DIY: Monitoring my chillies: The hardware

Earlier this week, I blogged about my plans to start monitoring my chilli plants in the first few weeks of their life. The primary goal is to know about the average soil temperature, as that needs to be high enough to allow the seeds to germinate well. In this follow up I wanted to dig a little bit deeper in the material and hardware that I have choosen to use for this project.

Chillies

The hardware

The photo above shows all parts that I plan to use in this setup. Let's have a look at the choices made.

  • Temperature Sensor: The little black dot in the center of the white tiny breadboard is the temperature sensor. I used the TMP36GT9Z by Analog Devices Inc. Update: Putting the temperature sensor together with the netduino in the red box turned out to be a bad idea, the netduino caused temperature to rise way to much to be valuable to my use case, so I've switched the temperature sensor for a sensor probe, the waterproof version of the DS18B20 by Maxim Integrated.
  • Microcontroller: The sensor is wired to a Netduino plus 2, which will act as the brains for this solution. It's a small microcontroller (384 KB storage, 100KB RAM, 168MHz cpu) that runs the .Net Microframework and has networking capabilities, including an HTTP stack, which will be required to send the measurements to the cloud. It has no native capability for battery power, so I need to wire it up, but in this scenario that's just fine.
  • Wireless Bridge: To enable wireless networking, I have opted to go with an external wireless bridge instead of a netduino shield. In this case the ZYXEL WAP3205 V2, which is actually a regular wireless access point that you can switch into client mode so that it acts as a bridge for the microcontroller. My choice to use a wireless bridge is mostly inspired by the need to demo this solution elsewhere. Now the netduino can simply be configured to use the DHCP provided by the bridge, so that I can change all network settings in the bridge whenever I come to a new location. Otherwise I would have to redeploy on every occasion to embed the wireless SID and credentials. Furthermore as the solution is only temporal, 6-8 weeks, I can use the device later as a regular part of my home network.
  • Housing: The sensor needs to get into the soil to measure the temperature there, but the soil will be wet and damp all the time, which is obviously not a great combo with electricity. My plan is to put both the netduino and the breadboard with sensor into a small plastic container, the red box, tape the box shut and bury the entire container in the soil. I just hope the netduino doesn't emit to much heat, so that it would influence the temperature sensor to much. This will require some testing to see if it will work out well and if adjustments need to be made to the software to deal with any difference that might be caused by the heat of the device. Update: Temperature rises significantly after putting the netduino in the container, I've used a probe instead. Still putting the netduino in the little red box though to protect it from soil or water spill.

Next time, we'll dive into the code deployed to the netduino, so stay tuned!

PS: I'm a software developer, not an experienced DIY'er, nor do I know a lot about electronics. If you see anything fatally wrong with this setup, please let me know about it before I fry my netduino!