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.

TMS320F28035-Q1: Interesting correlation between ADC and Flash

Part Number: TMS320F28035-Q1

Hi experts, 

In customer board and code, I found a very interesting correlation between ADC and Flash. In start-up codes, customer samples one ADC channel 1024 times and averages the results in order to find the offset of this channel. And for the first case, they have following calculated result: 1537 ADC code. And I also screenshotted the map file of the project as below: 

And in the VmaxPower_SCI file, I modified the codes as below, moved one NOP instruction

So this file's text size will decrease 1 as below fig shown. However, we can also find from the below figure that the ADC code is also changed (From 1537 to 1572): 

Similarly, I also modified the codes to let the file's text size increase 1 as below fig shown. And can also find from the below figure that the ADC code is also changed : 

Please help explain the reason behind the above phenomenon, thanks! 

BR, 

Will 

  • Hi,

    The subject matter expert is out of office and will get back ASAP. Please expect delay in response.

     

    Thanks,

    Saravanan

  • Will,

    Given that we are touching code in the SCI function, I think the most likely explanation is that there is some interference occurring when the SCI is transmitting/receiving data and the ADC is sampling. 

    Do we know which pins are being used for SCI comms, in terms of there PCB routes or proximity to the analog region the C2000/or the analog signal path of what ever channel is being converted?  There could also be coupling to the VREFHI/VREFLO or VSSA that are used by the ADC. 

    You could also try to pause the SCI while the ADC is converting to confirm the above, if that is possible.

    Best,

    Matthew

  • Hi Matthew, 

    It is a good idea, will try in customer side and let you know. 

    Best Regards, 

    Will 

  • Hi Matthew, 

    They have tried this, but no help. 

    Will Flash's operation affect ADC performance? 

    Best Regards, 

    Will 

  • We don't have alot of data around this, but there is potential for there to be some impact during heavy code execution, I would look at the goodness of the 3.3V supply and decaps near the device to help stabilize the analog supply from the digital supply.  Is customer using the min S/H window(ACQPS) for the ADC?  I'm wondering if an increased ACQPS would help average out any noise from the system.

    Best,

    Matthew