an4468 application note · 500 w fully digital ac-dc power supply based on the stm32f334...
TRANSCRIPT
June 2014 DocID026160 Rev 1 1/46
AN4468Application note
500 W fully digital AC-DC power supply based on the STM32F334 microcontroller
By Rosario Attanasio
IntroductionAC-DC converters designed for a wide variety of applications, from computer adapters to server and telecom systems, require high efficiency over the entire load range of operation and across the universal mains input voltage range. Given the demand for more efficient, smaller adapters, their design is becoming more challenging and new conversion approaches, rather than the standard designs based on analog ICs, have been investigated. In particular, while the standard approach is based on the use of a boost type PFC and a regulation stage, both of which controlled using analog PWM regulators, the new, fully digital approach relies on the use of microcontrollers to control both the PFC and the DC-DC stage. This approach is increasingly being used for high density, high efficiency power electronics systems.
Figure 1. 500 W digital power supply board.
This application note focuses on the design of a 500 W AC-DC switch mode power supply with full digital control based on the STM32 family of microcontrollers. The system consists of two power stages: an input semi-bridgeless power factor corrector (SBPFC) controlled by an STM32F051K8 and a regulation stage implemented with an LLC half-bridge with synchronous rectification (SR), controlled by an STM32F334C8 microcontroller. The operating principles, main features and design choices are discussed. Details regarding the main components used for the implementation of both the power and control stages are also provided.
www.st.com
Contents AN4468
2/46 DocID026160 Rev 1
Contents
1 System overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2 Bridgeless PFC stage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.1 Operating principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2 Semi-bridgeless PFC design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.3 PFC control algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11
2.4 PFC firmware overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3 LLC DC-DC converter design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.1 LLC converter firmware overview and control algorithm . . . . . . . . . . . . . 24
4 Experimental characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Appendix A Schematic diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
6 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
7 Revision history . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
DocID026160 Rev 1 3/46
AN4468 System overview
46
1 System overview
The block diagram of the 500 W digital AC-DC converter is shown in Figure 2. A two-stage architecture is generally implemented in this type of application.
Figure 2. Block diagram of the 500 W digital power supply
The first block, from left to right, is the EMI filter. In a standard AC-DC topology, the EMI filter is connected to the input of a diode bridge rectifier, which is then connected to the input of a PFC stage. However, this 500 W AC-DC implementation uses the so-called bridgeless PFC topology which has the advantage of lower conduction losses and higher efficiency by eliminating the need for the diode rectifier stage. This choice also has the advantage of reducing component count with respect to a standard PFC.
The input stage is typically controlled using an outer voltage loop for bus voltage regulation and an inner control loop to shape the current according to a sinusoidal waveform. The outer loop adjusts the current reference in order to maintain a regulated bus voltage independently from the load or input voltage variations. The output isolation and regulation stage is implemented using a half-bridge LLC topology operated with constant duty cycle and variable frequency control.
The DC-DC stage performs voltage step-down using an HF transformer with a primary-to-secondary turns ratio chosen to maintain good efficiency and regulation in the entire operating range. The transformer is supplied with a square wave voltage generated by the primary side active switches. On the secondary side this voltage waveform is rectified and then smoothed by the output filter. While on the primary side switching losses are reduced thanks to zero voltage switching (ZVS), on the secondary side synchronous rectification (SR) is used to ensure low conduction losses. The overall effect of these design choices is high system efficiency, in line with the stringent requirements of the power supply industry.
The system is controlled by two microcontrollers from the STM32 product family. On the primary side, an STM32F051K8 controls the bridgeless PFC by sampling the current of the two MOSFETs, the input AC voltage and the output bus voltage. Two control signals, PWM1 and PWM2, are then generated to drive two power switches. On the secondary side, an
Secondary MCUSTM32F334C8
Primary MCUSTM32F051K8
EMI Filter
HF Isolation
LLC HalfBridge
Synchr.RectifierPFC
Input Voltage 95 Vac-265Vac, 45Hz/65Hz
Sinewave 430V DC 12V DC
ADC CH9
Adv. TIM
Input voltage
Sensing & Conditioning
Sensing & Conditioning
Sensing & Conditioning
PFC MOS1 Curr.
Output Voltage
PWM1
PWM2
PWM3
PWM3N
Transf. Current
Output Voltage
Synch. MOS2 VDS
Synch. MOS1 VDS
Sensing& Condit.
PWM4
PWM4N
BidirectionalCommunication
GPI/O
LED
Opto COMP.
CT
GPI/O
LED
Opto
AUXPS
ADC CH1
ADC CH0
Sensing & Conditioning
PFC MOS2 Curr.
ADC CH8
ADC1/2 Ch3
ADC1/2 CH2
ADC1 CH1
HR TIM
System overview AN4468
4/46 DocID026160 Rev 1
STM32F334C8 microcontroller samples the output voltage and adjusts the frequency of the LLC bridge control signals to ensure stable operation in the overall load range. In addition, two channels of the ADC are used to sample the rising and falling edges of the drain-to-source voltage of the SR MOSFETs. The two microcontrollers exchange information about the status of the input and output power stage via bidirectional serial communication.
Both the power stage and control stage are supplied by an offline flyback circuit which provides a suitable regulated voltage to the microcontrollers, the gate drive ICs and signal conditioning circuits.
Table 1 summarizes the main specifications of the 500 W digital power supply.
The converter accepts universal input voltage and produces a 12 V regulated output. The continuous power rating of the unit is 500 W. Natural convection is used up to 300 W. Above this power level a cooling fan is activated to provide forced air cooling. The ambient operating temperature range is 0 ºC to 50 ºC.
The intermediate high-voltage DC bus is regulated at 430 V by the PFC which draws sinusoidal input current from the AC input maintaining high power factor and low current total harmonic distortion (THDI%). The LLC circuit converts this high DC voltage to low DC voltage proving isolation (4 kV) by means of an HF transformer, and high efficiency thanks to ZVS. Input and output current and voltage protection are also provided together with overtemperature protection.
Table 1. 500 W AC-DC converter specifications
Parameter Value
Input AC voltage 95 V AC up to 264 V AC
Input AC frequency 45 Hz up to 65 Hz
Output voltage 12 V DC
PFC output voltage 430 V DC
Output power 500 W
PFC switching frequency 60 kHz
DC-DC switching frequency 70 kHz up to 130 kHz
HF transformer isolation 4 kV
Peak efficiency 93.2%
Cooling Forced air cooling
Input short-circuit protection 10 A fuse
Output short-circuit protection Managed by STM32F334C8
Input under/overvoltage Managed by STM32F051K8
Input under/overfrequency Managed by STM32F051K8
Bus DC under/overvoltage Managed by STM32F051K8
Output under/overvoltage Managed by STM32F334C8
Overtemperature protectionManaged by STM32F051K8 (primary) and
STM32F334C8 (secondary)
DocID026160 Rev 1 5/46
AN4468 Bridgeless PFC stage
46
2 Bridgeless PFC stage
2.1 Operating principleThe bridgeless PFC is a high efficiency topology characterized by the absence of the diode bridge rectifier and by the use of only two semiconductors in the current conduction path during any operating interval. The basic scheme of the bridgeless PFC boost converter is shown in Figure 3. The boost inductor is connected directly to the input AC source and split between line and neutral connection. Each inductor is connected to the drain of a power MOSFET and to the anode of a fast switching diode. The cathodes of the two diodes are connected to the output filter capacitor which is then parallel connected to the load.
The operation of this circuit is very similar to that of a standard PFC. More in detail, during the positive half-cycle of the input voltage, the boost converter composed of L1, M1, D1 and C1 is active, while during the negative half-cycle of the input voltage the boost converter composed of L2, M2, D2, and C2 is active.
Although this circuit has a reduced component count and higher efficiency compared to traditional PFC topologies, it has some drawbacks which limit its use in such a configuration. The main one is that the AC source is not referenced to the PFC ground. As a consequence, some difficulties in input voltage and inductor current sensing are introduced. Common mode EMI filtering can also be an issue related to this circuit.
A variant of the bridgeless PFC, known as semi-bridgeless PFC, is characterized by the addition of two diodes, D3 and D4, is shown in Figure 4. The purpose of these diodes is to keep the negative phase connected to the PFC ground, thus resolving the EMI filtering issue mentioned above for the bridgeless topology.
