We are using op-amp (P/N: AD8422BRZ) to measure current using Rsense (10 ohms resistor). The calculations and simulations are matching but the actual measurement doesn’t.
For the given below circuit, With 270 ohms load, I = 86mA.
Drop across R50 = 860mV.
Which means 430mV should be across +IN and -IN pins but we measured 414mV. This 16mV error is acceptable.
As per transfer function.,
Vo= (1+ 19.6K/Rg)*Vdiff + Vref
= (1+19.6/15)*Vdiff + 1
= (2.32*0.414) + 1 = 1.96V
The simulation also shows same 1.96V as like calculations. But measured value in our board @ TP6 is 2.9V. Please let us know what is been missing at our end..
Resistor tolerance can affect the actual input voltage and also degrade the CMRR performance of the part. But I agree that measured value is too far from expected voltage.
To help us troubleshoot the problem, can you measure the voltage across these pins of the AD8422: Pin1, Pin4, Pin 5, Pin6, Pin7 and Pin8? This can give us clue where the error is coming from.
Thank you Emman for prompt response. Below are the measurements taken with respect to Ground..
Pin1 - 11.4V
Pin4 - 11.9V
Pin5 - 0V (GND)
Pin6 - 1V
Pin7 - 2.9V
Pin8 - 23.8V
Pin2 - 12.1V
Pin3 - 12.6V
What are the tolerances of the resistors in the dividers?
Are they discretes or an array?
One percent resistors can give you over 100 mV error, so 16 mV is actually pretty good.
Yes, 16mV error is acceptable as mentioned in my initial conversation. But the issue is instead of 1.96V, I am measuring 2.9V..
Thank you - Babu