• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 82 subscribers
  • Views 400 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.

CCS/TMS320F280025C: ADC value is not accurate/closer.

Monish27
Intellectual 636 points
Part Number: TMS320F280025C

Tool/software: Code Composer Studio

Hi,

I am working on internal ADC. I set reference voltage as internal reference(3.3V).

When the expected adc value is 482, the result that adc give is 490.

How to get the expected results?

I am using 80 MHz system clock. Also ADC clock is divided by 4.

Regards

Monish P

  • 0
    TI__Guru 60865 points

    Hi Monish,

    What are you using to source the input voltage?  What is it's accuracy (or the accuracy of whatever you are using to verify the input voltage).

    What is the source impedance of whatever you are using to drive the ADC?  Is it locally buffered on the board by a high-bandwidth op-amp?  What S+H (controlled by the ACQPS setting) are you using for this channel?

    You may want to sample at multiple points in the voltage vs. code transfer function to determine if the error is due to noise, offset error, gain error, or non-linearity.  Once you determine the error components, you'll want to compare this with the specifications of those error sources in the ADC section of the device datasheet.  For instance, the max gain error in internal reference mode is +/- 45 LSBs.  At an input voltage of 482/4095 = ~12% of FSR the max gain error would be 0.12*45 = 5 LSBs.  The offset error (the same at any input voltage) is max +/-5 LSBs. 

  • 0 in reply to Devin Cottier
    Intellectual 636 points

    Hi,

    In my case the read ADC value error increases when the voltage is increased. But the error percentage between read value and expected value remains same.

    Initially I set sample window value as 10, increased to 100 and 220. I get the same results.

    So every ADC pin have different error percentage. But error percentage remains constant for all the read ADC value for a specific pin.

    For example, A0 have 0.6% error, A3 have 0.5% error and A5 have 5.4% error. I have attached the reading below for A5.

    Note: This Pin measures the Current (I) value. Design has voltage divider , op-amp, RC  filter and  MCU.

    Input Voltage ADC read value Expected ADC value Error (%)
    0.057 74 71 4.225352
    0.168 220 209 5.263158
    0.28 366 348 5.172414
    0.558 731 693 5.483405
    0.726 951 901 5.54939
    0.839 1097 1041 5.379443
    1.008 1318 1251 5.355715
    1.117 1464 1386 5.627706

    The other pin A0 and A3 which have 0.6% and 0.5% error has design flow voltage divider, RC filter and MCU.

    Why I am not getting closer value.

  • 0 in reply to Monish27
    TI__Guru 60865 points

    Hi Monish,

    It looks like on this pin you have a ~5% gain error (~200 LSBs) which is well outside of the maximum datasheet spec' of 1% gain error with internal reference (which itself should only occur for outlying parts at extreme temperature).

    Considering there is such a large difference with the other channels and that the other channels have a different drive topology, I think your drive circuit is probably the problem.  Since you have an op-amp + RC, can you post the circuit?  A typical circuit here would be Cs = about 100pF to 300pf, Rs = about 10 to 100 ohms, op-amp BW = 10Mhz to 20MHz