Learning to use the Air602 WiFi Module

nsaspook

Joined Aug 27, 2009
13,086
After much experimenting, I'm about to finally install the module on one of my circuits so that a PIC16 will be able to interact with it. But now I've run into a conundrum... one of the module's pins (pin #2) is intended so that an antenna can be connected to it. I've been playing around with an evaluation board of said module, and I noticed that the antenna's connection is not as straightforward as one might think. In fact, after some measurements and verification, I found that the antenna is connected in series to what seems to be a ferrite (a small inductor whose resistance is about 0.015Ω), and also to a capacitor (1nF) connected to ground. Here's a couple of pictures, plus a diagram of the way I understand the antenna is connected to the module:


So I have three questions:
  • Are said components absolutely necessary? Or are they more of a recommendation to filter and refine reception/transmission?
  • Should the geometry of the antenna I plan to use be similar to the one present in the evaluation board?
  • Is the length of the antenna important?

Here's a clip from the datasheet that might be pertinent to my questions:


@nsaspook, this sort of thing seems to be right up your alley, any thoughts? @joeyd999 awhile ago I followed one of your threads in which you documented your progress with a Bluetooth module, maybe you'll find this thread to be of your interest.
Yes, the components are absolutely necessary (for that antenna design geometry) to transfer power from a source designed to produce X power into X resistive impedance. The LC (that can be electronic components or the equivalent physical values from conductors) matching network is much like a tuned transformer that optimizes power transfer from one set of driving impedance combinations (source to antenna) to another impedance combination called free-space and back again in receive mode. The antenna is designed to phase shift the fields generated across the conductors in respect to existing fields moving away that were previously generated across the antenna. The limited speed of 'light' means changing RF field effects happen at different times across space. At some size, length and shape (antenna geometry) of the electrical (modified by LC components if necessary) length of conductors RF power is matched to space and that RF power moves into space. The transfer of power from the antenna to space is seen as resistive by the power source so a closer resistive impedance match means more power transfer.
 

Thread Starter

cmartinez

Joined Jan 17, 2007
8,220
Thanks NSA... It's unbelievable how much there's to learn about antennas. I declare myself an un-willful ignorant on this subject... that is, I'm going to take this opportunity and try and learn the basics of antenna science... something tells me that this is the sort of craft that one never stops learning.
 

Thread Starter

cmartinez

Joined Jan 17, 2007
8,220
I've narrowed my antenna criteria to three options:
  1. MIFA design (Meandered Inverted-F Antenna)
  2. IFA design (Inverted-F Antenna)
  3. Use a chip antenna.

1609281392882.png

1609281418382.png


1609281831332.png

  • Options #1 and #2 are cheaper and easier to implement, but much harder to design, according to some of the literature I've gone through.
  • A chip antenna is easier to design.

All options mentioned require a matching network, and said matching network depends on the geometry specific to the PCB in which it will be installed. According to a document I read, the best practice (other than using a simulator software) is to start with approximate values for the network components, and then adjust them after measuring the antenna's performance using a Vector Network Analyzer... but the most affordable VNA's out there have a starting price of $10K ... lucky me...

Given the complexities involved in designing PCB antennas, I opted for the chip type and acquired a few of them for testing. Since the datasheet shows the following example, I bought a few 4.3nH inductors as well.

1609282100343.png


I'm going to start by installing those components and see how it goes... the question now is, is there an inexpensive method to measure the antenna's output and calculate and test adjustments as needed?
 

nsaspook

Joined Aug 27, 2009
13,086
Thanks NSA... It's unbelievable how much there's to learn about antennas. I declare myself an un-willful ignorant on this subject... that is, I'm going to take this opportunity and try and learn the basics of antenna science... something tells me that this is the sort of craft that one never stops learning.
Using good antenna design examples from the net will usually get you working antennas for most applications without adjustment.


It's one of the subjects that you need to forget almost everything you think you know about circuit theories assumption of instantaneous cause and effect and start thinking about fields, wavelengths (physical and electrical) and retarded (in time) effects across conductors not in traditional uniform serial current loops. When you see the antenna, you don't see the wire, you learn to see the electrical effects of the wire in space as changes of potentials across the conductors at instants in time during movement of the applied signal.

A very good older 'free' book with mainly HS level math is this one.
Transmission lines, antennas and wave guides by Ronold Wyeth Percival King
https://archive.org/details/in.ernet.dli.2015.177615/page/n219/mode/2up

https://en.wikipedia.org/wiki/Inverted-F_antenna
The inverted-F antenna was first proposed in 1958 by the group at Harvard led by Ronold W. P. King.[4] King's antenna was in wire form and was intended for use in missiles for telemetry.[5]
 

Thread Starter

cmartinez

Joined Jan 17, 2007
8,220
Here's an interesting question:

I plan to turn the Air602 module on and off using a low side nFet, as shown in the following diagram.

