Solar charging of meshtastic devices

Development Hardware
21

Thank you, I should have done some more reading of the forum first. Too bad about the CubeCell, I guess it is still that old tradeoff between power usage, performance, and also ease of use/programmability.

I’ve been building remote low-powered autonomous time-lapse cameras which definitely have a tradeoff between image quality, cost, and battery life. Having an island-wide autonomous mesh for telemetry would help me monitor the cameras via LoRa and know when I need to swap out batteries…

I’ll order some of the meshtastic supported boards and look at the ones best suited for low-power+solar LoRaWan relays.

Thanks for this excellent project. I’m super excited to test this out.

1 Like
22

Here is my experiment combining the output of four (red) equal panels, placed vertically
There is a shade tree on the east side above the two (East and S. East) panels.
Notice how the power jumps when the sun is lower in the afternoon and more
direct into the two west side panels
and how low the output is when at noon with high incident angles to the panels.

solar_2
solar_21941×1107 176 KB

1 Like
23

I’ve bought a couple of these to play with. They’re fairly cheap on clearance, and are a couple of dollars cheaper without the battery pack.
Solar Panel

1 Like
24

For one panel to Canada :mask:
image

25

Ouch! Sorry, not such a good deal if you’re outside the US. It’s free shipping here.

26

Something to remember for people in colder climates is that lithium batteries don’t charge when it’s below freezing. So that’s a challenge for making an autonomous repeater for winter use.

Now that I think about it, solar lights still work in winter so maybe the batteries charge very slowly and that’s fine. I guess it depends on how much power the devices need.

1 Like
27

Fortunately, there are miraculous batteries, (LiFePO4) that charge even when it’s cold and even when it’s hot, they cost even less than lithium batteries.
LiFePO4 batteries can work even in freezing conditions
And they can last between five to ten years, lithium ones last only two years.









temperature management board with relay for the carbon wire

non-flammable thermal insulator
2 Likes
28

LiFePO4 has a pretty similar charge temperature range to Li-Ion. In general, both are no good to charge below freezing from what I have read.

Lead Crystal might be a good option for those in extreme areas. Charge down to -40. 40% capacity at -40 as well.

1 Like
29

Hahaha -40 °C, in this case we need a nuclear reactor used in the Voyager probes for better performance😅.
The Voyagers also used the heat of the reactor to keep the instruments warm because they did not operate at low temperatures.
To withstand those temperatures you need insulation, and a small resistance to keep the batteries warm, maybe you use one battery to keep the other batteries operating the lora module warm.
-40 °C also the lora module cannot work, it needs to stay on zero degrees or more.

30

Straight out of the EByte E22 manual…

image
image809×43 4.6 KB

Some good info on pg 27-28 about charging and discharging lead crystal. https://indd.adobe.com/view/d39c6a34-2775-4f15-9b18-72c22c2eed17

1 Like
31

Nobody forbids you to use your beloved lead crystal batteries, I know those batteries and I know their performance without the need for data.
My proposal above of those materials listed was for a configuration more affordable for everyone, low cost and with the possibility of experimenting with them, also for compact and portable systems.
For low temperatures it is necessary to design specifically for that application.
However, LiFePO4 batteries work better than lithium ones in more aggressive conditions, and they have one more thing that they don’t burn, they are super stable and last five to ten years before they need to be replaced.
Try to do some configuration that works at -40 ° C starting from the batteries from the modules lora the insulation, low cost and everything else.
Everyone will be happy to have such setups from you here and some tests on real world and success.

1 Like
32

Which type of batteries are compatible with the charging circuit on the T-beams, or for fixed installation would we be better off not using the built-in battery and it’s associated charging and just supply power into the microusb connector ? What voltage would then be displayed ?

