1. I NTRODUCTION
1.6 T HESIS O RGANIZATION
In this thesis, we introduce the basics terms of the switched mode DC-DC power supplies in chapter 2. Some fundamental topologies of a switched mode DC-DC power supply like buck, boost and buck-boost are introduced. The modulation schemes of SMPS are briefly listed. The specifications of SMPS give us an evaluation standard of the performance of the SMPS.
In chapter 3 we demonstrate the advanced control and protection circuits in a monolithic current-mode buck converter. The newly developed strategies and circuits are discussed in this chapter. All of them can be easily adopted for integrated SMPS in different topologies and for different applications. The on-chip soft-start circuit, the dynamic partial shutdown strategy, the current sense circuit, the slope compensation
circuit and the over-current protection circuit are all realized in this monolithic converter. The performances are measured to show the practicability and effectiveness of them. We can see a compact size, high efficiency and well-protected converter illustrated in this chapter.
Chapter 4 presents the practical considerations in circuit design and silicon layout of a power converter integrated in single chip. Many practical issues such as switching noise, on-resistance of integrated power switches and some other protection considerations are discussed. These issues seem unimportant, but they may cause unstable operation, performance degradation and even destroy the converter or the system. In this chapter, we deal with these issues and give practical suggestions according real experience.
Chapter 5 concludes this work and tries to suggest for future works.
Chapter 2
S WITCHED M ODE DC-DC P OWER
S UPPLY B ASICS
2.1 R
EGULATEDDC P
OWERS
UPPLIESThere are two basic ways of maintaining a fixed output voltage in a DC power supply, i.e. by series and shunt regulators [5].
The principle of the series regulator is illustrated in Fig. 2.1 (a). Here you have an unregulated DC input, VIN, feeding through a series-control circuit, A, to give a fixed regulated DC output, VOUT. The series-control circuit works by adjusting the voltage drop from input to output to keep a constant voltage at the output. This variable series voltage drop takes up both variations in the input voltage, and varying voltage drops inside the system arising from changing load currents. The variable resistor symbol in the circuit block, A, emphasizes that the series regulator works like a variable resistor in series with the load, adjusting itself to keep the output voltage across the load constant.
The shunt regulator works on a different principle, illustrated in Fig. 2.1 (b).
Here an unregulated DC input feeds a current through a series resistor, RS, to give a lower DC voltage across the load resistor RL. The shunt-control element, B, draws current from the output rail, as does, of course, the load RL. The shunt-control element adjusts the current it takes so that the voltage drop across the series resistor RS varies
to keep the output rail voltage, VOUT, unvarying under changes of load resistance and unregulated input voltage, VIN. Once again the control element can be seen to work like a varying resistor as indicated by the symbol in the shunt circuit block, B, in the diagram.
Fig. 2.1 (a) Series regulated power supply.
Fig. 2.1 (b) Shunt regulated power supply.
In essence, the series regulator soaks up changes in the voltage drop from the input supply onward to the output, while the shunt regulator soaks up current variations in the load. In each case the net result is the same: the output voltage is regulated, i.e. held constant, at a fixed value.
In nowadays, there are mainly two types of regulated DC-DC power supplies.
One of them is linear regulated power supply and another is switched mode power supply.
2.1.1 Linear Regulated Power Supplies
Linear regulated power supplies is a simple, widely used DC-DC regulator. In linear regulators, the controlled resistor (series or shunt) elements described above continuously dissipate some power in keeping the output voltage constant under varying input voltage and output load conditions. Fig. 2.2 (a) and (b) show once again series and shunt regulator systems, but this time they emphasize that the control elements are linear amplifiers continuously dissipating power to carry out their functions.
Fig. 2.2 (a) Principle of linear series regulator.
The linear regulators have many desirable characteristics such as low output ripple, good line and load regulation, fast transient response and low electromagnetic interference (EMI). How low efficiency limits their application. When it comes to high volt-amp requirements or transformerless high-step-down voltage between input
and output, the major problem of design in linear regulators becomes largely a mechanical one of providing an adequate heat sink for the dissipating semiconductor devices.
Fig. 2.2 (b) Principle of linear shunt regulator.
