I'm just developing the second iteration of an 2.4Ghz sensor and receiver set. Because I can't afford the expensive development path of a big commercial project and because the environment these products get placed in is not constant or easy to stimulate, I'm trying to incorporate both an external antenna connector and a PCB antenna, with the option to select between the two.

To sum it up:

  1. Allow for switching between the PCB antenna and the external one. It's ok for this to be done by soldering/removing components.
  2. Have connectors (UFL) for both external antenna and VNA needed for antenna tuning.

Given the above, I was thinking of the following solution:

  • Transmitter ->
  • UFL with short stub for VNA ->
  • PI matching network ->
  • UFL connector for external antenna ->
  • 0 ohm resistor with two positions to select between PCB and external antenna

Does that sound like a good solution?


1 Answer 1


why the first UFL? that connector will bring in more inconsistencies than you'll fix by being able to measure at that point.

Also, I don't know your 2.4 GHz transmitter (IC, I guess), but that thing usually needs matching of its own.

So, in my world, this would look more like

  • 2.4 GHz interface of IC
  • matching circuit for 50Ω transmission line impedance
  • 50 Ω microstrip line up to 0 Ω "selector"
    • on one side of that, line, matching, and chip/PCB antenna
    • on the other, line, and RF connector

In manufacturing, often having different population options comes at a cost. Check whether a simple RF switch IC isn't the better option financially at the quantity of boards you'll produce, overall, and control that through a GPIO and some firmware on a microcontroller you probably have in your system, anyways, for example.

  • \$\begingroup\$ The first UFL is to allow measuring before the PI network. This is a strategy I've seen mentioned in a lot of articles and manufacturer notes on the subject. \$\endgroup\$
    – Alex
    Sep 13, 2020 at 12:30
  • \$\begingroup\$ @Alex I'd do that on a prototype board where there's nothing but the known-impedance transmission line and space for the matching circuit. Adding a U.FL for calibration might still be tolerable at 2.4 GHz (I don't know, feels dangerous), but the higher you go in frequency, the more the presence of that connector changes what you're doing – and leaving it open unterminated also makes it an antenna. \$\endgroup\$ Sep 13, 2020 at 12:50
  • \$\begingroup\$ So it should be ok for 2.4Ghz? That would greatly shorten the development time and cost. How would you do the selection circuit? Or just make it a really really short stub? \$\endgroup\$
    – Alex
    Sep 13, 2020 at 17:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.