. Digital control compared to analog control (of a feedback loop) does not really offer huge advantages; it is the combination of a digital control loop with digital management that creates a value that is more than a complete analog system. Single Inductor, Huge size Uses multiple MOSFETs in parallel Figure 1: Conventional PFC Relatively large capacitance
MCU based converters can measure and report the input power. The same input power measurement data can also be used internally to alter and optimize output voltage and switching frequency. Hence, with MCU-based converters, efficiency is optimized to the highest level and the efficiency curve can be flattened. Also, when demand for power is less than a certain threshold, one or more phases can be shut down to improve low load efficiency. Interleaving logic, typically an MCU or dedicated interleaved PFC controller Figure 2: Analog Interleaved PFC The cost of power components such as magnetics, MOSFETs and diodes, in higher power SMPS (Switched Mode Power Supply), is significantly higher compared to the cost of controllers. Hence, any saving in power components is easily paid off even if it requires higher cost in control components. Design for short circuit, power-up and power-down sequencing sometimes stresses power components beyond their normal operating level leading to the over-design of such components. With the help of the fuzzy logic of MCUs, it is very easy to implement timing and controls to minimize this over-design and reduce the overall cost of the system. An MCU based system also allows management of other functions in the SMPS that are beyond the realm of PFC, such as inrush current control and communication with the external world. A simple and cost-effective inrush current control can be implemented by using a resistor to limit the inrush current at start-up and then jumping the resistor out by means of a relay. A thermistor in contact with the resistor can be used to detect relay failures and prevent false startups of the circuit. One or more thermistors in contact with major power components can be used to sense internal temperatures of the system which can then be reported to the external world. Moreover, the same information can be used to implement a precise fan speed control system by means of a simple lookup table. With digital control it is very easy to set boundary conditions to set operating limits for the converter and reduce waste of over designing in the system for taking care of abnormal operating conditions. An MCU based system also reduces variations in the control module from one platform to another platform, compared with their analog counterparts, reducing engineering development time of future projects. 4 Digital PFC Block Diagram Output of the system is a gate drive signal for the MOSFETs of the two phases of the interleaved converter. These two phases are out of 180degrees. Inputs to the system, from the external world, are Voltage reference (V DCREF ), DC output voltage (V DC ), AC input voltage (V AC ) and AC input current (I AC ).
Figure 3: Digital PFC Block Diagram Voltage reference (V DCREF ) is set internally. Based on DC output voltage (V DC ), V ERR is calculated. This error signal is fed to Voltage PID loop which is outer loop. The output of Voltage PID loop is one of the inputs of Current PID loop which is inner loop. Figure 4: Functional Level PFC Block Diagram
6 ST Solution There are three main signals coming out from power board which are used in STM32 MCU daughter board for driving current loop (inner loop) and voltage loop (outer loop). There are two PWM (Pulse Width Modulated) signals out from STM32 MCU in interleaved manner which are 180 degrees out of phase to drive two MOSFETs. These are the four output signals from STM32: 1. 100kHz PWM with maximum duty cycle of 92% 2. 100kHz PWM 180 degrees phase shifted with first PWM with maximum duty cycle of 92% 3. One IO port for driving the relay 4. One IO for driving protection LED These are the nine input signals to STM32: 1. ADC (Analog to Digital Converter) input Q1 current sense 2. ADC input Q2 current sense 3. ADC input AC input current 4. ADC input Temperature sensor at input side 5. ADC input Output DC voltage 6. ADC input Heat Sink temperature sense 7. ADC input AC input voltage 8. ADC input 15V VCC_BAR 9. Protection Interrupt when Over Current and/or Over Voltage occurs STM32 daughter board PFC Main board Figure 5: Actual Hardware with ST devices
