1

I'm trying to devise a strategy to allow my system to run off lithium batteries. From what I understand charging a lithium in freezing temperatures is a big no-no, and my project will face these during winter, so I'm wondering about diverting solar charging energy towards a heating element (heat tracing wrapped around the battery?) during these conditions so that I can continue safely charging.

Here's a concept of what I've come up with:

enter image description here

A temperature sensor connected to the MCU determines when a freezing condition exists and controls the relays.

In a freezing conditions the battery is only allowed to discharge (in the diagram above it's shown in the wrong switch position - sorry), and the heating element is switched on, driven exclusively from the solar input.

Both relays would need to be bi-stable type so that

  1. The coil doesn't need to be constantly supplied current (this is a power-constrained device)
  2. If the battery fully drains and there's no solar input then with the MCU off the system maintains its last state.

My ESP32 and its connected LTE-M module draw ~150mW on average (with Tx/Rx spikes every 15 minutes). I'm looking for about a 15W 30V solar panel, and the battery I'm targeting around 20Ah (74Wh). I'll want my heating scheme to not add significantly to the power footprint when it's disengaged.

The application is for continuous acquisition of sound data from an I2S microphone, processing, and cloud upload of statistical data. The power numbers above are measured from my prototype.

Does this design make sense? Are there any obvious flaws I'm not seeing, or likely pitfalls? Am I reinventing the wheel here, ie are there any design patterns out there for this kind of thing?

Thanks in advance.

davegravy
  • 319
  • 2
  • 8
  • Can you post some links on what could happen if you charge the battery in freezing conditions? – Marko Buršič Jun 15 '21 at 12:34
  • @MarkoBuršič https://electronics.stackexchange.com/questions/263036/why-charging-li-ion-batteries-in-cold-temperatures-would-harm-them – davegravy Jun 15 '21 at 12:49
  • 1
    I posted a substantial answer regarding the use of insulation and water to allow low temperature battery use. Should be findable. ,(I'm on phone at1am at present) – Russell McMahon Jun 15 '21 at 12:53
  • 1
    Why would you go for Li-ion if it has such drawbacks? Why not using LiFePO4? – Marko Buršič Jun 15 '21 at 12:53
  • See here https://electronics.stackexchange.com/a/467545/3288 – Russell McMahon Jun 15 '21 at 12:54
  • 1
    If under freezing temps, Li-ion will have *significantly* less capacity. Pack battery in an insulator such as fiberglass or similar to keep it's heat in. Consider adding a [Li-SOCI2](https://www.batteryjunction.com/batteries-chemistry-lithium-thionyl-chloride.html) primary cell and diode to power the device if/when the Li-ion is depleted. Li-SOCl2 are good to -55°C. – rdtsc Jun 15 '21 at 13:43
  • What will be used for? 150mW is not big power, so you may have miscalculated things, IMO just the board dfr0535 will be consuming near this power, also 15W/30V solar cell is not even compatible with dfr0535. Edit the question with details. – Marko Buršič Jun 15 '21 at 17:01
  • @MarkoBuršič Edited question. DFR0535 specs state 7V ~ 30V solar input. Quiescent current <3mA – davegravy Jun 15 '21 at 17:49
  • @MarkoBuršič Reason for not choosing LiFePO4 is energy density: there is constrained space for the battery. Even with reduced cold weather capacity LiPo seems to have higher density (just inability to charge) – davegravy Jun 15 '21 at 17:55
  • What is the lowest expected operating temperature? – Bruce Abbott Jun 16 '21 at 02:43
  • @BruceAbbott The coldest it's ever been here is -33C. Most winters have at least one day where it gets down to -20C but it's not too common. -12 to -7 is common. – davegravy Jun 16 '21 at 03:45

0 Answers0