Jotrin Electronics
Beschreibung Menge Insgesamt (USD) Betrieb
Einkaufswagen Produkte
Einkaufswagen Produkte : 0
Startseite > EDA/IC Design > The working principle and proper use of capacitors in power supply des

The working principle and proper use of capacitors in power supply design

Updatezeit: 2022-07-13 18:07:31

The power supply is the most neglected aspect of the circuit design process. As a good design, the power supply design should be very important, and it largely affects the performance and cost of the whole system. The use of capacitors in power supply design is again often the most overlooked area of power supply design.

1. The working principle of capacitors in power supply design

In power supply design applications, capacitors are mainly used for filter and decoupling/bypass. Filtering is the operation of filtering out a specific band of frequencies in a signal, an important measure to suppress and prevent interference. The probability theory and method of estimating another random process related to it are based on the results of observing a random process. The term filtering originated in communications theory as a technique to extract a useful signal from a received signal that contains interference. The "received signal" is equivalent to the observed random process, and the "useful signal" is equivalent to the estimated random process.

Filtering mainly refers to external filtering noise, while decoupling reduces the local circuit noise interference to the outside. Many people tend to confuse the two. Let's look at a circuit structure.


The figure shows the power supply for A and B. When A needs a large current at a certain moment, if there is no C2 and C3, the voltage at the A end will become lower because of the line inductance. The voltage at the B end is also affected by the voltage at the A end and lower, so the local circuit A current changes caused by the local circuit B supply voltage, thus affecting the signal of the B circuit. The current change of local circuit A causes the supply voltage of local circuit B, thus affecting the signal of circuit B. Similarly, the current change of B will also form interference with A. This is "common coupling interference."

After adding C2, when the local circuit needs a momentary high current, the capacitor C2 can temporarily provide current for A. Even if the common circuit inductor exists, the voltage at A will not drop too much. The impact on B will also be much reduced. So the role of decoupling is played by the current bypass.

General filtering mainly uses large-capacity capacitors, the speed requirements are not very fast, but the capacitance value requirements are large. If the local circuit A in the diagram refers to a chip, and the capacitor is as close as possible to the power supply pin of the chip. Suppose the "local circuit A" is a function module. In that case, you can use ceramic chip capacitors or tantalum or aluminum electrolytic capacitors if the capacity is insufficient (provided each chip in the function module has decoupling capacitors - ceramic chip capacitors).

The capacity of the filter capacitor can be calculated from the datasheet of the power supply chip. If the filtering circuit uses electrolytic, tantalum, and ceramic chip capacitors, place the electrolytic capacitor closest to the switching power supply to protect the tantalum capacitor. The ceramic chip capacitor is placed behind the tantalum capacitor. This will give the best filtering effect.


The decoupling capacitor needs to meet two requirements, one is the capacity requirement, and the other is the ESR requirement. That means a 0.1uF capacitor decoupling effect may not be as good as two 0.01uF capacitors. Moreover, 0.01uF capacitors have a lower impedance in higher frequency bands. If a 0.01uF capacitor can meet the capacity requirement in these bands, it will have a better decoupling effect than a 0.1uF capacitor.

Many high-speed chip design manuals with more pins will give the power supply design requirements for decoupling capacitors. For example, a 500+ pin BGA package requires at least 30 porcelain chip capacitors for a 3.3V power supply and several large capacitors, with a total capacity of 200uF or more.

2. The correct choice of capacitors in various power supplies

As a basic component in the electronic circuit, the capacitor plays an important role in the traditional application, and capacitors are mainly used for bypass coupling, power supply filtering, isolation, and small signal in the oscillation delay, etc. With the development of electronic circuits, especially power electronic circuits have put forward different special requirements for capacitors in different applications.

The structure of the capacitor is stated. The simplest capacitor comprises pole plates at both ends and an insulating dielectric (including air) in the middle. When energized, the pole plates are charged, creating a voltage (potential difference), but the entire capacitor is non-conductive because of the insulating substance in the middle. However, such a situation is under the condition that the capacitor's critical voltage (breakdown voltage) is not exceeded. As we know, any material is relatively insulated. When the voltage at both ends of the material increases to a certain degree, the material is conductive, and we call this voltage the breakdown voltage.

