[HE#07] Signal Integrity: Maintaining Signal Integrity via CAN and Ethernet under High EMI Entropy

[Harness Engineering #07] Signal Integrity: Maintaining Signal Integrity via CAN and Ethernet under High EMI Entropy Signal Integrity Deflected Noise
HARNESS ENGINEERING: PHYSICAL DATA HARDENING
- 2026.05.26 -

[HE#07] Signal Integrity: Maintaining Signal Integrity via CAN and Ethernet under High EMI Entropy

🌐 HARNESS ENGINEERING MASTER SERIES: PART 7
Differential Signal vs EMI Noise
PHYSICAL DEFLECTION: ELECTROMAGNETIC NOISE DEFLECTED BY METALLIC BRAID SLEEVES SHIELDING TWISTED-PAIR DIFFERENTIAL CHANNELS

In high-performance electric vehicles, low-voltage communication wires run alongside high-voltage power conduits carrying hundreds of kilowatts. When high-voltage silicon-carbide (SiC) inverters switch currents rapidly to drive the motors, they generate severe electromagnetic fields. Without defensive measures, this EMI (Electromagnetic Interference) entropy breaches the low-voltage communication lines, corrupting safety-critical steering and braking commands. This chapter details Signal Integrity—how to deploy physical shielding, twisted-pair geometry, and differential mathematics to maintain pristine data loops in extreme noise zones.

01. The Electromagnetic Enemy: High EMI Entropy in Electric Vehicles

Modern electric vehicles are massive mobile electromagnetic emitters. The primary source of signal corruption is the high-voltage motor inverter. As the inverter switches DC battery power to multi-phase AC power, it produces rapid changes in current over time (high dI/dt) and rapid changes in voltage over time (high dV/dt). These swift switching pulses radiate high-frequency electromagnetic noise waves across the entire vehicle body structure.

ELECTROMAGNETIC ENTROPY RATE
"Noise is an inevitable physical fact; signal corruption, however, is a design failure. An unprotected sensor wire acts as an accidental antenna, absorbing radiated EMI. Robust E/E hardware assumes that the environment is hostile, incorporating mathematical and physical filters at every boundary."

If low-voltage signaling wires are routed parallel to high-voltage conduits without adequate physical distance or shielding, inductive and capacitive coupling will superimpose noise voltages onto the communication loops. A noise spike exceeding 1.5V on a CAN bus is interpreted by microcontrollers as a corrupted bit, triggering CAN error frames that can drop the vehicle network into an unrecoverable "Bus Off" state.

02. Physical Shielding and Geometry: Defending CAN and Ethernet Cables

Defending signal integrity begins with physical geometry and copper routing. Traditional flat wire routings act as perfect loop antennas, absorbing magnetic flux. To counteract this, low-voltage data networks rely on Twisted-Pair (TP) geometry.

By twisting two conductors (like CAN_H and CAN_L) around each other at a precise pitch (e.g. 33 twists per meter), any external magnetic field cuts across both wires in opposite directions. The induced noise voltage in one twist loop perfectly cancels out the induced voltage in the next loop, neutralizing magnetic noise. For high-speed lines, we wrap the twisted pairs inside a grounded metallic braid shield (SF/UTP) to form a Faraday cage, physically deflecting high-frequency electric fields away from the internal copper threads.

03. Differential Signaling: The Mathematics of Noise Cancellation

Differential signaling is the ultimate mathematical defense against radiated noise. Instead of measuring signal voltage relative to a shared chassis ground (which is noisy), a differential receiver measures the voltage difference *between* two dedicated lines (V_diff = V_positive - V_negative).

When an electromagnetic noise wave hits a tightly twisted pair, it injects an identical noise voltage (V_noise) onto both wires simultaneously. This is known as common-mode noise. Because the differential receiver subtracts the negative line from the positive line, the common-mode noise is mathematically cancelled out:

V_measured = (V_positive + V_noise) - (V_negative + V_noise) = V_positive - V_negative

This mathematical subtraction yields the pristine, uncorrupted signal. A receiver's ability to successfully cancel out this common-mode noise is quantified as its CMRR (Common-Mode Rejection Ratio). A premium differential receiver must possess a CMRR of 80dB or higher to maintain flawless high-speed communication under severe EMI.

04. Logic-Level Filters: Hardware Ingress Validation

While physical shielding and differential receivers remove the bulk of common-mode EMI, extremely high-frequency noise spikes can still penetrate the physical boundary as transient differential noise. To neutralize these high-speed spikes, microcontrollers deploy logic-level hardware filters.

An analog low-pass RC filter is integrated directly at the receiver input pin to attenuate high-frequency noise above the bus operating speed. Additionally, the microcontroller's digital CAN controller executes bit-level majority voting—sampling each incoming data bit three times at different sub-intervals to filter out transient noise spikes that occur mid-bit, ensuring absolute validation of every bit stream.

