INA181: About OUT accuracy

Hello Tsukui,

Thanks for considering to use Texas Instruments.  I tested out the circuit you have in TINA (shown below) and I get a little better performance than what you posted.  But that can probably be attributed to rounding.  If I compare the simulated result with the ideal result, I calculate a total error of about 1.4% error.  From looking at our specs this appears to be closer to a typical error and actually not the worst case error you might expect.  If we were to take the error from the worst possible offset voltage, the error could be as high as 2.5%.  If we look at the worst case possible error from the common mode rejection ratio, that error alone can be 3.8%.  Typically we expect some of our error sources to be correlated to each other and therefore use the RSS formula.  Worst case error using the RSS assuming error correlation, yields an expected error of 4.65% (shown below and attached), which is significantly higher than what you are observing. Consequently if you want a device with less error, you may want to look for something with a higher CMRR minimum and lower Vos maximum.

/cfs-file/__key/communityserver-discussions-components-files/14/INA181ErrorCalculation.xlsx

/cfs-file/__key/communityserver-discussions-components-files/14/INA181Circuit.TSC

Hello Yusuke-san,

You can show more digits of precision in TINA-TI by changing Analysis Options. Click Analysis, then Options, then in the new window increase the number for Numeric precision (max is 6).

Best regards,

Ian Williams

Hello Tsukui,

Im not sure about your customers' design requirements.  However they might consider looking at the INA190, INA240, and the INA186.  The INA190 should have the least error followed by the INA240.  As you were considering the INA181, I presume price was a major concern.  Consequently, I suggest the INA186.