Kenneth Wyatt

's profile
image
Principal Consultant

Kenneth Wyatt is president and principal consultant of Wyatt Technical Services LLC, as well as the senior technical editor for Interference Technology Magazine. He has worked in the field of EMC engineering for over 30 years. His specialty is EMI troubleshooting and pre-compliance testing and is a co-author of the popular EMC Pocket Guide and RFI Radio Frequency Interference Pocket Guide. He also coauthored the book with Patrick André, EMI Troubleshooting Cookbook for Product Designers, with forward by Henry Ott. He is widely published and authored The EMC Blog hosted by EDN.com for nearly three years. Kenneth is a senior member of the IEEE and a long time member of the EMC Society. He may be contacted at ken@emc-seminars.com or kwyatt@interferencetechnology.com. Check out his web site for more technical information, training schedules, and links: http://www.emc-seminars.com.


Kenneth Wyatt

's contributions
  • 06.26.2017
  • Electronics interference: LG displays and routers
  • Nice summary of the general issue of compatibility between various ITE products. This also is an issue among many disciplines. I anticipate some interesting internal and external system compatibility issues when it comes to multiple autonomous vehicles sharing the public roadways. I do believe LG resolved their EMC issue in short order.
  • 01.12.2017
  • Spread-spectrum clocking in PCI Express
  • Here's another good reference on spread spectrum clocking I wrote a few years ago, which explains the physics in a little more detail: https://interferencetechnology.com/spread-spectrum-clock-generation-theory-and-debate/
  • 11.09.2013
  • Is spread-spectrum clocking good or evil?
  • Ward is correct in that TVs - especially the digital transmissions used today - are quite immune to interference, such as SS clocking. This was shown to be true in later tests by Lexmark, following their initial analog TV tests. As for the accumulated RF pollution issue, I believe poorly-designed non-compliant switch mode power supplies, along with early design FCL and LED lighting are orders of magnitude worse than clocks within electronic products. In fact, the increasing noise level below 100 MHz has prompted the FCC to investigate solutions to the growing issue.
  • 12.28.2016
  • Use LT SPICE to analyze input filters
  • For higher frequencies, your point is well taken. Since the filter needed to perform well at just 500 kHz, I didn't bother adding in the parasitics. Bottom line: a quick and dirty analysis helped educate the client and the filter worked fine. For some simple applications, you needn't spend hours doing simulations and I think my clients appreciate that! More back-story on this simulation... At the time, I was working with a client based in Colorado, who had contracted out the design to a company in Los Angeles. When I received the call, my wife and I were enroute to her sister's place near San Antonio and driving through the back country of west Texas. While my wife was driving, I was sitting there with laptop doing the SPICE analysis and typing up a report. Fortunately, we had a strong internet connection along the way and I was able to deliver the solution (probably one of many "right" answers) within a couple hours. Later, the modification was implemented...it worked...everybody was happy. Ah, the life of a consultant! Ha!
  • 12.29.2016
  • Near-field scanning: useful or misleading?
  • In order for a signal to propagate, it must couple to some metallic structure that is starting to approach 1/4 or 1/2 wavelength at the frequency in question. Typically, this would be a PC board dimension, leaky shielded enclosure, or I/O or power cable (which is usually the most likely). There are four ways a HF signal can couple; conductive, radiative, inductive, or capacitive. The trick is usually determining the coupling path from the source to one of these radiating structures. At the PC board level, it is important to reduce the level of HF "noise", or actually HF harmonic (common mode) currents by proper board design. There are a number of ways to create common mode currents on a PC board, many of which can be researched in the literature. But one very common problem is one, or more, HF traces crossing a gap in the return plane. Another common error is separation of the power and ground return layers too far. They should be adjacent and ideally 3-4 mils apart with no other layers in between. As far as radiating structures associated with PC boards, common ones would be ribbon cables or other internal cables, poorly terminated cables (I/O) for example, or unfiltered I/O or power connectors. There are a number of good books on EMC and product design. A search on Amazon, etc., will reveal them.
  • 12.29.2016
  • Near-field scanning: useful or misleading?
  • You're so right, Bruce - great article. Near field probing can too easily throw you off-course when it comes to troubleshooting EMI. As you pointed out, there may be a strong H-field identified on a PC board, but the question should be, "does that internal field couple to some antenna-like structure?" By following up with current probe measurements on these "antenna-lie" structures, such as I/O and power cables, you can filter out just the harmonics potentially causing radiated emissions failures. I also advocate following up near field and current probe measurements with a closely spaced antenna (say, 1m away) that will help confirm which harmonics are actually being radiated. By identifying specific cables or enclosure seams that are culprits, you can then go backwards and determine how the internal fields are being coupled to these cables or seams.
  • 12.09.2016
  • Electronic products from hell
  • I just tossed my two year old $100 Weller WES51 soldering station into the trash yesterday. Apparently, the temperature sensor breaks connection within the soldering pencil and that shuts down the whole ball of wax (can of worms?). There's no way to fix it, other than replacing the pencil at half the cost of the station. Apparently plenty of others have had the same issue. Also, I'm told once Weller got bought out, the customer service really sucks. I guess some American companies still have quality lessons to learn! Surprising these days. Well, at least it didn't catch fire - Samsung!! You can see what replaced the Weller in this article: https://design-4-emc.com/2012/05/25/soldering-irons/.