Anyone out there have experience with using a a micro turbine for generating power for an IoT device. Something like this:
I have an application or two that would sure reduce some headaches if we didn't have to worry so much about the battery budget. It's in doors with limited lighting (i.e. so solar is out) and limited access to mains power. However, it does have plenty of access to water flow. Yes, batteries and leveraging deep sleep modes would suffice but the tradeoff is less granularity of data and even then the battery would need to be replaced or recharged maybe once a year or increased $$ for larger battery.
If we could generate just a little power then we could take much more frequent readings and there would be less hassle of what happens when the battery needs replacement or recharging. We already tap into the water line so the hassle to install is minimal.
Anyone out there try something like this or have any recommendations?
I really wish I had something meaningful to add, but I dont
Here's a long shot : If your device is ~ 200µA average current (maybe LoRa, etc), then indoor solar harvesting has greatly improved recently and might be an option after all.
Good luck with the turbine. I cant wait to hear about your progress.
Yeah all good points. Honestly, I’ve thought about that as well and yes this is LoRa + Particle.
While sleeping a node is about 50 uA if I recall. Or maybe 125 uA somewhere in that range. I’ve contemplated using the AB1805 external RTC and a deep power down sleep to save that remaining sleeping current but the complexity doesn’t seem worth the squeeze quite yet. The main power draw isn’t sleeping rather the awake time.
The sensor we are using unfortunately extends the “awake time” several seconds and consumes more power than just the MCU for the TX and RX short burst of power.
Not only that, ideally we’d have it take readings and report the data say every 5 or 10 minutes rather than every 1-2 hours. This way the battery budget isn’t a concern. The higher granularity of data provides better value to the end user.
I’ve also thought about selectively take readings and transmit or potentially take readings but buffer locally but once again tradeoff of complexity due to the battery budget. This could all be vastly simplified with some power generation.
In the end all tradeoffs. I’ll post here what we learn.
