During one of those nights i stumbled upon a solar charger on Aliexpress. [LINK]
It seemed to check multiple boxes and according to the description also provided enough juice and space inside to power my nodes.
There is a version with e remote to turn the panel on/off but the range was limited and turning it off did not work reliably.
I also added a INA219 Power meter to check on the consumption and battery level.
Instead of connecting the RAK19007 to the batteries, I chose to connect it to the Panels 5V DC output.
In theory it should cut power if the battery voltage gets too low and enable it, when the batteries are full again (however did not verify that this is the case).
Also the Panel has a power switch so the system can be power cycled if needed.
I also added the temperature module to the RAK19007 so keep an eye on the temperature just in case.
And I have to say it works like a charm! Without the power meter one could fit 4x18650 cells and that would last for a while.
With winter approaching I was thinking of adding a thermal switch between the solar cell and the charging circuit to prevent charging when the temperature is below 10°C and wait for the housing to warm up a bit until it is safe to charge.
After initial testing, the data seems to indicate that this works well for a solar powered router node
Data of the depicted solar router for a 24h duration:
I used hot glue to prevent the SMA connector from spinning, but as suggested i took some time today to seal the node with a proper sealant. Thank you for your suggestion!
I noticed that there is a slight gap between the solar panel and the case and patched that up. Would have been a shame to get the node damaged by water seeping in!
Having a gap between the solar cell and the box isn’t even that bad. I inserted 5mm plastic spacers. It then indirectly can act as a sun shield. By glueing it directly to the case it will heat up a lot because they are usually black.
Sure thing, so I got an INA219 board from DFrobot [LINK]
The thing is, that it uses a smaller shunt resistor as the ones configured in the firmware. Resulting in a way smaller current measured (x10). This can be fixed by adding the following to your board variant:
#define INA219_MULTIPLIER 10.0f
For the wiring i used the GPIO Pins on J12 on the RAK19007 [LINK]
The pins provide power and I2C.
As far as I know the power on those pins is not “switched off” if power save is enabled, so it is constantly powered. Not a huge issue in this case.
with the connector in the image going into the solar input of a solar power management board (I am trying this one Battery solar charger: Hailege CN3065 Mini Solar Lipo Charger Amazon.de but there are others) Update: this Hailege board did wake up sometimes when the device ran flat, but got stuck and stayed off sometimes, so the board does not get my thumbs up.
What kind of connection or software do you use to keep track of the stats like that? I’ve turned on telemetry and such and can fetch the data, are you connected directly? Or are you just monitoring e.g. mqtt, or perhaps even monitoring via another node that you directly connect with?
Just as an update, I ended up using another Meshtastic node I have and script it. I connect to the node, ask for telemetry over Meshtastic, once received I push it into HomeAssistant. Python script, running as a cronjob every 30 minutes. Will see how it impacts battery live overall. Based on current testing, a single 3400mah cell seems low. Based on estimations, it will not survive two days in a row with no to little solar on i.e. cloudy days - which happens frequently in my area. Time will tell, otherwise I’ll simply add another battery, as there is room for two total.
