Skip to content
Ohmic Audio

CAN Bus Physical Layer, Arbitration, And Fault Confinement

Canonical definition: A CAN network combines a protocol controller, compatible physical-layer transceiver, topology and termination system, bit timing, frame format, arbitration, error detection, fault confinement, and vehicle-specific application messages.

Controller Is Not Transceiver

application software
→ CAN controller: frames, arbitration, CRC, errors, counters
→ transceiver: logic-level TX/RX to differential bus signaling
→ twisted-pair network: topology, termination, stubs, connectors
→ other nodes and gateway

The ESP32-S3 includes one TWAI controller compatible with Classical CAN frame formats. It requires an external physical-layer transceiver. Its TWAI peripheral is not CAN FD-compatible, so it must not be attached to an FD network under the assumption that it will simply ignore the faster data phase.

Arbitration And Identifiers

CAN uses dominant and recessive bus states. Nodes compare the bus while transmitting; a node that sends recessive but observes dominant stops transmitting without corrupting the winning frame. The identifier participates in arbitration, but its application meaning is assigned by the vehicle or higher-layer protocol.

Classical CAN And CAN FD

FeatureClassical CANCAN FD
payloadup to 8 data bytesup to 64 data bytes
bit timingone nominal bit-rate regionnominal arbitration phase plus optional faster data phase
controller requirementClassical CAN controllerCAN FD-capable controller
application meaningstill vehicle- and protocol-specific

Physical-layer variants, transceiver capabilities, bit timing, sample point, propagation delay, and network topology must match the actual bus.

Error States

CAN nodes track transmit and receive errors. Fault confinement moves a repeatedly failing node through error-active, error-passive, and bus-off behavior. Bus-off is evidence of a fault or incompatible setup; it is not something to bypass until the wiring, timing, transceiver, and software are understood.

Safe Observation Workflow

  1. Identify the exact bus and gateway side from service information.
  2. Confirm Classical CAN versus CAN FD and the physical-layer variant.
  3. Use a correct transceiver, high-impedance instrumentation, and transmit-disable or listen-only safeguards.
  4. Verify termination and resistance only in the correct power and service state.
  5. Record bit rate, sample point, frame formats, bus load, errors, wake traffic, and vehicle state.
  6. Do not assign meaning to identifiers without supported vehicle-specific evidence.
No active vehicle-network experimentation

Develop arbitration, filters, transmit behavior, and bus-off recovery on an isolated bench network. Do not transmit into unknown propulsion, chassis, restraint, security, or driver-assistance networks.

Continue