1609364355216.png

Also, I plan to follow the example mentioned in the datasheet of the chip antenna that I'll be using and install a 4.3nH inductor connected to ground as its matching network.

The question is, to which ground should I connect the inductor to?
  • Should node (A) be connected to node (B), or to node (C)?
  • Or does it not make a difference whatever node (A) is connected to?

Something tells me that the right answer is to connect (A) to (B), so as to have the module's ground always at the same potential.
 

nsaspook

Joined Aug 27, 2009
13,086
The chip is half the antenna, the other half is the RF ground plane of the PCB to make a dipole. The module/antenna grounding and keep-out recommendations should be followed to the letter if you expect decent performance. The GND pins on the module must be directly connected to the properly designed PCB ground plane for the matching components and antenna to work.
 
Last edited:

Thread Starter

cmartinez

Joined Jan 17, 2007
8,220
The chip is half the antenna, the other half is the RF ground plane of the PCB. The module/antenna grounding and keep-out recommendations should be followed to the letter if you expect decent performance. The GND pins on the module must be directly connected to the properly designed PCB ground plane for the matching components and antenna to work.
Thanks for your observations. I have tried to follow the guidelines as close as I possibly can. But one problem I'm facing is that I can only (for now) build single sided boards. So unfortunately there's no ground plane beneath most circuitry. But maybe that's not too important since in the datasheet it's shown that no ground plane should be present under the chip antenna anyway.

And if I understand you correctly, it's not a good idea to split the circuitry's ground by using an nFet as I proposed. So maybe it's a much better option to use a high-side pFet to power the module instead?
 

nsaspook

Joined Aug 27, 2009
13,086
Thanks for your observations. I have tried to follow the guidelines as close as I possibly can. But one problem I'm facing is that I can only (for now) build single sided boards. So unfortunately there's no ground plane beneath most circuitry. But maybe that's not too important since in the datasheet it's shown that no ground plane should be present under the chip antenna anyway.

And if I understand you correctly, it's not a good idea to split the circuitry's ground by using an nFet as I proposed. So maybe it's a much better option to use a high-side pFet to power the module instead?
There's nothing directly under the chip but there has to be a ground plane Counterpoise for that quarter-wavelength chip antenna to work as a radiator of RF. A metal foil or coating in the enclosure is sometimes used but that's a design element that requires tuning.
https://www.digikey.com/en/articles/understanding-antenna-specifications-and-operation-part-2
Designing with a quarter-wave monopole antenna

A common pitfall for designers new to the wireless arena is the implementation of the ground plane. As stated earlier, the ground plane is the other half of the antenna, so it is critical to the final performance of the product. Designers have to get it right.

The ground plane is a solid copper fill on one layer of the circuit board that is connected to the negative terminal of the battery. This fill not only acts as the antenna’s counterpoise, but is also the ground connection for all of the components on the board. The problems arise when other components are added and the traces are routed to connect them.

It is a very rare and simple design that does not need to route a trace on more than one layer. Every trace that gets routed on the same layer as the ground plane can have a significant impact on the RF performance. It is best to look at the board from the perspective of the antenna connection. The goal is to have a low impedance path back to the battery or power connection. This is accomplished with wide, unobstructed paths.
High side power switching is always preferred with RF circuits.
 

Thread Starter

cmartinez

Joined Jan 17, 2007
8,220
There's nothing directly under the chip but there has to be a ground plane Counterpoise for that quarter-wavelength chip antenna to work as a radiator of RF. A metal foil or coating in the enclosure is sometimes used but that's a design element that requires tuning.
https://www.digikey.com/en/articles/understanding-antenna-specifications-and-operation-part-2


High side power switching is always preferred with RF circuits.
Thanks for those observations. I'll keep them in mind while designing the PCB.

I've watched these two videos about antennas, and found them very informative:




 

nsaspook

Joined Aug 27, 2009
13,086
Thanks for those observations. I'll keep them in mind while designing the PCB.

I've watched these two videos about antennas, and found them very informative:




The line 'kink' is Divergence and curl in EM equations.

The only thing I didn't like about the 'Understanding EM' video was the movement visualization of electrons in the wire. They don't travel from one end of the antenna to the other end as the E field polarity reverses. The massive numbers of free electrons in a good conductor shift slightly from their neutral positions as localized current giving rise to the B field component of the EM near-field wave. In the Dielectric resonator antenna design we obviously we don't have large numbers of free conduction electrons in parts of the physical antenna that are not galvanic conductors but we have do molecules that can be easily deformed by the E field causing the same degree of shift and generation of B fields in response to the applied potential.
 
Last edited:
Top