Capacitors are no exception. Capacitors are not insulators after the breakdown. However, at the secondary school level, such a voltage is not seen in the circuit, so they work below the breakdown voltage and can be treated as insulators. However, in AC circuits, the current's direction changes as a time function. And the process of charging and discharging the capacitor is time-dependent, and at this time, a changing electric field is formed between the pole plates, and this electric field is also a function of time.

Filter capacitor

AC power (frequency or high frequency) needs to be filtered by capacitors after rectification to smooth the output voltage, which requires a large capacitor capacity and generally uses aluminum electrolytic capacitors. The main problem of aluminum electrolytic capacitors is the relationship between temperature and life, which follows the law of 50℃. Therefore, in many cases where high temperature and high reliability are required, long-life electrolytic capacitors (such as 5000h or more, even 105℃, 5000h) should be used. In general, electrolytic capacitors of small size have a relatively short life.

For the DC/DC switching power supply input filter capacitor, since the switching converter draws power from the power supply in the form of a pulse, a large high-frequency current flows in the filter capacitor. When the electrolytic capacitor's equivalent series resistance (ESR) is large, it will generate large losses and cause the electrolytic capacitor to heat up. Low ESR electrolytic capacitors can significantly reduce the heat generated by ripple (especially high-frequency ripple) current.

Electrolytic capacitors used for output rectification of switching power supplies require that their impedance frequency characteristics not show a rising trend at 300kHz or even 500kHz. The common electrolytic capacitor shows a rising trend after 100kHz, and the output filtering effect is relatively poor for switching power supply. In my experiment, I found that the ripple and spike of 4700μF, 16V electrolytic capacitor of CDII type for switching power supply output filtering are not lower than that of 4700μF, 16V high-frequency electrolytic capacitor of CD03HF type, and the temperature rise of the ordinary electrolytic capacitor is relatively high. When the load is abruptly changed, the transient response of ordinary electrolytic capacitors is far inferior to that of high-frequency electrolytic capacitors.

Since aluminum electrolytic capacitors do not work well in the high-frequency band, ceramic or non-inductive film capacitors with good high-frequency characteristics should be used to assist. The main advantages are good high-frequency characteristics, low ESR, such as MMK5 type capacity 1μF capacitor, resonant frequency up to 2MHz or more, the equivalent impedance is less than 0.02Ω, much lower than electrolytic capacitors, and the smaller the capacity, the higher the resonant frequency (up to The smaller the capacity, the higher the resonant frequency (up to 50MHz or more) so that the output frequency response or dynamic response of the power supply will be very good.

How to choose filtering capacitors for switching power supplies

Filter capacitors play a very important role in switching power supplies. How to correctly select the filter capacitor, especially the choice of the output filter capacitor, is a matter of great concern to every engineer.

The common electrolytic capacitor in the 50 Hz frequency circuit has a pulsation voltage frequency of only 100 Hz and a charge/discharge time of milliseconds. To obtain a smaller pulsation factor, the required capacitance is up to hundreds of thousands of microfarads, so the goal of ordinary low-frequency aluminum electrolytic capacitors is to increase the capacitance, and the capacitor's capacitance, loss angle tangent, and leakage current are the main parameters to identify its merits. The output filter electrolytic capacitor in the switching power supply has a sawtooth voltage frequency of tens of thousands or even tens of megahertz. Currently, the capacitance is not the main indicator, but the "impedance-frequency" characteristic is the criterion to measure the superiority of high-frequency aluminum electrolytic capacitors. It is required to have a low equivalent impedance in the operating frequency of the switching power supply and a good filtering effect on the high-frequency spike signal generated by the semiconductor devices.

Many electronic designers are aware of the role of filter capacitors in the power supply, but the filter capacitors used at the output of the switching power supply and the filter capacitors selected in the frequency circuit are not the same; the pulsating voltage frequency is only 100 Hz, the charge and discharge time is milliseconds order of magnitude, to obtain a smaller pulsation coefficient, the required capacity of hundreds of thousands of microfarads, and thus the general low frequency with ordinary aluminum electrolytic capacitors, manufacturing. The goal is to increase the capacitance. Capacitance, loss angle tangent, and leakage current of capacitors are the main parameters to identify their merits.

In the switching power supply as an output filter electrolytic capacitor, the frequency of the sawtooth voltage is up to tens of kHz or even tens of MHz. Its requirements are different from low-frequency applications. The capacity is not the main indicator. The measure of its good or bad is its impedance a frequency characteristics. It is required to have a low impedance in the switching power supply operating band; at the same time, for the internal power supply, due to the semiconductor devices At the same time, the power supply internal, due to the semiconductor devices began to work up to hundreds of kHz of spike noise, but also can have a good filtering effect, the general low frequency with ordinary electrolytic capacitors in about 10 kHz, its impedance will begin to show inductive, can not meet the requirements of switching power supply.

