Bypass and decoupling chip capacitors, don't worry about it!

1. Bypass and decoupling

Let me talk about two more important concepts: bypass capacitors and decoupling capacitors.

As long as students who have designed hardware circuits are no strangers to these two words, there may not be many who really understand these two concepts.

Bypass: pass means to pass. By pass passes from a place close to the side. The main road does not follow the small road, and the main road does not follow the secondary road.

Decoupling: Couple, one pair. The verb is extended to mean pairing and connecting. If something (signal) in system A causes a thing (signal) in system B to appear, or vice versa, then we say that system A and system B are coupled (Coupling). De coupling is to weaken this coupling.

2. Examples of power supply bypass and decoupling circuits

Let's look at an example below, where the DC power supply supplies power to the chip.

a. If the power supply is interfered (may enter the power system through 220V mains, generally a signal with a higher frequency), then the interference signal will be conducted to the IC through the power line between Power and IC. If the interference is too strong, it may cause the IC The chip is not working properly. Add a capacitor C1 near the output of the power supply, because the capacitor is open to DC and has low resistance to AC, and interference signals with higher frequencies return to the ground through C1. The interference signal that would have passed from the IC now bypasses the IC and directly reaches the ground, so we call C1 a bypass capacitor (Bypass Capacitor), that is, bypass the IC.

b. The operating frequency of current integrated circuits is generally higher. When the IC starts up instantly or switches the operating frequency, it will cause large current fluctuations on the power supply wire. After this kind of fluctuation is transmitted to the power source in the reverse direction along the wire, it will cause the fluctuation of the power source. That is, the fluctuations of the IC are coupled to the power supply. When a capacitor C2 is placed close to the power port VCC of the IC, we know that the capacitor has the function of energy storage, which can provide instantaneous current to the IC and weaken the conduction of IC current fluctuations to the power supply. So we call C2 the decoupling capacitor.

Of course, we will find that the bypass capacitor C1 also has a decoupling function, and the decoupling capacitor C2 also has a bypass function.

3. The distance between theory and practice

Back to the puzzling question we started. We know the formula for calculating capacitance impedance:

Impedance Z=1/jωC

Capacitive reactance Xc=1/ωC=1/2πfC

The capacitive reactance is inversely proportional to the frequency and the capacitance value. The larger the capacitance, the higher the frequency, the smaller the capacitive reactance. Then the capacitive reactance of 0.1uF is 10 times smaller than that of 0.01uF. For an interference signal of a certain frequency, if it can be bypassed by a 0.01uF with a large capacitive reactance, it should be easier to bypass with a 0.1uF capacitor with a smaller capacitive reactance. Isn't it a waste to add a 0.01uF capacitor?

Due to the non-ideality of the lead and the medium, the actual chip capacitor has inductance and resistance characteristics in a capacitive device. For a specific capacitor, when the frequency is lower than a certain value, the component is capacitive, when the frequency is higher than this frequency, the original is inductive. This frequency is the self-resonant frequency of this capacitor. When we use a 0.1uF and a 0.01uF capacitor in parallel, it is equivalent to widening the filter frequency range.