LDC - AD8130/29 CMRR test circuit

Q

I made a very basic circuit with AD8130, aiming to determine it's wide
frequency range CMRR. I'm attempting to use the device as a high-speed
instrumentation amplifier. However, it turned out that the real CMRR of the
circuit at 10kHz and 100kHz test frequencies was only ~60dB while the minimum
CMRR specified in device datasheet should have been at least 93dB. The
difference is so significant that it cannot be explained by the input filter
asymmetry alone. I was wondering what sort of test circuit was used to come up
with the 100dB specification in the datasheet. Would you be able to provide
some insight.

A

The AD8130 CMRR measurement that is presented in the data sheet was made by
simply connecting the two inputs together and varying the signal to determine
the common-mode gain with the differential gain set to one.. It is very likely
that the differential filter is the cause of the degraded performance. The
series elements need to be matched to better than 10 ppm in order to get 100 dB
CMRR. Please check the CMRR by simply connecting the two inputs together.