expert insights
EMC BENCH NOTES
Some Starting Tools
By Kenneth Wyatt
A

s a product designer, one of the biggest issues you’ll face is radiated emissions. This month, we’ll describe the minimal set of tools to characterize and help mitigate radiated emissions right on your workbench.

Near Field Probes
Most designers may be familiar with near field probes but may not understand how to go from using them to identify “hot spots” of high harmonic energy within your board or system to actually mitigating the issues.

These can either be constructed as DIY projects or commercial probe kits may be purchased. Flexible or semi-rigid coaxial cables may be used. H-field (loop) probes may be constructed by soldering the center conductor to the shield, as in Figure 1. An E-field probe may be constructed by cutting away a short (5mm) portion of the shield, exposing the center conductor. Both types of DIY probes should be dipped in rubberized “tool dip” or otherwise insulated to avoid shorting out circuitry. See the example in red.

DIY H-field probes
Figure 1: Some DIY H-field probes. The smaller one may require a broadband preamplifier to observe usable harmonics
More details for commercial product choices can be found in Reference 1 at all price points. Each month, we’ll be using this basic equipment for some hands-on experiments we can do together.

Let’s start off with the most basic probe; E-field (voltage measurement) and H-field probes (current measurement). These are designed to be most sensitive to either E-fields or H-fields, respectively.

H-field probes are most sensitive to currents in wires or cables. While these DIY versions have an unbalanced geometry, which creates common mode currents flowing up the “handle” portion, they are still useful for general troubleshooting.

E-field probes are more sensitive to components that create large E-fields, such as heat sinks and any circuit switching large voltages. A good example of large changing voltages would be off-line switching power converters.

The advantage to commercial near field probes is that they are insulated and, being longer and thinner, can penetrate into narrow spaces (Figure 2). Commercial probes come as sets, usually three H-field probes in different sizes and an E-field probe. Beehive Electronics and Tekbox probe sets are about $350. Make sure to order the cable (sold separately) for the Beehive probes.

variety of commercial H-field probes
Figure 2: A variety of commercial H-field probes. Rohde & Schwarz, Com-Power, Tekbox and Beehive Electronics (top to bottom).
Com-Power probe sets include a capacitive-couple probe that can directly measure harmonic voltages on circuit traces and can also be used to inject signals useful for troubleshooting immunity issues. The Rohde & Schwarz probe set is unique in that, in addition to the usual H- and E-field probes, they also include some very tiny probes that may be more useful in today’s small wearable products.

Near field probes are most useful for identifying major energy sources on PC boards and internal cables. I use H-field probes for detecting high currents (as in ICs and circuit traces) and E-field probes for detecting high voltage swings (as in buck converters). A record of the harmonic spectrum for each major energy source should be recorded.

RF Current Probes
The reason most products fail radiated emissions is that their attached cables carry high-frequency harmonic currents, which tend to make them radiate as transmitting antennas. Note that it takes only 6 to 8 µA of high-frequency harmonic currents to exceed the FCC or EU limit for Class B (household) products!

By measuring and monitoring these RF currents, we can often perform troubleshooting right on our workbenches and mitigate emissions prior to taking the product to the compliance test lab. Reducing these RF currents will also reduce the cable emissions.

While I personally own several commercial current probes, those pictured in Figure 3 are the ones I started out with for a couple of years before I could afford a good set. Many of my clients are helped remotely, and I’ve had them make these DIY probes so I can guide their troubleshooting efforts.

DIY RF current probes made from standard clamp-on ferrite chokes
Figure 3: A couple of DIY RF current probes made from standard clamp-on ferrite chokes.
Choose a ferrite that has some impedance in the frequency range of most common mode currents: 30 to 200 MHz. A Fair-Rite choke made from type 31 material, or equivalent, should work well. The number of turns is not critical, and I usually use 5 to 7 turns. Terminate with the desired coaxial connector epoxied in place. Be aware the hinge on these DIY current probes won’t last forever, so be prepared to replace them occasionally.

The fact these DIY probes are uncalibrated is unimportant when used for troubleshooting purposes since we’re only looking for relative changes. For example, if we know we’re failing by 5 dB, then at the workbench, we’ll want to apply mitigations to reduce the harmonic amplitude by 10 to 15 dB for safety.

Eventually, you’ll want to purchase a calibrated RF current probe. Figure 4 shows an affordable commercial probe from Com‑Power. Similar affordable probes are available from Tekbox. Be sure to order one that can clamp around the wire or cable to be tested.

commercial RF current probe
Figure 4: An example of a commercial RF current probe that is calibrated from 10 kHz to 400 MHz. Photo, courtesy Com-Power.
As for the near field probes, you’ll want to record the harmonic frequency spectrum for each cable attached to your product. By examining the cable currents, you should find some correlation to the energy source or sources on the board or interior cables. Part of the mitigation process will be to identify and reduce the coupling between the dominant energy sources and radiating cables.
Nearby Antenna
Once your product is characterized using the near field and current probes, I usually switch to a nearby antenna placed about 1m from the product under test. The resonant frequency doesn’t matter much so long as you can observe the harmonics from the EUT. Monitoring the actual emissions while troubleshooting and applying mitigations in real-time is a very efficient way to resolve design problems.

The antenna I like to use is made by Kent Electronics and costs just $38. I show how to make the PVC fixture that attaches to a table-top tripod in Reference 2.

Spectrum Analyzers
In recent years, the cost of spectrum analyzers has dropped to very affordable levels. I started my career with a Rigol DSA815TG, which was about $1,500 at the time. Since then, this price has dropped to about $1,000. I’m currently using the Siglent SSA 3032X, with its larger screen. If you’re in the market for one of these, be sure to order the tracking generator and EMI options, as we’ll be using these for more advanced EMC characterizations.

There are several other good choices in analyzers, and Rigol and Siglent have captured much of the affordables market. Many other alternatives are described in Reference 1. The U.S. distributor for these models is Saelig Electronics.

spectrum analyzer
Figure 5: Monitoring the emissions from the product under test with a spectrum analyzer and simple antenna is a very efficient method for fixing problem harmonics.
Summary
This summarizes the most basic equipment needed for identifying the major harmonic noise sources and for characterizing radiated emissions. Next month, we’ll use these tools to help characterize some actual embedded processors and suggest some mitigations that would improve the designs.
References
  1. Wyatt, Create Your Own EMC Troubleshooting Kit (Volume 1, 2nd Edition), Amazon, 2022
  2. Wyatt, “PC Board Log Periodic Antennas,” EDN.
Share this story:
Back to Issue