Fundamentals tutorial provides an EMC overview
Technical conferences often hold tutorials and workshops on Mondays and Fridays with technical papers and exhibits on the days between. This year, I made it a point to arrive on Sunday, in time for the Monday tutorials at the 2014 IEEE EMC Symposium. This morning, I attended the first half of the full-day "Fundamentals of EMC."
Mark Steffka from the University of Michigan opened by asking "What is Electromagnetic Compatibility (and Why is it Important)?" He showed the difference between intentional and unintentional receivers and how an intentional receiver (a radio) becomes an unintentional receiver in the presence of a plasma ball. It also shows how the plasma ball is an unintended source of radiated emissions. He then showed a video of an oven that turned on when a cell phone that was in the kitchen would ring. Now that's unintentional.
signal to couple from a source
to an antenna.
Source: Silent Solutions, IEEE
EMC consultant Lee Hill of Silent Solutions explained basic EMI, but started by asking who of the hundred-plus people in the room were attending their first EMC Symposium. I'd estimate about one-third raised a hand.
Hill explained what constitutes an antenna— two pieces of metal with a voltage between them—which radiates emissions from a source. Of course, you also need a path from the source to the antenna for emissions to radiate. Hill described antenna length in terms of "EMI talk" where if the antenna is short enough, it's a "bad antenna" but that's good because little energy will radiate from it. At the "right length" or longer, a product will radiate too much energy and it will fail an emissions test.
Hill went in to cover coupling paths and how mutual inductance between two signal paths can cause a clock signal to couple into another loop. The loops form an antenna (Figure 1). "PCBs make emissions tests fail," said Hill.
Next, Hill showed how a square wave is built from sine waves using odd harmonics. You need to get to about the tenth harmonic to get a reasonable square wave. It's the higher harmonics that cause short antennas to be "good enough" to cause emissions test to fail.
Hill then covered Maxwell's equations and how engineers use them to calculate EMF in V/m in the far field. Knowing field strength can help you determine if a product will pass an emissions test. He then asked how many people think wires have inductance. A few raised hands and Hill told them that wires don’t have inductance? He said that it's the shape and position of a wire, forming a loop, that creates inductance. Do you agree?
Dr. Todd Hubing of Clemson University followed Hill, speaking about ground for 15 min. In that short time, he explained why referring to a signal's return path as "ground" is incorrect. "Ground is a system's reference voltage," he said. Analog and digital circuits should have separate return paths that are tied together to keep them at the same potential. "Half of EMC problems are caused by improper grounding," he said.
After a break, Hubing spent 30 min. discussing shielding. Here, he explained how electric and magnetic fields behave and how metal shields can block electric fields but not magnetic fields, for magnetic fields will bend around metal shields. Metal surfaces can, however, be used to move magnetic fields away from sensitive circuits.
You can use ferrites and other permeable materials (Figure 2) to block magnetic fields. Hubing recommended that engineers get information at the exhibits from makers of shielding materials, but "make sure the people you talk to understand the problem you needs to solve."
Figure 2. A permeable material can block magnetic fields from reaching sensitive circuits.
Source: Clemson University, IEEE.
Randy Jost of Ball Aerospace gave the morning's final presentation where he discussed emissions and immunity. "Be aware of EMI standards," he told the audience. "Some tell you how to measure, others tell you the limits." For example, MIL-STD-461F describes how to make measurements. Standards vary by country.
Jost also explained how conducted emissions affect radiated emissions. Take that into account when designing for EMI. He also warned against using too much shielding material such as gaskets and copper tape to control emissions. Instead, find the source and mitigate it there with ferrites and filters on cables to reduce emissions closer to the source. He also recommended using circuit-protect devices to improve susceptibility.
Jost continued by saying that when designing for immunity, you need to understand your product's intended environment. He used an MRI machine as an example because it generates high magnetic fields. Airports (Figure 3), on the other hand, have signals at frequencies from kilohertz to gigahertz.
Figure 3. Airports can have radio waves at many frequencies.