EMC isn’t black magic, but it’s easy to understand why it seems that way. When looking at a schematic like that in Figure 1, it looks like you’re only dealing with DC signals all across the board. There’s a 28 VDC input that goes through an EMI filter, then gets converted to 12 VDC power. Except in extremely rare circumstances involving equipment sensitive to magnetostatic fields, DC electricity will never be part of an EMC problem.
Figure 1: Power conversion stage of avionics unit
Harmonics
The biggest problem occurs when you use a switching operation to convert power from one form to another—something most modern electronics do all the time. For power and thermal efficiency reasons, we want switching waveforms to get as close to a perfect square wave as possible—the faster the rise and fall time, the less loss and heating you have in the operation. The problem is that there’s a trade-off: Faster rise and fall times necessarily generate high-frequency electrical signals. Then, if provisions aren’t made for controlling the high-frequency noise signals, they can escape to cause signal integrity and EMC problems throughout the board and system.
Let’s illustrate this with a simple function generator. The beginning setting is a 5 VDC square wave switching at 20 MHz—something you might see from a board’s clock signal. In Figure 2, we see a cable connected to the function generator channel 1 output. That’s the default cable that came with the function generator, with current being run through a 50 Ω resistor between the red and black leads.
To read the entire article, which originally appeared in the May 2025 issue of Design007 Magazine, click here.