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.

INA240: Non-Linearity Error Imapact

Part Number: INA240

Hi Team, Please let us know which parameter will be affected by non-linearity error. (for example: Current sense input, Gain,input offset voltage or reference input voltage). Suggest a formula to calculate output voltage by considering errors seperately (Instead of sum of squares method).

  • Hi,

    Other parameters are not affected by non-linearity error.

    You can consider errors separately; the process and individual terms are identical with what’s shown in the datasheet section 8.4.4.2.1, except for the very last square-sum-root part. You can substitute with another approach, such as simple addition if you want.

    You may also want to check out our training videos on error calculation. Most if not all error related questions can be answered by watching these videos.

    Regards, Guang

     

  • Hi,

    Thanks for your reply, it's really helpfull.

    We have some more doubts in INA240.

    Case1:

    When i check the training video session 5, they showed the below graph after including non linearity error with gain error.

    - How to get this characteristics in real condition (gain error=0.05%, Linearity=0.01% from datasheet).

    - As per the graph, output error is varying with respect to input voltage. but non linearity error is ±0.01%. how to relate non linearity denpendency with respect to input voltage?

         

    Case2:

    We have external series resistors connected to IN+ and IN-.

    Considering current flow direction as per below images, How to include Rs Resistor value for both below cases in the same formula, Since gain error factor formula only one Rs is mentioned. (In otherwords, What will be the impact of Rs value on gain error factor for charging and discharging cycles?)

     

    Regards, Vignesh

  • Hi Vignesh,

    Glad you found previous answers helpful.

    Case 1 – this has to do with how gain and linearity are defined and tested. Input/output pairs are obtained for the DUT at min and max input respectively. Then a straight line is fit through the two pairs, this is the actual gain curve (a line). At max input, the deviation of actual output from ideal can be found and is used to calculate gain error.

    Actual outputs are then measured for intermediate inputs. Depending whether it is characterization or final test, the number of intermediate inputs varies. A curve is traced through these actual outputs. The max % deviation divided by gain gives linearity. So linearity represents the worst curvature of the actual gain curve.

    You can similarly follow this process to produce this graph by yourself.

    Case 2 – Please follow the recommendations in datasheet section 9.1.1 regarding the impact of Rs on gain, specifically below section. Both situations are covered by this equation, and the gain error term is ± by default.

    Regards, Guang

  • Hi Guang,

    Thanks for your reply.

    In Case2, We still have some doubts.

    In the Circuit we have used two resistors.(One connected to IN+ and another one connected to IN-).

    In gain error factor formula, only one Rs is mentioned.

    Which resistor should i take for that formula??  (Please refer below image)

    Also please clarify how to consider the impact of Rs(Top) and Rs(Bottom) in the formula??

    Regards, Vignesh

  • Hi Vignesh,

    The two Rs should be matched to the best possible; the error equation assumes they are identical, which is obviously hard to achieve.

    If you want to go to such length as to consider the mismatch however, you can take the value difference and times nominal IB, which yields an additional Vos term. For example if a 100Ohm (recommend <10Ohm), 0.1% Rs pair is used, the mismatched value is 0.2Ohm max. The additional offset term is 0.2X(±90uA)= ±18uV.

    Does this answer your question?

    Regards, Guang

  • Hi Guang,

    Thanks for your reply,

    I understand that ±18uV should be add along with bus bar sense voltage becasue it impacts gain. As shown below,

    ((Iload x Rshunt) x Gain x Gain Error) ± 18uV  

    or it should be add with Vos total??  ( sqrt ( Vos^2 + Vos_cm^2 + Vos_ref^2) )

    Regards, Vignesh


  • Hi Vignesh,

    I would use the latter equation to add to Vos total because its effect is equivalent to input referred offset.

    Regards, Guang

  • Hi Guang,

    Thanks for your reply,

    As per your example, Additional offset term is added with Vos total.

    Is the below equation correct?? 

     or is it mere addtion to Vos total??   (Vos total + 18uV).

     

    Regards, Vignesh

  • Hi Vignesh,

    When Min/Max is used in the calculation, the Root_Sum_Square equation is a more realistic estimate of expected error. The simple addition is pessimistic.

    Regards, Guang