Solar Powered

I'm making a weather station that needs to be solar powered.
Currently using a Yun, dallas one wire temp and other sensors that don't use much power.
Currently I have this battery V25 USB Battery Pack | 6,400mAh 23Wh with a small cheap solar panel. Only test I did so far was using the Yun and the temp sensor. On a sunny day battery went from 4 bars (full) down to one bar. Doesn't seem like this setup will cut it.

I have been eyeing this kit, but it seems expensive.

V44 USB Battery

Capacity: 12,000mAh, 44 Watt Hours
Output: 5V/2A and 5V/1A USB (2 outputs)
Input: 5-6V, 1A
Battery Type: Li-Polymer
Protection: Short Circuit, Over Charge, Over Discharge, Over Current, Over Temperature

I have also seen some power sonic batteries, but they are 12 volts.

Any have and recommendations for powering a Yun via solar?

Maybe you would be better thinking about using a different Arduino that does not suck so much power or seeing if the Yun can be put to sleep.

You can do some reading here: Reducing Arduino Power Consumption - SparkFun Learn They talk about some good ways to cut power consumption when you don't need it. Put things to sleep, except when you want to take a reading and/or transmit data.
Hope this helps.


Any have and recommendations for powering a Yun via solar?


it appears the battery is Li-Polymer, which is very sensitive to heat. Is you battery in the shade and is it being kept cool?


Assuming your weather station is stationary just use a large lead acid battery.


Put things to sleep, except when you want to take a reading and/or transmit data.
Hope this helps.

Those would be good ideas, IF the main power drain were the cycles used by the '32U4 processor idling between updates. However, with the Yun, a majority of the power is used by the much faster AR3391 processor and the WiFi radio. I don't believe there is a way to put either to sleep or power them down.

Putting the '32U4 to sleep between updates is like fighting to stop the trickle of water leaking through a crack in the dam, while ignoring the torrent of water gushing through a major breach.

To get this to work will either require a bigger (and more expensive) battery and solar panel, or a more power efficient board. Perhaps an Ardiuno FIO which is specifically designed for battery power, and supports XBee form factor communications modules? (Since the Yun is being used, I assume some sort of communications is required.) The Yun is simply not designed for power efficiency and battery operation.

Thanks for all the comments.
I'm using a Yun as all the data is ending up on a web page and also being written to the SD card. I'm aware of how to put an an Uno to sleep and save power, but the Yun is a different story like ShapeShifter said.
Also the Yun, some sensors and the battery will be inside a Stevenson Screen. Although mine is more like a large birdhouse with lots of holes in it.

Still researching powering this.
I have been potentially looking at this battery as well.

Anyone know the difference between a battery like this and a deep cycle battery.
Would I damage either battery if it was drained all the way?

Any lead-acid battery will not be happy with a complete discharge. This is true for a deep discharge wet cell, and for your indicated sealed led acid. A deep discharge battery doesn't like being discharged below 50%, the SLA's threshold may be higher.

Just remember that in the long run, the solar cell needs to be able to provide more power in than the Yun takes out, or the battery will eventually be drained. The purpose of the battery is to ride out night time and stormy periods where you don't get full sun.

While taking the night/day cycle and the seasons into account, you need to figure out the average power that the cell can collect (not just the full sun output) and this average must be more than the Yun's consumption. For example, if the Yun is drawing an average of 300mA, that's about 7.2 amp-hours per day. Assuming a fully charged battery, your indicated battery will last about a day and half before it hits 50% discharged. In addition, your solar cell will have to average at least 7.2 Ah per day, even during a short stormy winter day. So, for example, during the shortest day of the year, you might only get 6 or 7 hours of usable sunlight, you would need to get at least an amp out of the solar panel into the battery.

These are just rough numbers, using an estimated current. You will need to determine the actual average current used by the Yun, and the power going into and out of your battery. I don't have real numbers, but my gut feeling is that you will need a bigger battery and a fairly large solar cell. It probably won't be cheap.

Don't be too quick to dismiss a lower power solution. Assuming that the weather station isn't out in the middle of nowhere, but is reasonably close to a power source (like in your back yard) rather than spending all of that money on solar cells and batteries, it might be cheaper to use something like a Fio and an XBee link. The Fio collects the data and sends it over the XBee radio to a Yun with an XBee shield. The Fio is battery/solar powered, while the Yun has a normal AC power supply. The Fio collects the data, powering down itself and the XBee between samples, while the Yun stores the samples to SD card and serves up the web pages. It's a little more equipment and coding, but is likely to save a lot of precious power out at the sensor station, resulting in a much smaller battery and solar panel. It could very well be significantly cheaper.

A deep discharge battery doesn't like being discharged below 50%, the SLA's threshold may be higher.

It's a little more complex - even if that rule of thumb is reasonable.

The major problem is that the actual capacity is probably only 50% or 60% of the sticker capacity - even for good quality batteries.

What wears out batteries is throughput (not surprisingly). If you take lots of amp-hours out (and put them back) every day the battery will have a shortened life compared to taking a few amp-hours out every week.

What damages batteries is not giving them a full top-up charge about once per week.

I try to limit the daily discharge of my batteries to about 33% of the sticker capacity - i.e. drawing 70 amp-hrs (and replacing most of it) from a 220Ah battery. And then I give them a full charge after 3 days. I want them to last at least 3 years.


Yes, I will probably draw about 300 mAh using the Yun.
Using a Fio & XBee, you can get it down to 330 uA when asleep according to
That is a huge difference and you may be able to go even lower.

I like the Fio idea, but it does mean more work.
I'm sure have a couple of old XBees around here somewhere.
I started using XBees for something a long time ago but switched to using JeeNodes as I found them easier to use.
Let me see if I can get them working.

Power up weather station environmental temperature should be count:

Lead acid
Charge temperature:–20°C to 50°C
Operating temperature:–20°C to 50°C

NiCd, NiMH
Charge temperature:0°C to 45°C
Operating temperature:–20°C to 65°C

Charge temperature:0°C to 45°C
Operating temperature:–20°C to 60°C

Lithium Battery (primary batteries)
Operating temperature:–40°C to 70°C

Lithium Battery is far better than other type at operating temperature.

Lithium batteries are disposable (primary) batteries.

Lithium batteries find application in many long-life, critical devices, such as pacemakers and other implantable electronic medical devices. These devices use specialized lithium-iodide batteries designed to last 15 or more years. But for other, less critical applications such as in toys, the lithium battery may actually outlast the device. In such cases, an expensive lithium battery may not be cost-effective.

Lithium primary batteries account for 28% of all primary battery sales in Japan but only 1% of all battery sales in Switzerland. In the EU only 0.5% of all battery sales

Panasonic’s CR123A lithium batteries have an operating temperature range of -40° to + 70°C.

PDF in Japanese

AR9331 which targeted lowend and lowcost home based router market is lack two key elements for battery operation:

  • Wake on Wireless LAN (WoWLAN)
  • Power Management Circuit (PMC)

To those of interest.

The video I'm posting is 1 hr and 13 minutes. I imagine most people don't have time for the whole thing, but I have not had time to index it. However, I do think within the first 35 minutes the important parts are covered. The important part is catastrophic failure.

The talk is given by Professor Jeff Dahn (Dalhousie University)


Why do Li-ion Batteries die ? and how to improve the situation?