CISPR 25: “Vehicles, boats and internal combustion engines - Radio disturbance characteristics - Limits and methods of measurement for the protection of on-board receivers”
CISPR 25 governs emissions related to automotive components, both for ground vehicles and boats. It includes both component level and vehicle level tests. The current version is the 5th edition published in 2021, and you can purchase it here. Some of the component level test methods in CISPR 25 are very similar to comparable methods in MIL-STD-461, but they are not at all similar to emissions testing done for the FCC via ANSI C63.4. (You can check out one of my presentations that dives pretty deep into that comparison.) And obviously MIL-STD-461 does not have an equivalent of the full vehicle testing found in CISPR 25.
More than anything, CISPR 25 is aimed at self compatibility. Historically this meant limiting emissions that would interfere with onboard radios (AM/FM/DAB) so as to avoid customer complaints. Today with vastly more complex vehicle electronics and the HV components of EVs, there are a lot more “victims” that can be affected by emissions than just AM/FM radio. [Aside: I once worked on troubleshooting an EV where a CISPR 25-noncompliant inverter interfered with the vehicle’s twisted/shielded CAN lines to the extent that whenever the driver stepped on the “gas” pedal, the battery control module shut down, requiring a full restart. In that case the solution was improving the shield terminations on the CAN lines.]
In general, vehicle level tests are done in four possible modes: key on, engine off; internal combustion engine (ICE) in driving mode (spinning wheels on a dyno); EV in charging mode, and EV in driving mode. (A plugin hybrid electric vehicle would need to test in all four modes.) For testing on the full vehicle, you need to disconnect each antenna from its receiving radio module, and feed the antenna coax into the measurement receiver, often through an impedance matching unit.
At the component level, CISPR 25 includes both conducted and radiated emissions tests. Conducted emissions can be measured via voltage measured from an artificial network (AN, equivalent to a LISN in the aerospace/defense world), or via a current probe measured at a minimum of two locations along the cabling. For radiated emissions the main test method is conducted in a semi-anechoic chamber, with an informative Annex F describing a stripline method. It currently does not provide for a reverb chamber method.
TIP:
Pay close attention to the flowchart in Figure 1 of CISPR 25. Because of the different limit lines for different detectors(peak, quasi peak, and average) and how they apply to narrowband vs. broadband noise sources, there are a lot more steps before saying something definitively “passes” or “fails” than in a method like MIL-STD-461 RE102.
TIP:
CISPR 25 is a critical standard for the HV components of electric vehicles as well as the traditional 12 Vdc systems. Take this as one consultant’s experience, but every EV on which I’ve done troubleshooting had problems with an HV component that failed CISPR 25 testing (particularly inverters and DC/DC converters). I have not yet had to troubleshoot on an EV where every component passed CISPR 25. It is a major challenge to design an HV inverter that passes CISPR 25 limits, but it can be done–and it’s likely worth investing the extra time to design a compliant system, or the extra money to buy one. You can also see my presentation on Noise Sources in EVs for more.
TIP:
Getting to the audio head unit (radio) to disconnect the AM/FM radio for vehicle level testing can be painful, and the sweeps can take a long time, especially if there are a lot of frequencies where quasi peak measurements are needed. CISPR 25 is not a regulatory test, so judgment of compliance is up to the manufacturer. It may make more sense to run the vehicle and tune through stations looking for audible problem areas, then only run the CISPR 25 sweeps on frequency ranges where there are issues. Making sure the audible test is legitimate has its own set of challenges, but once set up, it can run considerably faster than full CISPR 25 sweeps.