MSPM0G3519: MSPM0G3519 MCAN: How to achieve physical External Loopback for HW Verification?

Part Number: MSPM0G3519
Other Parts Discussed in Thread: SYSCONFIG

Hello TI Team,

We are currently working on a Design Verification (DV) process using the MSPM0G3519 LaunchPad and a custom board. Our goal is to verify the physical integrity of the CAN bus path (GPIOs + PCB Traces + Transceiver) using a single node setup.

We configured the MCAN module in "External Loop Back Mode" via SysConfig, expecting the signal to travel out of the TX pin, through the transceiver, and back into the RX pin. However, we observed that the communication remains successful even if we physically disconnect the CAN Transceiver or cut the RX trace.

Upon consulting the MSPM0 G-Series 80-MHz Microcontrollers Technical Reference Manual (Rev. C), specifically Section 26.4.11.1 (External Loop Back Mode) and Figure 26-11 (attached below), we found the following statement:

"In this mode, the MCAN module performs an internal feedback from the Tx output to the Rx input. The actual value of the RX input pin is disregarded by the MCAN module."

For our hardware verification tests, it is mandatory to validate that the signal is physically leaving the MCU and returning via the Transceiver. The internal feedback mechanism described in the TRM defeats the purpose of our physical layer test.

My Questions:

  1. Is there any configuration or "Test Mode" in the MSPM0 MCAN module that allows us to force the Core to sample the physical RX Pin while in Loopback mode?

  2. Since we are testing with a single node (no other device on the bus to provide an ACK), we cannot simply switch to "Normal Mode" because it results in immediate ACK Errors "lastErrCode = 3"

  3. Is there a recommended method to perform a "Physical Loopback" (TX Pin -> Transceiver -> RX Pin) on a single node without triggering ACK errors, or is this physically impossible due to the M_CAN IP architecture?

We need to confirm if the M_CAN IP is capable of ignoring ACK errors while strictly listening to the physical RX pin.

Thank you for your support.MCAN_Figure26-11_ExternalLoopback_Issue.jpg

  • Hello,

    As you've noted, the Loopback mode is to confirm out MCAN peripheral is working as expected.  It's validating our integrated hardware self test.

    For your test, how do you expect the TX signal make it back to the RX pin?  From the MCU perspective you want the MCU to behave as normal and you need the Transceiver to perform the loopback, right?  Does the transceiver have this functionality?    Currently, I don't understand how this could be an MCAN/MSPM0 feature.  

    Thanks,

    JD

      

  • Hello JD,

    Thank you for the clarification. I understand now that the MCAN's loopback mode is designed specifically for internal peripheral validation and, by design, disregards the physical RX input.

    We’ve noted that for our specific hardware path verification (TX -> Transceiver -> RX), a single-node loopback isn't supported by the M_CAN IP architecture due to the internal feedback and ACK requirements. We will move forward with a two-node setup to achieve the physical layer testing we need.

    Thanks again for your help in confirming the hardware behavior.

    Best regards,
    Fırat Can Özkök

  • Hello Firat,

    Of course.  Good luck with development and let us know if you run into any other questions.  

    Thanks,

    DJ