Portable cellular alarm button - Part 1

How do you design a portable cellular alarm button? What do you need to consider? What are the known pitfalls? How do I finally get the necessary radio licence? These are just some of the many questions that arise before development.

I chose the cellular alarm button as an example because this week a reader of my IoT M2M Cookbook wants to develop such an application with NB-IoT and LTE-M. The IoT M2M Cookbook already covers a lot of the basics. The book reflects my experience as a development manager for tracking devices, field application engineer for wireless technology and CTO of a distributor of wireless modules and antennas. The first step for any wireless IoT device should be to find the concept for the antenna and housing. Only when this has been tested and works can the experience be transferred to the device's PCB.

Designing the cellular antenna

The antenna concept and the enclosure concept are related. I told the designer of the case that I wanted a portable alarm button. The result is the sketches in the pictures. The rounded alarm button looks very interesting. Unfortunately, such a button is probably not feasible in the size shown. The second sketch is slightly larger and it is technically possible to make it into a mass-produced device. Whether the now 50 mm length can be realised without some RF tricks will be shown later in the test setup.

Once the approximate size of the housing has been determined, it can be printed in ABS. In the IoT M2M Cookbook the influence of the plastic on the antenna is shown in the simulations.

The emergency button project should be divided into several parts. One part is a functional PCB with all the components. The size of the PCB is not important. The antennas can be connected with cables. This test board proves that the chosen circuit concept works with all components. This board can then be used immediately to develop the necessary firmware.

The second part is to find the right antenna concept. The first step is to copy the Dual F antenna from the IoT M2M Cookbook, using the dimensions from the book. As this antenna has only two arms and therefore primarily two frequency ranges, this antenna is easy to understand. Add to this the free PCB tracking antenna study to give the alarm button designer a lot of experience before trying to downsize the PCB. The cookbook and study take the designer step by step through the process of integrating antennas. Any change in the antenna structure and any change in the proximity of the antenna will change some parameters of the antenna. To implement the examples you will need the IoT M2M Cookbook, the free PCB antenna study and a vector network analyser (e.g. MegiQ VNA). FR4 with two layers is suitable for the test PCB. The 50 Ohm stripline is connected to an SMA socket at the edge of the PCB. The dual F structure for the antenna is not etched, but glued on later with copper foil. This allows you to change the antenna structure several times and gain experience of what happens when you change the length or width of a conductor structure. What exactly happens is already described in the study mentioned above. However, the learning effect is much greater if you build the test board yourself and reproduce the possible changes yourself. The experience gained with the PCB antenna can be transferred directly to the small PCB for the alarm button.

Transferring the experience to the alarm button PCB

The IoT M2M Cookbook and the study show that the alarm button PCB cannot be made much smaller. The length of the PCB affects the lowest possible frequency for NB-IoT and LTE-M. The dual F antenna from the Cookbook is probably too large an antenna for the alarm button. Chip antennas give the impression that they can be used to make the PCB smaller. This is an illusion. Chip antennas have to follow the same laws of physics. The length of the PCB cannot be reduced at will. The antenna datasheets give some indication of what happens when the length of the PCB is made shorter than the reference PCB. The graph compares two chip antennas in the lower frequency range. It is clear that the efficiency of the antennas decreases as the PCB length is reduced. The effect of changing the width of the PCB, a nearby battery or a plastic case is not documented. You need to test these changes yourself close to the antenna.

The smallest PCB is obtained by moving the antenna structure into the third dimension. The antenna then floats above the main PCB. In the simplest case, the main PCB is extended, scored twice, folded twice by 90°. If the PCB is properly prepared, you can bend it several times before the material breaks. In mass production, the PCB is bent only once. Such an antenna is very cheap and does not have to be purchased and assembled separately. Another possibility is to use a layer of flex PCB. It is a little more complicated to stamp the antenna onto metal. An empty Coke can and a pair of scissors are sufficient for the test setup. Tinplate is easy to work with scissors. The antenna structure can also be printed on a plastic carrier using LDS or printed on the housing cover using LDS. There are many ways to move an antenna into the third dimension. If a budget is available, the possible antennas can be simulated.

Summary

The smaller the PCB becomes, the more effort is needed to develop the cellular antennas. By splitting the alarm button project into parts, software development can start immediately on a larger test board. The development of the cellular antenna in partial steps allows the experience with the dual F-antenna to be incorporated into the final antenna concept. Moving the antenna into the third dimension will probably result in the smallest PCB. However, this small concept cannot be scaled down at will. Therefore, a simulation with a subsequent test setup is recommended. If the budget for this is missing, only time, and patience with a lot of trial and error will help.

Outlook on the alarm button

In the first part of the article on the development of an alarm button, we only dealt with the cellular antenna and the enclosure. The GNSS antenna, the battery, the software concept, the radio approval and much more have not been mentioned yet. These topics will be dealt with in subsequent articles. I offer training on the MegiQ VNA remotely via video call. You can also ask me to simulate the antenna in the third dimension. And if you want to cut the antenna out of a Coke can, you can book a few hours of consulting. I am happy to share my knowledge. Enquiries are welcome to harald.naumann (at) lte-modem.com.

Sources:

Excerpt of the IoT M2M Cookbook: https://www.gsm-modem.de/M2M/m2m_iot_cookbook/

Free study on PCB antennas: https://www.akoriot.com/white-papers/

MegiQ VNA: https://www.akoriot.com/megiq-vna-0440/

Calibration of the MegiQ VNA on Youtube: https://youtu.be/8CV8cvbUTMo

Remote training on the MegiQ VNA: On request to harald.naumann (at) lte-Modem.com