• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 45 subscribers
  • Views 434 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.

ADS1120 ADC Linearity

Nick Price
Expert 2235 points
Other Parts Discussed in Thread: ADS1120, REF3312

Hi,


I now have my ADS1120 working , getting counts very close to the ADS1120-DEV board.

But I'm still finding it is a little unlinear.

I'm injecting three voltages into a pot down for a 0 to 10V dc input.

3V = 8688 counts

7V = 20375 count

5V = 14517 count

I'm expecting 14532 at 5V, so I'm 14 counts out.

I have another input, with is a resistance bridge which has a buffer out. This again is 14count out in the middle.

I'm checking well away from the ends of the supply range.

I'm using an external reference. 1.25V  (Ti REF3312)

PGA is OFF, but tested with PGA on (gain 1 ) and shifted the reading by a few counts, but still had the 14 count error.


If I test on the dev board, the error is about 1 count.


On my prototype the Dvdd and directly connected to the Avdd,  Would this account for the error?

Do you think it is a tracking error?

The input signal ground is the same ground at Vref ground. Dvdd ground is joined to signal ground at a star point about 3cm ways from the ADC.

Input welcome

Nick

  • 0
    Expert 2235 points
    Update.
    Have changed the setup to use the internal 2.5V reference instead of the 1.25V external (REF3312) and the linearity error is gone, now within ±1 count in the middle. Now the question is why?

    Tracking(ground) issue or oscillation?

    Nick
  • 0 in reply to Nick Price
    TI__Guru** 113765 points

    Hi Nick,

    It is not entirely clear how you are hooking this all together and applying the voltage.  Are you adjusting the pot? Or are you keeping the pot at one level and applying a different voltage to the pot?  What value of pot resistance are you using?  One thing to consider is the input is a switched capacitor circuit where an input voltage is sampled.  This is a fairly high impedance, but you must consider all current paths in your analysis.

    Many times with the reference you need to have some capacitance connected to the reference, but not too much.  Oscillation may occur if either too little or too much capacitance is connected.  Also the placement of the capacitance is important.  However, we don't know if oscillation is the issue.  If the reference is oscillating you should be able to see this by connecting a scope probe and observing the voltage at the input to the ADS1120.

    A couple of issues related to comments as you have made them in the past.  You cannot make a single-ended measurement with the PGA enabled as it places the PGA in an invalid state and will become non-linear. Another thing I have noticed is you have mentioned a couple of times about the internal reference being 2.5V.  It is actually 2.048V.

    Best regards,

    Bob B

  • 0 in reply to Bob Benjamin
    Expert 2235 points
    Thanks for reply Bob.
    I've finally found the problem, my external reference was supplied from a port pin on my MSP430 micro.
    I moved the supply from there to Avdd and all the errors have gone. I guess there was an error current being induced between the Dvdd and Avdd.

    Yes sorry, Typo on the internal ref, I meant 2.048V not 2.5V.

    Everything is working now. Case closed.

    Regards
    Nick