I. What is the function of a capacitor?
1) Bypass
Like a small rechargeable battery, a bypass capacitor can be charged and discharged to a device. To minimize impedance, the bypass capacitor should be placed as close as possible to the power supply and ground pins of the load device. This effectively prevents ground potential rise and noise caused by excessive input values. Ground potential is the voltage drop at the ground connection when a large current spike passes through it.
2) Decoupling
Decoupling, also known as decoupling, refers to the process of separating a circuit into a driving source and a driven load. If the load capacitance is large, the driving circuit must charge and discharge the capacitor to complete the signal transition. When the rising edge is steep, the current is large, causing the driving current to absorb a significant amount of power supply current. Due to the inductance and resistance in the circuit (especially the inductance on the chip pins), this current bounces back, creating noise that can affect the normal operation of the preceding stages. This is what is known as "coupling."
Decoupling capacitors act as a "battery," accommodating changes in the drive circuit current, preventing mutual coupling interference, and further reducing high-frequency interference impedance between the power supply and reference ground in the circuit.
Combining bypass capacitors and decoupling capacitors makes it easier to understand. Bypass capacitors also perform decoupling, but they generally refer to high-frequency bypass, providing a low-impedance discharge path for high-frequency switching noise. High-frequency bypass capacitors are typically small, usually 0.1μF or 0.01μF depending on the resonant frequency; while decoupling capacitors are generally larger, possibly 10μF or more, determined by the distributed parameters in the circuit and the magnitude of the drive current variation. Bypassing filters out interference in the input signal, while decoupling filters out interference in the output signal, preventing interference signals from returning to the power supply. This is their fundamental difference.
3) Filtering
Theoretically (assuming the capacitor is purely capacitive), the larger the capacitance, the lower the impedance, and the higher the frequency that can pass through. However, in reality, most capacitors larger than 1μF are electrolytic capacitors, which have a large inductive component, so the impedance actually increases at higher frequencies. Sometimes you'll see a large electrolytic capacitor connected in parallel with a small capacitor; in this case, the large capacitor filters low frequencies, and the small capacitor filters high frequencies. The function of a capacitor is to pass AC while blocking DC, and to pass high frequencies while blocking low frequencies. The larger the capacitor, the easier it is for high frequencies to pass through. Specifically, in filtering, a large capacitor (1000μF) filters low frequencies, and a small capacitor (20pF) filters high frequencies. Some netizens have vividly compared filter capacitors to "ponds." Since the voltage across a capacitor does not change abruptly, it can be seen that the higher the signal frequency, the greater the attenuation. A capacitor can be figuratively described as a pond, where the water level doesn't change due to the addition or evaporation of a few drops of water. It converts voltage fluctuations into current fluctuations; the higher the frequency, the larger the peak current, thus buffering the voltage. Filtering is essentially a charging and discharging process.
4) Energy storage
Energy storage capacitors collect charge through a rectifier and transfer the stored energy to the power supply output via converter leads. Aluminum electrolytic capacitors with voltage ratings of 40–450VDC and capacitance values between 220–150,000μF are commonly used. Depending on the power supply requirements, components may be connected in series, parallel, or a combination thereof. For power supplies exceeding 10kW, larger can-type screw-terminal capacitors are typically used.
II. What makes a good capacitor?
1. The larger the capacitance, the better.
Many people tend to use large-capacity capacitors when replacing them. We know that while a larger capacitor provides stronger current compensation for an IC, it also increases size and cost, and affects airflow and heat dissipation. Crucially, capacitors have parasitic inductance, causing the discharge circuit to resonate at a certain frequency. At this resonant point, the capacitor's impedance is low, resulting in minimal impedance and optimal energy replenishment. However, as the frequency exceeds the resonant point, the impedance of the discharge circuit increases, and the capacitor's current-providing capability decreases. A larger capacitance value results in a lower resonant frequency and a smaller effective frequency range for current compensation. Therefore, the idea that larger capacitors are always better is incorrect; general circuit designs use a reference value.
2. For capacitors of the same capacity, the more small capacitors connected in parallel, the better.
Voltage rating, temperature rating, capacitance, and ESR (equivalent resistance) are some of the important parameters of a capacitor. Naturally, lower ESR is better. ESR is related to the capacitor's capacitance, frequency, voltage, and temperature. When the voltage is constant, a larger capacitance generally results in a lower ESR. In PCB design, multiple small capacitors are often connected in parallel due to PCB space limitations. Some people believe that more parallel small resistors result in a lower ESR and better performance. Theoretically, this is true, but the impedance of the capacitor's solder joints must be considered. Therefore, using multiple small capacitors in parallel does not necessarily lead to a significant improvement.
3. The lower the ESR, the better the effect.
Based on the improved power supply circuit described above, the input capacitor should have a larger capacitance. The ESR requirement can be appropriately reduced relative to the capacitance requirement. This is because the input capacitor primarily serves to withstand voltage and secondarily to absorb MOSFET switching pulses. For the output capacitor, both the voltage withstand requirement and capacitance can be slightly lower. The ESR requirement is higher, as sufficient current throughput must be ensured. However, it's important to note that lower ESR is not always better. Low ESR capacitors can cause switching circuit oscillations. Oscillation damping circuits become more complex and increase costs. In board design, a reference value is generally used as a component selection parameter to avoid increasing costs due to oscillation damping circuits.
4. Good capacitors represent high quality.
The "capacitor-only" theory was once extremely popular, and some manufacturers and media deliberately used it as a selling point. However, in circuit design, the level of circuit design is crucial. Just as some manufacturers can create more stable products with two-phase power supplies than others using four-phase power supplies, simply using expensive capacitors doesn't necessarily result in a good product. When evaluating a product, it's essential to consider it comprehensively and from multiple perspectives; the role of capacitors should never be intentionally or unintentionally exaggerated.
