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.

INA237: INA237 checksum

Part Number: INA237

Hi,

We have trouble using the INA237 when trying to communicate with them.

Our software engineer mentioned this:

''They do respond, but send back 0x0000.  This apparently means a checksum error.

Bit 0 of 0x9 register is an indication of this issue. If it is set to 0, it means that chip has an issue.''

What are we supposed to do with an INA237 that has a checksum issue?

This is on a new revision of a board.  Previous versions had no problem with the INA237.  The new revision has no hardware change with regards to the INA237. 

In fact, when taking an INA237 from an old board and swapping it in the new board, it works.  This would indicate in IC problem.  

Thanks,

Nick

  • Hi Nick,

    Based on the description, it does sound like a genuine device problem. For the new build, how many boards have you tested, and what is the failure rate?

    You could return some of the bad units for failure analysis. The checksum error prevents the internal ADC from functioning, which in turn prevents the device from returning meaningful readings.

    Regards, Guang

  • Hi, Guang,

    Thanks.

    I have tested about 10 boards now.  2 per boards.  100% failure.  Still, a genuine device problem is surprising.  

    I have tried to put back a few of the faulty units in my board, still have 0x0000 returned.

    I was hoping to have some pointers as to what to do in software, or even hardware, to investigate this.  

    I will try to get some to TI, if I can get a case opened.

    Nick

  • Hi Nick,

    The checksum bit is a flag indicating whether there is an error in the internal memory. As end users, there is little we can do if the memory is corrupted. Of course, we have to confirm that it is indeed a memory issue, not something else. The factory will be able to tell for sure. It is hard to imagine 100% failure rate though.

    Meanwhile, you could do some read/write operation on the rest of the registers to see if anything strange stands out.

    Regards, Guang