How to simulate the impact of BALUN to WLAN's EVM?

For the WLAN chips, almost all of them use the differential structure for PA and LNA. To convert the balanced two ports to single-ended port, the type of LC Balun could be used. But recently I tried to tune the this kind of Balun and I found it impacted the EVM obviously. So I wonder if anyone knows how to simulate the impact of balun to EVM in ADS? Or anyone have the recommended books or papers related to this topic?

Thanks.

Could be a Group Delay distortion that affect the EVM, usually produced by parasitic resonances of the balun with some external components (capacitors).

You can apply a differential WLAN signal by changing the Phase between the signals.You should see "EVM versus Phase" characteristic.

Any effect should be due to impedance mismatch. Balun should not have effect on group delay because its bandwidth should be relatively wideband. LTCC baluns use low Q coupled dual resonant tank circuits which have narrower bandwidth but more likely effected by incorrect termination impedance (typically stray capacitance).

PA output termination will effect linearity. WLAN has high peak to average ratio therefore PA needs good linearity.