Ordinary low-frequency electrolytic capacitors start to show inductance around 10,000 Hz and cannot meet the switching power supply requirements. The high-frequency aluminum electrolytic capacitor for switching power supply has four terminals, with the positive aluminum terminal leading to the positive side of the capacitor and the negative aluminum terminal leading to the negative side of the capacitor. The current flows from one positive terminal of the four-terminal capacitor, through the inside of the capacitor, and then flows from the other positive terminal to the load; the current returning from the load also flows from one negative terminal of the capacitor and then flows from the other negative terminal to the negative terminal of the power supply.

The high-frequency aluminum electrolytic capacitor for switching power supply has four terminals, the two ends of the positive aluminum piece are led out as the positive terminal of the capacitor, and the two ends of the negative aluminum piece are also led out as the negative terminal. The current from the regulated power supply flows from one positive end of the four-terminal capacitor, passes through the inside of the capacitor, and then flows from the other positive end to the load; the current returning from the load also flows from one negative end of the capacitor and then flows from the other negative end to the negative end of the power supply. Because the four-terminal capacitor has good high-frequency characteristics, it provides an extremely favorable means to reduce the pulsating component of the output voltage and suppress the switching spike noise.

Switching voltage regulator with multifunctional integrated protection: In addition to the most basic voltage stabilization function, the regulator should also have overvoltage protection (more than +10% of the output voltage), Undervoltage protection (less than -10% of the output voltage), phase failure protection, short-circuit overload protection of the most basic protection functions. Sharp pulse suppression (optional): the grid sometimes has a high amplitude. The pulse width is a very narrow, sharp pulse, and it will break through the lower voltage electronic components. The anti-surge components of the regulated power supply can play a good role in suppressing such sharp pulses.

High-frequency aluminum electrolytic capacitors are also available in the multi-cell form, which divides the aluminum foil into shorter sections and connects them in parallel with multiple lead sheets to reduce the resistive component in the capacitive resistance while using low resistivity materials and screws as lead terminals to enhance the capacitor's ability to withstand high currents.

The stacked capacitor is also called a non-inductance capacitor; generally, the core of an electrolytic capacitor is rolled into a cylindrical shape, and the equivalent series inductance is larger; the structure of a stacked capacitor is similar to the book, and the magnetic flux generated by the flowing current is offset by the opposite direction, thus reducing the value of inductance and having more excellent high-frequency characteristics, this capacitor is generally made into a square shape, which is easy to fix and also can appropriately reduce the occupied volume.


Absorption and phase change capacitors

As the rated power of gate-controlled semiconductor devices becomes larger and larger, the switching speed becomes faster and faster, and the rated voltage becomes higher and higher. It is not enough for the capacitor of the buffer circuit to require only sufficient withstand voltage, capacity, and excellent high-frequency characteristics.

In high-power power electronic circuits, as the switching speed of IGBT is less than 1μs, it is normal to require the voltage change rate DV/dt "V/μs" on the capacitor of the absorption circuit, and some require V/μs or even V/μs.

For ordinary capacitors, especially ordinary metalized capacitors DV/dt "100V/μs, special metalized capacitors DV/dt ≤ 200V/μs, special metalized capacitors DV/dt ≤ 1500V/μs for small capacity (less than 10nF), and 600V/μs for larger capacity (less than 0.1μF), it is very difficult to absorb the voltage change rate in such huge, and high repetition rate peak current surge is very difficult to withstand. The phenomenon of damage to power electronic circuits.

At present, the capacitors for absorption circuits, i.e., metal foil electrodes, can withstand large peak current and RMS current shocks, such as smaller capacity (below 10nF) can withstand voltage change rate of 100000V/μs~455000V/μs, 3700A peak current and up to 9A RMS current (such as CDV30FH822J03); larger capacity (greater than 10nF, less than 0.47μF); and larger capacity (greater than 10nF, less than 0.47μF). The larger capacity (greater than 10nF, less than 0.47μF) or larger size can withstand shocks greater than 3400V/μs and 1000A peak current.

