Particle Flex Antenna & general RF question

I have 2 questions for RF folks out here.

First, a bit of background…

I have the Particle PCB ‘Flex’ 2.4GHz antenna that I use on my Photons. They offer improved reception over the chip antenna which is great. While looking around for shorter u.fl patch leads to replace the existing 25cm lengths, I noticed the following interesting bits:

I assume the horizontal straight strip on the PCB to be the groundplane. It is soldered to what I think is the outer conductor / shield of the 1.32mm coax cable. However, upon checking with my DMM, I discovered that the centre conductor is shorted to the outer conductor casing, most likely due to the soldering job on the PCB end (image below). Thing is, I’ve got 5 of these flex antennae, and they are all shorted this way.

From my own (very) limited RF knowledge, I always thought that the centre conductor shouldn’t be shorted to the groundplane. The u.fl bulkhead connector on the Photon seems to indicate this. However, the Flex antenna does work (it improves reception somewhat, compared to the chip antenna).

Here’s a closeup:

So, my questions are:

  1. For this particular Flex antenna, should the centre conductor be shorted to the outer casing?
  2. if I were to source for shorter coax patch leads (here’s one example from digikey, I am planning on cutting it in 2 and getting 2 cables as a result), how should I terminate the PCB end? Should I connect it the way it’s being done now, or should the centre/outer conductors be isolated?
  3. Are there any other considerations besides the 1.32mm OD and shielding when selecting a replacement coax cable for the antenna?

Perhaps another case of posting too hastily… upon closer inspection, I realised that the PCB traces are the ones connecting the shield and centre conductor, so it’s not the soldering.

Time to rephrase the question: in this antenna design, why is the shield of the coaxial cable connected to the centre conductor/antenna?

Hi @chuank

I have not studied this particular antenna closely, but what matters is that the antenna behaves at 2.4GHz, not at DC.

A good analogy would be measuring the resistance of an inductor: you would see a near dead short on an ohm meter and yet the inductor when used with AC signals has a certain impedance over frequency characteristic determined by its ability to maintain a magnetic field.

Another good thing to think about is a loop antenna such as was used for many years for UHF TV signals. At DC it is a direct short, but at UHF frequencies it is matched the impedance of the transmission medium (air in this case) so that a signal can be received.

With these 2.4GHz antennas, the shape of the conductors, the spacing between then and the dielectric material (PCB in this case) all contribute to the antenna design.


Much appreciated @bko! The analogies helped a lot, and it has helped crystallised other questions I had a while ago when I was figuring out matching circuits in antenna datasheets.

Would it be safe to say, generally, that antenna design is all about nailing the right impedance over a desired frequency then?

Yes I’m obviously not an engineer by training, but I’ve found the whole process learning about RF to be like voodoo and so, so intriguing…

Hi @chuank

Matching free-space impedance is important for good power utilization, but what matters most is the electric and magnetic fields generated or induced into the antenna. Software for designing antennas uses numeric approximation techniques to look at the complete design including the boundary conditions where the conductors meet the dielectric.

Simple antennas can be solved with fairly simple hand-written equations, but complex antennas like this one require computer simulation see the interaction between the electric and magnetic fields.


Thanks again @bko.