How To Make A Cactus – Multi Band Antennas

In the previous post, I discussed the fact that an Antenna should be designed from its stem, namely, its feeding point. I wouldn’t go as far as to say it’s easier to design this way, but it’s definitely better. I am going to review here a few aspects of antenna design in general, and multi band antennas specifically.

Hardest Question, The Specs:

When approaching clients of various kinds, be it in-house system engineers, creepy managers (not you, other managers) or as a free-lancer, one of the hardest questions to get proper answers to is the specifications of the design. Let’s talk about a few of these parameters:

  • Bandwidth – At what frequency range, or ranges, the Antenna should be well matched.
  • Reject bands – Just the opposite. Frequencies that the antenna should reject the signal altogether.
  • Gain – How much power is radiated and in what direction.
  • Size – How big (or usually, how small) do you want this antenna. Smaller antenna means smaller gain.
  • Target impedance – Here’s a shocker for you: You can’t always reach 50\Omega. Specifically, when interacting with various integrated circuits (ICs), this might not even be the goal, in the first place.
  • Polarization – Hard to explain in one sentence. I will probably dedicate a post to this.

There are many more parameters, but this is just to get us started. Usually, the Antenna isn’t a standalone product and has to operate within aggressive environments, so reaching these specs is usually more a negotiation, rather than a technical discussion.

Designing An Elementary Antenna

So kicking off from the (erroneous) feed presented previously, let’s match for the first band, namely 2.4_{GHz}-2.5_{GHz}. Let’s build a pole in the correct length, and viola:

Let’s look at the current distribution, as well

Fine and dandy. Let’s start approaching the second band.

Adding A Far Away Band – What Could Go Wrong?

An elementary pole will generally radiate close to its fundamental frequency and its next harmonics. A harmonic is (usually) a multiple of the fundamental frequency. So you can see that the antenna is matched in the expected band of the 2nd harmonic, as well. Not exactly, of course, but close enough.

However, the required band is close to that second harmonic, maybe even overlapping, and that is not good. Two resonances cannot exist together. Usually, each of them requires a completely different current distribution and they cannot co-exist.

Before talking about exactly how I solved this, let’s try to solve this problem like engineers, first.

Question:

Given an infinite area to solve the problem and not caring about anything apart from operating in two separate bands, how would you solve this? Take a moment…

Answer:

The simplest solution is the right one. Design two different antennas, and place a diplexer between them. A frequency diplexer, of course. These devices are usually available off-the-shelf, but are easy enough to design, as they are a low-pass (LPF) and high-pass filter (HPF) connected at a junction. I know it’s a tiny bit more complicated, but really, not that much. Diplexers can be designed with discrete components, T-Line filters, whatever.

Challenge:

So assuming we want to keep this design small enough, we need to integrate the Diplexer, somehow, into this Antenna. Step 1, add a LPF on the Antenna. For example, strategically placed open ended stubs.

I know, not perfect, but luckily engineering isn’t science. Let’s look at the current distribution now at the resonant frequency

and at the filtered frequency.

You can see that the power is not accepted by the antenna, and if anything, is dissipated at the LPF stubs.

Step 2, add second Antenna pole. Since the first antenna isn’t accepting signals anymore at the problematic bands, connecting a 2nd pole close to the feed should do the trick.

Viola! Cactus!

Not An Epilogue

I showed one approach here, to solve a specific problem. There are various ways to approach these problems. Capacitive coupling, different types of filters, multi-mode designs… If there was one solution, a lot of people reading this wouldn’t have a job. Experience is one factor here. Open-mindedness and a proper problem-solving approach is another.

Feed your antenna properly, eat healthy yourselves.

Cheers!

Leave a Reply

Your email address will not be published. Required fields are marked *