For a while now, I’ve wanted to have the capability to monitor isolated environments for changes in relative humidity and temperature over time. There are certain areas of our house that always seem hotter or more humid, especially in the summer. I’d like to have measurement nodes to confirm this by logging it over time, which will allow my to take some corrective action in the worst areas.
So the major requirements become:
- Must be able to measure RH + Temp.
- Must be wireless.
- Must be battery powered.
- Must allow remote data to be collected somewhere for future reference.
There are already available products on the market that do just this. One such product is the Wireless Sensor Tag System. These devices are small, relatively low-cost, battery powered, and wireless, with a backend for tracking and data. There are three things that I felt could use improvement, however: battery type/life, range, and the backend application.
Also, I like making things. So I decided to design my own. In addition to the above requirements, I had some additional ones:
- Uses a Li-Poly or Li-Ion type battery.
- Operates in the 900MHz ISM band.
- Has some local intelligence.
- Has a USB interface for battery charging, FW updates, and parameter changes.
I had the option of making the devices true IoT (Internet of Things) products by putting something like the TI SimpleLink WiFi devices on there, but I was concerned about battery life. Granted, the devices only need to wake up once a minute (or every five minutes, etc.,) transmit a data packet, and go back to sleep. Still, it takes time for a device to register on the network. By opting for just a low-bandwidth, low-power transmitter, the device can wake up and transmit a packet within several milliseconds, where WiFi can sometimes take 10’s of seconds to register on a network. According to the datasheet for the CC3100, TX mode can consume anywhere from 160-272mA.
The radio I decided to go with was the TI CC1175 transmitter. By comparison, the spec for transmit mode in the 900MHz band for a +10dBm output is 34mA. I chose this radio for several reasons: low-cost, low-power, flexibility of configuration, and ease of implementation. I’ve had experience with this series of radios before, so the interface feels familiar.
Full disclosure, I’m already pretty far along with this project, but I wanted to walk through the design process for everyone else. In the next post I’ll either detail the selection of sensor or micro.
The TX radio section of the wireless sensor module.