Meshtastic

Solar charging of meshtastic devices

Since I am in the lower 48, … we still get in excess of 120 inches of snow which affects how I place solar panels … they don’t work well under even a centimeter of snow… If I can get to them, I brush them clean, but that doesn’t help for the ‘repeaters’. That is why I am testing them in a near vertical orientation

Has anyone done the calculations of what it would take to run a repeater for three months without charging?

Swapping a fully charged battery pack in every few months could be the best overall solution in such harsh conditions.

Set it and forget it would really limit you on software updates unless over the air updating and config is added down the road.

Good point on snow accumulation. Have you tried leaving them for a few days and seeing if the snow stays? The snow usually melts off my array when the sun comes out, but I’m not sure how big the overall impact is. When that doesn’t work, wind usually strips it off. Once or twice a year I’ve climbed up to brush it off, but that’s mostly just because I can. Granted, my panels are usually almost vertical as well.

I suspect ice/snow might be more of a problem on panels with a rought texture. As with the rest of it, specific climate will make a difference too. I have pretty dry snow relatively.

I did a few consumption measurements of the TTGO T-Beam V1.1 I have, 150~180mAh transmitting very Long range “High Power” mode which I believe your node will use.
20mAh without transmission.
80mAh when the CPU is active doing its calculations and processes of the radio system.
Calculate a continuous average of 100mAh to stay inside with consumption and intensity of use during the day.
However I have a 2600mAh VTC5 battery and usually in my radio it lasts 24 ~ 28 hours.
I will charge the battery fully, measuring its exact capacity and leave my radio connected in the mesh to better calculate when that battery will run out to make more precise calculations.
I’ll let you know in two days when the radio runs out of battery. :wink::smile:

For relay use the GPS and Bluetooth should be disabled. Maybe even set the ESP32 to 80Mhz instead of 240.

Check this out:
https://www.ebay.com/i/313189536822?chn=ps&norover=1&mkevt=1&mkrid=711-117182-37290-0&mkcid=2&itemid=313189536822&targetid=935431405813&device=c&mktype=pla&googleloc=9059106&poi=&campaignid=10877149894&mkgroupid=107912403275&rlsatarget=pla-935431405813&abcId=9300400&merchantid=6296724&gclid=EAIaIQobChMI46msmKq86wIVnD6tBh33zwf8EAQYBSABEgKLR_D_BwE

That battery with its 3.2 ~ 3.6V can be plugged directly into the 3.3v line of the TTGO bypassing the power management and charging chip.
According to my calculations, you need 80Ah for 30 days of intense use of the network. A 100Ah battery is perfect, so you don’t drain it completely extending its life up to 10 years.
The nice thing about LiFePO4 batteries is that it keeps its voltage at 3.2 ~ 3.4 for almost the entire time of discharge

1 Like

Once the Bluetooth is turned off, and the GPS is configured to sleep 99% of the time and you are not using the built in power management I think we will see a dramatically lower power consumption.

NRF52 boards will be even better.

1 Like

Obviously it will certainly increase its performance to 40% ~ 70% but consider that the radio will wake up continuously if there is communication activity in the network. It will still be better than the ttgo currently available.

New real world battery life data from TTGO.
My TTGO V1.1 with a 2310mAh battery did exactly 36 hours, connected in the mesh and with the phone connected “in very long range mode”.

{(36 hours = 1.5 days)
(1.5 x 20 = 30 days ~ one month)
(2310mAh x 20 = 46.200mAh) ~ (46.2 Ah)}

For a node that acts as a repeater it is better to use a 70Ah battery or better a 100Ah battery to prevent it from being completely discharged in order to extend the life of the battery, especially if it is a LifePO4 battery connected directly into the 3.3v line of the TTGO.
In AliExpress I have seen 90Ah LifePO3 batteries for less than 60 €
I put only one image to give you an idea, I don’t leave links because shipping to your country could be very expensive, it can go as high as the price of the battery itself, but if you search on AliExpress you will find them immediately even with low shipping prices.

For safety reasons, please use a 100mA ~ 250v fast fuse.
Please do not connect the batteries in the 3.3v line of your TTGO without a fuse, or use a BMS 1S 2A protection.
Always take safety seriously

1 Like

Update: did some testing with a 3w non regulated and a 5w regulated solar panel. Both could run the TTGo by itself, but neither could charge the battery AND keep the TTgo…going.

Mesuring voltage and current drawn while plugged into my car, I discovered that the current drawn when charging a beefy 18650 could reach 0.9amps, which is not unexpected for a Lihium battery as they have a constant voltage/variable current(high at start) charging mode, and a constant (low) current variable voltage stand-by state.

This was far more than either panel could deliver, so the voltage from the panel dropped, the TTGO charger cut power to the board (I think) and all went dark.

Disconecting the panel and rebooting got it up and running again.

.
So, I guess that is that. :thinking:

Why not solder wires to the terminals of the 18650 holder and connect a separate charge controller that bypasses the PMIC on the T-beam. That way you can probably get away with smaller Panels and charge while the device is running.

