Troubleshooting EMI on your bench top

-May 19, 2014

If your product is failing radiated emissions at the test lab, it’s often more cost-effective to perform any detailed troubleshooting at your own facility where you can take time to methodically isolate the source and try out several potential fixes. Unfortunately, many companies don’t have the equipment or training to make these simple measurements. This article describes a easy method for measuring radiated emissions and providing a rough estimate of pass/fail.

Near field probing with E- or H-field probes will help identify potential sources of EMI on, or within your product. However, as I’ve pointed out in past blog postings (see additional links), not all structures or components will act as efficient antennas and therefore, are not ultimately causing emissions failures. To confirm which structures are radiating, I like to set up a “mini” radiated emissions range on a workbench or conference room table.

It’s best to work in an area relatively free from external signal sources or other operating equipment. A conference room or basement often works better. Basements are good because sometimes the ambient signals from broadcast, two-way radio or cellular phones that can cover up emission harmonics from your product will be lower in amplitude.

Try setting up your product at one end of a table or workbench and set up a simple antenna at the other end about 1m away. You may need to move it closer to better observe the harmonic emissions. Tape (or otherwise fix) the antenna in place, so inadvertent movement won’t cause a variable in your measurements. For general troubleshooting purposes, almost any antenna will work. Of course, it’s best if it is resonant at the frequency bands of interest. A simple “rabbit ears” TV antenna or PC board antenna, as shown in Figure 1, works well.


Figure 1 – General test setup for troubleshooting radiated emissions on your test bench. By setting up a nearby antenna, you can monitor the results of your proposed fixes in real time.

Connect a spectrum analyzer to the antenna and adjust it to the harmonic or harmonics of interest. Because we’ve moved from the test lab to the workbench, the measurement is no longer calibrated. So now you’ll want to establish a base line in your ad-hoc test setup, so you can tell if you make any improvements. If the analyzer has it, set a “display line” to the top of the highest harmonic. If you’re evaluating several harmonics simultaneously, you might also save a baseline trace on the analyzer display in which to compare during the troubleshooting process. This will serve as a reference point to help judge whether a fix has helped, or not. Be sure to sit in the same spot during times you’re recording measurements.

If the measurement test distance is set at 1m, you may use the adjusted radiated emissions limits in Figure 2 to estimate how the emissions might compare to 3m or 10m limits.

Caution: Because, the measured frequencies below 150 MHz may be in the near field, the results are ONLY a rough guess. Experience by Ott and Curtis (see References) has shown that adding an additional 6 dB on top of the extrapolated limits is often more accurate. Figure 2 accounts for this additional 6 dB increase. However, it’s always best to record a baseline and then troubleshoot from there.


Figure 2 – Adjusted FCC and CISPR radiated emission limits for a 1m test distance (plus a 6 dB correction factor). This will provide a rough estimate of pass/fail for emissions. Courtesy Henry Ott.

Caution: From experience, as you’re troubleshooting specific harmonics, there are two things to keep in mind. First, a 10 dB reduction with the antenna at a 1m distance will not necessarily translate to a 10 dB drop when measured at 3m or 10m. The primary reason for this is near-field and far-field effects. These effects make an inverse linear relationship of field strength with distance impossible when measuring as close as 1m. Second, consider reductions of 2 to 3 dB as significant improvements, but less than that is likely measurement error or maybe a change in your body position or equipment and cable orientation, which can affect the antenna or product emission pattern.

As a consultant, I’ve used this technique many times at client facilities with great success. With the ability to see results real time, it’s possible to run through many experiments or potential fixes quickly, saving time and money.

References:

Ott, H., Electromagnetic Compatibility Engineering, Wiley (2009).

Curtis, J., “Toil and Trouble, Boil and Bubble: Brew Up EMI Solutions at Your Own Inexpensive One-Meter EMI Test Site”, Compliance Engineering, July/August 1994.

See also:

Measuring resonance in cables

PC board resonance and the “balloon” effect

Questions on EMC pre-compliance testing for radiated emissions

Identifying emissions sources and propagating structures

PC board log periodic antennas

Loading comments...

Write a Comment

To comment please Log In

FEATURED RESOURCES