A tale of wearables and things
I’ve had a love-hate relationship with flexible printed circuits (FPCs).
On the one hand, terrible to manufacture, solder and maintain. On the other? They are so cool.
Extremely tough for RF designs, due to repeatablity issues. Then again, it’s flexible! Can’t bend Rogers’ materials, right?
Although it is not the goal of this post, in the future I’ll discuss further how to overcome the technical issues of FPCs and only keep the cool stuff. But for now, let’s make something with it!
Back when I started my way in this mess of an industry, a buzzword that used be thrown out a lot was IOT. Ahhhh yes, the internet of things. That undefined, unknown field that included nothing and everything, all at once. It was basically a code word for “I have a cool sensor, let’s connect to the cloud!”, as far as I understand.
One of the repeated sub-subjects that I came by was wearables. Wearable devices, patches, stickers, etc. Designing Antennas with wearables (or any device, for that extent) is tricky business. The dielectric properties of the skin, flesh and fat, vary slightly from person to person. Things like body hair and wrinkles can cause air gaps which mess everything up, as well. My oh my, I hate air gaps…
So one approach would be to design your Antennas and devices to work with a variety of media it may come in contact with. Generally, this is a very good idea, however it will inherently require several design-manufacture-testing cycles. Another approach is to separate the Antenna and the problematic material. How? A sheet of metal.
But wait! Wearable devices can’t be thick, and antennas too close ground have bandwidth problems! So how do you overcome that?
A Single Element Breaks, But A bundle Is Strong
Let’s take a reasonable stackup for an FPC (although a bit on the thick side):
And build a single patch on top of it:
Not nearly sufficient for anything, right? So let’s consider a strategy to make this wider. Now this may surprise you, but a resonator Antenna (e.g., patches and dipoles) are an EMC-nightmarish version of filters. They are good filters, but also happen to be good radiators. So why not approach this like a filter design problem? Let’s create multiple poles and connect them in a way that will allow a wider bandwidth!
Strategy #1: Brute force
Let’s create a slightly larger antenna. No thinking now, just doing.
Now all that is left is finding an appropriate trace to connect them together. For that extent, I’ll calculate the microstrip line parameters, due to this stackup. Some optimization, and…
Viola! Wider band! It is probably wise to make the lower-band antenna slightly smaller, but forget that for now. You can continue this logic further with more and more antennas, but I promise you that you are going to hit a wall eventually. A broader approach is necessary
Strategy #2: Start Thinking Like An Engineer (Again)
RF lovers, grab your Smith charts! Let’s look at the first antenna, in both resonance frequencies. Notice something?
You know it. The 1st antenna has high impedance at the 2nd resonance frequency. Hence, de-facto, it’s an open-circuit and only slightly loading our filter. Did I say filter? I meant your first log-periodic patch array!
Take this as point for though for the next lesson. How would you change your design approach?
See ya’ll next time!