Your cart is currently empty!
DFSS Instructor Prep · Module 8
Layer B — Tier 2 Domain Depth
EMC, ESD & Thermal Design for Automotive
The three invisible stressors. Electromagnetic compatibility, electrostatic discharge, and thermal management are the three “physics-driven” disciplines that every automotive electronic product must survive — yet they’re often treated as late-stage validation checks rather than design inputs. This module gives you instructor literacy across all three: the standards (CISPR 25 Ed 5, ISO 11452, ISO 7637, ISO 10605), the design principles, and the integration with DFSS DMADV.
What’s in this module
- The three pillars — EMC, ESD, Thermal — and how they relate
- EMC fundamentals — emissions vs immunity, conducted vs radiated
- CISPR 25 Edition 5 — the dominant emissions standard
- ISO 11452 series — component immunity
- ISO 7637 — transient immunity on power lines
- ECE R10 — vehicle-level type approval
- ESD fundamentals & ISO 10605
- EMI mitigation — design techniques that actually work
- Thermal design fundamentals
- Thermal validation & reliability
- DFSS linkage — DMADV through the three pillars
- Instructor facilitation pattern
- Self-check (10 questions)
1. The three pillars and how they relate
EMC / EMI
The product must not emit too much radio-frequency energy (emissions); must function correctly in the presence of external RF (immunity). Continuous, frequency-domain phenomena.
ESD
The product must survive electrostatic discharge events — short, high-voltage transients from human bodies, charged metal, or other charged objects. Single-event, time-domain phenomena.
Thermal
The product must dissipate the heat it generates and survive ambient extremes (-40 °C to +125 °C in engine bay, much wider for HV power electronics) over 15+ years.
The common pattern across all three
Each of these three is a noise factor in the DFSS P-diagram sense (Module 5 §2). The product’s control factors — PCB layout, shielding, grounding, thermal materials, package selection — must produce a response that meets specification across the full range of these noise factors. This framing turns each pillar from a separate engineering discipline into a unified DFSS robust-design problem.
Why these are commonly underweighted
EMC, ESD, and thermal problems often only surface late in the project — during DV testing, vehicle EMC chamber, or warranty data analysis. Fixing them then is 10–100× more expensive than designing for them upfront. The DFSS opportunity is to move these analyses into Analyze and Design phases rather than reacting in Verify.
2. EMC fundamentals — the 2 × 2 matrix
EMC literacy starts with knowing which of the four quadrants you’re talking about.
| Conducted | Radiated | |
|---|---|---|
| Emissions (product sources noise) | Noise on power/signal cables. Test: LISN (Line Impedance Stabilisation Network) + spectrum analyser. Frequency: 150 kHz – 108 MHz typical | Noise radiated as RF waves. Test: anechoic chamber, antennas at 1 m (CISPR 25). Frequency: 30 MHz – 2.5 GHz typical |
| Immunity (product survives noise) | Transients injected on power/signal lines. ISO 7637, BCI per ISO 11452-4 | External RF field illuminates product. ISO 11452-2 (absorber chamber), ISO 11452-11 (reverberation) |
Why 1 m measurement distance matters
CISPR 25 specifies radiated emission measurements at 1 m — much closer than the 3 m or 10 m of consumer EMC standards. The reason is automotive-specific: vehicle antennas (AM, FM, GPS, cellular) are physically close to electronic modules. The short measurement distance correlates better with the actual in-vehicle interference scenario.
3. CISPR 25 Edition 5 — the dominant emissions standard
If a participant says “we did CISPR 25”, they’re talking about the component-level emissions test that protects on-board receivers (AM/FM, GPS, Bluetooth, cellular) from interference. Current edition: 5 (2021); this edition added special test setups for EVs and hybrids during charging.
The five limit classes
1
Lenient
Low-risk scenarios, far from antennas
2
General
General components / modules
3
Minimum
Typical OEM baseline
4
Stringent
Near antennas / infotainment
5
Most strict
HV power electronics, safety-critical
Two practical truths
- The OEM decides the class per component — not CISPR 25 itself. The CSR specifies “Class 3 for component X mounted under the dash; Class 5 for HV inverter”.
