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.

DRA821U: DRA821U: Timeout during frequency handshake

Part Number: DRA821U

Hi,

We have a new revision (B) of our custom board based on the DRA821U. Previously, we posted about the bring-up of revision A HERE.

At the moment, we are uploading the boot stage through MCU_UART. Specifically, we are loading tiboot3.bin, and we are encountering the same issue as before:

U-Boot SPL 2025.01-ti-gac2433f99366-dirty (Oct 06 2025 - 12:51:40 +0200)
SYSFW ABI: 4.0 (firmware rev 0x000b '11.0.9--v11.00.09+ (Fancy Rat)')
EEPROM not available at 0x50, trying to read at 0x51
Reading on-board EEPROM at 0x51 failed -121
Timeout during frequency handshake
### ERROR ### Please RESET the board ###

In this revision, we are using the following DDR part number: MT53E512M32D1ZW-046 WT:B.

For DDR configuration, we are using the file k3-j7200-ddr-pnt-lp4.zip which worked correctly in revision A with the DDR part number MT53E512M32D1ZW-046 AUT:B. We reused the same configuration file for revision B; however, it is not working with the MT53E512M32D1ZW-046 WT:B device. This configuration file was generated as a result of the discussion in the previous post where we encountered the same issue. 

We also attempted to reduce both boot and operating frequencies to check whether DDR training would succeed at lower speeds, but the result remained the same. 

We would appreciate your guidance on the following points:

  • How can we verify whether the DDR is functioning correctly?
  • What messages or behavior should we expect if the DDR is not detected?
  • Assuming the DDR configuration file is correct, what debugging methods are recommended to further investigate the DDR initialization process and identify the root cause?

Thank in advance,
Santi

  • Hi,

    Based on the names of the two part numbers, it appears that the only difference may be the supported temperature range. I am not aware of any reason why a change in register settings would be needed to allow the DDRSS to complete initialization. 

    How many boards have you tested with the new memory? 

    Do you have access to JTAG on your board & Code Composer Studio?

    Regards,
    Kevin

  • Hi Kevin!

    Thanks for your quick response.

    We currently have JTAG available on our board and have installed Code Composer Studio.

    At the moment, we just have the XDS110 debugger that is embedded in the EVM.

    Should we get this DEV-BOARD for debugging purposes on our customs boards?

    Also, is there any way to use the debugger embedded in the EVM with our custom hardware?

    Finally, could you please clarify the recommended procedure for debugging DDR via JTAG using Code Composer Studio?

    Thanks in advance for your help.

  • Hi,

    We discovered an issue with the hardware: the DDR_RET pin was left floating. After connecting this pin to ground, the boot process starts normally.

    Regards,
    Santi

  • Thanks for updating the ticket with your findings. Glad to hear that your issue is resolved.

    Regards,
    Kevin