EMC Debugging Techniques – Part 1

In this post, I’ll share a powerful debugging technique you can use to identify the source of emissions in the 30 MHz to 300 MHz range. In this specific band, the test setup you use during EMC measurements is critically important. I’ve seen customers fail tests simply because of issues with how their setup was arranged. I’ve also had customers come to me in panic, thinking they needed a full redesign—only to discover the fix was quick and simple.

As usual, and in line with my style, I’ll walk you through both the theory and the practical steps so you can reach the same conclusions yourself. So... keep reading..

When I visit an EMC lab, the first test I usually request is the radiated emissions (RE) test.

This test is like an X-ray for your design. It gives you insights not only within the official measurement band (typically 30 MHz to 1 GHz), but also clues about what might be happening outside of that range. I’ll explain how and why in a moment.

Whenever I book time at an EMC lab, I always recommend starting with radiated emissions, followed by conducted emissions. Of the two, radiated emissions usually give you the most information about the general health of the design.

Every EMC issue leaves a unique emission fingerprint.

For example:

• Wideband emissions from 300 MHz to 1 GHz typically suggest poor PCB layout—often a 2-layer board radiating excessively.

• Sharp peaks at the 5th and 7th harmonics of a clock may mean you’ve done a good job with trace length and impedance—but you’re using a high-fanout clock that still needs attention.

• A bell-shaped emission pattern repeated across frequencies may indicate the need for a snubber on your DC/DC converter.

  
  

Using Radiated Emissions for Debugging

In this post, I want to show you how to quickly identify emissions that aren't coming from your PCB.

To do this, we’ll slightly modify the standard radiated emissions setup. But first, let’s review what a typical RE test setup looks like.

Typical Radiated Emission Test Setup

Radiated emissions are usually measured in a shielded chamber. The Equipment Under Test (EUT) is placed on a wooden table, 80 cm above a conductive floor. The measuring antenna is positioned 3 m away from the EUT.

• For vertical polarization, the antenna is set 1 m above the floor.

• For horizontal polarization, the antenna is raised to 4 m.

If the EUT has cables, they are typically 1.5 m long ish and exposed in front of the table, routed vertically. (See figure below.)

Diagnostic Trick: Leveling the Antenna Height

Now, here’s the debugging technique.

We ask the EMC lab to perform both the vertical and horizontal measurements with the antenna at the same height—specifically, 1 m from the ground.

Let’s start with the vertical test.

Suppose you see emissions at 250 MHz, 700 MHz, and 900 MHz (see diagram).

Now you rotate the antenna 90°, keeping it at 1 m, and run the horizontal measurement. If the emissions change significantly, you’re seeing a polarization effect—and that gives you information about the source.

For example we have this kind of emission:

What can we draw from our test?

The first think to notice is that we had an overall redution of emittions, in particular at the lower frequencies.

Since the EUT cables are in the length range of 1.5 m to 2.5 meter, we should expect that in the frequency range 30 MHz to 300 MHz an antenna gain such that any emissions from the cable is amplified by the antenna gain of the cable. Therefore, we should expect that when we do a vertical measurment any emittions coming from the cable is amplified by the antenna gain of the cable, and therefore, it is unluckly that we need a PCB relayout.

So, by just rotating the antenna and by looking at the RE plots, we can conclude that the emittion at 250 MHz, is due to the cable and can be easily fixed. while the emittions at 700 and 900 MHz, may require a different test to identify the issue.

Thank you for reading.