The Evolution Of Simulation Technology and Crosstalk
In Part 1 of this blog, I discussed the evolution of simulation as it applied to calculating impedance control. This article focuses on crosstalk which, along with impedance, is one of the more straightforward aspects associated with simulation. The following describes the various elements associated with crosstalk, the challenges associated with it and how to ensure that your PCB design meets crosstalk requirements.
Solving for crosstalk issues in any given design is a bit more complicated than the effort which is associated with impedance control. Crosstalk simulation requires a 3D representation of the traces so that the spatial relationship of the traces can be determined. And, a simulation analysis can’t be done until it’s been determined how much tolerance there is for crosstalk. And, that is a fairly complicated Exercise.
Reviewing the basics of what crosstalk is and how it functions is a good starting point. As terms that define unwanted and disruptive noise, crosstalk and coupling are used interchangeably. What occurs is that the electromagnetic field traveling on one transmission line interacts or couples with the EM field traveling on a neighboring line. The neighboring transmission line can be in the same signal layer or in an adjacent layer.
As it’s name suggests, the EM field has two components—the electric field and the magnetic field. Each of these two components can couple energy from a driven line into a quiet line. When the electric field is dominant, the stray capacitive coupling between the two transmission lines is larger than the stray magnetic coupling and it is often referred to as capacitive coupling. This form of coupling occurs when traces are run one over the top of the other in the adjacent signal layers of the PCB. The magnetic field dominates when the magnetic coupling between the two transmission lines is larger than the capacitive coupling and it is often referred to as magnetic coupling. It forms when traces run side-by-side in the same signal layer of a PCB. Both types of coupling are always at work even though each is discussed independently. Focusing on one type of coupling over the other is done to draw attention to the primary source of crosstalk as once it is under control the lesser form will be as well.
Capacitive crosstalk appears as though a very small capacitor has been connected between two signals. The faster the edges, the larger the coupled signal. With today’s very fast PCBs, a stray capacitance of only a few picofarads will result in enough coupling to exceed the noise budget of any logic family. In the past, a lot of effort was devoted to trying to develop an algorithm that would accurately predict how long one signal can run over the top of the other without creating excess coupling. No reliable predictive algorithm was established. Instead, controlling coupling between signals in adjacent layers is accomplished by imposing routing rules that force signals in adjacent layers to route at right angles to each other. In this manner, there is no detectable coupling either capacitive or inductive.
Inductive crosstalk occurs because the stray magnetic coupling is large enough to couple significant energy from one transmission line to its neighbor. This occurs when traces run side-by-side in the same layer. There is stray capacitive coupling as well but it is much smaller due to the fact that the two traces are “edge-on” with very little surface area in common.
Inductive coupling can’t be avoided because it is necessary to route signals in the same layer in order to fit all the wires into the as-designed PCB. As noted in previous articles along with the references at the end of this article, the coupling that occurs when transmission lines run in parallel takes two forms: forward crosstalk and backward crosstalk. Figure 1 illustrates these two forms of crosstalk.
Figure 1. Crosstalk in a Coupled Pair of Transmission Lines
As shown in Figure 2, backward crosstalk grows much more rapidly with coupled length than does forward crosstalk. In order for forward crosstalk to impact PCB designs, the two transmission lines would have to run in parallel in spaces that are much longer than those found in PCBs so, for board designs, we focus on backward crosstalk.
Figure 2. Forward and Backward Crosstalk vs Parallel Coupled Length
And, as cited in previous articles and the references at the end of this article, you can’t control backward crosstalk by limiting the length that two traces run in parallel. Instead, it is addressed by edge-to-edge separation and height of the trace above the nearest plane. It’s the combination of these two factors done in tradeoffs between PCB thickness, impedance and routability that result in a PCB that is most manufacturable while still meeting the design’s electrical requirements.
One of the key takeaways here is that the continuing trend to route PCBs with increasingly finer traces within increasingly smaller spaces has resulted in the risk of PCBs failing from crosstalk alone. But, just to make things a bit more complicated, there are a couple more twists to this story. There are other sources of noise, in addition to crosstalk, on a signal and they include:
• Ripple on Vdd
• Vdd and ground bounce in the package.
These four sources of noise--crosstalk, reflections, ripple and Vdd and ground bounce, consume the noise margin in a logic family. Until the amount of the other three is known, it is not possible to know how much crosstalk is allowable. So, a simple rule such as 3H will not assure satisfactory crosstalk levels. A thorough treatment of reflections, ripple on Vdd and Vdd and ground bounce in the package is contained in the references listed at the end of this article.
Even with all of the foregoing challenges, crosstalk simulation is still simple when compared to everything required for the successful design of power delivery systems.
The capabilities of today’s simulation tools have done a good job of keeping up with the impedance and crosstalk challenges found in the majority of today’s PCB designs. Of the two factors, crosstalk is the most challenging in terms of creating enough noise margin that will allow for the handling of it through the entire design.