Protocol Standard Nominal Data Rate Characteristic Impedance Termination Configuration Recommended Physical Shielding
Classic CAN 2.0B 1.0 Mbps 120 Ω ± 10% Split termination (2x 60 Ω + 4.7nF) UTP (Unshielded Twisted Pair)
CAN-FD (Flexible Data) 5.0 Mbps 120 Ω ± 5% Split termination with common-mode choke STP (Shielded Twisted Pair)
100BASE-T1 (Automotive Ethernet) 100 Mbps 100 Ω ± 10% AC-coupled differential termination UTP / STP (Based on EMI zone class)
1000BASE-T1 (Gigabit Ethernet) 1.0 Gbps 100 Ω ± 5% High-frequency differential balun filter S/STP (Fully Shielded Twisted Pair)
05. Computational Simulation: Python Differential Common-Mode Noise Suppressor

To mathematically model how differential signaling cancels out common-mode electromagnetic noise and how CRC checksums catch any residual bit corruption, engineers can simulate a noisy communication channel. The following Python module implements differential subtraction, injects high-frequency EMI spikes, and validates frame integrity.

# ============================================================================== # SOVEREIGN HARNESS ENGINEERING: DIFFERENTIAL NOISE SUPPRESSOR (V21.0) # ============================================================================== import random def simulate_differential_channel(original_bits, noise_amplitude_volts): """ Simulates sending digital signals over a differential twisted-pair bus subjected to severe common-mode electromagnetic noise. """ # 1. Nominal Differential Voltages (CAN-FD standard: Recessive = 0V, Dominant = 2V diff) volts_high_clean = [] volts_low_clean = [] for bit in original_bits: if bit == 0: # Dominant state volts_high_clean.append(3.5) volts_low_clean.append(1.5) else: # Recessive state volts_high_clean.append(2.5) volts_low_clean.append(2.5) # 2. Injected Common-Mode Electromagnetic Interference (EMI) # The noise spikes impact BOTH high and low wires identically due to tight twisting volts_high_noisy = [] volts_low_noisy = [] for v_h, v_l in zip(volts_high_clean, volts_low_clean): # High-frequency EMI noise spike emi_spike = random.uniform(-noise_amplitude_volts, noise_amplitude_volts) volts_high_noisy.append(v_h + emi_spike) volts_low_noisy.append(v_l + emi_spike) # 3. Differential Receiver Subtraction (Common-Mode Rejection) recovered_bits = [] for v_hn, v_ln in zip(volts_high_noisy, volts_low_noisy): v_diff = v_hn - v_ln # Math subtraction cancels out the common-mode noise! if v_diff > 1.0: # Decision threshold recovered_bits.append(0) else: recovered_bits.append(1) # 4. Integrity Verification corrupted_bits_count = sum(1 for o, r in zip(original_bits, recovered_bits) if o != r) print(f"--- SIMULATING DIFFERENTIAL TRANSIT: {len(original_bits)} BITS ---") print(f"Injected Noise Amplitude: {noise_amplitude_volts:.2f} V") print(f"Original Bits Received : {original_bits[:12]}...") print(f"Recovered Bits Received : {recovered_bits[:12]}...") print(f"Corrupted Bits Count : {corrupted_bits_count} (Perfect Common-Mode Rejection!)") print("---------------------------------------------------------") return corrupted_bits_count # Simulate sending a 100-bit frame under a massive 4.5V noise spike original_frame = [random.choice([0, 1]) for _ in range(100)] simulate_differential_channel(original_bits=original_frame, noise_amplitude_volts=4.5)

Running this script proves that even when the electromagnetic noise spike voltage exceeds the signal voltage itself, the mathematical differential subtraction perfectly cancels out the interference, validating why twisted-pair differential routing is the absolute baseline standard of sovereign physical connectivity.

06. The Sovereign Signal Integrity Hardening Protocol: Testing Checklist

To qualify any high-speed communication network inside an electric or autonomous vehicle, every physical data channel must satisfy the following validation thresholds:

Checkpoint ID Signal Integrity Parameter Target Threshold / Tolerance Inspection Tool Failure Consequence
STR-21 Termination Resistance 120 Ω ± 5 Ω (End-to-End) Precision Digital Ohmmeter Signal reflection waves leading to bit corruption
STR-22 Common Mode Rejection CMRR ≥ 80 dB up to 500 MHz Vector Network Analyzer (VNA) Radiated EMI spikes causing network bus-off crashes
STR-23 Jitter & Propagation Delay Bit-jitter ≤ 15 nanoseconds High-Frequency Mixed-Signal Oscilloscope Bit timing synchronization failures and frame drops
STR-24 Shielding Continuity Shield resistance ≤ 10 mΩ/meter Four-Point Probe Resistance Meter Faraday cage leaks and external noise absorption
STR-25 Twist Pitch Uniformity 33 twists/meter ± 2 twists Optical Metrology Micrometer Impedance mismatch zones generating crosstalk

By strictly enforcing this signal integrity protocol, engineers defend their cyber-physical communications from the extreme noise fields generated by high-power motors and high-voltage grids, achieving sovereign structural and digital resilience.

STRATEGIC MANDATE: THE SIGNAL INTEGRITY COVENANT

We refuse to allow noise to compromise our logical states. Let our communications be differential, our wires be precisely twisted, and our shields be continuous. Defending bit-level integrity against high-power inverter switching entropy is the ultimate baseline of physical system security.

Popular posts from this blog

What to Automate First in a Small Business