33

The LiFePO4 batteries are not good for the ttgo circuit, they have lower voltages that are not compatible, only those of Li-ion 4.2v
The TTGO circuit voltages are 3.0v ~ 4.2v end of the charge.
For LiFePO4 batteries you need to use either a small, very efficient 5v buck converter


Or the charge management board for solar panels that I put above in the lists of possible materials
34

Some thoughts on solar charging…

I recently talked to a salesperson in the solar industry regarding another project, and inquired about MPPT chargers for a single panel mobile device. I was told it was pointless for single panel, as MPPT requires much higher voltage (i.e. serial connected panels) to reap efficiency benefits over regular PWM chargers. (As far as I knew, it is the amps that change with the light, not the voltage, so I was a bit dubious). In any case that was out of the question for this particular project.

I would prefer to have any panels connected in paralell, as shading a single cell in a serial connected array drops the total power output by 75%. It is also wise to have a diode for each panel. Yes it lowers efficiency, but the panels tend to suck power from the battery at night if directly connected (at least the cheap, unregulated ones do). I don’t know if the TTgo charging circuit takes this into account.

If your panel has any circuitry for the USB output this may not apply, but for cheap unregulated panels it does, most definetly.

(I was on a hike and plugged in my phone at 50%, it went up to 60% before dark, but when I woke up the phone was completely drained.)

I see sugestions on using solar gardenlights as a possible powersource. An interesting and apealing idea, but… these devices charge all day, and discharge at night. We still want to power the device during day, which introduces a paracitic load on the charge circuit. depending on the size of the paracitic load vs charge load, this may lead to incorrect charging and premature battery failure.

(https://batteryuniversity.com/learn/article/how_to_charge_li_ion_with_a_parasitic_load)

The simplest way I can think of to avoid all that math, is to use 2 batteries and charge 1 while the other one runs the load, then switch at a predetermined discharge voltage. That keeps the charge circuit isolated from the load at all times.

If the TTgo charge circuit+a somewhat intelligent panel (a regulated, possibly of the type used to run wireless security cameras) is paired I think that would be the best solution, unless there are issues with TTgo’s charger I don’t know about. Anyone?

1 Like
35

I have given you ideas of possible materials to use.
You now choose the materials that you believe are the right ones.
I work in electronics and electrical engineering, it is easy for me to build solar systems with everything else needed to make a working system. Here I strongly believe that there are people who know a lot more about solar systems, maybe my suggestions give them an extra push to build something functional at low cost for everyone. I have little time to deal with this.

36

I’ll put my couple of cents on MPPT, for my home panels I was provided with multiple curves showing the rated voltage and current for a ‘given’ light intensity. What is interesting is it is NOT a linear relationship w/r to power, for example the unconnected (no-load) is around 40vdc, however at that voltage the panel will put out a maximum current of 7amp (so 40 x 7 =280 watts).
However, the point of maximum power transfer is at 27.5vdc and 10.5amps ( 27.5 x 10.5 = 288 watts)

My understanding is that internal resistance in the small tracings affect the limits. For me, the cost/benefit of using MPPT ($$$) versus PWM($) rarely justifies the higher cost.

37

I was merely pointing out that in general LiFePO4 isn’t good to charge below freezing as you suggested and that LoRa modules can operate below freezing according to the datasheet. If I lived in a climate that saw -40, it sounds like a good challenge!

Thanks for scoping out those parts for someone to have a crack at it.

1 Like
38

I think they will be useful for someone who likes challenges. :smile:
with a few tricks you can build a portable device that doesn’t have to go to -40 °C because things get complicated and expensive at those temperatures.
However, I’ll do some tests with various batteries at -40 °C ~ -50 °C to better understand their efficiency.
So I can post some more info here.
For someone in the South Pole it will be useful. :smile:

1 Like
39

Some info about LiFePO4 batteries from Andreas Spiess

40

I’m far from the South Pole. It regularly gets to these temps in Alaska. Lots of folks here travel remotely in use-cases which align with the stated goals of Meshtastic. As I’ve said it a bunch of times though, I also realize this climate is outside the fat part of the bell curve. All the same, there’s no comms infrastructure in a lot of the places people travel, and not everyone can afford InReach satellite subscriptions.

Happy to hear people continue to brainstorm for applications outside of the easy latitudes.

1 Like