• State Resolved
  • Replies 3 replies
  • Subscribers 30 subscribers
  • Views 124 views
  • Users 0 members are here
Support feedback
Options
Options
Related

TMP1826: TMP1826 Bus-Powered Mode – Skip Address Command Behavior with Write Lock

KIM JASEUNG
Prodigy 70 points
Part Number: TMP1826

Tool/software:

I am currently using the TMP1826 in bus-powered mode with three sensors connected in parallel to a single SDQ line.
Each sensor is assigned a unique short address by connecting the IO pins to Vcc through resistors.

To configure the short address and enable the Write Lock (Lock En = 1), I use the write command.
However, when I attempt to configure all devices simultaneously using the Skip Address command, only two of the three sensors respond correctly.
The remaining sensor always returns FFFFFF when read.

My questions are:

  1. In this configuration, is it possible to set all devices at once using the Skip Address command?

  2. If not, what is the recommended method to perform the initial configuration for each device individually in bus-powered mode?

Thank you for your guidance.

  • 0
    TI__Mastermind 43785 points

    Could you share your schematic?

    Is it the last device, furthest from the controller, that fails? 

    What is your pull-up resistor? Did you follow the guidelines 8.3.3 Bus Pullup Resistor in the datasheet?

    Are you performing only WRITE SCRATCHPAD-1 (0x4E) or does the failure occur when COPY SCRATCHPAD-1 (0x48)?

    What you describe sounds like a brown-out condition where the last device does not receive enough voltage to complete its operation when all of the devices are triggered to do it at the same time. This is expected to be a problem when COPY commands are issued to all devices at once using skip address. You should be able to communicate with individual devices using the short address after WRITE SCRATCHPAD-1 (0x4E) and issue each one an individual COPY SCRATCHPAD-1 (0x48) command. Otherwise, the only way to perform individual communications on your bus would be to first discover the 64 bit IDs using the SEARCHADDR (0xF0)

    thanks,

    ren

  • 0 in reply to Ren Schackmann
    Prodigy 70 points

    I would like to confirm whether the COPY SCRATCHPAD-1 (0x48) command is always required.

    My current sequence is as follows:

    Skip Address → WRITE SCRATCHPAD-1 → wait → Reset → Answer to Reset → FLEX → Send Short Address → CONVERT → wait → Reset → Answer to Reset → FLEX → Short Address → READ SCRATCHPAD-1

    In this flow, is it still necessary to perform the COPY SCRATCHPAD-1 (0x48) command?
    Or is COPY only required when storing the configuration permanently into non-volatile memory (EEPROM)?

    For reference, I have also attached a simple schematic of my setup.

    Thank you for your clarification.



  • +1 in reply to KIM JASEUNG
    TI__Mastermind 43785 points

    COPY is only used to write to EEPROM. You don't need to use COPY at all to use the devices. The only benefit of writing to EEPROM is that the configuration will be restored on power-up/reset. You could configure the address mode on every power-up, and the devices will work just the same for your use case.

    ren