QWe are using AD8220 for current measurement in our application (In normal
measurement conditions a 100R shunt is placed between Vin+ and Vin-). In order
to calibrate the offset of the AD8220, we are shorting Vin+ and Vin- of AD8220
and check the Vout. We have following issue:
Vout depends of Vin voltage whereas Vin+=Vin- (shorted). In other words, when
Vin+ = Vin- = 0V then Vout = 0V. But when Vin+ = Vin- = +12V then Vout = 400µV
(Vin+ & Vin- still shorted together!) Is it normal? If so, could you explain
AThe difference in the offset voltage depending on the common-mode input voltage
can be attributed to the common-mode rejection ratio (CMRR) of the AD8220.
The definition of CMRR is: CMRR (dB) = 20*log10(A_d/A_cm) where A_d is the
input to output differential gain, and A_cm is the input to output common-mode
For a change in the output voltage by 250µV in response to a change of 5V in
common-mode voltage, or A_cm = 50 µV/V, the CMRR amounts to
CMRR (dB) = 20*log10(2 / 50µV/V) = 92 dB for a gain of 2 (A_d = 2).
This is well within the A grade specification and equal to the B grade
specification of the AD8220. Kindly compare with the Specifications section of
the www.analog.com/AD8220 datasheet, page 3. For your reference I am attaching
the screenshot below.
If CMRR is the main error in the circuit and the bias current is not as
critical, AD8422 and AD8421 are the best offerings that we have, closer to
~20µV/V_cm change in the output voltage.