Micro-Coaxial Cable Effect in Antenna Design?

[niholan]

Hi All,

I am currently studying compact antenna designs for small hand-held devices.

One of my initial thoughts was to grow my antenna pattern onto a piece of FR4 board and fed it with a piece of 50ohm micro-coaxial cable, which was then connected to the U.FL connector soldered on the main board of the terminal.

In this configuration, the antenna is connected to the device only through the coaxial cable.

After some quick experiments, I found something quite puzzled.
If I made a PCB antenna in any form and connect it to the system ground with a piece of coaxial cable, the return loss changes hugely while the cable is routed in different way or position. It even creates some unexpected deep responses on S11 which are not supposed to be there.

It gives me an strong impression that this piece of coax is becoming part of my antenna pattern now (w.r.t. the ground plane of the device).

Based on this, I have a few questions and hope you can help me clarify some of them. Many thanks in advance for your help!


1. Is this what you are supposed to see if you do not stick your antenna onto the system ground plane but connect them with a coaxial cable instead?

2. I heard someone said this coax won't be a big issue if you design your antenna in dipole or monopole( provided that it has a ground plane area which is large enough on the PCB for the antenna ). In other words, the antenna should be of balanced form or given a large ground plane for itself. Is that right?

(Although my quick experiment rejected the dipole one, I am afraid the fact that I didn't use a balun in-between the dipole and coax might be ruining the result)

3. I found some so-called "ground plane independent" PCB antenna on google search, and most of them are connected through coaxial cables as well. They seem to be of monopole form with a small block serving as a local ground plane on the PCB.
**broken link removed**
Will these types of designs suffer from the changes for coaxial cable length or connecting locations?

4. Currently, patch is my best option without taking the huge size and valuable bw into account. Any other suggestions or thoughts on this?

Lastly, if you read the whole post till this line, you have proven that you are an antenna expert or you really enjoy discussing anything about antenna. Please allow me to express my gratitude on meeting you and sincere appreciation on any of your advices, suggestions, or discussion.

 

[tonyoxy]

Hi,

you are right: in my opinion in your case the cable is an active part of the antenna.
In other words the "antenna" is formed by the radiating structure (we design with electromagnetic simulator) and the coaxial cable (that is, in principle, an umpredictable structure).

There are a lot of confusion on design of WIFI antenna expecially because the electromagnetic simulator feeds the radiating element by an exitation that is self balanced.
In the real word we know that the source are single ended so we need a balun to separate the radiating structor from the feeding port.
Generally the patch antenna design doesn't take into account of the balun and the result is that the radiating properties depends on the feeding port (i.e. the coaxial cable).

 

[biff44]

Yes, antenna cables are often part of the antena, with rf currents running along the outside shield.

If you have the room, you can make a qurterwave choke on the outer shield to keep the rf currents off the cable shield. You can use a ferrite bead to somewhat help reduce the cable shield currents.

In some cases, you WANT the cable to have rf currents, as it affords a bigger ground plane, which can improve antenna efficiency.

 

[niholan]

Hi Both,

First, thank you very much for your replies! It's great to have a discussion with someone rather than scratching my head all day long for the whole month... lol

to Tonyoxy,

Initially, balun was one of my few known methods for preventing RF current flowing backward through the outer shield of the coaxial cable.

However, after a quick survey, most of the definitions available on-line say "balun is designed for connecting an unbalanced device to a balanced one", in which mostly between the coaxial cable (unbalanced) and a dipole antenna (balanced). To me, it sounds like balun could be helpful only if a balanced-type antenna was used, not for unbalanced types. Is this correct? If so, anything else might be helpful for unbal. to unbal. configuration (other than providing a "local" ground plane on the unbal. antenna end)?



To Biff44,

For the lamda/4 choke you described, it reminds me of a sleeve dipole, which is considered to be ground plane independent by itself. I guess the reasons behind are quite similar - to minimize the backflow current on the cable?

When you said sometimes ppl do want to have some backflow current on the coax to enhance the efficiency, do you mean the radiation from the coaxial cable actually can be used as part of the design? In other words, if a low band antenna is desired, could I reduce the actual radiating element length by using the coaxial cable of the right length as part of the design? (Though, I have to admit that I haven’t encountered any techniques about reducing the antenna size by adapting a piece of coaxial cable as part of the antenna design.)

Besides, according to my personal experience, the BW for those resonance caused by counting in the coaxial cable part is quite narrow, and the efficiency is quite low compared to S11 at the corresponding frequency, as if those resonances were generated in-between the two different currents flowing on the inner and outer shields. (inductance effect) ?


===================================================

I read a document written by an antenna company who sells these types of PCB antennas, where it says they provide the so-call "detuning" services, because the S11 could be different after their pre-designed antenna is connected onto a terminal. This really makes me wondered if these kinds of design all end up in customized form... :|

 

[biff44]

yes I was thinking of a sleve choke.

There are plenty of handheld devices that benefit from the user holding it in their hand, and thereby forming a "ground" for the antenna to work with.

I have tested a whole bunch of purchased small antennas and chip antennas, and my own printed antennas,. They all look like garbage after the pcb gets into the act. My approach is to make the garbage at least smell nice!

 

[niholan]

Hi Biff44

Thanks for your sharing of valuable experience~

Looks like I was hoping too much to turn those garbage into eatable. lol