- The standard’s frequency range is 150 kHz – 2.5 GHz, but emerging wireless (UWB, 5G mmWave) operates beyond this — a known gap the next CISPR 25 edition is expected to address.
3.1 What an EV-specific Edition 5 setup looks like
CISPR 25 Edition 5 (2021) added test configurations for the special EV/HEV cases that previous editions didn’t cover:
- Vehicle parked but charging — connected to AC or DC charger; emissions still must protect on-board receivers
- HV components stand-alone — inverter, OBC, DC-DC, BMS slave modules tested independently with HV harness
- Different operating modes — charging (AC/DC), driving (motor running), traction-off (key on, motor stopped)
This is structurally important because the HV power electronics in EVs (Module 6) are the noisiest electrical devices ever put into vehicles — switching kilo-amps at tens of kHz. Without explicit test setups, an EV-class component could pass CISPR 25 in a “driving” mode but fail catastrophically when charging.
4. ISO 11452 series — component-level immunity
The complement to CISPR 25. While CISPR 25 measures what the product emits, ISO 11452 measures what the product can withstand.
| Part | Test method | What it simulates |
|---|---|---|
| ISO 11452-2 | Absorber-lined shielded enclosure (ALSE) — RF field illumination from an antenna | Vehicle driven near a transmitter (AM/FM tower, cellular base station, radar) |
| ISO 11452-3 | TEM cell — controlled E-field exposure | Cost-effective alternative for small components |
| ISO 11452-4 | BCI — Bulk Current Injection on the harness | Currents induced in the wiring by external RF (1 MHz – 400 MHz) |
| ISO 11452-5 | Stripline — uniform field over a length | Long-cable scenarios |
| ISO 11452-9 | Portable transmitter close to the product | Mobile phones, hand-held radios near electronics |
| ISO 11452-11 | Reverberation chamber | Statistically uniform RF environment; faster than absorber chamber |
Test severity levels (e.g. 100 V/m, 200 V/m) are specified by the OEM in the CSR. Field strengths up to 200 V/m are common for safety-critical components.
The four performance criteria
ISO 11452 tests record one of four outcomes:
- Class A — functions normally during and after exposure
- Class B — minor degradation during exposure; self-recovers
- Class C — temporary loss of function; recovers only after manual reset
- Class D — permanent damage; not acceptable
5. ISO 7637 — transient immunity on power lines
The 12 V battery in a vehicle is a remarkably nasty power source. Voltage transients from inductive loads (relays, motors, ignition coils), load dumps, and alternator switching create well-characterised pulse shapes the product must survive.
5.1 The ISO 7637-2 pulse menagerie
| Pulse | Source | Severity |
|---|---|---|
| Pulse 1 | Inductive load disconnect on supply line | -50 to -150 V, 2 ms |
| Pulse 2a | Inductive load disconnect on parallel circuit | +25 to +50 V, 50 µs |
| Pulse 2b | DC motor turn-off | +10 V, 0.2 to 2 s |
| Pulse 3a / 3b | Switching transients on supply | -150 V / +100 V, 100 ns rise, 100 ns to 0.1 ms duration; fast repetitive |
| Pulse 4 | Starter motor cranking — voltage drops | Down to 4–6 V for tens of seconds |
| Pulse 5a / 5b | “Load dump” — alternator disconnect with battery loose | +87 V to +120 V (centralised clamped to lower in newer cars), up to 400 ms |
Why “load dump” is the legendary one
The classic load-dump scenario: alternator is charging the battery, the battery cable comes loose at high RPM, the alternator’s stator inductance dumps its energy into the rest of the electrical system — up to 120 V on what should be a 14 V bus, for hundreds of milliseconds. Every connected electronic module must survive this without permanent damage. Modern transient-voltage-suppressor (TVS) diodes and central load-dump clamping have made this less brutal than it once was, but the test (Pulse 5) is still part of every automotive electronics qualification.
6. ECE R10 — vehicle-level type approval
While CISPR 25 and ISO 11452 are component-level, ECE R10 is the vehicle-level type-approval regulation in Europe, India, and most non-US markets. India’s CMVR refers to ECE R10 for vehicle-level EMC.
