Below are the questions asked during the live event, along with their respective answers.
Q: What are the most common mistakes you see people make when installing EMI filters?
A: One problem is the improper grounding of filters by using a ground wire which typically has about 20nH of inductance per inch of length. Even using a heavy braid, which is often sold as an EMI ground strap, can easily have 5-10nH of inductance per inch. Which does not provide a low impedance above 1MHz. But the MOST common mistake is not shielding the filter input wiring from the output wiring. I hear questions like, “The wires are separated by four inches. Don’t you think that’s enough?” I like to remind people that their cell phone routinely communicates with cell towers miles away, so no, four inches may not be enough separation. Additional shielding is often needed.
Q: Can you tell from a device’s emission profile what’s wrong with a filter installation?
A: With system-level emissions, there are just so many variables that it’s nearly impossible to see from just looking at the data. The only time I feel confident is when measuring conducted emissions on the input power lines, and you see the low frequencies start high, slope down into spec, stay low for a few hundred kilohertz, but then there is a sharp, distinct peak somewhere between 2 and 20MHz. This telltale sign usually means there is another cable that is insufficiently filtered and the power supply emissions are coming out this second cable and being picked on the input power cable as they return through the power lines or the power input ground wire. If this is the case, ferrite beads on these secondary cables or the system input ground wire will lower the emission peak. If you do not identify these emissions as originating from secondary cables, cables that you are not actually measuring, you will waste significant time and effort removing and modifying the filter internals and yet see no reduction of these emissions.
Q: As a filter manufacturer, do you think clip-on ferrites are overrated?
A: Feedthrough filters and custom power line filters often require great effort to install in a system that has not been specifically designed for them. Ferrite beads, especially the split-core clip-on types, are often the quickest and easiest types of filters to install while in the lab. I find them one of the most useful items in the EMC engineer’s bag of tricks.
Q: Do you have any general recommendations for engineers working on new design?
A: Get the power supply designed and running in the EMC Lab for emissions ASAP. One test is worth a thousand expert opinions!
Q: What types of devices cause the biggest headaches to filter designers?
A: Historically, many digital devices struggled to meet radiated emission requirements. Drastic size reductions, the abundance of multi-layered PCBs, and improved I/O interconnects like Ethernet and USB, have greatly improved system-level radiated emissions. The biggest headaches I see are switch-mode power supplies causing conducted and radiated emission problems.
Q: Have you seen issues with HDMI cables for video transmission? Do you have any recommendation to reduce the EMI caused by HDMI cables?
A: Yes, I have seen radiated emission issues from HDMI cables. Shielding and ferrite beads are likely the only things that could be added to cables to improve EMI on HDMI cables since low-pass filters would attenuate the video signals excessively. Any other solutions would have to be at the circuit level on the PCB.
Q: What are some buzzwords to listen for or good questions to ask engineers that will help us uncover whether they’re experiencing challenges with EMC that utilizing these resources will be helpful?
A: The high-performance filters manufactured by APITech are critical for meeting RE102, or radiated emissions, of MIL-STD-461 for “Tactical”, “Mobile”, “Ground” and “Helicopter” applications, due to the very strict EMC requirements of these platforms. There are EMC requirements for ships, space, aircraft, submarines, etc. but these platforms have RE102 requirements that are much less stringent. Equipment with Radiated Susceptibility requirements at 200 volts per meter or higher will also have a need for the performance of APITech products. And any type of radio transmitting and receiving equipment used in military or commercial environments would also be a good fit for our products.