Asymmetric differential rf output in nRF51822

[fontikos]

We have designed a differential antenna for the nRF51822 CEAA chip.

Each of the RF output connects through a matching circuit to each one of the two 50-ohm to ground end of the antenna. The transmission lines on the board and matching circuit components are completely symmetric.

However, using a spectrum analyser, we are measuring an approximate 8dB difference in the output power of each port, as shown the picture. The spectrum analyser is connected to the board via u.FL connectors.

Shouldn't the differential output be symmetric and give equal power levels on the two sides?

Would it be possible to provide us with more information about the differential amplifier of the radio?

Is there any particular reason why in all Nordic reference antenna designs the differential output is not used and, instead, it is transformed into a single 50-ohm end?

[Asbjørn]

In theory yes, they are symmetrical. This will be affected by your matching implementation. Parasitic effects and small difference could change the output by several dB.

We don't have any further information on the interface that we share as this is a part of advantage towards customers.

The reason for why we recommend using a single ended antenna is that most antennas for 2.4 GHz are single ended 50 Ohm. For example wifi and regular bluetooth use this as a reference. Also the interface toward spectrum analyzer are often 50 ohm and this makes the testing easier. We just want to make it simple for our customers to implement our radio and antenna into designs and at the same time it to be relatively easy to do output power measurements. Secondly, the matching network also filters the output, so in order to meet regulatory requirement around the world, we recommend to follow our guidelines.

[fontikos]

Thanks for the info.