DS280DF810: What is the maximum number of devices available for programming a single EEPROM Part?

Part Number: DS280DF810

Tool/software:

I am using the Sig Con Architect 2.0 EVM Module and looking to program a set of registers. I am seeing that the software is limiting the amount selectable (no. of device field) to four. Is this the actual limit for a single EEPROM? The part specification lists a maximum of 16. The part number for the EEPROM is 24LC16B.

  • Hi Matthew,

    For devices requiring individual channel configurations, a maximum of 3 DS280DF810 devices can be programmed from a single EEPROM. If common channel configuration is used, a maximum of 16 devices can be programmed from a single EEPROM. The issue you are seeing is a soft limitation of SigCon Architect, and configuring more devices with a single EEPROM could be done outside of the software. However, it must be noted that the common channel configuration offers a less flexible channel set-up. 

    Best regards,

    Greg

  • Thank you for the response Greg! Do you have a guide to link for programming outside of SigCon Architect? The common channel configuration is adequate for our application. 

    Thanks,

    Matthew

  • Hi Matthew,

    The SNLA244 Application Report details EEPROM programming for 25-28G retimers. Please let me know if you have any further questions about the document.

    Best regards,

    Greg

  • Okay, so I should externally generate the hex file according to the address map given, then use Sig Con to apply the hex file to the EEPROM part?

    Thanks,

    Matthew 

  • Hi Matthew,

    The Ethernet SigCon GUIs do not have the ability to write the file to EEPROM. We would recommend to make the EEPROM, and then to apply it to the device using Flash Center or something comparable.

    Best regards,

    Greg

  • Hi Gregory, the documentation states a need for sequential registers, however with our design we have to skip x36. Is there a recommended route for handing the addresses:
    x30 x32 x34 x38 x48

    Would I set this up at filling in each intermediate device - even the skipped ones? Per documentation 16 max
    The hex file ranging x30 x32 x34 x36 x38 x40 x42 x44 x46 x48

    Would you be able to attach an example file for this case at mode 1 CTLE for all? Common lane configuration for all 8.

    Thank you,

    Matthew

  • Hi Matthew,

    It should be possible to create the EEPROM with non-consecutive device addresses. I will look into this and get back as soon as possible with an example EEPROM for this case.

    Best regards,

    Greg

  • Thank you Greg, additional parameters are 20.4 Gbps, divide by 1 in CDR for generation

  • Hi Matthew,

    I am currently working on the EEPROM and have a few clarifying questions. Are you using 8 devices or 9? Wanted to make sure as there are 10 addresses in the list you sent, including 0x36. Also, are you using a 25MHz or 30.72 MHz calibration clock? This will affect configuration settings.

    Best regards,

    Greg

  • Hi Greg,

    I only need these 5 devices addresses programed - I was unsure if the format required sequential ordering and would need all addresses in the space.

    x30 x32 x34 x38 x48

    25 MHz Calibration Clock

    CDR Divide by 1, 20.4 Gbps

    5 devices

    Mode 1 CTLE

    Thank you,

    Matthew

  • Hi Matthew,

    Thank you for the clarification. I will test the EEPROM on my side by flashing it to the device and will confirm if it properly configures the relevant device addresses.

    Best regards,

    Greg

  • Hi Greg, 

    Let me know when the EEPROM is correctly configuring the device addresses or if you need any additional details to proceed. 

    Thank you,

    Matthew

  • Hi Matthew,

    I will let you know when it is working on my side as soon as possible. Thank you for your patience on this.

    Best regards,

    Greg 

  • Thank you for the update.

    Matthew

  • Hi Gregory, 

    Are there any updates to share? Is this use case possible to create with non consecutive addresses? 

    Thanks,

    Matthew

  • Hi Matthew,

    I am currently unable to test the EEPROM with 5 devices, but testing with one DS280DF810 device configured to the desired addresses produced the correct results for all 5 device addresses. I attached the EEPROM here - please let me know if it works in your setup. 

    Best regards,

    Greg

    https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/138/5DevEEPROM.hex

  • Hi Gregory, the checksums were invalid for the file you sent me - the line by line checksums. The start address (given x0A vs x33) did not match the sig con generation. This is what I regenerated with those changes, but am not seeing the programming go through within the setup. 

    :2000000074001000330000330000330000330000330000330000330000330000330000335E
    :2000200000003300003300000000000000000000000000003300003300003300003300008E
    :200040000000000000000000000000000000000000000000000000000000000000000000A0
    :20006000000000000000000000000000000000003300003300003300003300007304063007
    :20008000200300A0F46C102DC9E5600015A2BB182A202310FC7B60001A40400040508090DA
    :2000A000C0D0D1D5D8EAF7FDEEEFFF00B300B3FFF4428000000521344813A9001F9FF1222E
    :2000C00048C300900F00000000000000000000000000000000000000000000000000000076
    :2000E000000000000000000000000000000000000000000000000000000000000000000000
    :200100000000000000000000000000000000000000000000000000000000000000000000DF
    :200120000000000000000000000000000000000000000000000000000000000000000000BF
    :2001400000000000000000000000000000000000000000000000000000000000000000009F
    :2001600000000000000000000000000000000000000000000000000000000000000000007F
    :2001800000000000000000000000000000000000000000000000000000000000000000005F
    :2001A00000000000000000000000000000000000000000000000000000000000000000003F
    :2001C00000000000000000000000000000000000000000000000000000000000000000001F
    :2001E0000000000000000000000000000000000000000000000000000000000000000000FF
    :00000001FF

  • Hi Gregory, 
    Would we be able to communicate via email to get this closed out? If we could have assistance from the author of the documentation it would be helpful too.

    Thank you,

    Matthew 

  • Hi Matthew,

    My team prefers to keep queries on the E2E platform. If there is sensitive information you would prefer to share privately, we can use the E2E private messaging system too. 

    Regarding my hex file, the checksums were indeed incorrect. I must have sent a version of the file that had not been updated with the correct ones. However, the start address used is correct, 0x0A indicates the location of the beginning of the common channel register data. In the regenerated version you sent, this address should be 0x7B. If you make this change in your file for all of the device address map headers, does the setup work?

    Best regards,

    Greg