A project I ran 2016 but when I saw the t3chflicks teams project and noticed people wanting to chime in I thought I'd disclose my project as well to either share ideas between the projects or simply to get some energy back into the project. I ran into issues with sensor inaccuracy and pause the project. Actually started another project I'm more actively developing at the moment. But if people are interested to join the WaveBee effort please let me know and we will take if from there :-)
Some calculations made today trying to size solar panel and LiPO battery. Alot of work but finally settled on the panel size and as well a needed amp hours to survive a full winter.
The amount of parameters looked at would take too much time to go over but this is the forecasted lifetime after plotting two years. If I go-live in January I wont experience a full winter from a solarpowered perspective since only Jun-Aug could be considered summer months. But with a 6000mAh I make a full year and summer seems to be able to charge it back to 100%
They are pretty even. But small adjustments makes a hugh impact on the overall consumption.
One thing I hadn't realize is that adding a larger battery which seems like an obvious solution to prolong the life of the node is that increase the amount of totalt energy self-discharged over the year which needs to be produced in order to restore a full charge. Might not even be possible. Just masks the issue.
This is a 2000mAh LiPO battery
And this is a 6000mAh LiPO battery. Since the battery contains more energy that 5% of self discharge will now increase in terms of Joule. Since the production side of things (solar panel size) has not grown it could be that we couldn't even keep up with that loss. So here we see that LDO waste and the self-discharge adds up to about a third of the power/energy usage.
Single event consumption
Looked quickly at the power usage from an IOT event-perspective.
MCU awake period is ~3 seconds. LoRA transmit 0.5 seconds and I pretty much expect to measure the full 600 seconds in between sending data.
Plotted an event. Stacked line graph. ~256 mJ consumed during each event which would equate to the area.