Figure 3. Basic scheme of a bridgeless PFC converter.
L1
L2
D1
M1
C1
LOAD
D2
M2
Vac
Bridgeless PFC stage AN4468
6/46 DocID026160 Rev 1
Figure 4. Semi-bridgeless PFC with addition of diodes D3 and D4
The operating principle is briefly described as follows:
• During the positive half-cycle
1. The input voltage diode D3 is off and D4 conducts the current returning to the source. MOSFET M1 is switched ON and OFF while MOSFET M2 is idle.
Figure 5. Current path during positive half-cycle when M1 is ON
2. When MOSFET M1 is ON the current flows as highlighted in Figure 5.
3. When MOSFET M1 is OFF the current flows as highlighted in Figure 6.
DocID026160 Rev 1 7/46
AN4468 Bridgeless PFC stage
46
Figure 6. Current path during positive half-cycle when M1 is OFF
4. The return path of current is offered by the parallel of diode D4 and the body diode of M2.
• During the negative half-cycle of the input voltage:
1. Diode D4 is OFF and diode D3 conducts the current returning to the source. In this case M2 is switched ON and OFF while M1 is idle.
2. When M2 is ON current flows as shown in Figure 7.
Figure 7. Current path during negative half-cycle when M2 is ON.
3. When M2 is OFF, the current flows as shown in Figure 8
Bridgeless PFC stage AN4468
8/46 DocID026160 Rev 1
Figure 8. current path during negative half-cycle when M2 is OFF
4. The return path of the current is offered by the parallel of diode D3 and the body diode of M1.
Note that in the previous analysis it has been supposed that during the positive and negative half-cycles one of the two MOSFETs is switching and the other is idle. There are actually two additional control strategies besides the one mentioned above.
In the first, during the positive half-cycle of the input voltage M1 is switched ON and OFF while M2 is kept ON. During the negative half-cycle of the voltage, M2 is switched ON and OFF while M1 is kept ON. This control strategy allows the return current to flow across the channel of the MOSFET rather than across the body diode and, therefore, some improvement in efficiency is possible.
In the second, both MOSFETs are controlled synchronously, with the same PWM signal applied to each gate. Again, the benefit is lower power dissipation when current flows through the MOSFET during the return phase. In addition, only one driver can be used for both MOSFETS.
Due to these advantages, the PFC topology chosen for the 500 W digital power supply implementation described in this application note is the semi-bridgeless PFC with synchronous control.
2.2 Semi-bridgeless PFC designThe practical implementation of the proposed 500 W bridgeless boost PFC rectifier is detailed in this section. The converter is designed to operate in DCM. In low and medium power applications, discontinuous and critical conduction mode are often used to obtain numerous benefits such as reduced switching losses, low diode reverse recovery current and small inductor size.
The PFC was designed starting from the specifications shown in Table 2.
DocID026160 Rev 1 9/46
AN4468 Bridgeless PFC stage
46
The design steps are reported as follows:
• Maximum input RMS current value
Considering the minimum input voltage and the minimum desired converter efficiency, the maximum RMS value of the input current can be calculated.
Equation 1
• Maximum input peak current value
Equation 2
• Boost inductor value
The inductance value of the two input chokes has to be carefully selected in order to ensure DCM operation across the universal input voltage range and load range. This is guaranteed if the inductor is selected according to equation 3:
Equation 3
Where M=Vout/Vin is the converter voltage gain, R=Vout²/Pout is the load resistance and T is the switching period. Note that in equation 3 the minimum value of peak input voltage has been considered for the calculation of the value of the input inductors.
Two 100 µH, 10 A RMS inductor chokes, characterized by a 33 mΩ DC resistance have been selected for this application. The peak-current rating of the inductor must be well above the expected peak current of the converter to avoid core saturation. A gapped ferrite with core material characterized by low core losses at the design switching frequency of 60 kHz has been used.
• Inductor peak current
The inductor peak current can be calculated with equation 4:
Table 2. Specifications of the 500 W Bridgeless PFC
Parameter Value
Input AC voltage 90 V AC up to 264 V AC
Input AC frequency 45 Hz up to 65 Hz
Output voltage 12 V DC
PFC output voltage 430 V DC
Output power 500 W
PFC switching frequency 60 kHz
Efficiency >95%
Bridgeless PFC stage AN4468
10/46 DocID026160 Rev 1
Equation 4
Equation 5
Where it is assumed that the duty cycle is limited to a maximum of 0.9
• Power MOSFET selection
The choice of power semiconductors is fundamental to meet the efficiency requirements of the application. In a boost PFC, when the switch is on the current is equal to the inductor current. The peak current calculated in equation 4 is also the peak current of the switch. When the switch is off, the drain-to-source voltage is the output voltage. Therefore, the MOSFET is selected with a rated voltage greater than the output voltage and rated current greater than the maximum inductor current. Since the boost PFC operated in DCM is mainly affected by conduction losses, it is very important to use a power MOSFET with low drain-to-source resistance in order to ensure high efficiency. The output voltage of the device value is chosen according to equation 6:
Equation 6
To meet the design requirements, the STP57N65M5 N-channel power MOSFET has been used for M1 and M2. This device is characterized by a minimum breakdown voltage of 650 V and a maximum on resistance of 63 mΩ at 25 °C. The total gate charge is 98 nC.
• Rectifier diodes selection
The average diode current can be calculated as:
Equation 7
Two STPSC1006D 600 V, 10 A diodes are used as rectifier diodes. The forward voltage drop at 1.16 A is about 0.75 V at TJ=150 °C.
• Output capacitor selection
The output capacitor bank value is selected to limit the output voltage ripple to 1% of the nominal output voltage. Equation 8 has been chosen to define the value of the output capacitor bank:
Equation 8
Where ω is the mains angular frequency and ΔV0 is the output voltage ripple. Four 450 V, 100 uF electrolytic capacitors from Vishay have been parallel connected on the PFC output. In addition, a 600 V, 560 nF ceramic capacitor with low ESR and ESL is connected in parallel to the each electrolytic capacitor.
• Input filter capacitors
DocID026160 Rev 1 11/46
AN4468 Bridgeless PFC stage
46
The need for a high power factor introduces a limit on the maximum capacitance that can be placed across the line. The maximum capacitance is a function of how much phase shift that can be tolerated. This phase shift angle can be calculated using equation 9.
Equation 9
Choosing Cin=3 uF the total displacement is 7°.
• Current sensing
The current flowing across each power switch is sensed by means of CTs. These are placed between the inductor and the MOSFET. Due to this placement, only the rising part of the inductor current is available to the control algorithm. The current signal is always sampled at the mid-point of the gate control PWM. Then, the current samples could be digitally corrected in order to calculate the average inductor current. In fact, contrary to the CCM case, when the PFC is in DCM the current sensed at the mid-point of the MOSFET triangular current waveform is no longer equal to the inductor average current.
The CT must withstand the peak current calculated in equation 4. The selected current sense transformer is the 53040C from Murata. This is characterized by a magnetizing turns ratio of 40.
2.3 PFC control algorithmThis section contains an overview of the PFC digital average current-mode control method in discontinuous conduction mode (DCM). This control method allows good input current shaping in terms of harmonic distortion (low THD) and synchronization with input voltage (PF close to one). The proposed control technique estimates the average value of the inductor current in each switching cycle. The average inductor current can be precisely estimated by the digital controller modifying the sampling instant within the switching period and calculating a correction factor. Sampling the inductor current more than once in a switching cycle, with a high-performance A/D converter, also helps to achieve a precise average inductor current estimation.
The control is implemented by means of a conventional current sensing circuit and a 32-bit microcontroller, the STM32F051K8 from the STM32 family, which features a fast 12-bit A/D converter. As in traditional average current mode control implementations, this algorithm uses two control loops: an outer voltage loop, performed at 100 Hz providing the regulation of the bus voltage output at the reference value (430 V) and an inner current loop, performed at 20 kHz, to ensure that the input current follows a sinusoidal reference in phase with the input voltage. While the voltage control is achieved with a single linear regulator, the current control is based on a PI controller plus a feed-forward controller.
