Two-Wire LIN networking with Atmel (Part 1)

Current-gen vehicles are packed with hundreds of sensors used to monitor and display parameters such as temperature and pressure. In most instances, these sensors are remotely located within a vehicle far away from the host microcontroller (MCU) responsible for monitoring and processing the sensor data.

As such, these sensors typically do not directly connect to a network (such as CAN or LIN) due to the vehicle wiring overhead associated with connecting to the network. One such method for overcoming this wiring limitation is to convert the standard three-wire LIN network to a two-wire implementation where the LIN slave nodes harvest power directly from the LIN bus master communication wire, thereby eliminating the need for an individual battery supply wire to each slave node.

linnetworkfigure1

As Atmel engineering rep Darius Rydahl notes, a standard LIN bus consists of a master node and up to 15 slave nodes connected to a single network. The physical LIN network is a three-wire configuration consisting of power (vehicle battery), ground and the LIN bus communication line. A pull-up resistor, RLIN, typically 1kΩ, is required on the master’s LIN bus line. Under normal LIN bus operation, this pull-up resistor provides a voltage bias on the LIN bus line to the slave nodes on the LIN network. It does not power the LIN slave nodes, rather slave node power is derived from the battery input to the device, as shown in Figure 1.

“It is possible to use a non-standard LIN network architecture that simplifies to two wires. This approach relies on the harvesting of power by a connected slave node directly from the LIN bus line, thus eliminating the need for an independent slave node battery supply line (see figure 2),” Rydahl told Bits & Pieces. “With the battery supply line removed, all that is required to power the slave node is a blocking diode, VDS and buffer capacitor, CVS_S, large enough to sustain the slave node supply voltage during the transmission of LIN data packets, which periodically pulls the LIN signal to ground.”

In this series, Bits & Pieces will outline the implementation of this two-wire approach and identify the inherent system-level tradeoffs that must be considered to fully realize a functional two-wire LIN network.

According to Rydahl, the key to successfully implementing a two-wire LIN network centers around the power requirements of the connected slave node. Simply put, the slave node must be supplied with sufficient power to maintain communication at the minimum system operating voltage: typically 9V. If this condition cannot be met, it is unlikely that the two-wire LIN implementation will be a viable solution. Key parameters that affect the slave node’s performance in a two-wire implementation include LIN bus power supply, slave node current consumption, slave node buffer capacitance and LIN Bus data protocol.

linnetworkfigure2

“In terms of the LIN Bus power supply, the two-wire LIN network is limited by the power supplied from the master to the slave node over the LIN bus line. Meaning, the supply to the LIN slave in this configuration will be dictated by the LIN bus master pull-up resistor, RLIN (see figure 2),” Rydahl continued. “The slave node has a fixed minimum input voltage operating requirement of 5.5V (reference: the Atmel ATA6624 LIN transceiver). In order to meet this minimum operating voltage requirement, the load current drawn by the slave node must not cause the voltage drop across the LIN master pull-up resistor to increase to the point at which the input voltage to the slave node drops below 5.5V.”

As Rydahl points out, this is the minimum operating voltage threshold for slave node voltage regulator operation. Indeed, figure 3 shows the maximum load current available to the slave node at the minimum supply voltage of 5.5V at different LIN master pull-up resistances.

linnetworkfigure3

“The 1kΩ master pull-up resistor specified in the LIN standard specification cannot be used in the two-wire configuration. The resistor is too large and, as a result, is unable to properly source the slave node load,” he said. “As such, the pull-up resistor must be reduced in size to the smallest value possible without exceeding the current limitation specification of the LIN driver. In the case of the typical Atmel LIN transceiver, the ATA6624, the recommended minimum pullup resistor value is 220Ω. Resistances lower than this could result in excessive current flow through the LIN transceiver when the LIN bus is asserted low.”

Interested in learning more about Two-Wire LIN networking with Atmel? Be sure to check out part two of this series here.

3 thoughts on “Two-Wire LIN networking with Atmel (Part 1)

  1. Pingback: Two-Wire LIN networking with Atmel (Part 2) | Bits & Pieces from the Embedded Design World

  2. Pingback: Two-Wire LIN networking with Atmel (Part 3) | Bits & Pieces from the Embedded Design World

  3. Pingback: Two-Wire LIN networking with Atmel (Part 4) | Bits & Pieces from the Embedded Design World

Leave a comment