Issue link: https://iconnect007.uberflip.com/i/666328
April 2016 • The PCB Design Magazine 13 delivery? No! Knowledge of circuit theory does not hurt your cause but, by itself, is not the so- lution. What you need to know is elemental elec- tromagnetic field theory. The problem is not about voltage and current; it is about EM waves moving through transmission lines. Folks who understand this have their arms around the real problem. Thanks to guidance from friends like Ralph Morrison and Dan Beeker, I have understood this for many years and now know the real so- lutions. Once this knowledge settled into my little brain, I suddenly realized exactly what causes problems in PC boards and what to do to avoid them. Everything became crystal clear! Where do you stand? Consider the following: • Current in a transmission line forms in the copper, but the energy of the line is not in the copper. • Energy in transmission lines moves through the plastic insulation of the board, the space between the copper features. • Current formed in vias is all on the outside of the via barrel. No current flows on the inside of a via. Those three simple statements tell us a lot about why problems occur in some boards and not in others. If any of those comments sound foreign to you, whether you are a circuit engi- neer or a board designer, you must read a good " What you need to know is elemental electromagnetic field theory. The problem is not about voltage and current; it is about EM waves moving through transmission lines. " Working With circuit Design engineers It's Hard Out Here for an EE There is one thing that frustrates many cir- cuit engineers: They are usually held responsible for the entire success of a system. Even though they do not do the packaging design or the board layout, they are held liable when hard- ware-related problems occur that are caused by those design features. EMI is almost always the result of a physical entity causing resonance and radiation, yet circuit engineers are often not allowed to control these physical items. They want to put in their two cents' worth, and rightly so. Most board designers under- stand that, but they believe, "If you are going to put in your two cents, make sure you actually know what you are talking about." It is OK to be heard, but it's not OK to tell the board designer to do this or that when you really do not under- stand the problem. Far too many engineers do not understand the cause of the problems that designers see every day. This goes for board designers as well. If you are going to push back at your engineer's sug- gested layout ideas, make sure to base your thinking in physics, not in some half-baked idea that came from an app note or a poorly designed evaluation board (this is fodder for an- other article). It is very important to know, schematically, how a circuit will function, "what drives what," "what is dependent on what," and what timing issues are critical. These are things you should understand whether you are a circuit engineer or a board designer. Not having an EE degree does not relieve designers from the absolute need to understand basic circuit behavior. If you do not know, ask! However, knowledge of these things, alone, will not eliminate the risk of SI or EMI problems. There is much more to this issue. So, what is the solution to good PC board de- sign? One thing has held true over the decades: Understanding how to design to maximize manufacturability, testability, ease of repair, and good thermal transfer has not changed and probably never will. Board designers will always need to know about these issues. Of this, most designers and engineers agree. Is knowledge of electronic circuit theory necessary to lay out sensitive transmission lines and complex power