R10 tests cover:
- Vehicle radiated emissions (protects external receivers — TV, FM at the road side)
- Vehicle conducted emissions during charging (EV-specific)
- Vehicle immunity to external RF (drive past a TV transmitter — vehicle must function)
- Electrostatic discharge to vehicle
- Transient emissions/immunity on charging interface
For your Indian cohort specifically
Vehicle type-approval in India requires R10 compliance. Component compliance with CISPR 25 and ISO 11452 is a necessary but not sufficient condition. The interaction between multiple compliant components in a real vehicle can still produce R10 non-compliance — which is why vehicle-level integration testing exists at the OEM and supplier-collaborative level.
7. ESD — the single-event killer & ISO 10605
Electrostatic discharge is the most common single source of electronics damage in the field. A person walking across a carpet can build up tens of kV; touching an exposed pin discharges through whatever’s connected. Connector pins, switches, USB ports, charge inlets — all are entry points.
Three discharge models are commonly used:
Human Body Model (HBM)
Models a charged person touching a pin. RC circuit: 1.5 kΩ + 100 pF discharges through the device.
Typical spec: 2 kV minimum on every pin; 8 kV+ on connector pins
Machine Model (MM)
Models a charged tool or fixture. Lower R, higher peak current. Sometimes deprecated in favour of CDM.
Typical: 200 V
Charged Device Model (CDM)
Models the device itself being charged and then making contact (e.g., during automated assembly). Very fast (sub-ns) but lower voltage.
Typical: 500 V – 1 kV
ISO 10605 (automotive)
The automotive-specific ESD test — different RC values than HBM (330 Ω + 150 pF, or 2 kΩ + 330 pF for handler-during-service). Both contact and air discharge.
Typical: ±8 kV contact, ±15 kV air (powered); ±15 kV / ±25 kV (unpowered/handling)
Why the automotive RC values differ
Human-body capacitance and resistance vary by scenario (person standing vs sitting; wearing leather vs cotton; metal tool in hand). ISO 10605 captures the specific scenarios relevant to automotive: a person at the dashboard, a service tech with an arrester tool, a child poking a USB port. The 330 Ω + 150 pF model is a deliberately harder pulse for in-vehicle scenarios.
A real-world DFMEA failure mode
ESD damage from a service technician touching a connector pin during service is a recurring warranty pattern. A robust DFMEA includes the failure mode “ESD strike during service” with appropriate protection (TVS diodes, shielded connector body, “do not touch the gold pins” service procedures). This is also a moment to point at Module 3’s “lookup-table failure mode” anti-pattern: writing “ESD damage” is incomplete; “loss of communication on CAN_H due to permanent gate-oxide damage from ESD strike on connector pin 3 during service operation” is the physics-statement version.
8. EMI mitigation — design techniques that actually work
EMC problems are almost always traced to one (or a combination) of these design factors. Knowing the catalogue lets you ask the right diagnostic questions.
8.1 The classic EMC mitigation hierarchy
- Suppress at source. Slow switching edges (lower dV/dt and dI/dt), spread-spectrum clocking, soft-switching topologies in power converters. The cheapest place to fix EMC.
- Contain by shielding. Metal enclosures, conductive gaskets, finger-stock around lids. Effectiveness depends on aperture size relative to wavelength.
- Filter on cables. Common-mode chokes, feed-through capacitors, ferrite beads where cables exit the enclosure. The cable harness is the dominant antenna for radiated emissions.
- Ground correctly. Star ground for low-frequency analog, ground plane for high-frequency digital. Separate analog and digital grounds, connected at a single point.
- Route carefully on PCB. Loop areas of high-frequency current paths minimised; return paths kept under signal lines (controlled impedance); high-speed signals away from edges.
- Differential signalling. LVDS, FPD-Link III, GMSL, automotive Ethernet — all use balanced pairs to cancel common-mode noise. Critical for HUD video links, ADAS camera links.