Bridgeless PFC stage AN4468
12/46 DocID026160 Rev 1
Figure 9. PFC control scheme block diagram
A block diagram of the control scheme is shown in Figure 9. The average inductor current is forced to follow the reference current, which is proportional to a rectified sinusoidal voltage synchronized to the input voltage and reconstructed by a lookup table, so that unity power factor is achieved. The lookup table provides the sinusoidal values in a [0°, 90°] range with 1.15 fixed-point format (1 bit for the integer part and 15 bits for the decimal part), with the amplitude given by the output of the voltage PI regulator.
The STM32 microcontroller calculates the duty cycle in every three switching cycles based on the error between the feedback current and the reference current values. The two switches are then controlled with these duty cycle values to achieve unity power factor and sinusoidal input current waveform. The whole digital average current mode control includes:
1. bus voltage and MOSFETs current sampling
2. voltage error calculation
3. voltage PI regulation
4. reference current calculation
5. current error calculation
6. current PI regulation
7. duty cycle feed-forward generation
8. final duty cycle computation.
The final calculated duty cycle value is therefore composed of two terms: an open loop, feed-forward term, which depends on the instantaneous input and output voltage values, and a closed loop term which depends on the applied load. The input voltage value of the feed-forward term is obtained by means of the same lookup table used to calculate input voltage amplitude and phase. This choice is effective for noise reduction and for compensation of sampling delays.
Considering only half the period of the input sinusoid, the expression of the duty cycle for the boost converter in CCM is:
DocID026160 Rev 1 13/46
AN4468 Bridgeless PFC stage
46
Equation 10
where Vo is the output bus voltage, and Vin sin ωt is the mains voltage. However, this converter is designed to work in DCM and therefore the correct expression of the duty cycle is:
Equation 11
where D is the constant duty cycle in absence of modulation. Since the implementation of this formula is quite complex, the feed-forward duty cycle term has been calculated using the following expression:
Equation 12
Where kff is a gain depending on the load and the input voltage and m is a constant modulating index.
The gain kff increases with load and decreases with input voltage. Since the output of the PI voltage regulator depends on the load, the expression chosen for feed-forward duty cycle gain is the following:
Equation 13
where Vin is the amplitude of mains voltage, VPIout is the output of the voltage PI regulator and kg is a gain that varies with the input voltage amplitude according to the diagram shown in Figure 10.
Bridgeless PFC stage AN4468
14/46 DocID026160 Rev 1
Figure 10. variation of the Kg parameter with the input PFC voltage
2.4 PFC firmware overviewThe execution of the PFC firmware is based on the implementation of a state machine made up of six states. The diagram in Figure 11 provides an overview of the firmware architecture. As soon as the microcontroller is supplied, all the peripherals are configured and the firmware is in PFC IDLE state. Here the system waits for the AC mains insertion and verifies that no faults are present. If no faults are detected, the new state is PFC INIT. In this state the integral terms of PI regulators are reset and the PM8834 (PFC MOSFET driver) is enabled. The PFC output reference voltage is initialized to the last sampled value. After this, the new state is PFC START, where the output voltage ramp function is executed. When the ramp is completed and the output voltage is equal to 430 V DC, the LLC L6491 driver is enabled and a serial command is sent to the secondary microcontroller to confirm that the PFC startup is completed. Then, the feed-forward duty cycle control is enabled and the next state is PFC ON. If in any of the above states a fault condition is verified the system is forced into PFC STOP state.
0
50
100
150
200
250
300
350
0 50 100 150 200 250 300
Kg
Voltage [V]
DocID026160 Rev 1 15/46
AN4468 Bridgeless PFC stage
46
Figure 11. PFC firmware state machine overview
The possible faults are:
• Input voltage under/overfrequency (F< 45 Hz; F>65 Hz)
• Input under/over voltage (Vin< 95 V RMS; Vin>264 V RMS)
• Bus DC undervoltage (320 V)
• Bus overvoltage (Vbus>470 V)
In any of these cases both the PFC and LLC PWM modulation are disabled. The MOSFET driver enabling signal is low and the system is in PFC WAIT state for 5 seconds before the system enters the PFC IDLE state.
If during PFC operation, the bus voltage is greater than 450 V but lower than 470 V, the PFC operates in BURST MODE. In this case, the PFC is forced into the STOP state and only the PFC modulation is disabled while the LLC converters continue to operate. The WAIT state is bypassed and the system is forced into IDLE state. The PWM modulation is re-enabled when the PFC bus voltage is below 430 V.
As described above, the control firmware performs several tasks, with different timings and priorities to ensure both output voltage regulation and input current regulation. The main
Inner current
loopperformedat 20kHz
Bridgeless PFC stage AN4468
16/46 DocID026160 Rev 1
tasks and functions are summarized in Table 3, together with the execution time and priority level.
After the initialization of the main peripherals such as system clock, ADC, DMA, timers, UART and I/O ports, which are executed in the main loop, the fault check function and serial communication tasks are performed. These tasks have lower priority level compared to those executed in specific interrupts.
As soon as the 500 W power supply is connected to the mains supply, the PFC is in idle state and a fault check routine is executed to ensure that no fault conditions are present. During this phase, the input voltage amplitude and frequency are calculated in order to verify that the mains parameters are within the converter specified operating window. In this case, the PM8834 gate driver, used to drive the two PFC MOSFETs, is enabled and the main PFC control algorithm is performed. A soft-start routine is first executed to linearly increase the DC bus voltage up to 400 V. When the soft-start is completed, the primary STM32F051K8 microcontroller sends the information of the correct PFC status to the secondary STM32F334x microcontroller, via serial communication. After having sent this information, the primary microcontroller starts the control loop routine, updating the control regulator parameters and enabling a second voltage ramp-up to increase the bus voltage to 430 V.
The algorithm for the computation of the input voltage amplitude and frequency is performed when the update event of timer 6 (TIM6) occurs (every 500 µs). The voltage control loop is executed every 10 ms when the update event of timer 14 (TIM14) occurs. Also, the calculation of new feed-forward duty cycle control parameters is triggered by the same event.
The current control loop is the highest priority task and is performed when the interrupt of the ADC end of conversions is triggered. All the acquisitions are synchronized with timer 1 (TIM1) which generates the output PWM signals (CH1 and CH2 of TIM1) as shown in Figure 12. A single sampling in single period technique is adopted.
Table 3. Function names and tasks of the PFC firmware
Function Task Execution frequency Priority
ADC1_COMP_IRQHandler()
Current control loopSinusoidal reference calc.
Input voltage sign detection
Average of last 3 current samples
20 kHz60 kHz60 kHz
60 kHz
High
TIM14_IRQHandler()Voltage control loop
Update control parameters100 Hz Medium
TIM6_DAC_IRQHandler()Computing of Vac frequency and
amplitude2 kHz Medium
main()
Fault checksState machine
Serial communication
LED
Low
DocID026160 Rev 1 17/46
AN4468 Bridgeless PFC stage
46
Figure 12. Timer 1 utilization for PFC PWM generation and ADC conversion trigger
Figure 13. ADC utilization and acquisition sequence
The switch current is sensed at the midpoint of the PWM ON time which corresponds to the average inductor current in CCM. A sample correction algorithm can be implemented to take into account DCM operation. This is generally done, when necessary, to improve the input current THD. The ADC samples the MOSFET current first, either across MOS1 or MOS2, depending on the sign of input voltage sine waveform. Then, the input voltage is sampled and immediately after the output bus voltage is sampled. The acquired samples are then transferred into a buffer by DMA, keeping CPU resources free for other operations. The acquisition sequence (which depends on the sign of the input voltage) and ADC channels utilization is shown in Figure 13.
If a fault condition occurs, the PWMs are stopped, the PFC gate driver is disabled and a serial message is sent to the secondary side microcontroller in order to stop the LLC control algorithm.
Table 4 summarizes the resources required for the implementation of the digital control of the 500 W semi-bridgeless PFC.
