Post navigation

MeshNet

Key technologies: RF, 802.15.4, battery-powered sensors, C, FreeRTOS

After building Wimic, I wanted to try another, more advanced project. I had heard a lot about ZigBee, but I found that the ZigBee stack was large, complicated, and resource-intensive. I wanted to design my own simpler protocol that would use less power and work on smaller microcontrollers.

I designed the hardware over the summer of 2010. Towards the end of the summer, I started a blitz to get the hardware finished. Unfortunately, my brother and dad needed my help on a 48-hour road trip out to Idaho. We had a mattress in the back of our 15-passenger van, and I spent several hours on the trip sprawled out building schematic symbols and footprints. I got the hardware wrapped up and sent out for PCB printing before I went back to college.

This project taught me the importance of soldermask. I ordered the PCBs without soldermask to save money (I used the “Barebones” offer from Advanced Circuits). When the PCBs arrived, I had an enormously difficult time soldering the components. I had routed traces under my 0603 passive components, and the solder was forming bridges underneath my capacitors. I was able to get the boards working, but I have since vowed never to order boards without soldermask.

On this project, I also tried out a PCB-trace antenna. This was much cheaper than using a chip antenna. The antenna worked very well; I was able to get a few hundred feet of range.

The hardware was successful, and I began writing the FreeRTOS applications to run the radio. The radios worked, and I was able to transmit data between sensors.

Unfortunately, the software for this project became a much larger task than I had expected. I began to understand why the ZigBee protocol was as large as it was, and I eventually stopped work on the project due to other priorities.

Leave a Reply

Your email address will not be published.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