It can be seen that although the same non-inductance capacitor, metallization, and metal foil capacitor will have different performances when applied in an absorption circuit, the similar shape but different specifications are not interchangeable here. The size of the capacitor will affect the DV/dt and peak current withstand of the capacitor, generally speaking, the larger the length, the smaller the DV/dt and peak current.

The capacitor in the absorption circuit is characterized by a small duty cycle of high peak current and not very high RMS current. Similar to this circuit, there is a phase change capacitor in the thyristor inverter. Although the required DV/dt of this capacitor is smaller than that of the absorption capacitor, the peak current and RMS current are larger, and the common capacitor cannot meet the requirement in terms of current.

In some special applications, energy storage capacitors are required to be discharged repeatedly and rapidly, and the discharge circuit resistance is meager. The parasitic inductance is very small, in which case the absorber capacitors can only be used in parallel to ensure long-term reliability.

Resonant Capacitors

Resonant converter, such as resonant switching power supply and resonant capacitor in the resonant circuit of thyristor IF power supply, often flows a large current when working. Another example is the electronic ballast resonant capacitor specifications are not properly selected. There will be capacitors on the voltage that did not reach the breakdown voltage due to the flow of a large resonant current and damage to the phenomenon.

In the circuit containing a capacitor and inductor, if the capacitor and inductor are connected in parallel, it may appear in a very small period: the voltage of the capacitor gradually increases while the current gradually decreases; at the same time, the current of the inductor gradually increases, while the voltage of the inductor gradually decreases. At the same time, the current of the inductor gradually decreases, and the voltage of the inductor gradually increases. The voltage increase can reach a positive maximum. The voltage decrease can also reach a negative maximum, and the same direction of the current in the process will also occur in the positive and negative direction of change. At this time, we call the circuit oscillation of electricity.

The circuit oscillation phenomenon may gradually disappear or may continue to be maintained unchanged. When the oscillation is continuously maintained, we call it equal-amplitude oscillation, also known as resonance.

Resonant time capacitor or inductor two forging voltage change a period of time is called the resonant period, the inverse of the resonant period is called the resonant frequency. The so-called resonant frequency is defined in this way.

In summary, in modern power supply technology, different applications require different performance capacitors that can not be mixed, abused, or misused in order to eliminate as much as possible should not appear damaged and to ensure product performance.

3. Example of capacitor buck power supply design

The conventional method of converting AC mains power to low-voltage DC is to use transformers to step down and then rectify and filter. When limited by the size and cost of factors, the simplest and most practical method is to use a capacitor step-down power supply.

Capacitor buck power supply circuit principle

The basic circuit of the capacitor step-down power supply is shown in Figure 1. C1 is the step-down capacitor, D2 is the half-wave rectifier diode, D1 provides a discharge circuit to C1 during the negative half of the mains, D3 is the voltage regulator diode, and R1 is the charge-discharge resistor of C1 after the power is turned off. The circuit shown in Figure 2 is often used in practical applications. When a larger current is required to be supplied to the load, the bridge rectifier circuit shown in Figure 3 can be used. The unregulated DC voltage after rectification is generally higher than 30 volts and fluctuates greatly with changes in load current, which is due to the large internal resistance of this type of power supply and is therefore not suitable for high current supply applications.

Resistive buck circuit device selection principles

(1) When designing the circuit, the exact value of the load current should be measured first, and then the capacity of the bulk capacitor should be selected by referring to the example. The excess current will flow through the regulator, and if the maximum allowable current Imax of the regulator is less than Ic-Io, it is easy to cause the regulator to burn up.

(2) To ensure the reliable operation of C1, the withstand voltage should be more than twice the supply voltage.

(3) The selection of the drain resistor R1 must ensure that the charge on C1 is discharged within the required time.

Design example

In Figure 2, C1 is known to be 0.33μF, the AC input is 220V/50Hz, and the maximum current that the circuit can supply to the load.

The capacitive resistance Xc of C1 in the circuit is: Xc = 1 / (2 πf C) = 1/ (2*3.14*50*0.33*10-6) = 9.65K The charging current (Ic) flowing through capacitor C1 is: Ic = U / Xc = 220 / 9.65 = 22mA.

Usually, the relationship between the capacity C of buck capacitor C1 and the load current Io can be approximated as C = 14.5 I, where the capacity unit of C is μF and the unit of Io is A.


Vorherige: Challenges and Optimization of Data Center Connector Systems

Nächste: What is the ripple and noise of the power supply?