CAN interface

Hardware

The O3R has a built-in CAN-bus interface, with the CAN-High and CAN-Low lines on pin 4 and 5 respectively (cf. hardware diagram). Note that cables will need a terminating resistor like the E11589.

Software

CAN support requires a firmware version 1.4.30 and above.

Please note that the CAN interface is only accessible within Docker when using the --network host option.

Before utilizing the CAN interface, it needs to be set up. To verify whether the CAN interface is active and with which bitrate it operates, there are two possible methods: using the ifm3d CLI or Python scripts.

  1. Using the ifm3d CLI:

    $ ifm3d ovp8xx config get | jq .device.network.interfaces.can0
    {
    "active": false,
    "bitrate": "125K"
    }
    
  2. Using Python script: see the script below or download it from ifm3d-examples.

Note

New network settings will only be applied after reboot. The Python script performs a reboot. Using the ifm3d CLI a reboot can be performed by ifm3d reboot.

After activating the CAN interface and rebooting the VPU, we can verify the status of the CAN interface again using the ifm3d CLI as follows:

$ ifm3d ovp8xx config get | jq .device.network.interfaces.can0
{
  "active": true,
  "bitrate": "125K"
}

Available bitrates

Set the bitrate to one of these string values:

  • "1M"

  • "500K"

  • "250K"

  • "125K"

Example: Interfacing with the DTM425 RFID antenna using Docker

Step 1: Connect the DTM425 to the O3R and both to power.

Step 2: Create a minimal Dockerfile (filename: Dockerfile) like shown below (the example can be downloaded here):

FROM arm64v8/alpine

RUN apk add --no-cache python3 py3-pip

COPY requirements.txt /home/ifm/
# Create a venv and install python dependencies
RUN python3 -m venv /home/ifm/venv \
    && /home/ifm/venv/bin/pip install --requirement /home/ifm/requirements.txt

# Copy the can_example.py example.
COPY can_example.py /home/ifm/
# Copy the eds file to the container
# You can download the DTM425.eds file from https://www.ifm.com/us/en/product/DTM425?tab=documents
COPY DTM425.eds /usr/local/share

# Make the script executable
RUN chmod +x /home/ifm/can_example.py

# Activate virtual environment and run the script
CMD ["/home/ifm/venv/bin/python", "/home/ifm/can_example.py"]

This Dockerfile installs Python and the CANopen library for Python. The example script is then installed into the image and set to automatically execute when the container is run.

Step 3: Use the example Python script below (or download the script here) and download the required EDS file (filename DTM425.eds). Place the files in the same location as the Dockerfile.

# -*- coding: utf-8 -*-
#############################################
# Copyright 2024-present ifm electronic, gmbh
# SPDX-License-Identifier: Apache-2.0
#############################################
import time

import canopen


def connect():
    nw = canopen.Network()
    nw.connect(channel="can0", bustype="socketcan")
    nw.scanner.search()
    time.sleep(0.05)

    device = nw.add_node(nw.scanner.nodes[0], "/usr/local/share/DTM425.eds")
    device.nmt.state = "OPERATIONAL"
    time.sleep(0.05)

    return (nw, device)


def disconnect(nw, device):
    device.nmt.state = "PRE-OPERATIONAL"
    nw.disconnect()


def write_tag(device, data):
    memory_size = device.sdo[0x2182][0x4].raw

    if len(data) < memory_size:
        data = data + b"\x00" * (memory_size - len(data))

    for offset in range(0, memory_size, 8):
        length = 8 if offset + 8 <= memory_size else memory_size - offset
        device.sdo[0x2380].raw = offset
        device.sdo[0x2381].raw = length
        device.sdo[0x2382].raw = data[offset : offset + length]


def read_tag(device):
    memory_size = device.sdo[0x2182][0x4].raw
    data = b""

    for offset in range(0, memory_size, 8):
        length = 8 if offset + 8 <= memory_size else memory_size - offset
        device.sdo[0x2280].raw = offset
        device.sdo[0x2281].raw = length
        data = data + device.sdo[0x2282].raw

    return data


def main():
    nw, device = connect()
    data = b"\xDE\xAD\xBE\xEF"
    print("Writing tag:", data)
    write_tag(device, data)
    print("Reading tag:", read_tag(device))
    disconnect(nw, device)


if __name__ == "__main__":
    main()

The script writes the hex-value 0xdeadbeef to the RFID tag and reads the data from the tag. When scanning for the device, it is assumed that the RFID antenna is the only CAN device on the bus, besides the VPU itself.

Step 4: Build, deploy and run the Docker container:

docker build . -t dtm425_example
docker save dtm425_example | ssh -C oem@192.168.0.69 docker load
ssh oem@192.168.0.69 docker run --network host dtm425_example

Note that --network host is required to access the CAN interface.

The output of the last command should look like this:

Writing tag: b'\xde\xad\xbe\xef'
Reading tag: b'\xde\xad\xbe\xef\x00\x00\x00[...]\x00\x00\x00'

For more information on necessary setup steps for building and deployment, please see the linked pages.

Sample point

As defined by the CAN standards, the sample point is where the CAN signal is evaluated as dominant (0) or recessive (1). For example, if the sample point is set to 0.5, this means that the signal sample is taken at 50 % of the bit width. The bit width is equivalent to the inverse of the bitrate, meaning at a bitrate of 250 kbps, the bit width is (1/250kbps) = 4 us. With a sample point of 0.5, the measurements will be taken at half the width of the 4us bit, which will be at (2us + 4us*n), where n represents the bit count (0 for the first bit, 1 for the second bit, and so on).

The sample point does not have to be at the 50 % mark of the bit time; its optimal position depends on the characteristics of the signal. However, placing the sample point in the middle of the bit time is often more stable. At the beginning of the bit time, the signal may not have fully settled, and towards the end, the signal may start to transition to the next bit. Therefore, sampling in the middle helps in capturing a more stable and accurate signal.

Typically, a default sample point of 0.875 (or 87.5 % of the bit time) works well for most signals. However, if the signal is noisy, meaning the rising and falling edges are not clean, or other signal degradations are visible on the signal scope, adjusting the sample point might be necessary to ensure accurate data reception.

Changing the sample point

If CAN communication is unstable, adjusting the sample point may help improve signal robustness.

Note

Sample-point tuning is an advanced troubleshooting step. If you suspect a CAN timing issue, please contact us for guidance.