# AD9361 MDS, Sensitivity, and EVM

How do I determine the MDS and sensitivity of the AD9361 and in general a receiver based on a non-ideal IQ demodulator?  For instance sensitivity S=-174 +NF +10log(BW) + S/N of detector to meet BER.  But this assumes an ideal IQ demodulator.  So if my receiver requires a detector SNR of 5 to meet my BER with an ideal IQ demodulator, how do I now determine the sensitivity if I factor in an 17.8 % EVM (SNR of 15)?  And if I use a digital quadrature error correction algorithm (QMC) can I recapture or come close to achieving the sensitivity of the receiver with an ideal IQ demodulator?

• QEC error correction algorithm takes care of this imbalance between IQ. For AD9361 Quadrature gain and phase error are 2% and 2 degree typ respectively. The impact of this will be much less than a dB in terms of SNR.

• >>The impact of this will be much less than a dB in terms of SNR.

But in terms of 'unwanted sideband' the impact is -40 ... -34 dB

• Sorry my mistake.

For AD9361 Quadrature gain and phase error are 0.2% and 0.2 degree typ respectively and image rejection is ~ -50dBc

• According the formula for the receiver's sensitivity:
S=-174 +NF +10log(BW) + S/N, the BW here is the bandwidth of the signal being received and sampled to the IQ data which is the same as a bandwidth of the channel selected by the filters inside AD9361 or the BW refers here as an analog RF front-end bandwidth before an AD9361 RF input?

• The bandwidth here depends on the standard that is used, which is actually taken care by channel filtering in baseband.