• State
  • Locked Locked
  • Replies 1 reply
  • Subscribers 33 subscribers
  • Views 83 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: Statistical bring up issue with Linux driver

eytan dahan
Prodigy 80 points
Part Number: TCAN4550

Tool/software:

Hey Team , 

We have some statistical bring up issue with the TCAN4550.

its appeared that the Linux drive not success to detect the TCAN4550 correctly and then the initialization is failed with the below printing from dmesg 

tcan4x5x spi0.0: interface-name can2
tcan4x5x spi0.0: no clock found
tcan4x5x spi0.0: Not a tcan device ffffffff

We thought that issue around the crystal can be the route cause , and we tried to reproduce it by removing the crystal itself and connected OSC2 to gnd ( without external clock) but we got different printing from the dmesg as below 

then its look that the crystal is not the rout cause of "Not a tcan device ffffffff "

do you have any idea on what may cause the detection issue ?

  • 0
    TI__Mastermind 33340 points

    Hi Eytan,

    eytan dahan said:
    We have some statistical bring up issue with the TCAN4550.

    By "Statistical" do you mean that sometimes it comes up and is detected, and other times it is not detected?

    eytan dahan said:
    We thought that issue around the crystal can be the route cause , and we tried to reproduce it by removing the crystal itself and connected OSC2 to gnd ( without external clock) but we got different printing from the dmesg as below

    Do I understand you correctly that you grounded the OSC2 pin but did not provide a single-ended clock signal to the OSC1 pin?

    If so, then the digital core will not have a functional clock to respond to any SPI Read/Write transaction to a register or MRAM address location greater than 0x0800. 

    The Device ID and SPI Interface FIFO registers with addresses in the range 0x0000 to 0x000C do not require the OSC1/2 clock and only require the SPI clock to return a value.

    If the OSC1/2 clock is missing or disrupted, the digital core and MCAN controller will not have a functioning clock and SPI communication to registers in the 0x0800 to 0x083C and 0x1000 to 0x10FC range will fail.

    eytan dahan said:
    do you have any idea on what may cause the detection issue ?

    Can you verify the VSUP and VIO voltage levels at the time of bring up when you detect these failures?  The device is powered from VSUP with exception of the OSC1/2 clock circuit that is powered from VIO.  VIO is also used as the voltage reference for the digital IO level shift circuits.

    Verify the crystal has a stable oscillation and has appropriate load capacitor and series dampening resistor components.  More information can be found in the TCAN455x Clock Optimization and Design Guidelines Application Report (Link).

    Also verify the SPI format meets the datasheet requirements.  If you are able to read the Device ID to detect a TCAN device, then you likely don't have an issue with your SPI driver, and if so, the issue is likely related to either the supply voltage or clock.

    Regards,

    Jonathan