IN-AMP and DIFFAMP: CMRR Test Conditions

Document created by analog-archivist Employee on Feb 23, 2016
Version 1Show Document
  • View in full screen mode

Why for in-amps, CMRR testing will use 1kOhm unbalanced signal source, but
differential amp don’t?
How you can get the 1kOhm unbalanced signal source? Is it normal signal source
specs, e.g. some models from Agilent or we have to produce special board to get


The instrumentation amplifiers like the AD8221 have a very high input
impedance, and in fact the input CMRR is based on the transistor matching of
the input pair. This matching is excellent, leading to a very high CMRR,
typical >110 dB. When we do our characterization over frequency, we add a 1kOhm
resistor on one input, and then sweep a common mode signal over frequency. This
allows us to provide the customer with a worst case spec on the specifications
Adding the 1Kohm also simulates what would happen in the real application,
where customers will most likely connect the inputs of the AD8221 to a
resistive Wheatstone bridge, which has some impedance and will look like an
imbalance to the AD8221.
So, we add the 1Kohm to provide the customer with a worst case number, and also
try to simulate real life conditions.
The diff amps on the other hand have a resistive input. So, CMRR performance
varies on the matching of those input resistors. The applications are different
as well, like for example current sensing, where the inputs of the diff amp are
connecting to a very small shunt resistor, which will not cause an imbalance
with the input resistors of the diff amp. In fact, with diff amps, we ask
customers not to add external resistors to the inputs, since they lead to
accuracy issues and mismatch with our trimmed precision resistors internal to
the part.