Bridgeless PFC stage AN4468
18/46 DocID026160 Rev 1
Table 4. Microcontroller resources for PFC digital control implementation
Signal Type MCU resource Resolution
Input voltage Analog ADC1 CH8 12 bit (1MSPS)
Output voltage Analog ADC1 CH9 12 bit (1MSPS)
MOS1 current Analog ADC1 CH1 12 bit (1MSPS)
MOS2 current Analog ADC1 CH0 12 bit (1MSPS)
ADC trigger Digital TIM1 CH4
Gate drive 1 Digital TIM1 CH1 (PWM1) Dres=0.125% @ 60kHz
Gate drive 2 Digital TIM2 CH2 (PWM2) Dres=0.125% @ 60kHz
RS232-RX Digital UART1 (PB10)
RS232-TX Digital UART1 (PB6)
Temp. sense Analog ADC1 (CH6)
LED Digital GPI/O (PB2)
PM8831 enable/disable Digital GPI/O (PA12)
L6491 enable/disable Digital GPI/O (PF1)
DocID026160 Rev 1 19/46
AN4468 LLC DC-DC converter design
46
3 LLC DC-DC converter design
The purpose of the DC-DC stage is to step the PFC output voltage down to 12 V. The topology used for this conversion stage, the LLC half-bridge, is shown in Figure 14. The power stage of the LLC converter is formed by the input and output capacitors C2 and C3, MOSFETs M3, M4, transformer T1 and resonant capacitor Cr. The resonant inductor Lr and the magnetizing inductor Lm depicted in Figure 14 are integrated in the high frequency transformer. The output rectification stage is implemented with two MOSFETs, M5 and M6, used to exploit the advantages of synchronous rectification. The main advantages of the LLC topology are:
• ZVS at turn-on for the primary side switches
• ZCS at turn off for the secondary side switches
• Very good load regulation
The main drawback is the high sensitivity to input voltage variations which requires some design trade-offs to optimize the circuit in a wide input voltage range. The specifications used for this design are reported in Table 5.
Figure 14. Basic schematic of the LLC DC-DC converter
LLC DC-DC converter design AN4468
20/46 DocID026160 Rev 1
The voltage gain of the LLC converter resonant tank can be expressed as in equation 14.
Equation 14
where:
Under no-load conditions (Q=0) equation 14 can be written as:
Equation 15
In this operating condition a second resonant frequency can be defined:
Equation 16
Or in normalized form
Table 5. Main specifications of the LLC converter
Parameter Value
Input voltage 420 V
Minimum input voltage 400 V
Maximum input voltage 440 V
Nominal output voltage 12 V DC
Efficiency >96%
Output power 500 W
Switching frequency 80 kHz @ full load, 420 V DC Input
Maximum switching frequency 90 kHz
Stray capacitance 350 pF
Dead time 350 ns
is the ratio between the resonant tank inductance and the transformer magnetizinginductance;
is the normalized frequency, ratio between the switching frequency and theresonant frequency;
is the quality factor;
is the equivalent load resistance.
DocID026160 Rev 1 21/46
AN4468 LLC DC-DC converter design
46
Equation 17
In fact, in no-load conditions (the rectifier is not conducting) the total primary inductance (magnetizing plus leakage inductance) resonates with the capacitor. At this frequency the no-load gain characteristic tends to infinity.
For an infinite value of the normalized frequency the no-load gain MOL tends to M∞:
Equation 18
The design steps are summarized in the following part of this document:
• Transformer turns ratio:
This is calculated assuming that the required gain at nominal input voltage is unitary:
Equation 19
The transformer turns ratio has been chosen equal to 18.
• Calculation of minimum and maximum voltage gain (assuming 5% margin):
Equation 20
• Maximum input current
Equation 21
• Equivalent load resistance
Equation 22
• Calculation of maximum normalized operating frequency:
Equation 23
• Inductance ratio:
LLC DC-DC converter design AN4468
22/46 DocID026160 Rev 1
Equation 24
• Calculation of the maximum value of Q to operate in ZVS at full load and minimum input voltage
Equation 25
• Calculation of the maximum Q value to work in the ZVS with no-load condition and maximum input voltage:
Equation 26
The value of QZVS is chosen equal to QZVS1.
• Calculation of resonant tank components:
Equation 27
• Resonant Tank
The selected values differ from the calculations as a result of the efficiency optimization process done at an input voltage of 430 V DC (selected output PFC voltage). The resonant tank values are:
Cr=30 nF
Lr=140 uH
Lm=500 uH
The transformer has been designed to integrate both the series inductance Lr and the magnetizing inductance Lm. The resonant capacitance is realized by parallel connection of two 15 nF, 1 kV polypropylene capacitors. With the values reported above, the resulting resonant tank gain is depicted in Figure 15.
DocID026160 Rev 1 23/46
AN4468 LLC DC-DC converter design
46
Figure 15. Resonant tank gain at full load, half load, no load
The frequency at full load and 430 V DC input is 77.7 kHz. The minimum operating frequency at full load and minimum input voltage is 68.3 kHz. The maximum operating frequency is obtained with no load and maximum input voltage and is equal to 96.3 kHz.
• Half-bridge MOSFETs selection:
The power MOSFETs used on the primary side of the LLC converter have 600 V breakdown voltage and drain current of 6.9 A at 100 °C case temperature and 0.38 Ω maximum drain-to-source resistance.
The part number of this device is STP13NM60ND. It is characterized by intrinsic fast-recovery body diode produced using the second generation MDmesh™ technology. This ensures very low on-resistance and optimal switching performance making them ideal for this application.
The breakdown voltage is selected considering the following equation:
Equation 28
where 20% derating is considered.
• SR MOSFETs selection:
The power MOSFETs used on the secondary side for synchronous rectification are characterized by 30 V breakdown voltage and 32 A drain current at 100 °C PCB temperature. They feature extremely low drain-to-source resistance, ideal to keep secondary side conduction losses low thanks to the 7th generation of STripFET™ DeepGATE™ technology. The part number is STL160N3LLH6 which is packaged in a Power Flat 5X6. Two of these devices are parallel-connected on each side of the rectifier circuit in order to minimize conduction losses.
The LLC converter parameters resulting from the design are summarized in Table 6.
LLC DC-DC converter design AN4468
24/46 DocID026160 Rev 1
3.1 LLC converter firmware overview and control algorithmThe LLC firmware is implemented on the secondary side MCU which is the STM32F334x, a 32-bit microcontroller from the STM32 family. The control strategy generates the gate signals for both primary and secondary side MOSFETs in order to ensure precise output voltage regulation at 12 V DC.
The secondary side MOSFETs are used to reduce the conduction losses produced by standard output rectifier diodes. The SR MOSFETs are properly controlled by the MCU implementing a synchronous rectification control strategy which improves the efficiency and compensates for delays caused by hardware components and parasitic elements.
The block diagram of the control scheme is shown in Figure 14.
Table 6. LLC converter parameters
Parameter Value
Magnetizing inductance 500 uH
Resonant inductance 140 uH
Transformer turns ratio 18
Resonant capacitor 30 nF
Resonant frequency 77.7 kHz
Maximum switching frequency 96.3 kHz
Minimum switching frequency 68.3 kHz
Input voltage operating range 400 V to 440 V
Output voltage 12 V
Primary side MOSFETs STP13NM60ND
SR MOSFETs STL160N3LLH6
DocID026160 Rev 1 25/46
AN4468 LLC DC-DC converter design
46
Figure 16. Block diagram of the LCC converter control
A simple voltage control loop, based on the use of a PI regulator whose input is the error between the reference voltage and LLC converter output voltage, is adopted.
The STM32F334x high resolution timer HRTIM generates the driving signal pattern to control the primary and secondary MOSFETs. The HRTIM is specifically designed to drive power conversion systems. It is characterized by a modular architecture and can generate up to ten digital signals with either independent or coupled waveforms.
The HRTIM has timing measure capabilities and links to built-in ADC and DAC converters. It features light load management mode and is able to handle various fault schemes for safe shutdown purposes. The HRTIM can be partitioned into several sub modules:
• The master timer
• The timing units (timer A to timer E)
• The output stage
• The burst mode controller
• An external event and fault signal conditioning logic
• The system interface
The HRTIM is configured as follows to drive the LLC stage with synchronous rectification:
– The master timer is used for synchronization of Timer A, C and D
– Timer C is used for primary half-bridge modulation
– Timer A and Timer D are used for secondary synchronous rectification
– Timer C is configured in half mode and drives two complimentary outputs CHC1 and CHC2 with 50% duty cycle and a suitable dead time.
Timer A output CHA1 and Timer D output CHD1 are used to generate the two synchronous rectification MOSFET modulation signals. The ON and OFF time of each driving signal generated by Timer A and Timer D is dynamically adjusted and depends on the drain-to-source voltage of the synchronous rectification MOSFETs. These voltages are acquired by two 5 MHz ADC converters, namely ADC1 and ADC2. The utilization of the two A/D converters is shown in Figure 17.
