Bench-Top Conducted Emissions Testing for Automotive and Vehicle Components

Conducted emissions testing on a bench-top setup is a valuable pre-compliance method that helps engineers assess and debug EMC performance before formal lab submissions. In automotive applications, this test setup follows the CISPR 25 standard, where the ground plane represents the vehicle chassis and requires direct LISN bonding to the test plane to achieve accurate results.

Previously, we discussed how to set up a conducted emissions test for a consumer product. This requires a ground plane, an 80 cm-tall table, a spectrum analyser or receiver, and a LISN. An isolation transformer is often also required to prevent excessive leakage current being drawn through the LISN, which could otherwise cause a circuit breaker to trip.

In this article, we will discuss how to set up conducted emissions testing for automotive applications. Note that defence and avionics conducted emissions set-ups are largely based on the same principles as automotive testing; we generally refer to this category as vehicle component EMC testing. In this type of test, the ground plane represents the vehicle chassis, whereas in commercial or industrial EMC testing, the ground plane represents earth.

Understanding this difference allows engineers to appreciate why, in this case, the LISNs must be directly bonded to the test ground plane. For automotive EMC testing, we typically follow the CISPR 25 set-up. The LISNs used have a defined RF impedance from 150 kHz up to approximately 110 MHz, which is the frequency range under test.

What You Need for a Bench-Top Conducted Emissions Setup 

In summary, what you need:

• A clean DC power supply (12 V, 24 V, or 28 V depending on the application). Note that high-voltage electronic units are outside the scope of this article.

• A ground plane, which can be copper, steel sheet, or aluminium. We recommend copper or galvanised steel sheet.

• Two CISPR 25 LISNs. The key difference between these and commercial/industrial LISNs is the frequency range and inductance value: CISPR 25 LISNs use 5 µH rather than 50 µH. The 5 µH inductance roughly represents the impedance of a 5 m-long cable (e.g. between a 12 V battery and a fuse box in a vehicle), whereas 50 µH represents approximately 50 m of wiring (e.g. between a wall socket and a building transformer).

• A 50 mm-tall insulating support between the device under test (DUT) and the ground plane.

CISPR 25 Bench-Top Test Configuration (Automotive)

Engineers often ask why the DUT must be supported 50 mm above the ground plane. After all, in reality, the unit is typically bonded directly to the vehicle chassis, and the cable run is often laid directly on the chassis surface. The short answer is that this requirement is historical in origin. Readers who wish to understand the reasoning in depth should refer to my colleague Ken Javor’s excellent article on the history of LISNs.

If your DC power supply is of high quality, the LISNs will often sufficiently filter out noise generated by the supply itself. However, if you are unsure, it is advisable to use a battery instead, or place an additional filter between the DC power supply and the LISNs.

As always, it is essential to follow the test standard exactly. For example, under CISPR 25, the wiring length between the LISN and the DUT is limited to approximately 20 cm. Wiring longer than this can lead to inaccurate test results.

Curious engineers will inevitably ask why this cable length must be so short. Unfortunately, this cannot be fully explained in a short article. In brief, to understand this requirement, one must first understand the purpose of LISNs in conducted emissions testing. LISNs provide a defined RF impedance over a defined frequency range. Cables, however, can exhibit resonance when they become electrically long—in other words, they behave as transmission lines. Once a cable is electrically long, its impedance becomes unstable due to resonance. Using such a cable in conjunction with a LISN defeats the very purpose of using the LISN. A 20 cm cable ensures that the wiring remains electrically short at the highest test frequency (approximately 110 MHz), thereby avoiding cable resonance, so the LISNs control the impedance.

Figure 1 CISPR 25 LISN set-up

Once you examine the test set-up diagram defined by CISPR 25, it becomes straightforward to replicate the conducted emissions bench set-up. See the figure below.

Figure 2 Two LISNs are used for the set-up, note the unmeasured LISN needs to be terminated with a 50 ohm

Next, you will need to configure EMC test software (such as EMCview) to control the spectrum analyser during the sweep. Alternatively, you may use a receiver, which will most likely include CISPR 25 test functionality. You might also wonder why EMC software is necessary. If you examine the test results shown in Figure 3, you will notice a “jump” at 30 MHz. This occurs because, above 30 MHz, the resolution bandwidth RBW must be changed in accordance with the standard, resulting in a shift in the noise floor. EMC software handles this automatically, saving significant time and effort.

Figure 3 Benchtop pre-compliance conducted emission results vs lab results

The Devil Often Lies in the Details

Because this type of test is often performed by electronics design engineers rather than specialist EMC engineers, subtle mistakes are frequently made. One of the most common errors is forgetting to terminate the other LISN (the unmeasured one) with a 50 Ω load. This oversight can result in measurement errors of several dB.

Another important topic is the 1 µF input capacitor on the LISN. Some commercially available LISNs include a switch that allows this capacitor to be enabled or disabled. The capacitor is useful for conducted emissions testing but must be switched off during transient testing. Failure to do so can result in the capacitor inadvertently shorting the transient.

It is worth noting that the LISNs shown in this article do not include a built-in 1 µF input capacitor. The manufacturer recommends that users install this capacitor externally to achieve the correct test set-up.

Using a Reference Noise Source

It is always beneficial to use a reference noise source when measuring noise levels with a spectrum analyser, as this allows you to verify the integrity of the test set-up. In accredited test laboratories, it is standard practice to validate set-ups using a reference signal source. I came across a Texas Instruments evaluation board (TPS54561EVM-555) with published conducted emissions results and have adopted it as my personal reference noise source.

A video demonstration can be found here.

https://www.youtube.com/watch?v=KHxbk4eToXs