I must admit the more I think about this the more i’m tempted to just oversize everything…
£10.99 for ‘7w’ (probably nearer 5w in reality) solar panel
For the sake of £24 seems like it would be worth it to have it -just work- through a cold and dingy winter?
Nice finding ! I should have found it as it is used in Seedstudio Stalker, the kit I bought with the solar pannel 
Regarding my test, it was not charging yesterday with a charging current set to 100mA.
After searching specs of my solar panel I found that it is a 100mA@5.5V. So I tried to set the TP4056 at 50mA, but does not work today too.
I suspect that it is not linked to constant current setpoint but to the current injected in constant voltage mode as my battery is currently @4.15V
Unfortunately the current required is not in the datasheet I might try to measure it this weekend
By the way temperature inside the enclosure grew up to 60°C, which is usually the max specified in18650 datasheets.
Wow, useful data !
So it is stated that the charger does not start above 4V, my battery is at 4.15, so I will discharge it or try another one 
for big batteries you need a bigger charger/discharger … I made this one today from an old PC power supply and an RC hobby charger, 10 A ! 
I’am waiting for my ’ LiPo RC Battery Guard bag Explosion Protection Charge Sack 240x110/300x230 mm J before switching it on.
No… can’t wait for a explosion protection enveloppe 
discharging for the last 24 minutes @ 2A (and still 3.82V !)
This morning I woke up wit an crazy idea.
I’ve read some articles about tracking antenna’s for drones and solar tracking (big panels)
And I’m thinking … ‘is it possible (and worth the effort) to make a solar tracker for a small 2W 330 mA panel’ ?
From some experiments with these little panels, I think you can increase efficiency by at least 100% compared to a static mounted solar panel.
But you need GPS to automatically start a the right point, fine tuning is done by four LDR’s
a solar dome node
fortunately you can find a lot of information on the web
Even if it should be better, how much it improves harvested energy vs energy consumed by servomotors ? Maybe just one axe might be enough ?
I never really read the page but a friend of mine did it with LDR only, without GPS 
Here in French
Yes that’s a valid question, have to find out the hard way.
By testing (moving by hand a little solar panel connected to a multimeter) you will see how much energy you ‘miss’ during the day.
It’s even possibly to gain more, like 140% 
but my Laird RM186 DEV kit can arrive any moment, so i’m afraid this solar thing will be last on the ‘todo’ list
…or when it gets stuck in the early morning or late night position
ah there’s the sun, time to test this one on an empty 750 mA LiPo
indoor behind glass clouded day test 
you think 6V - 2W - 330 mA = huge ?
I put this setup yesterday afternoon in the sun - it filled the LiPo during that time with 360 mA
but I have some smaller panels coming soon
0.6W 6V 80 x 55mm
1W 6V 110 x 60mm
2W 6V 330mA 136 x 110mm
A small node with a very small solar panel looks nice, could work very well in certain countries with many sun hours, but during winters and a clouded weeks you’ll need a bigger battery and a bigger solar panel to charge it faster.
All depending on how much ‘juice’ your node&sensors wants to consume of course.
It depends of your power requirement. An RFM95 transmitting data + ATmega328 is about 35 mA. About 1.5 second is required to transmit 10 bytes in SF12, 0.5 sec for the application. Data are send every 5 minutes, which gives 352/360012=0,233 mA/hour, I add 4 uA for standby power consumption.
With 1000 mA, it can last 175 days. If the device is connected to TTN with fair policy restriction it will be a lot more. (times 12 ?)
Of course this simple figure does not take into account self-discharge nor temperature factor but lot of modern uC are less hungry than the ATmega328, so I was just curious about your application 
Also, Charles spoke about ESP8266 which I don’t know very well the power consumption…
The figures I used are from my sensor connected to solar panel, even if I still need to measure all of them. Also, I tinker around the solar panel yesterday and it seems that my boost converter is far from efficient ! I hope to have some time to characterize it after my vacations but connecting the solar panel directly to the TP4056 works where using the boost did not. By the way, I can’t wait to receive my CN3065 to improve the efficiency
they seemed to work very well 
@BoRRoZ, I think I missed one important information how consumption can be can be 35mA ? RFM95 datasheet says (several so not sure what the real value)
One page => 120mA
Another One => 90mA
If I understand what you’re writing, does this mean Lora never use PA_BOOST?
I use LTC3105 for solar controller (because need small device and panel ~ 40x15 mm)
Real consumption for RF part ~ 80mA and small battery + supercap provide 2-3 seconds to transmitte. Device size ~ 35x20 mm with sensors, gps.
@charles : True I made a mistake, but we should transmit at 14 dB max in EU, so around 30mA I assume. I will record it after my holydays, I only checked standby current to be sure that the module enters in sleep mode.
@x893 : nice ! Did you post some data somewhere about them ? Which duty cycle ? I wanted to try a supercap only…
@Oliv
oh yes 14dB max, that’s the trick, and so why it’s limited to ~30ma which in fact is perfect for me, got a node that can’t go over 90mA
thanks for sharing
received the board and a little disapointement is, that the LiPo connector is 0.1" which is not standard for all my batteries I have.
That should be corrected or at leased two solder holes added.
I would also like 2 solder points/holes close to the micro-usb as input from the solar panel
otherwise the board is of high quality!