The textbook insight
About 80% of automotive EMC problems come from cabling, grounding, and shielding implementation — not from the active circuit itself. This is why “EMI is a wiring-harness problem” is a fair starting hypothesis from a Tier-1 perspective. Your WH cohort should think of EMI as part of their core design responsibility, not a customer problem.
8.2 Harness-specific EMC considerations
- Twisted-pair routing — for CAN, FlexRay, Ethernet. Twist pitch chosen for frequency band.
- Shielded cable — for HUD video links, HV power, ADAS sensor links. 360° shield termination (Module 6 §11) is critical and frequently a defect source.
- Routing relative to vehicle ground (chassis) — distance from ground affects coupling.
- Loop area minimisation — signal & return wires close together.
- Separation of victim/aggressor cables — keep low-level signals away from high-current HV cables.
- Drain wire termination — even shielded twisted pair fails EMC if the drain wire isn’t terminated correctly at both ends.
9. Thermal design fundamentals
The third pillar. Thermal stress accelerates every Module 2 failure mechanism — Arrhenius (chemical degradation), Coffin-Manson (solder fatigue), polymer ageing. Doing thermal design well in DMADV’s Design phase prevents a long warranty tail.
9.1 The four heat-transfer mechanisms in automotive electronics
| Mechanism | How it works | Where it matters |
|---|---|---|
| Conduction | Heat flows through solid materials proportional to thermal conductivity (W/m·K) | Die → leadframe → PCB → heatsink path; thermal interface materials (TIM) |
| Convection (natural / forced) | Air movement carries heat away from surfaces. Natural for sealed enclosures, forced for cooled modules. | Inverter cooling, BMS slave modules, ECU enclosures with fan or coolant |
| Radiation | IR emission proportional to T⁴ (Stefan-Boltzmann). Significant only at hot surfaces (>100 °C). | Power-electronics surfaces, HUD PGU under solar load |
| Liquid cooling | Coolant loop (typically water-glycol) absorbs heat at the device, dumps it at a heat exchanger. | HV inverter, OBC, DC-DC, traction motor, increasingly common |
The 10°C rule (Module 2 connection)
Arrhenius says — a rough rule — every 10 °C reduction in operating temperature roughly doubles the time-to-failure for thermally-driven mechanisms. Knocking 20 °C off a hot-spot temperature can quadruple field life. This is a powerful “small change, large outcome” story for the DFSS Verify phase.
9.2 Thermal-design vocabulary every senior conversation uses
| Term | Meaning | Typical unit / value |
|---|---|---|
| θJA (junction-to-ambient) | Total thermal resistance from semiconductor junction to ambient air | °C/W; depends on package, PCB, airflow |
| θJC (junction-to-case) | Resistance from die to package case (cooled side) | °C/W; lower is better |
| Tj (junction temperature) | Semiconductor die temperature — the engineering limit | Typical max: 150 °C automotive, 175 °C extended, 200 °C SiC |
| TIM (thermal interface material) | Material between die package and heatsink — grease, pad, phase-change, gel | Conductivity 1–8 W/m·K typical; gap-fillers higher |
| CFD (computational fluid dynamics) | Simulation of fluid flow and heat transfer — Ansys Fluent, Star-CCM+, Simcenter Flotherm | The Shared Service Thermal/CFD specialist owns this |
| De-rating | Operating components below their absolute max ratings for life | Typical: 80% of max electrical and 50% of max junction temperature |
10. Thermal validation & reliability
Two strands of testing — measure actual temperatures, then accelerate to predict life.
Measurement
- Thermocouples embedded at key components during dyno or in-vehicle running
- IR thermography — non-contact mapping of surface temperatures
- Junction temperature estimation via package temperature + θJC + load
- Internal die-temperature sensors on power semiconductors (most modern SiC/IGBT)
Acceleration tests (links to Module 2)
- Thermal cycling (TC) — typically -40 °C to +125 °C, 5 °C/min ramp, ~1000–2000 cycles. Tests solder fatigue (Coffin-Manson).
- Thermal shock (TS) — faster transition (2-zone chamber), more severe than TC.
- Power cycling (PC) — heating from device’s own dissipation rather than chamber. Tests wire bond fatigue, die-attach degradation. More relevant than TC for power semiconductors.
