As the shift towards electrification gains momentum, an increasing number of companies are venturing into the development of products and systems used in electric vehicles or compact electric aerial vehicles like unmanned drones. These products, often referred to as electric control units (ECUs) in the automotive industry, typically operate on a DC supply voltage of less than 60 V (12V, 24V, and 48 V). Unless you’re dealing with high-power conversion, the current draw is usually below 10 A. This makes it quite feasible to establish an affordable benchtop conducted emission setup during the product’s research and development phase.
The advantage of having a pre-compliance conducted emission test setup lies in its ability to enable design engineers to identify potential design issues early on, thereby averting costly last-minute modifications. Conducted emission tests can provide reasonably accurate results and also serve as a reliable indicator of radiated emissions, as some of these emissions propagate through cable wiring.
Setting Up the Benchtop Test
Fortunately, there are manufacturers that offer low-cost line impedance stability networks (LISNs) characterized up to the 120 MHz range, as illustrated in Figure 1. In this setup, we follow the CISPR 25 standard for EMC testing. Given that the specifics of an EMC test setup often hinge on parasitic elements (like stray capacitance, which affects the common mode current return path), the key elements in this arrangement are:
- LISNs ground connection to the test ground plane: I typically use a galvanized steel plate, readily available at the local tool shop. The LISNs need to be bonded firmly to the plate.
- Wiring connection between the device under test (DUT) and the LISNs: The length should be around 20 cm, in accordance with the standard.
- Insulation support: Its height needs to be 5 cm above the test ground plane, as this dictates the parasitic capacitance between the DUT and the ground.
Minimizing the Ambient Noise
To perform the conducted emission test, we require an electromagnetically quiet environment. Typically, in the R&D workspace, such interference originates from:
- Noise generated by the benchtop power supply;
- Noise generated by nearby equipment, which radiates out and couples with the cables of the benchtop power supply and the wiring connection between the DUT and the LISNs; or
- Local radio transmitter signals coupled to the wiring.
It is generally considered good practice to install a DC filter between the benchtop power supply and the LISNs. Alternatively, depending on the noise characteristics of the supply, it is often possible to mitigate the noise by applying multiple-turn ferrite cores. During my assistance with clients in setting up tests in their offices, I observed that placing a two-turn ferrite core on the mains input cable to the power supply also significantly aids in noise suppression. This is particularly effective because switched-mode power supplies generate common-mode noise, so addressing noise from both ends of the power supply is advantageous.
Suppressing signals measured in the FM band due to local radio transmitters is often impossible without an EMC tent or chamber. However, it is worth noting that the characteristics of the radio signal spectrum in this band are distinctive and can be readily identified.
I have a video link that demonstrates how to minimize ambient noise [1].
The Devil Often Lies in the Details
Since, in this case, the test is often performed by electronics design engineers rather than skilled EMC engineers, unnoticed mistakes are often made. One of the most commonly seen mistakes is that the test engineer forgets to terminate the LISNs using the 50‑ohm termination. This can lead to measurement errors of up to a few dBs.
Another topic I would like to discuss is the 1 µF input capacitor to the LISNs. In certain commercially available LISNs, there’s a switch designed for toggling the 1µF input capacitor. This capacitor proves useful in conducted emission tests but must be switched off when conducting any form of transient test. Failing to do so might result in the capacitor inadvertently shorting the transient.
It’s worth noting that the LISNs featured in this article do not come equipped with the 1µF input capacitor. The manufacturer recommends that users install it themselves for the correct setup.
Utilizing a Reference Noise Source
It’s always advantageous to employ a reference noise source when assessing noise levels with a spectrum analyzer. This enables you to check the setup’s integrity. In accredited test labs, it’s common practice to verify setups with a reference signal source. I came across a Texas Instruments small evaluation board TPS54361EVM-555 [2] with conducted emission test results [3], and I’ve adopted it as my personal reference source.
