(AN5592, ST25R) Differential Antenna with cable

[yeji_jang]

Hello ST Community,

I’m working with the ST25R3911B on a NUCLEO‑NFC08A board and would like to drive a custom differential antenna via two separate coaxial cables. Here is what I’ve done so far and the questions I have:

• I disconnected the onboard PCB antenna/cable and wired two 40 cm jumpers to the PCB antenna pads. (X-NUCLEO-NFC08A board)

• I was able to run the reader in differential mode, but significant decrease in Rx power due to impedance mismatch caused by wires.

• My goal is to design a custom differential PCB antenna and connect it to the ST25R3911B board using coaxial cables, use differential mode to maximize the delivered RF power.

1) Dual‑Coax Differential Connection
I am considering fitting an SMA connector on each of the chip’s RF Output 1 and RF Output 2 pins and likewise two SMAs on my differential antenna. I would then use two 50 Ω coaxial cables to link them.

• The AN5592 datasheet only shows a single‑ended coax example.

• Is it acceptable to run the ST25R3911B in true differential mode using two separate 50 Ω coax cables and SMA connectors? Any tips or concerns?

2) Alternative: Matching Tool & Non‑Coax Wiring
If the above approach is not recommended, I plan to:

• Use ST’s Antenna Matching Tool to design a matching network.

• Wire my antenna to the board with simple wire (not coax) and measure the antenna + wire characteristics with a VNA.

• Apply the matching network based on those measurements.

• Does this workflow make sense? 

Thank you for any guidance or suggestions you can provide!

[Travis Palmer]

Hello Yeji,

It would be interesting to find out why your solution did not work as expected. Maybe you can share a few more details: (smith chart, voltage on the input of the cables, voltage on the output of the cables, voltage on the RFI.

You stated that the Rx power was decreased due to mismatch of the cables. Why do you consider that the Rx is more impacted? You should build up the same system without cable and compare the Tx power at the antenna and Rx signal. In this way you can asses the impact of the cable. Can you share the schematic used for your prototype?

In my opinion you need to adapt the results of the single ended calculation of the AMT (antenna matching tool) and simulate the system you want to build. 

A coaxial cable is mostly chosen to have a defined characteristic impedance of the cable. A twisted pair cable is more likely to deviate from cable to cable and also within the cable the impedance could change. A coaxial cable should be independent on the placement (metal close to the cable). A twisted pair cable has the advantage of being symmetrical. The differential signal can proper gate through one twisted pair cable instead of two coaxial cables. But bending and placement can influence the (characteristic) impedance of the cable. 

I expect that a twisted pair cable works nicely for a prototype, but might be hard difficult to implement on a series production. 

BR Travis