Estimating power consumption

I’ve been trying to estimate the power consumption for Aqla. I’d like to order parts so I can start building a prototype. There are some expensive parts that needs dimensioning; the battery and the solar panel. The closer I can get a good estimate for these parts, the more likely it is I can reuse them to build the final device. And since it is a hobby project, I don’t want to splurge needlessly.

There are three design parameters to consider for powering the system:

1. Time spent in different modes (how often the sensor is sampled and how often data is uploaded to the cloud)
2. How large battery is needed (needs to run on battery of a given amount of time without recharge from solar)
3. How large solar panel is used (determines how fast the battery can be recharged)

There will be two tasks that will draw power; measuring and uploading. (Idle will be any time not spent executing the tasks) How often they are active determines how much power will be required.

From the table we can see that the current consumption is about 3 mA on average. At 3.3 V voltage supply, the power consumption is around 10 mW. The battery will need to last at least 5 days without charging, in case of bad weather. ² That gives a capacity of ≥1.2 Wh. The battery should recharge within 2-3 hours (0.5C), so the solar panel should be about 2W.

The time spent for each task is not fixed. This allows for some wiggle room to adjust the average power consumption, if it is found that the hardware won’t support it sufficiently. It can also be acceptable with reduced performance; maybe 100% of the samples doesn’t need to be captured. Maybe we can drop 10% and still have a useful system. Or the 5 days on battery power is overly harsh; maybe it only occurs once per season and then its fine to loose some data.

Once the prototype is assembled, it will be very interesting to see how these estimates plays out.