+-
Fsampling=Fsw
PI
Vref
Vout_DCDC_fbk ADC
DC/DC
DCDC CurrentSense
DCDC Output Voltage
LOAD
Voltage Control Loop Bandwidth
Fsw VDs1 Synch. Rect
VDs2 Synch. Rect
Input Bridge and Synch. Rect. Control
LLC DC-DC converter design AN4468
26/46 DocID026160 Rev 1
Figure 17. STM32F334x ADC1 and ADC2 utilization
The hardware is configured to acquire some other measurements such as heatsink temperature, auxiliary power supply output voltage, LLC transformer primary current and its average value.
ADC1 CH1 acquisition is triggered by the update event of the master timer, while CH2 and CH3 acquisitions are triggered by either timer A or timer C compare events. Similarly, ADC2 CH2 and CH3 are triggered by either timer C or timer D compare events. The injected acquisition sequence and trigger events are shown in Figure 18 while the HRTIM timing configuration is shown in Figure 19.
Figure 18. ADC1 and ADC2 injected acquisition sequence and trigger events
DocID026160 Rev 1 27/46
AN4468 LLC DC-DC converter design
46
Figure 19. HRTIM timing configuration
The voltage control loop consists of a standard PI regulator and is executed at the update event of the master timer. The regulator output is the new value of the primary MOSFET driving signal period in terms of number of clock cycles (new value of HRTIM_MPER register). In this way, the frequency can vary from 70.3 kHz up to 130 kHz, while the duty cycle is fixed at 50%. Together with the period, the new rising and falling edges of the SR signals are calculated. The calculation includes a blanking time to take into account the opto-isolator delay and the delay due to the sampling and acquisition of SR MOSFETs drain-to-source voltage.
The SR control algorithm tries to maintain the two drain-to-source voltage ADC samples (each drain-to-source voltage is sensed twice) under/above a reference value to detect the correct turn-on and turn-off of SR MOSFETs. If the first and second samples are not under/above the preset reference value, the on-time is made longer or shorter.
The LLC converter firmware structure and its main tasks are highlighted in Figure 20 and,Table 7 respectively.
Figure 20. LLC firmware structure overview
LLC DC-DC converter design AN4468
28/46 DocID026160 Rev 1
The STM32F334xx devices embed 3 comparators that can be used either as standalone devices (all terminals are available on I/Os) or combined with the timers. They can be used for a variety of functions such as wakeup from low-power mode triggered by an analog signal, analog signal conditioning and cycle-by-cycle current control loop when combined with the DAC and a PWM output from a timer.
In this design an internal comparator is used to implement the output overcurrent protection. The configuration is shown in Figure 21 where the inverting pin of the microcontroller internal comparator COMP2 is connected to CH1 of DAC1 and the non-inverting pin is connected to the output of the current sense circuit used to measure the LLC converter primary current.
Figure 21. STM32F334x internal comparator configuration for overcurrent protection
Table 7. LLC firmware tasks
Function Task Execution frequency Priority
HRTIM_MASTER_IRQHandler()Voltage control loop
Synchronous rectification
Burst mode
HRTIM master frequency
Medium
HRTIM_TimerC_IRQHandler()Output short-circuit
protectionComparator trigger High
USART1_IRQHandler() Acquisition of PFC status On acquisition Low
main()
State machine
LED Serial communicationTemperature check
Low
DocID026160 Rev 1 29/46
AN4468 LLC DC-DC converter design
46
When the sensed current is higher than the preset threshold, a fault event is generated to set the Timer C output channels in idle state. In addition, the HRTIM Timer C interrupt is generated to set the state machine in LCC_STOP state. After such an event, the state is updated to LLC_WAIT and, after this, the new state is LLC_OFF. If no faults are detected the LLC converter can be started again after the PFC bus voltage ramp-up.
Burst mode operation is implemented for light load management. The main purpose is to increase the efficiency of the converter by reducing the number of transitions on the outputs and the associated switching losses. The STM32F334C8 burst mode controller allows having the outputs alternatively in IDLE and RUN state, by hardware, so that some switching periods are skipped with a programmable periodicity and duty cycle.
The burst mode operation is enabled when the output of the voltage control PI regulator is lower than a predefined threshold (switching period) corresponding to a switching frequency of 115 kHz. Similarly, it is disabled when the PI regulator output is greater than a threshold corresponding to 110 kHz. In this way, the output voltage is regulated at 12 V ± 100 mV during no-load and light load operation.
Experimental characterization AN4468
30/46 DocID026160 Rev 1
4 Experimental characterization
This section includes the validation results of the 500 W digital power supply design. It includes the efficiency and power quality test results at different input voltage values and the main waveforms of the two converters.
The power supply efficiency measured at different mains voltages, namely 120 V AC and 230 V AC., is shown in Table 8 and Table 9.
The efficiency was measured in the following conditions:
1. The board was operated at half load for 30 minutes.
2. The efficiency results were taken in the following output load points: 6 A, 12 A, 18 A, 24 A, 28 A, 34 A, 42 A.
3. The input and output voltages were measured directly at the input and output connectors of the board.
4. All the measurements were taken using the Voltech PM6000 Universal Power Analyzer
5. The power measurements do not take into account the power consumption of the fan which will be activated above 300 W.
6. The board was tested with open frame at an ambient temperature of 25 °C.
Table 8. Efficiency measurements at 120 V AC input
VIN (V) IIN (A) PIN (W) VOUT (V) IOUT (A) POUT (W) Efficiency (%)
120 0.712 82.6 12.07 6 72.76 88.08
120 1.33 157.98 12.03 12 144.58 91.51
120 1.96 234.01 12.01 18 216.49 92.51
120 2.61 312,34 12.02 24 288.74 92.44
120 3.05 365.2 12,03 28 337.48 92.40
120 3.72 445.24 11.99 34 409.37 91.94
120 4.63 554.2 12.01 42 505.69 91.24
Table 9. Efficiency measurements at 230 V AC input
VIN (V) IIN (A) PIN (W) VOUT (V) IOUT (A) POUT (W) Efficiency (%)
230 0.45 81.71 12.1 6 72.62 88.87
230 0.73 156.37 12.04 12 144.47 92.38
230 1.04 232.65 12.01 18 216.56 93.08
230 1.37 309.26 12.04 24 288.63 93.32
230 1.59 360.98 12.03 28 336.62 93.25
230 1.94 441.15 11.99 34 410.26 92.99
230 2.40 546.81 12.01 42 505.5 92.44
DocID026160 Rev 1 31/46
AN4468 Experimental characterization
46
The peak efficiency is obtained at 230 V AC input is equal to 93.2%. For 120 V AC input the peak efficiency is reduced to 92% due to higher input to output voltage ratio.
Table 10 and Table 11 report the measurement of current THD% and power factor both at 120 V and 230 V AC input voltage. It can be noted that the total harmonic distortion is still low even for high input voltage values. The power factor is close to unity for both low and high input voltage values.
Table 10. DSMPS power factor and current THD% at 120 V AC input
Output load at 120 V AC PF THD%
6 A 0.965 5.4
12 A 0.984 5.4
18 A 0.992 4.2
24 A 0.994 3.5
28 A 0.996 3.2
34 A 0.996 3.2
42 A 0.997 4.3
Table 11. DSMPS power factor and current THD% at 230 V AC input
Output load at 230 V AC PF THD%
6 A 0.78 13.6
12 A 0.925 11.4
18 A 0.964 9.7
24 A 0.977 8.9
28 A 0.983 7.9
34 A 0.986 7.6
42 A 0.989 7.4
Experimental characterization AN4468
32/46 DocID026160 Rev 1
The PFC inductor currents are shown together with the input voltage in Figure 22 and Figure 23. It is possible to highlight that only one branch of the bridgeless PFC circuit is switching during each half period of the input voltage.
The PFC efficiency in the whole operating range is highlighted in Figure 24 and is as high as 98%. The power factor for 120 V and 230 V AC input is depicted in Figure 25 for an output load varying form 70 W up to 500 W. In this operating range the power factor is always higher than 0.9.
