• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 45 subscribers
  • Views 135 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

ADS122C04: Deviation with higher gains (Strain Gauge Simulator)

Michael Gröne
Prodigy 10 points
Part Number: ADS122C04

Hello,

I try to get my head around this but I cannot find an issue.

I see depending on the gain different amount of deviations:

Following is the wiring diagram. To the ADC a strain bridge simulator is connected, simulating the steps 0mV/V ... 3mV/V. I already tried removing the input filter but it does not change anything.

Reference measurement made with a calibrated 6.5 digits multimeter.

Is this in the expected range? Anything I can check? How would I calibrate this out best?

  • 0
    TI__Guru** 113765 points

    Hi Michael,

    Welcome to the E2E forum!  I have to make a few assumptions.  One is the external reference is connected across the excitation supply at the bridge simulator.  Secondly, that you have properly configured the ADS122C04 to use the external REFP/REFN inputs.

    Ideally the measurement should be ratiometric where the excitation for the bridge is also the same for the ADC reference.  In this way any changes or drift in the excitation is removed from the measurement as the voltage is common to both.

    All ADCs will have some gain error.  The gain error specification is listed in the datasheet under System Performance.  As to your graph, you have 5 points but they are not evenly spaced and that is why you are seeing a hump between points 4 and 5.  Point 5 should be 2.5mV and point 6 should be 3mV.

    It is unclear as to how you are actually computing the error.  The error should be based relative to full-scale range of the ADC.  What appears in your graph is higher gains are worse than the lower gains.  This is most likely due to the scaling as the resolution of one code at a gain of 1 (3V/2^23 = 358nV) is a much larger value than at a gain of 128 (3/128/2^23 = 2.79nV).

    I would suggest taking a look at A Basic Guide to Bridge Measurements.  This document contains a lot of good information including a section on calibration.

    Best regards,

    Bob B

  • 0 in reply to Bob Benjamin
    Prodigy 10 points

    Hi Bob, thank you for your reply.

    I read many documents and almost every answer in this forum already but was not able to track down the issue.

    1)

    The reference is connected at the bridge, right. The ADS is also configured for this. I also tested with AVDD+AVSS and the result is very similar. So yes, the reference should not be an issue.

    2)

    I do not expect the gain deviation to be this high. Here you see it in an XY plot - it's linear now. the grey one is probably not really valid because there is too much noise with that low amplification.

    I also changed the deviation relative now to the FSO (which should be about 3.3V/128=26mV    divide by the supply voltage about 7,8mV/V FSO.

    Following are my absolute numbers.

    Measured Val (calc) [mV/V] 6.5 digit multimeter 0,0000 0,4980 0,9980 1,4975 1,9977 2,9969
    DUT
    Reading in [mV/V] -0,0002 0,5015 1,0034 1,5050 2,0068 3,0100
    Deviation, absolute [mV/V] -0,0002 0,0035 0,0054 0,0075 0,0091 0,0131
    Deviation, relative to ref [%] 0,71% 0,54% 0,50% 0,46% 0,44%
    Deviation, relative FSO [%] 0,05% 0,07% 0,10% 0,12% 0,17%

    I will reread the document you linked, but am not sure if I finally can solve it.

  • 0 in reply to Michael Gröne
    TI__Guru** 113765 points

    HI Michael,

    Can you send me the register configuration settings you are using for all registers?  And also, can you send me the raw codes being returned from the ADC (preferably in hex) instead of the calculated values?  Your excitation/reference voltage is 3V and not 3.3V and for your computation of codes to volts you should measure the reference voltage at the inputs to the ADC to remove any voltage drop error from circuit path.

    There does appear to be some gain error, but this appears to be fairly linear.  So you should be able to calibrate this out following the calibration section in A Basic Guide to Bridge Measurements that I referred to earlier.

    Best regards,

    Bob B