Image Plane of PCB (I) Definition of Image Plane

tmily

Image Plane of PCB (I) Definition of Image Plane [Copy link]

An image plane is a layer of copper conductors (or other conductors) that is located inside a printed circuit board (PCB). It may be a voltage plane, or a 0V reference plane adjacent to a circuit or signal routing layer. The concept of image plane became commonly used in the 1990s and is now an industry standard term. This article will explain the definition, principles, and design of image planes.

Definition of Image Plane

RF current must return to the current source via a previously defined path or other paths; in short, this return path is a kind of image plane. The image plane may be a mirror image of the original routing, or another path located nearby - that is, crosstalk; the image plane may be a power plane, a ground plane, or free space. RF current will couple with any transmission line in the form of capacitance or inductance as long as the impedance of this transmission line is smaller than the impedance of the previously defined path. However, in order to meet EMC standards, free space must be avoided as a return path.

Although single-sided PCBs can reduce costs, this simple structure may not meet EMC standards. Most 2-layer or 4-layer PCBs have relatively high signal integrity and can pass EMC tests. High-density (multi-layer) PCB stacking can provide approximately 6dB to 8dB of RF suppression for each pair of image planes, which is due to the effect of eliminating magnetic flux. There is a simple rule that can be used to determine when a multi-layer board should be used: when the clock rate exceeds 5MHz, or the rise time is faster than 5ns, a multi-layer board must be used.

Definition of Inductance

Both traces and copper planes have a finite amount of inductance. When voltage is applied to a trace or transmission line, these inductances prohibit the flow of current, so the two conductors become unbalanced common-mode radiation, and the magnetic flux cannot be reduced. In the circuit board structure, there are three different types of inductance:
● Partial inductance: The inductance present in the conductor or PCB trace.
● Partial inductance of the self: The inductance from a conductor segment relative to an infinitely long segment.
● Common partial inductance: The effect of one inductor segment on a second inductor segment.

Compared with capacitance and resistance, inductance is the most difficult to measure. Inductance represents the dynamic characteristics of a closed current loop. Inductance is the ratio of the magnetic flux through the closed loop to the current that generates the magnetic flux. Its mathematical expression is: Lij=Ψij / li, Ψ is the magnetic flux, and I is the current in the loop. In a closed loop, the inductance value is related to the shape and size of the loop. When designing PCBs, engineers often ignore the inductance of the traces. Inductance is always related to the closed loop. The inductance effect of a closed loop can be described by the effect of partial inductance and common partial inductance.

Partial Inductance

The internal inductance of a conductor is generated by the magnetic flux inside the conductor. The sum of the partial inductances of a closed loop is equal to the sum of the partial inductances of each segment, that is . The Li of each segment is equal to Ψi / li, Ψi represents the magnetic flux coupled to the loop by the i-th segment, I is the current in the i-th segment, and Li is the partial inductance. Therefore, different loops will have different values of partial inductance. We are concerned with the partial inductance value, not the total inductance of the trace. Moreover, the common partial inductance can be derived using the partial inductance.

Common Part Inductance

The main factor that can eliminate the magnetic flux in the image plane comes from the "common partial inductance". After the magnetic flux is eliminated, the magnetic lines of force can be connected and the best return path for the RF current can be found. The own partial inductance refers to the inductance of a specific loop segment and has nothing to do with other loop segments. Figure 1 shows an own partial inductance. The current in a routing loop is I, and Lp is the own partial inductance of the routing segment. Assume that this routing extends from a finite end to an infinite end.

In theory, although the partial inductance of a conductor has nothing to do with the adjacent conductors, in practice, adjacent conductors with a small spacing will change each other's partial inductance. This is because one conductor will interact with other conductors, making the current distribution over the entire length of the conductor no longer uniform. This is especially true when the ratio of the spacing and radius of the two conductors is less than about 5:1.
　

Figure 1: Partial inductance of the device itself

　
There is a common partial inductance between two conductors. The common partial inductance Mp is based on the spacing (s) between parallel lines or conductor segments. Mp is the ratio of the magnetic flux generated by the current in the first conductor (through the second conductor to a far distance) to the current generated by the first conductor. Figure 2 shows a common partial inductance. Its equivalent circuit is shown in Figure 3. The mathematical expression of this circuit is as follows:

Figure 2: Common partial inductance

　

　Figure 3: Common partial inductance between two conductors

　

Now, let's consider the transmission of a signal, such as a clock signal, in the circuit of Figure 3, using the concept of common partial inductance. V1 is on the signal path, and V2 is on the RF current return path. Assume that these two wires form a signal path and its return path, so I1 = I and I2 = -I. If there is no common partial inductance, these two wires will not be able to couple with each other, and this circuit will not work properly and will not form a closed loop. The voltage drop in Figure 3 will become:

From the above formula, we can know that if we want to reduce the voltage drop, we must increase the common partial inductance value (Mp).

The simplest way to increase the common partial inductance is to move the RF return current path as close to the signal trace as possible. The best design method is to use an RF return plane near the signal trace, and the distance between them should be as small as possible within the achievable capability.

Partial inductance always exists in the wire, it is like a preset value. Therefore, it is equivalent to an antenna with a specific resonant frequency. "Common partial inductance" can reduce the effect of "partial inductance". By reducing the distance between two wires, their individual partial inductance can be reduced, which can meet the requirements of EMI compatibility standards.

In order to maximize the effect of the common partial inductance, the currents in the two wires must be equal in magnitude but opposite in direction. This is why the image plane (or ground wire) is so effective. There is a common partial inductance between two parallel wires, and the value of this inductance varies with the distance and length of the two wires (see the technical specifications of the wires). When the distance and length of the two parallel wires are both minimum, their common partial inductance will be maximum.

What role does the "common partial inductance" play if the power and ground planes are separated by a dielectric material? Again, as long as the spacing between the two planes is small, the common partial inductance value will be large. In this case, the RF signal current measured on the power plane should be zero because it is offset by the RF return current of equal magnitude and opposite direction.

Furthermore, it should be noted that if the common partial inductance between the two conductors is reduced, not only will the image plane effect be reduced, but the capacitance between the two planes will also increase.