The MGM210P is based on the EFR32MG21 SoC that has 2 RF ports: RF2G4_IO1 and RF2G4_IO2.
The RF2G4_IO1 port seems to be connected to the internal antenna and the RF2G4_IO2 port seems to be connected to pin 29 of the module throught a matching circuit.
When using the module as a NCP application with the EmberZNet SDK, how the RF port can be selected?
How the RF2G4_IO2 can be enabled if we would to use an external antenna instead of the module internal antenna?
Is-it the setting in the "hardware configurator" of Simplicity Studio in Radio / "Antenna Diversity"?
In the EZSP reference guide, there is also a value "EZSP_VALUE_ANTENNA_MODE".
An update on this discussion to describe the behavior of multiple modules:
On Series-2, like MGM210P, RAIL_AntennaConfig_t.defaultPath (aka ant0Loc) can be:
• 0 (RAIL_ANTENNA_0) selecting F2G4_IO1, the built-in chip or trace antenna
• non-zero (e.g. RAIL_ANTENNA_1) selecting F2G4-IO2, the U.FL connector
RAIL_ANTENNA_AUTO (255 or -1) is treated the same as RAIL_ANTENNA_1 on xGM210x, while on xGM220x this field is completely ignored. On ZGMx30x, RAIL_ANTENNA_AUTO means internal path diversity.
Note that the value 2 on MGM210P thus selects the U.FL connector, but it is not the exclusive value to achieve this output.
What is the proper way to change the code, so that RF2G4_IO2 is enabled?
Changing it directly in the SDK doesn't seem right. I was expecting something like changing hardware config file.