• State
  • Locked Locked
  • Replies 5 replies
  • Subscribers 35 subscribers
  • Views 1919 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TCAN4550: TCAN4550 - CAN bus ringing

Josko M
Intellectual 950 points
Part Number: TCAN4550

Hi,

currently we are performing a compare test between TCAN4550 and competitor's MCP2517FD-H/JHA + MCP2542WFDT-E/MF combination. 

Both, TCAN and MCP are connected on the same CAN BUS. We have noticed something very strange, if the competitors CAN PHY is disabled, the signal on the BUS goes crazy if TCAN trys to send something. 

Figure1: both CAN PHYs, TCAN and MCP are active, and TCAN send test mesage (the CAN BUS is very short,  ~50 cm and it was unterminated at both ends). 

 

Figure 2: the MCPs CAN PHY is disabled, (Pin EN is high) and TCAN trys to send a test message (the CAN BUS is very short,  ~50 cm and it was unterminated at both ends). 

 

Unfortunately, we haven't managed to make third case where TCAN internal PHY will be disabled , in order to see, what's going on if MCPS CAN PHY trys to send a test message. 
The TCAN is controlled over SPI from Linux running on the Sitara processor. 

Q1: What would be the easiest way to disable TCAN PHY and set pins 10, 11 to high-impedance state? 

Q2: Do you have maybe any hint, why does the signal has a lot of ringings?  

If the schematic will be needed, I can share it over our local TI's office in our country. 

Br, 

Josko

  • 0
    TI__Guru 74252 points

    Josko,

    Thanks for bringing this issue to our attention.

    Q1: Setting the device into sleep or standby mode via bits [7:6] in register 0x0800 disables the CAN bus driver and receiver (besides WUP detection). You can also disable the CAN bus by putting the TCAN4550 in controller-only mode, but I think that might be overkill for what you're trying to do.

    Q2: Typically this is seen on a physically long bus with improper termination. In this case, it looks like something is also pulling the recessive level down as it starts at the expected ~2.5V level then decays down to ~1.6V. When the MCP device is disabled, are the bus pins biased to ground?

    Also, a schematic would be very helpful. You can also email it to me directly. My email can be found by click on my username.

    Regards,

  • 0 in reply to Eric Hackett
    Intellectual 950 points

    Hi Eric,

    thanks on your write back. 

    It seemes we have perheaps figured out what's going on. The STDBY pin was interpreted wrongly and also used as well. :-) 

    If the STBY pin is set to high level, the MCP PHY will not go into high impedance state and as we perviously tought, but rather into low-power sleep mode. As the TCAN starts to send message , the MCP PHY will figure out that can bus is active and it will wake up from sleep mode. After a wake-up event was detected, the CAN controller gets interrupted by a negative edge on the RXD pin. The CAN controller must put the MCPback into Normal mode by deasserting the STBY pin in order to enable high-speed data communication.

    In our case the STBY pin was controlled by Sitara and it was kept in high level whole time, which means it has never been deaserted. It means that MCP PHY gets wake-up event from the CAN bus and at the same time the sleep mode was requested through the STDBY pin. It means MCP PHY switchs steadily from high-speed data communication state into sleep states. It means further that MCP PHY activates and deactivates the transmitter and the high-speed part of the receiver. This switching ON-OFF-ON-OFF loads CAN BUS at every transition from recesive to dominant state. 

     Is it possible that switching from sleep mote into normal mode can cause this ringnig? 

    Br

    Josko

  • 0 in reply to Josko M
    TI__Guru 74252 points

    Josko,

    The device isn't actually switching from standby mode to normal mode on the WAKE event from the bus. What's happening is the transceiver recognizes the pattern from the bus, then as you said, asserts the RXD pin low to have the controller deassert the STBY pin. Since this is never happening, the device stays in standby mode with the CANH and CANL pins biased to ground through a high-impedance resistor network. Though this impedance isn't high enough to have no effect on the bus. 

    What I think is happening is since the device is staying in standby mode, it's affecting the overall impedance of the bus enough to cause ringing while the other device(s) communicate. 

    To get the effect you were originally intending, I believe you'd have to completely unpower the transceiver.

    Please let me know if you have any other questions or issues.

    Regards,

  • 0 in reply to Eric Hackett
    Intellectual 950 points

    Hi Eric,

    thanks on your explanation. 

    My last question, if the MPC CAN PHY would be replaced with TCAN1042GV, is to expect the same behaviour?

    I am going to order some samples of this TI's CAN PHY in order to perform the same test our board. 

    Br

    Josko 

  • 0 in reply to Josko M
    TI__Guru 74252 points

    Josko,

    The behavior between both of these devices is similar, but I'm not sure of the impedance value with the GND bias when switched into standby mode. So this may not show the same issue you're seeing with the MCP CAN PHY.

    Regards,