RN42 Antenna

Introduction:

I am planning on making a remote control robot using a Microchip PIC24 microcontroller and an RN42 bluetooth transceiver chip. Although the RN42 is sold as a kit or module by various sites, I wanted to make my own printed circuit board (PCB) for practice along with having a full analysis and design of the antenna for the RN42.

I begin the robot project by first analyzing the various PCB antennas used for Bluetooth. I will first analyze the PCB antenna used by Microchip on the RN42¹ module. Then I will compare the results to various other antennas including the inverted F antenna (IFA), wire monopole antenna, and other IFA configurations.

In addition, I will analyze the use of microstrip line from the RN42 Tx output to the antenna for efficiency benefits.

Microchip RN42 PCB Antenna (RNA):

When analyzing the RNA one should take into consideration Microchip's documentation^1,2. The Tx output line from the RN42 is approximately Z₀ = 50Ω.

This however, is not relevant for the antenna analysis itself. The RN42 datasheet² in pages 8 and 9 states that the antenna must have a clearance of 30mm of clearance on 3 sides (although for some reason it says 31cm on the sides of Figure2-3 which I believe is a typo). So our model must have this clearance when modelling in OpenEMS³.

Figure 1 shows the model via OpenEMS. The red color is the ground plate, the blue is FR4 substrate with dielectric constant of ε_r = 4.3. The small dark green square is the feed port to the antenna which is modeled as a lumped element voltage signal with a real impedance of R = 50Ω.

Figure 1: OpenEMS Microchip PCB Antenna Model

The pcb design considerations have now been met by the OpenEMS model.

Unfortunately Microchip provides little and vague information on the RNA antenna itself. Luckily the open source OpenEMS is available to give us the input impedance and a better visualization of the far field pattern along with the efficiency so we can better compare this MA with other antenna configurations.

Side Note: In addition, it seems that most if not all PCB antenna information is scattered across the internet without a real general rigorous solution or analytical method available to use. I believe this is because pcb antennas are very difficult to solve analytically and are usually analyzed numerically and computationally. However, as I go along with this project I will most likely be collecting various information about pcb antennas which one can hopefully find useful rather than having to spend hours on google hoping to find the correct information.

RN42 Antenna (RNA) Results

As one can see in Figure 2, the resonance frequency is for the RNA is:

F_resonance = 2.586GHz

as calculated by OpenEMS. The center frequency of the RN42 is 2.45GHz which is very close and so we expect to not have much problems.

Figure 2: Reflection Coefficient S11 at RNA input

Figure 3 shows the input impedance of the RNA structure, which OpenEMS calculated at 2.45GHz to be:

Z_in = 7.5313 - j7.9730

Figure 3: Input Impedance Zin

Figure 4 shows the far field of the RNA (please click on video). It is a donut shape but overall it is very isotropic which is what is needed for a remote control robot. In addition OpenEMS gave a maximum directivity of:

D₀ = 1.9478 = 2.895dBi

Again, which tells us that the antenna is not very directive and has the desired isotropic characteristics.

Figure 4: Far Field Visualization of RNA

OpenEMS also calculates the efficiency of the RNA structure. At the center frequency of 2.45GHz we obtain an efficiency of 98.05% while at the resonance frequency of 2.586GHz we get 99.25%, which isn't much of a difference as expected. However, I am surprised this structure is so efficient as one will see that other structures, although more simpler, have a worse efficiency.

OpenEMS IFA Tutorial Results:

OpenEMS comes with an IFA example, as shown in Figure 5, analyzed at a frequency of 2.5GHz, close to what we want.

Figure 5: OpenEMS model of IFA

OpenEMS calculated the resonant frequency to be:

F_resonance = 2.42GHz

Figure 6 shows the input impedance, which was calculated to be (at 2.42GHz):

Z_in = 52.7196 + j1.8494

Notice that Z_in has a real part of near 50Ω. This is desirable in order to obtain less reflectance at the input.

Figure 6: Input Impedance Zin of IFA

Figure 7 shows the far field visualization (please click on the video).

Figure 7: Far Field Visualization of IFA via OpenEMS tutorial

Additionally, OpenEMS calculates that the directivity is:

D₀ = 1.8007 = 2.5544dBi

And the efficiency being 99.33%.

The IFA obviously has less directivity than the RNA, which is what we need. In addition, it seems that it also has better efficiency.

So far it seems that the IFA is a better choice than the RNA.

Next on the Project:

• Modify the IFA to fit with the RN42 Tx output
• Find more PCB antennas for Bluetooth
• Compare all results
• Design microstrip line from RN42 Tx output to antenna to check for better efficiency.

Citations:

1. http://ww1.microchip.com/downloads/en/DeviceDoc/50002328A.pdf
2. http://ww1.microchip.com/downloads/en/AppNotes/RN-42%20Antenna%20Characteristics.pdf
3. OpenEMS. (2015, October 10). openEMS, . Retrieved 02:29, October 26, 2015 fromhttp://openems.de/index.php?title=OpenEMS&oldid=934.