Figure 22. PFC inductor current at 120 V input, full load
Figure 23. PFC inductor current at 230 V input, full load
Figure 24. PFC circuit efficiency Figure 25. PFC input current power factor
90919293949596979899
100
0 100 200 300 400 500 600
Efficiency
[%]
Output Power [W]
PFC Efficiency
0.88
0.9
0.92
0.94
0.96
0.98
1
0 200 400 600
PF
Output Power [W]
PFC Power Factor
120V AC 230V AC
DocID026160 Rev 1 33/46
AN4468 Experimental characterization
46
Figure 26 and Figure 27 show the inductor currents and PFC MOSFETs drain-to-source voltage at full load for 120 V operation. Figure 28 and Figure 29 show the same waveforms for 230 V operation. In both cases the PFC is operating in DCM.
The LLC converter main switching waveforms are shown in Figure 30 and Figure 31 for full load and no-load operation, respectively. The MOSFET current is negative before the device is turned on thus providing zero voltage switching.
The switching frequency at full load and nominal input voltage is 77 kHz while during no-load operation the burst mode is enabled and the switching frequency can vary in the 115-130 kHz range. The operating switching frequency for different load currents is reported in Figure 32, while Figure 33 shows the LLC converter efficiency in whole operating range at nominal input voltage.
Figure 26. PFC inductor L2 (blue) and L1 (green) current and MOSFET Q1 drain voltage
(purple) at 120 V, 60 Hz input, full load
Figure 27. PFC inductor L1 (green) L2 (blue) and current and MOSFET Q2 drain voltage
(purple) at 120 V, 60 Hz input, full load
Figure 28. PFC inductor L2 (blue) and L1 (green) current and MOSFET Q1 drain voltage
(purple) at 230 V, 50 Hz input, full load
Figure 29. PFC inductor L1 (green) L2 (blue) and current and MOSFET Q2 drain voltage
(purple) at 230 V, 50 Hz input, full load
Experimental characterization AN4468
34/46 DocID026160 Rev 1
The impact on the power supply regulated 12 V output voltage when the load is increased from 10% to 90% of the nominal value is shown in Figure 34, Figure 35 and Figure 36. In every case the output voltage is tightly regulated and the steady state operation is restored in less than 5 ms.
Figure 30. LLC converter, MOSFET Q12 Gate signal (yellow), drain current (blue), drain voltage (purple) at full load (42 A output)
Figure 31. LLC converter, MOSFET Q12 Gate signal (yellow), drain current (blue), drain
voltage (purple) at no load
Figure 32. LLC switching frequency vs. output current
Figure 33. LLC converter efficiency
0
20
40
60
80
100
120
0 10 20 30 40 50
[KHz]
Output Current [A]
LCC Converter: Switching Frequency Vs Ouptut Current
70
75
80
85
90
95
100
0 200 400 600
Efficiency
[%]
Output Power [W]
LLC Converter Efficiency
DocID026160 Rev 1 35/46
AN4468 Experimental characterization
46
The typical converter waveforms at startup are shown in Figure 37. The bus DC voltage is initially regulated at 400 V. As soon as the output current has reached the steady state value, the bus DC voltage is increased to 430 V.
Figure 34. Load transition from 10% to 90% max load (max load=42 A) with 120 V, 60 Hz AC input
Figure 35. Load transition from 90% to 10% max load (max load=42 A) with 120 V, 60 Hz AC input
Figure 36. Repetitive dynamic load variation 10% to 90% max load
Figure 37. Startup at full load (42 A), 230 V, 50 Hz input voltage; PFC bus voltage (yellow), 12 V output (purple); output current (green and
blue)
Experimental characterization AN4468
36/46 DocID026160 Rev 1
Finally, Figure 38 and Figure 39 show the input current waveforms at 120 V and 230 V AC input at full load.
Figure 38. Input current waveform at full load, 120 V AC input
Figure 39. Input current waveform at full load, 230 V AC input
DocID026160 Rev 1 37/46
AN4468 Schematic diagrams
46
Appendix A Schematic diagrams
Figure 40. Schematic diagram (1 of 6)
BBRR CCOONN
TT
DDRRAAII
NN
SSOOUURR
CCEE
CCOONNTT
RROO
LL
VDD
FB 42
8 1
35
7
U2
VIP
ER
27H
1 2
4 3
OP
TO1
PC
817
D
P1 1 2
Hea
der2
6V V
iper
-OU
TPF
C
+3.3
Vpr
i
GN
D p
ri
+Vcc
GN
D se
c
GN
D2
GN
D2G
ND
2
GN
D2
R2
3.3
R11
33
KR
120
R13
47
KR
14
4.3K
R10
16
0K
R9
16K
R8
1.8K
R7
nc
R1
1R
5
2.4K
D3
STTH
1L06
D2
1.5K
E250
97
24 15 T
R1
86
WU
RTH
ELEK
TRO
NIK
7508
7101
2
A 51K3
2 NCNCVREF
4
TS43
1ILT
U
3
D1
STPS
2L40
C10
ncC
11
22nF
GND 1Vin3
Vou
t2
U4
LD11
17
GNDVin 13
Vou
t2 V
out
4
U1
LD11
17D
4
1N41
48W
1469
425
100u
F16
V17
8061
5 C
6
C8
100u
F25
V17
8061
5
C13
100u
F16
V17
8061
5
+ O
UT
PFC
C9
120n
F
C7
82nF
D6
BAR
46D
5 BA
R46
+3.3
Vse
cond
ary
C2
470uF16VYXF
C4
22uF
35V
189
9657
C5
1uF
Schematic diagrams AN4468
38/46 DocID026160 Rev 1
Figure 41. Schematic diagram (2 of 6)
F1 FUS
E
Q1
STP
57N
65M
5
Q2 STP
57N
65M
5
G_P
FC1
-OU
TPF
C
+ O
UT
PFC
PFC
VA
C
PFCCS1+3
.3 P
rimar
y
GN
D1
GN
D1
GN
D1
+3.3
Vpr
i
+3.3
Prim
ary
GN
D p
ri
PFC
CS2
R39
1K
R41 1K
R38
R32
10K
10K
R40
2000
8n2 R43
5K1
R21
6.
8
R15
22
GN
D1
GN
D1
GN
D1
13
24
MU
RA
TA53
020C
(165
6055
) T5
13
24
T4 MU
RA
TA53
020C
(165
6055
)
C22
1nF
C15
4n
7 Y
2 11
6648
6
4n7
Y2
1166
486
C18
3 2 1
Hea
der 3
JP1
31
D7
STPS
C10
06D
31
D8
STPS
C10
06D
L1
YU
JIN
G Q
P-39
30H
115
uHL2
YU
JIN
G Q
P-39
30H
115
uH
G_P
FC2
R16
22
C20
10
0nF
R22
6.
8
D11
BAR
43AS
not
conn
ecte
d
C83
nc
GN
D1
+3.3
Prim
ary
C65
1 2
D9
1N40
07
C35
GN
D1
GN
D1
GN
D1
EAR
TH
C1
100p
F19
0223
8
C46
100p
F19
0223
8
EAR
TH
R18
nc
GN
D1
R50
nc
GN
D1
GN
D1
R3
R
4
R6
470K9241108
C53
560nF1890496
C64
560nF1890496
C16
560nF1890496
C12
560nF1890496
C17
C19
STPS
1L30
990
7491
D
14
C24
1uF
1781
894
12
D10
1N
4007
12
D13
1N
4007
12
D12
1N
4007
C14
C36
C34
C25
C23
R49
0
R51
0
4
V- 2
INN++
1
INN--
3
V+5
OU
T
U6
TSV
611
470K
9241
108
470K9241108
R27
470K
9241
108
R31
470K
9241
108
R36
470K
9241
108
R26 470K
9241
108
R30 470K
9241
108
R34 470K
9241
108
C76
68
nF
GN
D1
C71
68nF
GN
D1
STPS
1L30
990
7491
D
19
1 342
L6
EAR
TH
C21
2.