- HTOL (High-Temperature Operating Life) — continuous operation at elevated T (e.g., +125 °C) for 1000 hours. Tests Arrhenius mechanisms.
- HAST (Highly Accelerated Stress Test) — combined T + humidity + bias.
1000 hours
Typical HTOL at +125 °C — proxy for 15 years field life
2000 cycles
Typical TC -40 to +125°C — proxy for 15 years thermal exposure
Tj < 150 °C
Automotive Si junction temperature limit (175°C extended, 200°C SiC)
11. DFSS linkage — DMADV through the three pillars
| DMADV Phase | EMC / ESD / thermal content that lands here |
|---|---|
| Define | Customer EMC requirements (CISPR 25 class, ISO 11452 severity, ISO 7637 pulses), ESD class per ISO 10605, thermal operating range, life target. CSR explicit. |
| Measure | Operationalised CTQs: emissions margin (dB below CISPR 25 limit); immunity Class A at 200 V/m; ESD ±15 kV air, ±8 kV contact powered; Tj < 130 °C with 20 °C margin to 150 °C max; θJA target. |
| Analyze | Concept selection: shielded vs unshielded; passive vs liquid cooling; central vs distributed transient protection. EMC simulation (Ansys HFSS, CST), thermal CFD (Fluent, Flotherm). DFMEA captures EMC, ESD, thermal failure modes. |
| Design | P-diagram with EMC/ESD/thermal as noise factors. Robust design DOE: vary shielding, TIM material, ground topology. Tolerance design on package thermal resistance, shield aperture size, harness twist pitch. Pre-compliance EMC scans during design iteration. |
| Verify | CISPR 25 emissions test; ISO 11452-2/-4 immunity; ISO 7637-2 transient pulses; ISO 10605 ESD; thermal cycling -40 to +125 °C; HTOL; vehicle-level ECE R10 (if applicable). All to OEM CSR. |
Cross-module integration
An HUD product (Module 7) integrates with this module in multiple places:
- The PGU board generates RF emissions on its video links → CISPR 25 Class 5 (high stringency near IVI)
- The LVDS/FPD-Link III cable shield termination drives EMC outcomes
- Solar load through the HUD aperture creates a thermal extreme that drives PGU thermal design
- ESD on the touch buttons and connector pins requires ISO 10605 protection
12. Instructor facilitation by function
| Function | EMC / ESD / thermal angle that lands |
|---|---|
| Shared Service — Thermal / EMI / CFD lead | This is their core domain. Treat as cohort SME. Engage on CFD methodology, thermal CTQ flow-down, EMC simulation. |
| Testing Center Manager | Owns CISPR 25 chamber capacity, ISO 7637 generator, ISO 10605 ESD gun. Capacity planning for EV-era loading. |
| EI — System Eng Lead / AGM-EI Software | EMC-immunity software response — should the module reset on a Class B disturbance? ISO 11452 performance classification mapping. |
| EI — Sensor Developer / AR HUD PM | Differential signalling (LVDS/FPD-Link III/GMSL), shield termination, solar-load thermal management at HUD aperture. |
| WH (LV/HV) leads | Twist pitch, shield termination, victim/aggressor separation. EMC is “wiring harness physics” — they own most of it. |
| CDDC — AGM & connector leads | 360° shield-terminating connectors, EMC robustness of HV connector designs, ESD protection in service. |
| Shared Service — Advance Materials | Thermal interface materials, conductive gasket materials, polymer thermal stability. |
| SD Coordination / Project Mgmt | EMC budget allocation across components; coordinating multi-supplier integration for vehicle-level ECE R10. |
A diagnostic question for any electronics DFSS project
“What CISPR 25 class are you targeting, what’s your Tj design margin to the 150 °C limit, and which ISO 10605 strike scenario is in your DFMEA?” Three concrete numbers that test EMC, thermal, and ESD literacy simultaneously.
Instructor self-check
Ten questions across EMC, ESD, and thermal — calibrated to the level of conversation you’ll be in.
