• State Answer Suggested
  • Locked Locked
  • Replies 12 replies
  • Answers 1 answer
  • Subscribers 33 subscribers
  • Views 2373 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.

Can GNDA be at different level from GND?

David Chance
Genius 4775 points
Other Parts Discussed in Thread: TM4C1294NCPDT

Hi there,

I am using TM4C1294, and I have a need to change the analog ground to 0.8V above GND (in TI app notes, GNDA is the same as GND), I don't see anything in the datasheet against it, but I figure I ask TI FAE guru's opinion just in case


Thanks!

  • 0
    Genius 4775 points

    I can raise VDDA if necessary

  • 0
    TI__Guru**** 244400 points
    Hello David,

    I would like to understand the need for changing GNDA = GND +0.8V first?

    Regards
    Amit
  • 0 in reply to Amit Ashara
    Genius 4775 points
    The input analog signal swings between 0.8 to 2.7 volt, by raising it to 0.8, I thought I could get better resolution.

    But after some thinking, I believe it was not a good idea, it will screw up the PLL and digital controls of analog circuit due to the level change.

    I think I will find another path
  • 0 in reply to David Chance
    TI__Guru**** 244400 points
    Hello David,

    I would have suggested using the VREFA+ and VREFA- pins for this specific purpose, but then on the TM4C1294NCPDT (or the 128 pin package) the VREFA- pin is not available.

    Regards
    Amit
  • 0 in reply to Amit Ashara
    Genius 4775 points
    I noticed that :)
  • 0 in reply to David Chance
    Guru 47900 points

    David Chance said:
    thought I could get better resolution.  

    Should you (really) require higher resolution - might your search include (this) vendor's higher resolution ADCs?   These higher resolution ADCs are often more expensive than your MCU - and most always "avoid" the digital switching & high-frequency PLL circuits - found in most all modern MCUs.

    MCUs are classed, "Mixed signal devices" and that "mix" is not always welcomed by feeble analog signals - which must "walk" that noisy digital path...

    Independent ADCs exist for good reason and are the "device of choice" when "serious" analog measurements are required.   They likely warrant your consideration...

  • 0 in reply to cb1
    Intellectual 710 points
    David, I'm glad you talked yourself out of this one. Especially because 2.7/(2.7-0.8) is only 1.4:1, so you didn't even improve your resolution by one bit! I expect the TIVA might be able to handle +-0.5V between GND and GNDA, but it may not be reliable.

    Have you tried averaging? Average 4 samples to improve your resolution by 2 (=sqrt(4)). If you have the time, you can keep increasing the number of samples until you hit a systematic noise floor defined by the hardware. As cb1 mentioned, this technique might need an A/D better than the one built into the TIVA.
  • 0 in reply to John Duquette
    Genius 4775 points

    Even though 1.4:1 doesn't seem to improve the resolution by a single bit, the readings will have a lot of missing codes when you scale it.

    We tried cascaded inegrator-comb filter on it, and it does improve the resolution a lot, with the sacrifice of lower frequency.

    We didn't want to add a better ADC because of the budget, but we are good at this point, again, thank you all for the inputs!



  • 0 in reply to David Chance
    Guru 10570 points
    Another option is to include the scaling in your signal conditioning. Offset and scaling should be fairly straight forward.

    Robert
  • 0 in reply to David Chance
    Guru 20670 points
    At this point of your prorudct/project development, it's probably to late , but here it goes:
    unless the TM4C1294 is really necessary (maybe you need the ethernet or the 120Mhz), you could maybe use a cheaper TM4C123 with a better ADC so the price stays about the same. There are many good stand alone ADCs that combo-ed with the TM4C123 would still be cheaper than a TM4C1294.
  • 0 in reply to Luis Afonso
    Guru 117855 points

    My firm "likes" your logic, Luis.   As poster has "resolution" concerns - and as the world moves toward, "more conversions/second - NOT fewer" the weaknesses (inherent) in such mixed-signal devices weigh heavily.   (i.e. we "burn" conversions in the (attempt) to render a weak (ADC) peripheral more usable.)

    Yet there's another key advantage to selecting a "proper" ADC - as MCUs (continually) "leap-frog" each other - developers escape the necessity to master (yet another) MCU's ADC "care/handling."  

    Profit most always lands at those who arrive @ market early - and "reuse" of a "high performance ADC" (rather than another "cobbled" MCU version) speeds, eases & enhances product introductions...   Repeatedly...   Excelsior...

  • 0 in reply to Luis Afonso
    Genius 4775 points

    Thanks for the suggestion, but this is a budget constrained project, and both USB and Ethernet are required