2uF
100µF 450V 1166652
100µF 450V 1166652
100µF 450V 1166652
100µF 450V 1166652
560nF 1890496
560nF 1890496
560nF 1890496
560nF 1890496
DocID026160 Rev 1 39/46
AN4468 Schematic diagrams
46
Figure 42. Schematic diagram (3 of 6)
ENB2
O
UT1
V
CC
O
UT2
8 7 6 5
ENB1
PW
M1
GN
D
PWM
2
1 2 3 4
IC1
PM88
34
VC
C3
GN
D4
HIN
2
LIN
1
BOO
T8
HV
G7
OU
T6
LVG
5
U7
L638
7E(n
otm
ount
ed)
Brid
geH
S_P
Brid
geLS
_P
GD
_1
OU
T
GD
_2
GD
R_1
PFC
PWM
1
PFC
PWM
2
LLC
Driver
section
-Primary
Rectification
-Secondary
LOWSIDEMOSFETDRIVER
PFC
Driversection
SR2
GN
Dse
c
GN
D2
+VccEN
DR
PFC
GN
Dpr
i
+Vcc
R55
82
R61
330
R53
360
R57
360
C31
100p
F(n
otm
ount
ed)
C30
100n
F C
73
D15
BAT7
54
D16
BAT7
54
ENB2
O
UT1
V
CC
O
UT2
8 7 6 5
ENB1
PW
M1
GN
D
PWM
2
1 2 3 4
IC2
PM88
34
GD
R_2
LOWSIDEMOSFETDRIVER
SR1
GN
D2
12V
ENSY
NC
REC
T
R60
330
R11
80
R11
70
G_P
FC1
G_P
FC2
C32
100p
F(n
otm
ount
ed)
C28
4u7
(not
mou
nted
)C
29
100n
F(n
otm
ount
ed)
C33
100n
F
D17
BAT7
54
6VV
iper
D18
BAT7
54
1uF
C72
1uF
(not
mou
nted
)
C86
1uF
U8 LI
N
SD/O
D
HIN
V
CC
D
T SG
ND
PG
ND
L649
1D
1 2 3 4 5 6 7
CP+
CP-
LVG
10 9 8
BOO
TH
VG
OU
T
14 13 12B
ridge
HS_
P
Brid
geLS
_P
OU
T
+Vcc
C85
10
0pF
C82
10
0pF
C77
4u7
C79
10
0nF
C78
1uF
R23
0
+Vcc
C81
nc
R20
10K
1KR
25
SDL6
491D
R54
3.3
R58
3.3
R59
82
R62
10
K
R63
10K
GN
D2
R64
10K
Q5
PMB
F170
1081
473
R65
nc
R67
330
R68
nc
R66
30
0+3
.3V
pri
+3.3
Prim
ary
+3.3
Prim
ary
+Vcc
Schematic diagrams AN4468
40/46 DocID026160 Rev 1
Figure 43. Schematic diagram (4 of 6)
12
4 5 6
3
U9
AC
PL-P
484
2070
254
12
4 5 6
3
AC
PL-P
484
2070
254
U
10
12V
GD
_1 GD
_2
GD
R_2
GD
R_1
USA
RT1
RX
Prim
ary
R10
1
USA
RT1
TX
Prim
ary
3.3V
pri
USA
RT1
TX
Sec
onda
ry
39/2
00fo
r 357
NT
nc/8
20fo
r 357
NT
USA
RT1
R
X S
econ
dary
GN
D2
GN
D2
GN
D2
GN
D2
GN
D2
+3.3
Vpr
i
GN
D p
ri
3.3V
pri
GN
D2
GN
D se
c
3.3V
sec
+Vcc +V
cc
Brid
geH
S_P
Brid
geLS
_P
Brid
geH
S_P_
MC
U
Brid
geLS
_P_M
CU
Brid
gecu
rren
t
OU
T
+ O
UT
PFC
-OU
TPF
C
VD
rain
1
1234G
ND
5
Vo
6 7
Vcc
U11
8
PS98
21/K
B357
NT
PS98
21/K
B357
NT
1 2 3 4G
ND
5
Vo
67
Vcc
8
U12
GN
D2
R77
30
0
R80
300
R79
300
R82
0
R83
R81
0 R84 39
/200
for 3
57N
TR
860
R74 R52
R56
20 20 20
R85
0
C4551nF1kVwithMAGNETICA1892.0039
C84
C48 100n
F
C39
1500uF16V1551414
D23
1N4148W1469425MMSD914
D20
1N4148W1469425MMSD914
C80
4n7
4kV
GN
D2
C47
100n
F C49
10
0nF
C50
100n
F
C43
220pF1216461
C44
10uF
2112
705
Cap
LLC
STP1
3NM
60N
D
Q3
Q4
nc
Q12
3.3V
sec
+3.3
Vse
cond
ary
+12V
30nF1kVwithYUJINGLP-4925
T2
Tran
s CT
1:1+
1EF
D15
119
0583
Prim
ary:
1 tu
rn,A
WG
#29
0.29
mm
12w
ires
GN
D2
Seco
ndar
y:2
turn
s (1
+1 in
serie
s w
ithm
idta
p),A
WG
#25
0.45
mm
1w
ireD
33
1N41
48W
1469
425
D36
1N41
48W
1469
425
R33
39
GN
D2
R37
1K
C89
22
pF
BAT7
54 R
29 1KC87
3uF
GN
D2
curr
ent_
AV
G
curr
ent_
cycle
D37
BA
T754
3.3V
sec
C40
1500uF16V1551414
C41
1500uF16V1551414
C42
1500uF16V1551414
C90
10uFX7R2112722
C88
10
nF
STL1
60N
3LLH
6 Q
8
STL1
60N
3LLH
6 Q
6Q
7
STL160N3LLH6
Q10
STL160N3LLH6
nc
Q11
STP1
3NM
60N
D
R42
39R
44
47R35
47
CT
3
17 15 11TR
AN
SFO
RM
ERET
D44
14
T1
12V
CT
3
C91
10uFX7R2112722
C92
10uFX7R2112722
C93
10uFX7R2112722
9 7 5 3 1
10 8 6 4 2
P6
14
1312
11
Hea
der7
X2H
155
1388
P7 1 2 3
BLD
CFA
N-P
Cty
pe
GN
D2
D28
3.3V
sec
Q9
PMBF
170
1081
473
GN
D2R94
100
(163
2521
)
R97
Disc
harg
e
R10
00
R10
4 0
nc/8
20fo
r 357
NT
R10
3
19 17 12TR
AN
SFO
RM
ERET
D49
15
T6
Q15
ST
N4N
F03L
R11
1 330
GN
D2
FAN
switc
h
STH
320N
4F6-
6 Q
17
STH
320N
4F6-
6Q
16
VD
rain
1
VD
rain
2V
Dra
in2
VD
rain
2
VD
rain
2
R11
5
0 (n
c)G
ND
2
6
TRA
NSF
OR
MER
LP45
25H
15 13 1012
T10
D31
1N41
48W
1469
425
D40
STPS3045DJFnotmounted
D39
STPS3045DJFnotmounted
17 15 1114
T3 MA
GN
ETIC
A 1
892.
