Bonding to the Ground Plane

expert insights

Standards Practice

By Karen Burnham

f you usually spend your time testing to one standard, it can be easy to forget about some of the little things that differ from standard to standard. One thing that’s easy to overlook is the different ways that equipment under test (EUTs) are required to be bonded (or not bonded) to a ground plane. Let’s start by looking at the difference between Figures 2 and 3 of MIL-STD-461 Rev G.

In both cases you have the EUT sitting directly on the surface of the test bench. In the first, the surface is metallic (usually copper), and in the second, it’s non‑conductive except for a small area where the LISNs sit.

Technical diagram illustrating a standard EMC test setup for conducted emissions, showing the EUT (Equipment Under Test), LISNs, Ground Plane, and interconnecting cable lengths

Figure 1: Figure 2 of MIL-STD-461 Rev G “General test setup”

Technical diagram of an EMC conducted emissions test setup showing the EUT on a non-conductive table, with LISNs, a Ground Plane, and detailed measurement dimensions

Figure 2: Figure 3 of MIL-STD-461 Rev G “General test setup for non-conductive surface mounted EUT”

This is a classic example of the Test Like You Fly (TLYF) principle. Whether the EUT is housed in metal or plastic, the assumption is that it will be mounted onto some platform, and the test setup should mimic the material of that platform. This has a huge impact on test results, as you can imagine, so you want to make sure you’re testing in the correct environment.

What’s mentioned in the text but not shown is the inclusion of more subtle aspects of mounting. It’s not uncommon for a metal-housed unit to be installed on a metal chassis but conductively isolated from the chassis by some kind of standoff, either for thermal or mechanical reasons. If that’s the case, those standoffs should also be represented in the test setup, along with a bond strap to the chassis/surface if and only if a bond strap will be included in the final installation. (If unsure, or if there’s a chance a bond strap may be removed from the installation for cost or weight concerns, I’d recommend testing without one and only adding one if poor test results require it. That way, you’ll have solid data to either support or argue against the need for a bond strap in the installation.)

One thing that’s easy to overlook is the different ways that equipment under test are required to be bonded (or not bonded) to a ground plane.

You can see the same Test Like You Fly/Drive principles at work in automotive testing. Looking at Figure 9 of CISPR 25, 4th edition, you’ll see the EUT is isolated from the conductive test bench surface by 5 cm of dielectric. This represents the common situation where a unit housed in plastic is mounted near, but not directly to, the metal chassis of a vehicle. There’s a dotted line connection from the EUT housing to the surface, indicating that a bonding connection should be included if the EUT is housed in metal and if it will be bonded to the chassis in installation. Similar considerations are included for the load simulator used in testing.

Finally, looking at ANSI C63.4 (the test method used for FCC compliance testing), there’s no ground plane at all. The most common case for FCC testing is a stand-alone electronics unit that will be used in a residential or office environment, on a non-conductive desk or handheld, with no fixed installation and no assumption of nearby metal planes.

The presence or absence of a 5 cm gap between EUT and table surface or the inclusion/removal of a short bond strap can be easy to gloss over during testing. However, making sure that your test setup accurately captures the salient features of the EUT’s final installation will ensure that your testing is the most efficient and effective it can be.

Share this story:

Back to Issue