The problem with very small panels, type 5w when connected directly to the battery they consume the battery when they do not have the right power.
Usually a silicon diode is put in, to avoid the reverse current.
If the current return back from the battery to the solar panel, it can react like a big infrared led, consuming the battery.
Another problem with silicon diodes is their 0.55v voltage drop.
If a panel produces 5v, with the diode it will only produce 4.45v.
Another negligible problem for all is the efficiency of the panel itself, for example the solar panel when not connected can produce 5v but under load it can only produce 0.9v at ~ 0.9Amp or 0.5v at ~ 0.2 Amp ~ (just random data, they need a verification on real world application)
To charge everything efficiently you need at least a panel of 25 ~ 35w to cover days without direct sun.

https://youtu.be/-SJbdPvgQnE

good guide for battery/solar power

Oooh, I didn’t know huge LiFePo from AliExpress were a thing. (I’m guessing they’re not a “delivered quickly by airfreight” thing, I have had awful experiences getting anything except small LiPo batteries out of China, last time I ordered 5000mAH sized drone batteries, then ended up going seafreight to Austria then airfreight to Sydney. I’d totally given up on them arriving at all, and was pleasantly surprised to have them show up ~6 months later. By which time I’d crashed that big drone and settled on much smaller ones. Ended up using those batteries in an electric bicycle for a year or two…)

One other thing to keep in mind (with all bettery chemistries, but specifically for LiPo/LiFePo) is the capacity degradation as the temperature drops.

To a first approximation, you lose ~10% of capacity dropping from 30C to 0C, then another ~15% of capacity for every 10 degrees below zero. By -40C you’re lucky if you get 20% of the rated capacity. (WHich is better than a standard lead acid battery, which will probably have frozen solid, split the cases, and be sitting there waiting for spring to leack sulphuric acid everywhere…)

You might need 5 of those 90AH cells to run a repeater for a month in winter in some places…


This is a simple cheap solution for concurrent charging. Getting 41mA in the shadow outside without directing the panel at the sun. It’s a 6V 1A 9dollar panel from ali with a CN3791 charger connected to the battery terminals directly.
Even in this almost worst case scenario you get useful energy from this small panel. 40mA is twice the current draw of the nrf52-based boards.
I think when deploying repeaters you might have to hide them somehow in a tree or so and therefore there might not be direct sunlight most of the time.

3 Likes

With the sun going down at 16:30 on Sept. 19. I was getting 900ma when directing it towards this huge fusion reactor.

1 Like

The question is to keep the batteries warm with the capacity that the ttgo does not use.
We don’t have to run the batteries below 10 degrees.
To keep the batteries warm, they must be insulated with fireproof thermal insulation and an internal heating system must be created that keeps the battery temperature under control.
Batteries that work well at -40°C do not exist, if they do exist, they are very expensive and no one needs them

I will admit to having almost no personal experience with running any sort of batteries at what some people here would consider “low temperatures”. Im in Sydney, we do not insulate our houses much, and complain when it drops below about 10C outside… The closest I’ve ever been to properly cold is local Australian or New Zealand ski trips back in the 90s when my camera batteries kept running out, and the February I spent in New York once where I realised I needed to keep my iPod in a pocket inside my jacket to get it to last all day…

I’m a motorcycle rider. “Below freezing” translates to “unfit for human habitation” in my mind… :slight_smile:

It’s nice to be a motorcycle rider
I always say that if I have to choose between an F16 and a motorcycle, I would choose the motorcycle :grin:
Returning to the question of batteries.
Lithium batteries aren’t great in low temperatures, but they have one thing that shouldn’t be overlooked, their capacity compared to weight.
If a 50Ah battery is enough for a month for the ttgo, a 100Ah battery would be enough to keep the battery warm with good insulation.
I recommended LiFePO4s because they keep their voltage constant when plugged into the ttgo’s 3.3v line, bypassing the ttgo’s charge management chip, so that not so much energy is wasted.
One more reason to use LiFePO4 is their price and the fact that they don’t catch fire if something goes wrong.
Remember that LiFePO4 batteries should not be charged if their temperature drops to zero degrees.
Despite this, LiFePO4 can provide energy even at minus ten to minus twenty degrees
Another effective way to keep batteries warm is to put them underground.
Making a hole in the ground and insulating around with insulation such as Polystyrene and a container that does not enter the water inside, perhaps insulating everything with silicone for gutters

@bigiain

So, I have an idea for arctic batteries.


this is basically a thermos with a solar heat input. It is vacuum isolated, (which is the best kind) I will not guarantee that the inside does not drop below 0 at night, but there won’t be any solar charging going on at night anyway. You may need som LCD blinds to stop it from overheating in the day though.

If you are REALLY handy, you could make one yourself, and coat the inner tube with
white thermocromatic paint that goes black at +10 and below, that way you have a self regulating totally passive battery heating unit.

SunRocket solar kettle
thermocromatic paints

Optionally you could throw in some TEG units to convert excess heat to electricity. This will work better than solar panels on cloudy days.

1 Like