Anyone working on automotive hardware agrees: EMC is the most painful part of R&D.

If an antenna won't tune, at least you have a Smith chart. If RF power is low, you can swap the PA. EMC fails – you're tracing every single PCB trace, checking every shielding‑can seam, even wondering if a ground via was drilled slightly off‑centre. True story: I've seen a team spend three weeks tracking down a radiation emissions failure caused by a 0.1µF decoupling capacitor shifted by less than 2mm.

1. What Does EMC Actually Test?

EMC breaks down into two things:

·EMI (Electromagnetic Interference) – how much noise your product radiates.

·EMS (Electromagnetic Susceptibility) – how much external interference your product can withstand.

EMI is you talking too loudly and disturbing your neighbour. EMS is how well your house is soundproofed – whether thunder outside wakes you up. Certification requires passing both – neither can be skipped.

Under EMI: conducted emissions and radiated emissions.
Under EMS: conducted immunity, radiated immunity, electrostatic discharge (ESD), and electrical fast transients (EFT).

It sounds like a long list – but it all comes down to one principle: your product must not interfere with others, and must not be interfered with by others.

Why automotive is more brutal

The electromagnetic environment inside a vehicle is several orders of magnitude worse than consumer electronics. Ignition systems, motor drives, DC‑DC converters, in‑vehicle radar – all strong interference sources.

A T‑Box mounted behind the dashboard – body controller on one side, on‑board charger on the other, sunroof motor above, high‑power audio amplifier below. Take consumer‑grade EMC design and put it in that environment – you're basically naked.

  2. EMC Core Standards

·CISPR 25 – the emissions ruler

Internationally, automotive EMC has three main standards. CISPR 25 covers component‑level radio disturbance.

·Conducted emissions (CE): measures noise on power and signal lines – 150 kHz to 108 MHz.

·Radiated emissions (RE): measures electromagnetic fields – the 2021 version already extends to 6 GHz – not just OEM‑specific extensions.

·1‑metre vs. 3‑metre

Consumer‑electronics people are used to 3‑metre or 10‑metre chambers. Automotive uses 1‑metre – the antenna is closer. In ideal free‑space far‑field, the field strength at 1 metre is about 9.5 dB higher than at 3 metres – but that's a theoretical value. CISPR 25 component testing is performed in an ALSE (Absorber‑Lined Shielded Enclosure) – near‑field effects plus cavity reflections – you cannot directly convert using far‑field formulas. 1‑metre testing must be measured separately.

·ISO 11452 series – the immunity toolkit

ISO 11452 is the main component‑level immunity standard:

·11452‑2: radiated immunity – absorber‑lined chamber method.

·11452‑4: Bulk Current Injection (BCI).

·11452‑9: portable transmitter method.

·11452‑3: TEM cell; 11452‑8: harness radiation.

OEM requirements vary – check your project SOR.

·ISO 7637 and domestic standards

ISO 7637 covers power‑line transient disturbances.

·7637‑2: conducted transients along power lines – 12V system pulses 1 to 5b – each with different waveforms, amplitudes, and internal resistances.

·24V commercial vehicles: standard pulse 5a amplitude 174V, 5b (with clamping) 131V – not 200V+ – that's old OEM standards or extreme conditions.

Domestic:

·GB/T 18655 – modified adoption of CISPR 25 – not identical – some local adjustments.

·GB/T 33014 series – corresponds to ISO 11452.

·GB/T 21437 series – corresponds to ISO 7637.

  3. How It Differs from Consumer Electronics

People used to FCC Part 15 and EN 301 489 are often confused the first time they see automotive EMC. Same radiated emissions – FCC only requires 3‑metre or 10‑metre – CISPR 25 requires 1‑metre – antenna polarisation, frequency bands, and limit lines are all much stricter.

The fundamental difference is in the limit philosophy. Consumer electronics limits are set to avoid interfering with neighbouring devices. Automotive limits are set to avoid interfering with the vehicle's own safety systems. ABS, ESP, ADAS – if EMI affects them, it's not a user‑experience issue – it's a life‑safety issue.

Conducted transient testing – consumer electronics basically doesn't have it. In automotive, it's a core item.

BCI: ISO 11452‑4 does not specify a mandatory 200mA level – 200mA is an OEM internal severity level – not an international standard requirement.

CISPR 25 is a test method + limit standard – not the text mandatorily referenced by UN R10 for whole‑vehicle type approval – whole‑vehicle type approval follows UN R10.

  4. How Testing Runs – Where Things Go Wrong

Testing is performed in a third‑party automotive‑grade chamber.

·Sequence matters

Many engineers run emissions first, then immunity – reasoning that if emissions fail, the design has fundamental issues – fixing emissions first avoids rework. Don't do this for formal certification. Immunity tests – ESD, EFT, BCI – can damage samples or alter component parameters – making subsequent emissions data invalid. If you only have one set of samples – run all immunity first, then emissions – or better – use multiple sample sets for different tests.

·Radiated emissions failures

DC‑DC converter harmonics conducted along input lines then radiated; DDR and eMMC clock harmonics; poor PCB stack‑up creating excessive return‑path length causing differential‑mode radiation – it's the same few issues recurring. Solutions: add filtering, add shielding, control loop area. Three sentences to say – but doing it will make you bleed.

·BCI failures

BCI injects interference through current probes onto wiring harnesses. Most common failures: CAN/LIN communication dropping out, or sensor signals jumping. Root causes: insufficient interface filtering, poor harness impedance matching, or sensitive analogue signal traces too close to the connector on the PCB. MCU runaway or reset is also common – a classic BCI failure mode.

·Power transient failures

7637 pulse 5b load‑dump energy – an SMB‑package TVS failing is normal. Power stages need SMC‑package (DO‑214AB) high‑power TVS, combined with power resistors or power MOSFETs for current limiting. PPTC resettable fuses are too slow – don't use them as the primary protection for pulse 5b – I've seen more than one company learn this the hard way.

  5. Is Pre‑Testing Worth It?

Yes – and the earlier, the better.

If certification fails and you have to retest – one day of chamber rental plus engineer time is painful enough. Pre‑testing doesn't need to run the full certification suite – just scan the vulnerable bands.

For example, for a T‑Box with a 4G module – focus on radiated emissions around 800 MHz and 1800 MHz – these bands are easily interfered with by DDR clocks and power‑supply switching harmonics.

Note: you're scanning for interference from your own digital circuits – the module's own communication carriers are not counted as EMI noise – don't scan the wrong target.

Full pre‑scan for multi‑projects: RMB 30,000–50,000. Simple product local‑band pre‑scan: significantly less.

For EMC testing, contact BlueAsia at 13534225140 (King) or email king.guo@cblueasia.com.