The regulated output is converted into digital using ADC peripheral. The error is calculated based on predetermined voltage setpoint value. And this error is adjusted by PID (Proportional, Integral and derivative) algorithm which is called outer loop or voltage control loop. The basic fundamental of PFC design is that the input current has to follow input AC voltage. So, the error between voltage envelope and the input current is made 0 by applying another PID (Proportional, Integral and derivative) algorithm which is called inner loop or current PID loop. The PID constants for both current loop (inner loop) and voltage loop (outer loop) (Proportional constant (K p ), Integral Constant (K I ) and Derivative Constant (K D )) are experimentally derived. Figure 6: Digital Control Block Diagram The block diagram shows the three critical inputs used in the control blocks of the PFC circuit. These three inputs include voltage on ADC1_CH0, the boost output voltage on ADC1_CH1, and the current feedback signal on ADC2_CH2. The measured input voltage is used in two control functions: the first one is used as one of the inputs of the multiply calculation that along with the error term is used to shape the current waveform. The second one is used to determine the average input voltage used in figuring the maximum allowable current so that the maximum allowable power can be kept constant as the input voltage range changes. This is done by Voltage Feed Forward Compensation circuit. The output voltage is sensed by an ADC and compared to a numerical constant in the code. Based on that, an error term is calculated and multiplied with the instantaneous input voltage to shape the current waveform to match the input voltage waveform. By further dividing this by the square of the average input voltage, a current
reference (I ACREF in the diagram) is established and is used to program the current compensator. The current compensator uses the last critical input, which is the current feedback level converted on ADC2_CH2. Implemented as a PID loop, the current compensator uses the I ACREF level as the set point and the ADC2_CH2 level as the process variable. The output of the loop programs the duty cycle of the MOSFET switches via peripheral TIM1 and TIM4 which produce 180degrees out of phase PWM of 100 khz. The inputs ADC1_CH3 and ADC1_CH4 measure the currents flowing in the boost MOSFETs of the two phases. With both converters being fed from the same voltage source, having equal value inductors and driven with the same value duty cycles, the average currents in the two phases share the output current almost equally. The measured values are compared relative to each other and if there is a large difference between the two currents, indicating trouble, the outputs are shut down and an error is reported via the communications link. The inputs ADC1_CH5 and ADC1_CH6 measure the temperature at input side before rectifier and heat sink temperature at chopper circuit respectively. If there is any unwanted increase in these temperatures, indicating trouble, again the outputs are shut down and an error is reported via the communications link. The ST solution using 32-bit microcontroller STM32F103C6T6 was designed to run at the 100W power level, 100 KHz switching frequency, with universal input range, and an output voltage of 400V. Because the timers of the STM32 can be triggered from one to the next with a precise time offset, it is possible to do interleaving, with as many phases as needed, limited by the number of timers on the device. Our testing showed good efficiency even with the overhead of startup power regulators needed to boot up the MCU first. Based on the design value of power components, including boost capacitor, we see a significant cost saving at a typical power of 1000W. 6 Conclusion The high-end server system, which holds over 100 CPUs, consumes tens of kilowatts of power. For mission-critical applications, communication between modules and system controllers is critical for reliability. Information about temperature, current, and the total harmonic distortion (THD) of each module will enable the availability of functions such as dynamic temperature control, fault diagnosis and removal, and adaptive control, and will enhance functions such as current sharing and fault protection. The dominance of analog control at the modular level limits systemmodule communications. Digital control is well recognized for its communication ability. Digital control will provide the solution to system-module communication for the DC power supply. The PFC converter is an important stage for the distributed power system (DPS). As described above, using full digital control, the power supply systems become flexible and can also realize complex control arithmetic which is difficult for analog control to perform. The digital approach will help designer to design different types (in terms of power level, features etc.) of power supply without changing the MCU. So, this helps the design cycle of a final product. The MCU based power supply provides new methods for design of power electronics and thus, provides high level control and communication capability required for SMPS. The above described ST solution uses STM32F103, the 32-bit microcontroller, to perform the input power factor correction and interleave average current control mode control with excellent efficiency, low cost and design flexibility.