We are using the Johanson balun specific to the CC2520. We have tried different matching inductors, from 1.5 to 3.9nH. 3.3nH gives us our best results for a good sinX/X spectrum, but at certain power levels the spectrum is distorted. Is this because the CC2520 conjugate match requirements are different at different power levels? Please see below for an example of the distortion, seen at 0dBm output setting. Thanks.