Another Project14 Challenge complete - sort of..... As always, some goals were met and some were not. I thought that I would summarize where I got to.
Blog Post Index
- LoRa Experimental Environmental Sensors (LoRaXes) - Introduction
- LoRa Experimental Environmental Sensors (LoRaXes) - Need to use a bigger case
- LoRa Experimental Environmental Sensors (LoRaXes) - GPS is Working
- LoRa Experimental Environmental Sensors (LoRaXes) - Case Layout
- LoRa Experimental Environmental Sensors (LoRaXes) - Temperature Sensor Working
- LoRa Experimental Environmental Sensors (LoRaXes) - Initial RSSI Tests
- LoRa Experimental Environmental Sensors (LoRaXes) - Case Assembly Part 1
- LoRa Experimental Environmental Sensors (LoRaXes) - Assembly Complete
I started this project to find a use for a Things LoRa Gateway and two MKR WAN 1300 boards that I had won in a previous challenge. This was my first experience with LoRa and the Things Network.
- Connect gateway to Things Network - there were only five LoRa gateways within a 50 km radius of my location that were connected to the Things Network, so I've added access for sensors near me.
- Build two portable battery operated sensor units - I've only fully completed assembly of one of the units, but I have tested the electronics for both.
- Connect both units to the Things Network - I have unique applications running for each MKR Wan 1300 on the Things Network console.
- Provide GPS functionality for one of the units - the fully assembled unit has the GPS module and I have used it with the mapping integration on the Things Network to display the unit's location.
- Measure RSSI in the proximity of my house - units have good operational margin on my property.
Goals Not Achieved:
- Determine the effective LoRa range of my gateway/sensor setup - I need to find the time to get some data walking the GPS unit away from the gateway. I should be able to correlate the RSSI data associated with the GPS data packets and use -120 dBm as a range metric (or I guess the real metric is when packets are not received - use the last location received). My accuracy may not be that good if I don't do some averaging. I've noticed that over time I can get geolocation outliers of 50-60 meters even though the clustering is more like 5-10 meters (with the unit in a static location).
- Measure/optimize battery life - I sort of gave up on this one when I realized that I have older MKR Wan 1300 boards that do not power down the Murata LoRa module in sleep mode (fmilburn pointed out this issue). The achievable sleep mode current is ~2 mA. I have seen implementations of full power shutdown using the TI TPL5110 ultra low power timer. I think I will give that a try. I can make room on the proto board to handle that. And I'll need to make sure that the GPS handles save and restore of the tracking data correctly. I guess even if I had a newer MKR Wan 1300 (released the end of 2018) I would still have had to power down the GPS since its power save mode draws ~11mA. I might need to look at other GPS units.
- Determine the reliability of my system by analyzing uptime - this is difficult to quantify in the prototype stage since I've been constantly power cycling the units. I have noticed that my gateway reboots at least once daily. I probably need to automate some processing on the Things Network console.
- The sensors won't have any long term utility unless I can implement a solution to extend battery life, so I need to do that first. I'll try to implement a TPL5110 power control. Then I'd like to add some other sensors to the unit without the GPS. I'll move the BME280 outside the case and add a PIR and an ambient light sensor and put the unit in the backyard for monitoring and data collection.
- I'd also really like to get the effective range data. I've seen posts for setups like mine that indicate the range may be as short as 0.5 km for an inside antenna. I'll need to research what types of outside antennas others are using with their gateways. It would be fun to track my walks with the dog.