2.1.2 Switched Mode Power Supplies
The switched mode power supply gets round the low efficiency of the linear regulator by using controlled on-off switching of the power supplied to the load to keep the output voltage constant.
In series form of the switching regulator illustrated in Fig. 2.2 (c), the series on-off switching control element, SWA, inserted between the supply input and output is switched on and off by a controlled on-off duty-cycle generator circuit X. As a result, current from the input is released in pulses, which, after smoothing, provide a controlled DC voltage output level. The switching duty cycle of X adjusts itself so that the mean output voltage remains constant irrespective of input voltage or load current variations.
Fig. 2.2 (c) Principle of switching series regulator.
The same principle of rapid on-off “chopping” of the supply also appears in the shunt form of switching regulator shown in Fig. 2.2 (d). Here the controlled on-off duty cycle generator circuit Y switches on and off the shunt switching circuit SWB to bleed off current from the supply away from the load. Once again the chopped DC is smoothed and fed to the load. The duty cycle control of Y is such that the output voltage is held constant under varying input voltages and load currents.
Fig. 2.2 (d) Principle of switching shunt regulator.
The smoothing circuit used in the SMPS usually consists inductors and capacitors. The inductors and capacitors can be viewed as energy storage components.
Depending on the arrangement of switches and energy storage components, the output voltage can be generated that is greater than or less than the input voltage.
Switching power supplies can be more compact than equivalent linear ones because of the high efficiency of the switching mode. However, due to the switching operations, large ripples can be observed at the output. Large voltage and current swings also induce EMI issues. These disadvantages make it less favorable for some applications such as communication and audio equipments. Additionally, control circuit for SMPS is usually more complex than linear ones. But the transient response is slower due to the limit of the switching frequency. Table 2.1 lists some typical values of linear and switching regulators [12], [38].
2.2 B
ASICC
ONVERTERT
OPOLOGIESA major decision that must be considered at the beginning of a SMPS design is which basic topology to use. The term topology refers to the arrangement of the power components within the SMPS design. There are more than ten different topologies can be used in DC-DC conversion [1], [12], [39]. Here we limit our introduction to three basic non-isolated topologies of DC-DC SMPS: buck, boost and buck-boost.
2.2.1 Buck
A more detailed discussion of the buck regulator as opposed to the other topologies is presented due to popularity of the buck regulator.
Fig. 2.3 Buck converter topology and related waveforms.
For the buck converter of Fig. 2.3, the output voltage VOUT is less than the input voltage VIN, hence the name buck. When the switch is closed, input current flows through the filter inductor, the filter capacitor, and the load. When the switch is opened, the voltage across the inductor reverses since VL becomes a voltage source (VL = L × di/dt), and the energy stored in the inductor is delivered to the load. Since the current in the inductor cannot change instantaneously, the current flowing through the switch at the time the switch is opened now flows through the inductor, the capacitor, the load, and the diode. When the switch is again closed, the current, which
The average output voltage is
VOUT = VIN × D (2.1)
Where D is the duty cycle. The duty cycle is the ratio of the switch on-time to the period T. Since
VIN × IIN = VOUT × IOUT (2.2)
The average input current is
IIN,AVG = IOUT × D (2.3)
We can see that when D = 100 %, VOUT = VIN and IOUT = IIN. Conversely, when the duty cycle approaches 0 %, the output voltage becomes very small and the peak input current becomes very large.
2.2.2 Boost
For the boost converter of Fig. 2.4, the output voltage is greater than the input voltage, hence the name boost. With input voltage applied, the current flows through the inductor, the diode, the capacitor and the load. When the switch is closed, the current flows through the inductor and switch and in effect, the voltage across the inductor is the input voltage. When the switch is opened, the induced reverse voltage in the inductor is then in series-adding with the input voltage to increase the output voltage, and the current which was flowing through the switch now flows through the inductor, the diode, the capacitor, and the load. The energy stored in the inductor is transferred to the load. When the switch is again closed, the diode becomes reverse
biased, the energy in the capacitor supplies the load voltage, and the cycle repeats.
Fig. 2.4 Boost converter topology and related waveforms.
