I want to check a data acquisition input or overall CMRR of a signal chain roughly by a simple method.
Normally for a daq input CMRR is defined as:
CMRR = 20×log10(Vin/Vout)
where Vin is the common mode voltage Vcm applied at both inputs.
But in practice I hesitate to use the above formula. It is because when I short the inputs together the Vout is not zero. It has some offset with standard deviation.
So first I short the inputs I obtain x at the output. x represents output voltage and xrms is the rms value of x.
Then I apply purse sinusoid common mode voltage by a function generator. I call the rms value of common mode input as Vrms and the output as a.
Finally I double the pure sinusoid input common mode voltage as 2Vrms and I call the output as b.
I tried to illustrate these by the following drawing:
Is the following approach to estimate the CMRR mathematically correct?:
CMRR = 20xlog10(ΔVcm_rms/ΔVout_rms)
ΔVcm = 2Vrms - Vrms = Vrms
ΔVout_rms = (a_rms - x_rms) - (b_rms - x_rms)
CMRR = 20xlog10[V_rms / (a_rms - b_rms)]
Is the above a valid estimation to verify the claimed CMRR?