I started nonlinearity tests of 2 AD215BY.
On a test board, I follow the figure 18 for offset adjustment, an figure 17 to set unity gain in inverting configuration.
With a voltage calibrator, I adjust the output : 0 to 8.000V exactly for 0 to -8.000V input. From 0 to +10V the nonlinearity is under +/-0.01%, but from 0 to -10V up to +0.07% nonlinearity measured.
If I change the gain adjustment inverting input polarity, meaning : adjust the output 0 to -8.000V exactly for 0 to 8.000V input, from 0 to -10V the nonlinearity is under +/-0.01%, but from 0 to +10V up to -0.07% nonlinearity measured.
Actually the AD215 seems no symetrical. Idem on the both AD215BY we bought.
When you read the datasheet : +/-0.005% typical for +/-10V Output swing, G = 1V/V the sepcified nonlinearity appears to be for whole dynamic from -10V to +10V.
If I try to trace the gain nonlinearity vs output voltage curve like figure 3 in the datasheet, the curve is close (a V form) but it is not the same scale in ordinate.
The power supply is OK +/-15.5VDC and the current consomation is +40mA / -20mA
Is there something I don't understand ?