The average output voltage is
VOUT = VIN / (1-D) (2.4)
Where D is the duty cycle. The duty cycle is the ratio of the switch on-time to the period T. Since
× I × I (2.5)
The average input current is
IIN,AVG = IOUT / (1-D) (2.6)
We can see that when D = 0 %, VOUT = VIN and IOUT = IIN. Conversely, when the duty cycle approaches 100 %, the output voltage does not necessarily approach infinity because the conducting operation of the semiconductor switch produces a peak current which will quickly exceed the safe operating area (SOA) limit. In other words, we can say that the conduction loss will limits the output voltage when D approaches 100 %.
2.2.3 Buck-Boost
The buck-boost converter shown in Fig. 2.5 is referred to by many names. The buck-boost terminology will be used since the output voltage may be less than, or greater than, the input voltage. The converter is sometimes referred to as a flyback converter. The flyback designation is appropriate due to the inherent action of the inductor. This action in itself is sometimes referred to as a ringing-choke regulator.
Also, the topology is sometimes referred to as an inverting regulator, since the output voltage polarity is opposite the input voltage polarity.
When the switch is closed, the current flows through the inductor since the diode is reverse biased. When the switch is opened, the current, which was flowing in the switch, now flows through the inductor, the diode, the capacitor and the load. The energy stored in the inductor is transferred to the load. When the switch is again closed, the current, which was flowing through the diode, now flows through the switch, and the diode becomes reverse biased.
Fig. 2.5 Buck-boost converter topology and related waveforms.
The average output voltage is
-VOUT = VIN × D / (1-D) (2.7)
Where D is the duty cycle. The duty cycle is the ratio of the switch on-time to the period T. Since
VIN × IIN = VOUT × IOUT (2.8)
The average input current is
IIN,AVG = IOUT × D / (1-D) (2.9)
We can see that when D equals 50%, -VOUT = VIN and –IOUT = IIN.
2.2.4 Synchronous Rectification
In the above topologies, we can see a master switch and a diode served as a slave switch. We can use a controlled switch (synchronous switch) instead of a diode in the above topologies, i.e. synchronous rectification. Synchronous rectification is used in DC-DC converters when low output voltage and high current is needed. Synchronous rectification utilizes power MOSFETs instead of rectifying diodes. These MOSFETs are synchronized to the converter frequency and perform more efficiently the rectification of the output voltage than rectifying diodes due to the low I × R drop through the channel. [40]-[41]
2.3 M
ODULATIONT
ECHNIQUESIn this section, we briefly discuss about three modulation techniques.
2.3.1 Pulse Width Modulation (PWM)
Fig. 2.6 Switching signal of pulse width modulation (PWM).
The pulse width modulation (PWM) maintains a constant switching frequency and varies the duty ratio according to output voltage and load current. Take voltage mode control for example. The switching signal of PWM is shown in Fig. 2.6. The error voltage is modulated by a saw-tooth waveform generated by an oscillator. The switching frequency is determined by the oscillator and the pulse width is controlled by the voltage error. This modulation scheme provides high efficiency at medium to heavy load conditions. At light load condition, the reversed inductor current (may appear in synchronous rectification) and the switching losses degrade the efficiency.
Because the switching frequency of PWM is fixed, the noise spectrum is relatively narrow. We can use simple low-pass filter to greatly reduce the peak-to-peak voltage ripple. For this reason, PWM is popular to the noise sensitive communication applications.
2.3.2 Pulse Skip Modulation (PSM)
Fig. 2.7 Output waveform and switching signal of pulse skip modulation (PSM).
A feature offered in many modern switching controllers is pulse skip modulation (PSM). Skip mode allows the regulator to skip cycles when they are not needed,
system, there exists an oscillator as a timing reference. Each pulse of the switching signal starts with the clock signal. As long as the converter output is below the reference voltage, the PSM pulses continue to run the converter switch. Once the converter output reaches or exceeds the target, the PSM pulse is skipped. This operation will result in decreasing pulse density as the converter output reaches its target, or as the output loading decreases. When the converter output falls below the target, or as the output loading increases, the PSM pulse density will increase.
However, the inductor selection is complicated, the peak-to-peak voltage ripple can be quite high, and the noise spectrum will vary greatly with the load.