0039
5 3 5 4 7
Cap
LLC
300
DocID026160 Rev 1 41/46
AN4468 Schematic diagrams
46
Figure 44. Schematic diagram (5 of 6)
VD
DPF
0/O
SC_i
n PF
1/O
SC_o
ut
NR
STV
DD
a PA
0 PA
1 PA
2
+3.3
Prim
ary 1 2 3 4 5 6 7 8
PA3PA4PA5PA6PA7PB0PB1PB2/BOOT1
910111213141516
PA14
PA
13
PA12
PA
11
PA10
PA
9 PA
8 V
DD
24 23 22 21 20 19 18 17
PB8BOOT0
PB7PB6PB5PB4PB3
PA15
3231302928272625
U13
STM
32F0
51K
xU6
PFC
CS1
Brid
gecu
rren
t
PFCVAC
USA
RT1
RX
Prim
ary
USA
RT1
TXPr
imar
y
+3.3
Vpr
i
+3.3
Prim
ary
+3.3
Prim
ary
C57
100n
F
+3.3
Prim
ary
+3.3
Prim
ary
END
RPF
C
GN
Dpr
i
R92
+3.3
Prim
ary
R45
470
+3.3
Prim
ary
T9te
stTe
st p
oint
1P4
1H
eade
r4
2 3 4
test 1T8 Test
poi
nt
+O
UT
PFC
PFC
CS2
R93
620
R96
3K
6
R91
18
0K
C51
100n
F
R87
10
K
D24
LED
1058
385
C56
1uF
C55
10
0nF
C58
10
nF
L4
D25
BA
T754
PFC
PWM
1
PFC
PWM
2
180K
R95
18
0K
+3.3
Prim
ary
D38
BAT7
54
C94
47
pF
C95
47
pFI2C1_SCL
I2C1_SDA
P3 1 2 3 4
Hea
der4
SDL6
491D
P2 DA
Cou
t
GN
D1
+3.3
Prim
ary
1
C67
10
0nF
U5
LM19
CIZ
+VS
VO
G
ND
2
3C
27
100n
F
R17
470
R28
470
C52
10nF
C68
22nF
GN
D1
D21
BA
T754
D22
BA
T754
R10
8
330
R10
9
330
+3.3
Prim
ary
+3.3
Prim
ary
Schematic diagrams AN4468
42/46 DocID026160 Rev 1
Figure 45. Schematic diagram (6 of 6)
GN
D2
GN
D2
GN
D2
GN
D2
GN
D2
GN
D2
10K
R99
GN
D2
GN
Dse
c
GN
D2
+3.3
Vse
cond
ary
VC
C
VC
C
VC
C
VC
C
VC
C
VC
C
GN
D2
USA
RT1
RXSe
cond
ary
USA
RT1
TXSe
cond
ary
12V
ENSY
NC
REC
T
VC
C
GN
D2
6VViper
P5 1 2 3 4
Hea
der4
test 1T12
T13
Testpoint
test 1Testpoint
L5
GN
D2
GN
D2
VCC
Vbbaa
tt PCC
1133PCC
1144//OO
SSCC3322
iinnPCC
1155//OO
SSCC3322
oouutt
PDD00//
OOSSCC
iinn
PDD11//
OOSSCC
oouutt
NRR
SSTTV
SSSSa
VD
Da
PA0
PA1
PA2
1 2 3 4 5 6 7 8 9 10 11 12
PA3PA4PA5PA6PA7PB0PB1PB2PB10PB11VSS_2VDD_2
131415161718192021222324
VD
D_3
V
SS_3
PA
13
PA12
PA
11
PA10
PA
9 PA
8 PB
15
PB14
PB
13
PB12
36 35 34 33 32 31 30 29 28 27 26 25
VDD_1VSS_1
PB9PB8
BOOT0PB7PB6PB5PB4PB3
PA15PA14
484746454443424140393837
U14
STM
32F3
34
VC
C
GN
D2
GN
D2
GN
D2
VC
C
VC
C
GN
D2
GN
D2
R98
200
R10
70
R11
40
R11
2 36
0
R10
6 36
0D
30
R11
02k R
113
2k
R10
5 1K
0.1%
12
Y2
8Mhz
HC
49/U
S
D26
LED
1058
385
C62
20pF
D27
BA
T754
D32
BA
T754
R10
2
4K3
0.1%
C61
20pF
C63
1uF
C3
100n
F
C66
1nF
C69
10
0nF
C75
47
pF
C70
47
pF
C60
100n
F
C59
10
0nF
C74
10
0nF
D29
BAR
43S
D35
BAR
43S
BAT7
54
D34 BA
T754 V
Dra
in1
VC
C
VC
C
SRle
g1
R69
0R
46 0
R47
0
GN
D2
C37
10
0nF
GN
D2
C38
10
0nF
GN
D2R
24
470
R19
470
Tem
pSR
out
P8 1 2 3 4
Hea
der4
GN
D2
C54
10n
F
TempSRoutcurrent_cycle
R70
0
R48 0
R71 0
R72
0
SRle
g2
R73
0
R76
0R
78 0
R88
0R
89 0
R90
0
R75
0
SR1 Brid
geH
S_P_
MC
U
Brid
geLS
_P_M
CU
SR2
current_AVG
DischargeV
Dra
in2
VC
C 1
123
P9I2
C
GN
D2
FAN
switc
h
U15
LM
19C
IZ+V
SV
O
GN
D
2
3
C98 n c
(47
to10
0nF )
GN
D2
tom
ain
GN
D2
DocID026160 Rev 1 43/46
AN4468 Conclusion
46
5 Conclusion
This application note describes the design of a 500 W fully digital switch mode power supply. The system is based on a two-stage architecture consisting of a semi-bridgeless PFC circuit and an LLC half-bridge DC-DC stage. The control architecture is designed around two 32-bit MCUs from the STM32 family of microcontrollers. The control algorithm for both the PFC and the LLC converters is highlighted throughout the document. The experimental evaluation shows that high efficiency (93.2%), near unity power factor, and low THD% can be achieved under wide input voltage and load current conditions thanks to the performance of the ST products used and to the implementation of suitable control strategies on high-performance 32-bit microcontrollers.
References AN4468
44/46 DocID026160 Rev 1
6 References
• AN2450: LLC resonant half-bridge converter design guideline
• High frequency transformer for 500 W LLC converter: Magnetica 1892.0039
• Inductor chokes for 500 W bridgeless PFC: Yujing Technologies 11999-126H400210 QP-3930H
DocID026160 Rev 1 45/46
AN4468 Revision history
46
7 Revision history
Table 12. Document revision history
Date Revision Changes
19-Jun-2014 1 Initial release.
AN4468
46/46 DocID026160 Rev 1
Please Read Carefully:
Information in this document is provided solely in connection with ST products. STMicroelectronics NV and its subsidiaries (“ST”) reserve theright to make changes, corrections, modifications or improvements, to this document, and the products and services described herein at anytime, without notice.
All ST products are sold pursuant to ST’s terms and conditions of sale.
Purchasers are solely responsible for the choice, selection and use of the ST products and services described herein, and ST assumes noliability whatsoever relating to the choice, selection or use of the ST products and services described herein.
No license, express or implied, by estoppel or otherwise, to any intellectual property rights is granted under this document. If any part of thisdocument refers to any third party products or services it shall not be deemed a license grant by ST for the use of such third party productsor services, or any intellectual property contained therein or considered as a warranty covering the use in any manner whatsoever of suchthird party products or services or any intellectual property contained therein.
UNLESS OTHERWISE SET FORTH IN ST’S TERMS AND CONDITIONS OF SALE ST DISCLAIMS ANY EXPRESS OR IMPLIEDWARRANTY WITH RESPECT TO THE USE AND/OR SALE OF ST PRODUCTS INCLUDING WITHOUT LIMITATION IMPLIEDWARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE (AND THEIR EQUIVALENTS UNDER THE LAWSOF ANY JURISDICTION), OR INFRINGEMENT OF ANY PATENT, COPYRIGHT OR OTHER INTELLECTUAL PROPERTY RIGHT.
ST PRODUCTS ARE NOT DESIGNED OR AUTHORIZED FOR USE IN: (A) SAFETY CRITICAL APPLICATIONS SUCH AS LIFESUPPORTING, ACTIVE IMPLANTED DEVICES OR SYSTEMS WITH PRODUCT FUNCTIONAL SAFETY REQUIREMENTS; (B)AERONAUTIC APPLICATIONS; (C) AUTOMOTIVE APPLICATIONS OR ENVIRONMENTS, AND/OR (D) AEROSPACE APPLICATIONSOR ENVIRONMENTS. WHERE ST PRODUCTS ARE NOT DESIGNED FOR SUCH USE, THE PURCHASER SHALL USE PRODUCTS ATPURCHASER’S SOLE RISK, EVEN IF ST HAS BEEN INFORMED IN WRITING OF SUCH USAGE, UNLESS A PRODUCT ISEXPRESSLY DESIGNATED BY ST AS BEING INTENDED FOR “AUTOMOTIVE, AUTOMOTIVE SAFETY OR MEDICAL” INDUSTRYDOMAINS ACCORDING TO ST PRODUCT DESIGN SPECIFICATIONS. PRODUCTS FORMALLY ESCC, QML OR JAN QUALIFIED AREDEEMED SUITABLE FOR USE IN AEROSPACE BY THE CORRESPONDING GOVERNMENTAL AGENCY.
Resale of ST products with provisions different from the statements and/or technical features set forth in this document shall immediately voidany warranty granted by ST for the ST product or service described herein and shall not create or extend in any manner whatsoever, anyliability of ST.
ST and the ST logo are trademarks or registered trademarks of ST in various countries.Information in this document supersedes and replaces all information previously supplied.
The ST logo is a registered trademark of STMicroelectronics. All other names are the property of their respective owners.
© 2014 STMicroelectronics - All rights reserved
STMicroelectronics group of companies
Australia - Belgium - Brazil - Canada - China - Czech Republic - Finland - France - Germany - Hong Kong - India - Israel - Italy - Japan - Malaysia - Malta - Morocco - Philippines - Singapore - Spain - Sweden - Switzerland - United Kingdom - United States of America
www.st.com