• Locked Locked
  • Replies 7 replies
  • Subscribers 33 subscribers
  • Views 383 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.

TCAN1043G-Q1: CANFD Timing Parameters

Zehui Li
Intellectual 540 points
Part Number: TCAN1043G-Q1
Other Parts Discussed in Thread: TCAN1043-Q1

Is Bit Time the only timing parameter we care when we measure CANFD signal? For classical CAN, there are other parameters like Tr and Tf. But I didn't notice these in CANFD.

  • TI__Guru 53555 points

    Hi Zehui,

    During the data portion of a CAN FD frame, there is only a single transmitter on the bus (arbitration is over so any simultaneous transmission is complete). Therefore the propagation times from the transmitter are no longer the limiting factor of the system and are thus not characterised as closely as they were for HS CAN devices. The more critical characteristic during this phase is the bit timing and symmetry as the differences dominant vs recessive edge become significant at the speeds used for CAN FD. 

    Let me know if you have any more questions.

    Regards,
    Eric Schott

  • in reply to Eric Schott1
    Intellectual 540 points

    Eric，

    I read ISO 11898-2. But I couldn't find any clue about what TI called as DRIVER SWITCHING CHARACTERISTICS in the table below. And I found these parameters were different in the 1043 devices of other company. Thus, I think the parameters like propagation times are the parameters for TI's own device. It may not be a common specifiction. While the BIT TIME which is published in ISO 11898-2 is all the same between different devices.I don't know whether it's correct.
    Since TI offers the typical values rather than MAX or MIN value, how should I use these to evaluate my test result?

    What do you mean by "the differences dominant vs recessive edge become significant"? Is the rising or falling edge too different in every bit period to be normalized?

  • in reply to Zehui Li
    TI__Guru 53555 points

    Hi Zehui,

    I cannot see the image you attached. Please use the "Insert -> image" function to share images. 

    The ISO 11898-2 Physical Layer CAN Standard defines timing requirements for propagation delay and bit width. TI provides both of these parameters in the timing characteristics portion of the datasheet. 
    The datasheet also includes additional timing parameters such as the differential rise/fall time that is not required by the ISO standard. These parameters give extra information that can give insight on how the device will behave in a typical network. If the system requirement is only to meet the ISO standard, these parameters do not need to be evaluated. 

    Zehui Li said:
    What do you mean by "the differences dominant vs recessive edge become significant"? Is the rising or falling edge too different in every bit period to be normalized?

    CAN uses an a-symetrical drive state method to transmit data, so dominant bit-times are commonly longer than recessive bit times. The protocol layer of the standard accounts for this so it is not a significant issue in the asynchronous protocol. However because the protocol needs to be able to correct for this behavior, the physical layer standard imposes limits on how large in magnitude this difference may be with the bit width specification. 

    Regards,
    Eric Schott

  • in reply to Eric Schott1
    Intellectual 540 points

    Eric,

    I post two pictures from chapter 8.8 of Tcan1043-Q1 Datasheet. Maybe you can find them for sake of not showing.

    There are FD Timing Parameters and other pamameters like DRIVER SWITCHING CHARACTERISTICS. FD Timing Parameters are for CAN FD only,literally. Are the others for classical CAN and CAN FD, or for one of them only? Another question is I don't see anything about data rate in the test condition, so does the device perform the same no matter how fast the data rate is?

  • in reply to Zehui Li
    TI__Guru 53555 points

    Hi Zehui,

    The FD timing parameters are a part of the ISO 11898-2 CAN specification and are required for any compliant transceiver. This mainly relates to bit symmetry (timing of a dominant bit vs a recessive bit) which as mentioned is only a limiting factor at CAN FD rates. The data rates for the CAN FD timing parameters are implied by the bit time in the test conditions - 500ns bit time for 2Mbps, etc. All other timing parameters apply for all data rates. 

    Regards, 
    Eric Schott 

  • in reply to Eric Schott1
    Intellectual 540 points

    Eric，

    Thank you.

    I have last one question about parameters. I can measure one parameter in my board, for example, I measure tR, the differential output signal rise time. I may not get the same result as TI post even in the same test condition, because of the different layout design, value shift of capacitances and resistances, etc. Because TI offers a typical value rather than a range, I don't konw how to evaluate my result. If I get tR = 70ns, which is larger than 45ns, what does it mean? Is it unacceptable,or fine, or other?

  • in reply to Zehui Li
    TI__Guru 53555 points

    Hi Zehui,

    Correct, different board characteristics such as capacitance with change the measured transition time of the differential signal. As these are not critical parameter for CAN, it is sufficient to have them as measurement points without requirements. If some other test or parameter results in failure, the transition time may be good extra information to help diagnose the issue. However by itself it would not represent a risk to data integrity if all other parameters were met. 

    Regards,
    Eric Schott