2.3.3 Pulse Frequency Modulation (PFM)
Fig. 2.8 Output waveform and switching signal of pulse frequency modulation (PFM).
As shown in Fig. 2.8, pulse frequency modulation (PFM) is somewhat different from pulse skip modulation. They all reduce the pulse density in light load conditions and hence the efficiency can be boost under light load operation. The major difference of PFM to PSM is that the PFM does not need a clock signal. The pulse width is variable and the pulse repetition rate is varied in accordance with load current and input/output voltages. As soon as the output voltage reaches the bottom value of regulation, the main switch turns on until the inductor current reaches the peak current limit or until the predetermined time is up. The major drawback of PFM control is still
its varying switching frequency.
2.4 P
ROTECTIONS OFSMPS
In the design of a power supply it is prudent to provide protection circuitry to protect against extreme and abnormal operating conditions that will inevitably occur when the supply is in use. These can occur in the form of output short circuits and excessive loads or high voltage transients on the input supply line. Many of the components in a power supply are handling powers greatly in excess of their dissipation capability. Under fault conditions it is quite possible that they may start to dissipate this power, leading to their rapid failure. The power supply designer has no control over these faults and therefore must incorporate circuitry to accommodate them safely. This falls into five broad categories: (1) over-current protection; (2) over-voltage protection; (3) inrush protection; (4) device protection; (5) over temperature protection. [4], [42]
2.4.1 Over-Current Protection
In order to provide current limiting, some means of sensing over-current conditions must be provided. In an SMPS the commonest method of achieving current limiting is to control the switching activities of the switching transistors. Under fault conditions the transistors can be switched off. Cycle-by-cycle protection is a useful method of output current limiting in an SMPS.
The purpose of current limiting is two-fold: firstly to limit the dissipation in the power supply components to safe values and thereby prevent damage to them, and secondly to provide some protection to circuits and systems being powered by the
supply.
2.4.2 Over-Voltage Protection
Over-voltage protection must deal with three possible situations: (1) reverse voltage on output; (2) external over-voltage on output; (3) internally generated over-voltage. The first two situations are reasonably easily dealt with by placing
“catcher” diodes on the output. Fore reverse polarity protection, a normally reverse biased diode can be placed on the output. Normal over-voltage protection can be provided by a zener or avalanche diode whose voltage is in excess of the normal operating voltage of the power supply. Diodes for both types of protection must by amply rated to cope with the anticipated fault conditions.
It is important that a power supply does not give out an abnormally high voltage under fault conditions. If it did so it could easily damage the circuitry that it is powering. We can monitor the output voltage by an over-voltage control circuit. This is usually some form of comparator that is set to trigger under over-voltage conditions.
Its output is used to stop the switching activity thereby shutting down the power supply; in much the same way as was done for cycle-by-cycle current limiting.
2.4.3 Inrush Protection
In most SMPS designs it is desirable to introduce a certain delay during start-up, in order to avoid inrush current and output overshoots at turn-on. Circuits that are employed to perform this task are called soft-start circuits. In general they control the modulation circuitry to make the output to increase from zero to its operating value very “softly”.
Fig. 2.9 shows how a soft-start circuit may be implemented in a PWM control circuit. At time t = 0, when the power supply is just turned on, capacitor C is discharged and the error amplifier output is held to ground through diode D1, thus inhibiting the comparator output.
Fig. 2.9 A typical soft-start circuit used in a PWM control circuit aids the gradual increase of the PWM signal to its operating value.
At time t = 0+, the capacitor starts to charge through resistor R with a time constant determined by τ = RC toward the charging voltage VSUPPLY. As capacitor C attains full charge, diode D1 is reverse biased, and therefore the output of the error amplifier is isolated from the soft-start network. The slow charge of capacitor C results in the gradual increase of the PWM waveform at the output of the comparator, and consequently a “soft start” of the switching element is initiated.
Diode D2 is used to bypass resistor R in order to discharge the capacitor C fast enough in case of system shutdown, thus initiating a new soft-start cycle even during
Diode D2 is used to bypass resistor R in order to discharge the capacitor C fast enough in case of system shutdown, thus initiating a new soft-start cycle even during