• State
  • Locked Locked
  • Replies 9 replies
  • Subscribers 46 subscribers
  • Views 642 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.

TSW1400EVM Data Overwritten

Thayne Henry
Prodigy 145 points
Other Parts Discussed in Thread: ADS5562EVM, ADS5562

Hello,

I am having an issue where if I take a long data file using the TWS1400EVM with an ADS5562EVM the data is periodically overwritten. 

I have attached a data capture for reference.  For this data capture I used an arbitrary waveform generator to generate a 1KHz sine wave with a decaying ramp from 250mV to 0V over the 3 second capture length of the file.  I used this waveform to stay above the 30Hz break frequency of the input AC coupling while also proving a ramp waveform that would be indicative of the time that was being overwritten in the file.  As can be seen from the capture, every 16.777216M samples (~210ms) the file overwrites itself.  That is to say, that samples 1-16,777,216 are over-written by samples 16,777,217-33,55,44,32, and then this repeats periodically throughout the data capture with every other period being the correct waveform.  This occurs independent of sample rate (just a function of the number of samples) and therefore seems to be a memory issue.

Has anyone seen anything like this before?

Thank you in advance,

Thayne

  • 0
    Prodigy 145 points

    For added information:

    The 16,777,216 samples is 1/16th of the maximum number of samples that can be taken, which is 268,435,456 (this is the maximum number of samples allowed by the software).  

    Also, since I did not explicitly say so in the last post, every other section of data is the expected waveform, with the section immediately before it being identical (for example, section #2 is correct and section #1 is an exact copy of it, section #4 is correct, and section #3 is an exact copy of it, and so on...)

    Thanks again,

    Thayne

  • 0 in reply to Thayne Henry
    TI__Guru* 79134 points

    Hi Thayne,

    I will take a look and get back to you later this week, early next week the latest.

    -Kang

  • 0 in reply to Kang Hsia
    Prodigy 145 points

    Hi Kang,

    Thank you for the response.  I have taken lots of data to characterize the issue and can forward any of that data or screen shots of the results if it would be useful.  At this point I am almost certain it would have to be something in the process of storing the data into memory and reading it out on the TWS1400, but it is hard to tell where it the chain the error might be occurring.  

    I have characterized the issue across 3 different TWS1400 boards and two different laptops running the high speed data converter pro software and seen the same results.  Also, I am sure this is not a post processing issue because the error shows up in the high speed data converter pro GUI itself.

    Thanks,

    Thayne

  • 0 in reply to Thayne Henry
    TI__Guru* 79134 points

    Hello Thayne,

    We have narrow down the issue to be the memory of the TSW1400. You may notice that the current memory on the TSW1400 is labeled as 2RX16 (dual rank). We have running some tests with 1RX8 memory and it seems to improve the capture accuracy of another ADC. I need to run more tests with the ADS5562 to confirm this.

    Would it be possible to ship you an 1RX8 memory for you to try out and to help us with correlating the results? Thanks.

    -Kang

  • 0 in reply to Kang Hsia
    Prodigy 145 points

    Hello Kang,

    Absolutely, my shipping address is:

    shipping address removed from the forum for privacy purpose.

    Thank you,

    Thayne

  • 0 in reply to Thayne Henry
    TI__Guru* 79134 points

    Hi Thayne,

    I have done some testing with the 1Rx8 module and I was not able to see the sample glitches anymore. We will send the memory module to you via priority overnight package. Please let us know if you don't receive it within the next day or so.

    I will pm you the tracking number once available.

    -Kang

  • 0 in reply to Kang Hsia
    Prodigy 145 points

    Kang,

    That is excellent news!  I will let you know when I have received and tested tomorrow.

    Thank you,

    Thayne

  • 0 in reply to Thayne Henry
    Prodigy 145 points

    Kang,

    I tried the new memory module and the issue went away!  I have attached the same 3 second ramp data file as before.

    I have 3 of these boards total.  Would it be possible for you to send another 2 of the correct memory modules for the other boards as well?


    Thank you,

    Thayne

  • 0 in reply to Thayne Henry
    TI__Guru* 79134 points

    Hi Thayne,

    Thanks for the update. We currently only have two of these memory modules left and need them for internal testing purposes. We are having a hard time sourcing these modules from any memory vendors, and the earliest lead time is about 6 to 8 weeks. If you can wait, we would be happy to send them once we received them.

    If you need these modules sooner, we will need to recommend you to purchase them from online sources like ebay. They are fairly low cost - typically about $10. You will need to make sure you have the following part number with 1RX8 on the label

    MT8HTF12864HZ

    -Kang