• State
  • Replies 1 reply
  • Subscribers 63 subscribers
  • Views 19 views
  • Users 0 members are here
Support feedback
Options
Options
Related

UCD3138128: UCD3138128

Raphael_w
Intellectual 305 points
Part Number: UCD3138128
Other Parts Discussed in Thread: UCD3138, USB-TO-GPIO2

Tool/software:

Hi sir,

I have a large quantity of UCD3138128 chips on hand, and I’m encountering errors when reading their version numbers.

I’ve confirmed that many of these chips are brand new. When I enter BootROM and send 0xEC to read the version number, the IC always returns 0xFF 0xFF 0xFF 0xFF.

With previous chips such as the UCD3138/3028, I’ve never encountered this situation.

Could you please advise what could cause the IC to return an abnormal version number, or whether it is possible that this chip’s version number is actually 0xFF?

Regards,

Raphael

  • 0
    TI__Expert 8680 points

    Hello Raphael,

    In BootROM, the UCD3138x devices should all have the PMBus address 0x11. This should not change. Are you saying that the UCD3138128 is returning an address of 0xFFFF FFFF? That is abnormal. Are you sure that your connection to the device is correct? Can you also probe the PMBus line using a logic analyzer to see what the actual PMBus signals look like? What hardware are you using the connect with the UCD3138128? Are you using the USB-TO-GPIO2 (2022) adapter? What version of the Fusion GUI are you using?

    See the below snippet on page 459 in the UCD31xx Technical Reference Manual.

    We do have one issue where an invalid PMBus read of the UCD will cause the UCD to return erroneous data instead of just a NACK. This is found in pmbus_common.c in pmbus_invalid_read() for most UCD3138x projects. If you want the UCD to just return a NACK instead of erroneous data, then you can just change line 455 to  pmbus_number_of_bytes = 0

    Regards,

    Jonathan Wong