I am not aware of a software work around for this issue. Can you post a snapshot of the matching network you have for the radio and the Antenna? Having a low pass matching network (with series L and shunt C) reduces the spurious very much.
Thank you for your reply
My circuit does not have a low pass filter circuit only the antenna matching,
Now that the product is in volume production, I want to know if there are other ways to improve this issue apart from modifying the PCB circuit.
Thank you again
The 6.8 nH shunt, 1.5 pF in series followed by 1.2 pF shunt you have after the ANT pin is actually meant for PSoC4 BLE radio. This is something that you could alter to get it matching with a low pass components.
We have experimented with a firmware workaround for reducing this spurious during the product validation. We got a setting that reduced the spur by 6 dB. But taking up this setting means a full round of characterization (of testing all the parameters across temperature, voltage and process corners on tens of device and ensure that all the other parameters are still intact. The easier solution was to contain in through the matching networks (of the radio and optionally the antenna too).
Can you try the following modification.
Replace L1 with 1 pF
Replace C31 with 3.3 nH
Replace C32 with 1.5 pF
Thank you for your guidelines, I will modify it according to your suggestion,
Thank you very much.