PCB design: Power supply plus magnetic beads (I)

The position of the capacitor: "The filter capacitor has its own filter radius, so the focus is on the distance between the positive end of the filter capacitor and the chip power pin".

Generally speaking, the capacitor has its filtering radius, and the filtering radius of the low-frequency capacitor is large, so the layout can be put slightly further away.

And generally speaking, the low-frequency capacitors acted by simple filtering should not be stacked up and placed evenly.

The filter radius of medium and high frequency capacitors is small, and it needs to be placed strictly close to the chip pin, not too far away, otherwise the capacitor will "not work".

This statement, as our filter capacitor layout design guide, no problem, is still the correct guiding principle, this is not the design error we will discuss.

But as the preamble says, it's all about getting to the bottom of what's going on with the filter radius of the capacitor?

First of all, the filtering radius is based on the famous quarter wavelength theory.

(The theory of quarter wavelength appears repeatedly in various articles . In order to facilitate everyone's understanding, we will discuss various issues of quarter wavelength, which will not be repeated here. If you have any problems in understanding this article, you can discuss with Mr. High Speed separately.)

According to the theory of capacitance decoupling radius, when the distance between the position of the capacitor and the device (pin) to be filtered is just 1/4 wavelength, the phase difference between the compensation current and the signal noise current of the capacitor is just 180 degrees, and the filtering fails.

Therefore, in order to ensure the filtering effect of the capacitor, the distance between the position of the capacitor and the device (pin) to be filtered is required to be less than 1/10 of a quarter wavelength. Of course, more stringent requirements want to be less than 1/20 of a quarter wavelength.

The filter radius of the common filter capacitor.

(To simplify the calculation, we assume that the ESL of the capacitor is 0.4NH and the mounting inductance of the capacitor is 1.5NH.

Since the decoupling capacitor is only a part of the power supply network, the high-frequency noise will be more dependent on the flat plate capacitor of the power supply, the filter capacitor in the package or even the DIE capacitor to filter out.

1 NF capacitors of 10 NF or less will play less and less of a role in power filtering systems.

In most designs, the 0.1uF capacitor is the highest "high frequency" capacitance used in the power filtering system for the board level system design.

For example, the decoupling radius of a capacitor, which everyone knows, and then everyone knows that the decoupling radius of a "small" capacitor (high frequency capacitor) is very small and needs to be strictly close to the chip pin.

The design principle of no wrong, just when the area of digital consumer products due to the smaller, lower cost and to meet the requirements of structure (e.g., thickness), we need to single layout, the capacitance on and BGA on the surface of the same time, a lot of people confused, capacitance so far away from the chip power pin (sometimes also consider 3 mm or 5 mm spacing of repair, went further),Does the capacitor still work?

This leads to another problem, which is related to our topic this time. Let me give you a preview: when we do design, we like to use the same kind of digital power supply to separate the power supply of different chips with magnetic beads, hoping to avoid the interference of the same power supply between different chips.

With capacitors in such a wide range, wouldn't the beads also block the "imaginary" interference from the power tracks between the chips?

At the end of this section, I'll show you another picture of the decoupling radius and range of the capacitor.

The quarter wavelength of the resonant frequency we talked about earlier, there are some limitations, first of all the maximum frequency at which the capacitor acts is not just the resonant frequency.

Secondly, the scope of action also has to consider the filtering efficiency you set.

The theory is complicated, and there is a long way to go. The more we move forward, the clearer the truth will be.

The conclusion is that the filtering radius of our commonly used 0.1uF capacitor is much larger than we thought.