Q1. A participant says their module passed CISPR 25. The most useful follow-up question is:
A. “What’s the model number?”
B. “Was it tested at room temperature?”
C. “Which class — and how much margin to the limit? Class 3 with 0 dB margin is brittle; Class 5 with 6 dB margin is robust”
D. “What colour?”
Correct — CISPR 25 has five classes; OEMs specify per component. Margin to the limit line is the engineering quality measure, not just “pass”.
Q2. CISPR 25 Edition 5 (2021) added what to previous editions?
A. Eliminated the radiated emissions test
B. Specific test setups for EV and HEV components, including vehicle-while-charging scenarios
C. Replaced the standard with ECE R10
D. Made all classes more lenient
Correct — Edition 5 was a significant update for the EV era, recognising that charging introduces emissions scenarios previous editions didn’t address.
Q3. The “load dump” pulse (ISO 7637-2 Pulse 5) is generated by:
A. A driver pressing the brake
B. The engine starting
C. Closing the door
D. Alternator disconnect at high RPM with battery loose — the alternator’s stator inductance dumps energy into the bus, creating a transient up to ~120 V on the 14 V system for hundreds of milliseconds
Correct — load dump is the canonical worst-case automotive supply transient. Every electronics module must survive it.
Q4. Approximately 80% of automotive EMC problems trace back to:
A. Cabling, grounding, and shielding implementation — not active circuits
B. The active circuit design
C. The CISPR 25 standard itself
D. Customer test fixtures
Correct — this is the well-known pattern. The wiring harness is the dominant antenna for radiated emissions and the primary aggressor-victim coupling path. Yazaki owns this physics directly.
Q5. ISO 10605 is the automotive-specific ESD standard. A typical specification for a powered electronic module is:
A. 0.5 kV contact, 0.5 kV air
B. 2 V contact only
C. ±8 kV contact and ±15 kV air discharge, typically with the 330 Ω + 150 pF in-vehicle network
D. ESD is not tested on automotive
Correct — ±8 kV/±15 kV is the typical powered-module spec; higher levels for service / unpowered scenarios.
Q6. The “10 °C rule” from Module 2’s Arrhenius model says:
A. Components stop working at 10 °C
B. Roughly, every 10 °C reduction in operating temperature doubles the time-to-failure of thermally-driven failure mechanisms
C. Thermal cycling is only valid at 10 °C amplitudes
D. 10 °C is the lowest allowable storage temperature
Correct — this rough heuristic from Arrhenius is the single most useful thermal-design talking point. Knocking 20 °C off a hot spot can quadruple field life.
Q7. The four ISO 11452 immunity-test performance classes are:
A. 1, 2, 3, 4 — by frequency
B. Red, yellow, green, blue
C. AM, FM, DAB, GPS
D. Class A (normal function), B (minor degradation, self-recovers), C (temp. loss needs manual reset), D (permanent damage) — safety-critical typically requires Class A
Correct — the A-D performance framework is how OEMs specify what counts as “passing” immunity for a given function.
Q8. A power-electronics module passes CISPR 25 but fails ECE R10 at vehicle level. The most likely explanation is:
A. CISPR 25 is a fake standard
B. Component-level compliance is necessary but not sufficient — coupling between multiple compliant components, or harness routing in the actual vehicle, can produce vehicle-level non-compliance
C. The vehicle was the wrong colour
D. ECE R10 is older than CISPR 25
Correct — vehicle-level EMC integration is a genuine engineering activity beyond component-level testing. Tier-1 / OEM collaboration is required.
Q9. The Si automotive junction temperature limit is typically:
A. 150 °C standard / 175 °C extended; 200 °C for SiC
B. 30 °C
C. -40 °C
D. 1000 °C
Correct — Tj limits are central to power-electronics thermal design. SiC’s higher limit is one reason it’s preferred for 800 V inverters.
Q10. The most diagnostic three-part question for any electronics DFSS project is:
A. What’s the colour, the size, and the cost?
B. AC or DC?
C. What’s your target CISPR 25 class, your Tj design margin to the 150 °C limit, and which ISO 10605 strike scenario is in your DFMEA?
D. Connector or solder?
Correct — these three test EMC, thermal, and ESD literacy with concrete numbers. A team that can answer all three is operating at the right level.