• State TI Thinks Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 1 answer
  • Subscribers 48 subscribers
  • Views 219 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

INA239: Read VBUS at 0x0000 some rare times

Sebastien HAROUET
Prodigy 30 points
Part Number: INA239
Other Parts Discussed in Thread: SYSCONFIG

Hi,

I use INA239 to get VBUS and current.

I used Sysconfig.

Some rare case, I read 0x0000 on VBUS register. 

The ADC is configured as continuous.

I read the register asynchronously.

Should I check CNVR bit ?

If yes, how it works because it is configured in continuous mode ?

If not, what appends if the register is updated during the reading ?

Regards,

Sébastien

  • 0
    TI__Mastermind 27380 points

    Hi Sébastien,

    There should be no need to check CNVR bit, the register stores previous result.

    The registers are buffered to prevent conflict like you described, ie, if conversion finishes during a reading communication, the register will be updated after the reading operation completes. This prevents bogus readings.

    I would make sure VBUS pin connection is solid. It might also be a good idea to inspect the SPI waveforms. This can help rule out potential math issues in the software.  

    Regards, Guang

  • 0 in reply to Guang Zhou
    Prodigy 30 points

    Hi Guang,

    Thanks for your answer.

    I have 5 prototypes and I have the same problem on all boards. So it is not a connection problem.

    It happens on different voltages, so it is not on a specific voltage of VBUS.

    Do you have examples of bad use to obtain a value of 0x0000 on the VBUS register in continuous mode?

    It is very hard to inspect SPI bus during errors because it happens a couple of times on each 24 hours.

    Regards,

  • 0 in reply to Sebastien HAROUET
    TI__Mastermind 27380 points

    Hi Sébastien,

    After talking to colleagues, I found no similar cases. There have been one case where all-zeros were read but the device was damaged and it has other failures.

    Can you read other registers such as Shunt, and DIAG_ALRT to see if results/flags are as expected?   

    Regards, Guang

  • 0 in reply to Guang Zhou
    Prodigy 30 points

    Yes, I can read the current register without any problem.

    I can read VBUS register with the right value most of the time (problem is seen 1/50000 reads)

    So it might be a timing problem, but how and why ?

    I have no interrupt on my firmware.

  • 0 in reply to Sebastien HAROUET
    TI__Mastermind 27380 points

    Hi Sébastien,

    Can you share your configuration and ADC settings? Maybe we could do an experiment on our end to see if we can duplicate the issue.   

    Regards, Guang

  • 0 in reply to Guang Zhou
    Prodigy 30 points

    The configsys is on the first post. Do you need something else ?

    I use the code generated by SysConfig.

  • 0 in reply to Sebastien HAROUET
    TI__Mastermind 27380 points

    Hi Sébastien,

    We experimented with the same exact setting on an EVM, and didn’t find any erroneous reading after capturing a large amount of data.

    I understand it may not be easy, but I still think it is a good idea to capture the communication waveform when zero-reading happens. This might be the only way to tell if the issue comes from the IC itself.   

    Regards, Guang