• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Subscribers 31 subscribers
  • Views 322 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.

LDC1101: Loses conversion data when entering sleep mode

Arthur Thompson27
Expert 1800 points
Part Number: LDC1101

Tool/software:

Hello,

I wanted to check something I think is an undocumented behavior with TI.  We configure the LDC1101 to generate an interrupt when the conversion is ready.  As a power optimization, we tried to put the sensor into sleep mode as soon as we got the interrupt.  The thinking was that we could immediately stop driving the coil while we wait for our SPI transactions to  read out the conversion data.  This might save a little bit of power.

When we read the 0x3B register after entering sleep mode, we see a value of 0x01, which means LHR Data not ready.  This disagrees with the signal from the interrupt, which told us that LHR data was ready, just before we entered sleep mode.

Is there an explanation for this behavior?  Is entering sleep mode causing the LDC to retain its "configuration" registers, but clear its "conversion data" & "Status" registers?

Thanks,

Arthur

  • 0
    TI__Mastermind 43730 points

    Hi Arthur,

    Thanks for posting on the forum.  I am looking into this and will get back with details shortly.

  • 0 in reply to Eddie LaCost
    TI__Mastermind 43730 points

    Hi Arthur,

    Sorry for the delay.  I was not able to reproduce this behaviour using the EVM functionality, but I can check with the design group to see if they can clarify whether the LHR data registers should also be retained in sleep mode. I will let you know what I am able to find.

  • 0 in reply to Eddie LaCost
    TI__Mastermind 26335 points

    Hello Arthur,

    Eddie's is out on paternity leave, so I will try to support this.  From my tests today, I have confirmed that I will not be able to test that behavior with our EVM, so tomorrow I will attempt to test by another means.

  • 0 in reply to Patrick Simmons
    TI__Mastermind 26335 points

    Hello Arthur,

    Turns out I overlooked something yesterday and could test with the EVM. Per the tests I conducted today, it looks like I am getting similar behavior to what you observed.  My test consisted of the steps below.  Per your test and mine, it looks like data registers are cleared when sleep mode is entered and the status register updates to confirm that data is no longer valid.

    1. Open the Register Configuration tab in the EVM gui
    2. Set register 0x0B (START_CONFIG) bit0 to 0 to put the device in active conversion mode to gather a measurement then write.
    3. Read All.
    4. Verify registers 0x23 and 0x24 show some sort of nonzero value.
    6. Set register 0x0B (START_CONFIG) bit0 to 1 to put the device in sleep mode and write.
    7. Read All.
    8. Look at 0x23, 0x24, 0x3B contents.

  • 0 in reply to Patrick Simmons
    Expert 1800 points

    Hi Patrick,

    Thank you for investigating.  I did review the datasheet and it seems to claim the register contents are maintained in Sleep mode.

    It sounds like you and I are both seeing differently.  Will you be able to confirm with the design team if this is a bug?

    Thanks,

    Arthur

  • 0 in reply to Arthur Thompson27
    TI__Mastermind 26335 points

    Hello Arthur,

    I will check with them.  However, as the original designers are no longer here and the original supporting team is not here, im not optimistic we will get a definitive answer.

  • 0 in reply to Patrick Simmons
    TI__Mastermind 26335 points

    Hello Arthur,

    I reached out to our digital design manager on this.  He needs to get access to the database where hopefully it was properly archived.  Will provide an update if we get any meaningful information on this.

  • +1 in reply to Patrick Simmons
    TI__Mastermind 26335 points

    Hello Arthur,

    One of the digital designer's confirmed the behavior we both observed is to be expected.  Looks like a typo or important word may have been left out that clarifies what type of registers are preserved.