NRF21540 setting +13dBm without SPI, expected RX sensitivity with CODED_PHY

[Jason]

My current product uses the NRF523840 with CODED_PHY periodic adverts. I want to improve the range.

I'm considering a +14dBm (SAR limited) product variant with the NRF54L10 + NRF21540

The NRF54L10 has fewer pins than the nrf52840. I'm aware that the TX gain varies considerably according to the frequency and temperature and there is support for compensating for this in NCS FEM driver. My product will need this to pass SAR testing.

I want to know 2 things

1. Is it necessary to connect the SPI pins of the FEM to specify +13dBm TX power? I imagine it might be possible for the FEM driver to vary the NRF54 TX power up to +3, and lock the FEM Mode to +10. Is this how it works?
2. Does the NRF21540 also improve RX sensitivity of CODED_PHY periodic adverts (S=8, 125kbps)? Has this been evaluated? Is there any data on this? 

Kind regards,

-Jason

**EDIT**:   I have found the header files for the MPSL FEM, and I can see comments that seem to support the idea that the SPI interface is not required to set the TX power to +13dBm 

**EDIT2** I've also found this post which re-confirms that the SPI isn't needed.  Setting TX power for nRF5340+nRF21540 

I guess the advantage of using SPI would be that the FEM is more power efficient. If so then it might be worth trying to factory set +13dB in the nrf21540 UICR

I would still like to know if the 21540 improves the RX for CODED_PHY if there's any info available on that.

Thank you!

[Jason]

I think GPT5-pro answered my question

Do you think it's right?

[Asbjørn]

Hello Jason

Q1: There's a combined value to be added. So if you're using the MPSL FEM  drivers you set the wanted output power and then it set the appropriate levels for nRF54L10 and for the nRF21540 to achieve the wanted +13dBm. I see you've found your information on this already, so that is all good.

Q2: Regarding RX sensitivity on CODED_PHY. It looks like the GPT-5 summary is good and would be along the lines of what we would say as well. Perhaps we would focus more on the "it depends" bit, because often there is noise in the 2.4GHz band and you will see something between the theoretical best and the no gain scenario. Most likely you'll see a bit of gain on RX with coded phy, but there's also a higher risk for collisions as it takes you longer to send packages/data with the 125kbps mode. 

Best regards

Asbjørn

[Jason]

Great! thank you!