How to choose a battery for your DIY project in a frost?

All kind time of day!

The challenge is developing a research device (within the thesis), which is characterized by the following parameters:
1) Composition of the device - a microcontroller (probably Arduino) sensor, receiver (in principle come ZigBee, but in fact requires a minimum of GSM GPRS modem).
2) Voltage 5V, quiescent current of 4 mA (in pause) to 1000 mA (peak in the case of GSM modems), the average power consumption with reserve in this case will be about 40 mAh or 1000 mAh per day.
3) working Conditions - thermobox in the woods, possibility of connection to the mains is missing. It is planned to use solar panels and store energy during the day. Possible care to sleep at night (and day) with shutdown transceiver with insufficient stored energy for daylight hours. Period of operation - year round, minimum 2 years without physical maintenance.

In fact, the whole structure of the device, its layout and finish, the case was for small - to choose a battery for energy storage with solar panels.
Based on energy consumption, plus reserve for cloudy days, I would like to have a battery capacity of not less than 6000mAh (more is better), and given a peak current of at least 1.5 A.
Actual also the possibility of trickle charging in cloudy watch not to lose the opportunity charge.

At the beginning of the development somehow did not think on this subject and wanted to put the usual powerbank with li-ion batteries, but then I thought, because they are frost and heat tolerant is absolutely not (and if the heat still the device can be hidden in the shadow, from the cold no escape). Then came the idea to use superministry, which in principle is characteristics fit easily, but considering how much is required to be set and comparing with the financing fell into a perfect down.
Now consider the options up to this, to bury a separate hermoupoli with a li-ion battery under the earth, where some of the conditions are adequate during the cold weather, and controller with sensor and solar panels to make the surface...
I will time the community Council in this difficult matter :)

PS: it would be ideal to put personal portable RTGs... But engineering has not yet reached safe individual radioisotope power sources :)
June 10th 19 at 15:53
6 answers
June 10th 19 at 15:55
Conventional lead - acid battery with a decent reserve capacity will be just right.

and if the heat still the device can be hidden in the shadow, from the cold no escape
Well, why so sad? Are you in the far North is the tundra going? If you are in the permafrost - take a shovel, dig a hole a few meters deep, throw the battery and added dropwise.
In the end he works at a constant temperature of +3градуса.
June 10th 19 at 15:57
From Panasonic has a series Сadnica GT -40 to +60
June 10th 19 at 15:59
Well here is that LiFePO4 will fit. Not sure how well his charge in the cold, but at least the current will give

By the way, modern NiMh rechargeable batteries of the LSD type (low self discharge) will also fit, just will be harder and more of them will need
Here, LiFePO4 is the most difficult in this regard ))
At low temperatures the ability of the charge is cut off on the controller, if charge - in the trash. - earl.Weissnat commented on June 10th 19 at 16:02
June 10th 19 at 16:01
had a similar job in the company - through trial and error came to the lead-acid-acid. oak as boots, excellent vesch for the backend to solar panels. only to defuse much is not necessary, the price is not much put with a reserve of capacity and the socket pobolshe :).
by the way did not work normally start dc/dc with mppt, so I put the simplest dc/dc c separate input uvlo and a resistor divider, just ask static operating point.

himh brings big dance with a tambourine when charging, like a piggy Bank with solar panel vaasche not an option, although analeptiki has bought :)

lithium sucks is charged by cons, EEMB presented low-temperature accum LC series , but they even at -10 is guaranteed 70% of the dial capacity at a current of 0.2 C.
by the way did not work normally start dc/dc with mppt
What's the problem? - earl.Weissnat commented on June 10th 19 at 16:04
when it exceeds a certain current (less than the maximum allowable) generation break down and the controller goes into a complete blackout, which he leaves with baaalshoy scratch.
I thought there would be a current limit or something, but no such trinkets. - earl.Weissnat commented on June 10th 19 at 16:07
Well cheap mppt far from the stated characteristics, and hanging with a large current, it is a common thing. While hands will not reboot figures released.
Here it is necessary normal to buy, or take with a good store on a current.
Especially in winter problem - there's often current more than face value sometimes. - Sandra_Kautzer42 commented on June 10th 19 at 16:10
mikroshemka LTC4142 it's cheap or not ?? I do not modules you take :)
the solar panel according to the calculations in the best case will current, previousy its parameters.
and she's a bitch buggy. - earl.Weissnat commented on June 10th 19 at 16:13
I do not know, used willing.

the solar panel according to the calculations in the best case will current, previousy its parameters
In the winter I am going to give, and much more. - Sandra_Kautzer42 commented on June 10th 19 at 16:16
I have now a winter sky all the clouds.
In the summer on a Sunny day for a test shot current-voltage characteristic, according to calculations issued 1.5 W in the mppt.
stated - 3.5 W, I don't know how it should fry the sun to panel went 3.5 W. so 3.5 W took the maximum, according to calculations this was more than enough.
5a2fa9fdb5a3c902864989.gif - earl.Weissnat commented on June 10th 19 at 16:19
All depends on the situation and the quality of the solar panel.
If the bar is not unknown nonam, it needs to be tested.
Usually test them at a certain temperature and light - this is a stated feature.

The actual current that will generate a panel depends on two major factors -
1)the Amount of solar radiation incident on the panel at the moment.
The more claim 1 and claim 2 is smaller the greater the current.

If you increase the illumination panel three times - the current will grow three times.
This can be achieved through a system of mirrors.
Or pure white snow.
Well, the panel temperature in winter is quite low - Sandra_Kautzer42 commented on June 10th 19 at 16:22
quite agree with your theories - you're smart, it shows.
just do not understand what it is now ??
my existing panel will not give no more than 3.5 W, the question I have in this. had managed to plant in every any pack-a-punch is meaningless, because the party will be in a few hundred pieces at least. - earl.Weissnat commented on June 10th 19 at 16:25
I mean that in winter always knocks the controller from excessive current flow, if not to take them with a good supply of current. - Sandra_Kautzer42 commented on June 10th 19 at 16:28
OK , will take note.
poke your finger into the sky and not static if they break ??
in a good snowstorm with frosty dry weather static electricity on the metal items well so typed. almost to the foot of the lightning jumping can. - earl.Weissnat commented on June 10th 19 at 16:31
June 10th 19 at 16:03
If the device is stationary in the woods, pricopie it.
At least the battery.
So the battery will be warmer, even in permafrost areas.
And Yes, lead maintenance-free battery fit. The selection is large.
June 10th 19 at 16:05
tell us the results? What batteries choose? how to work out the system?

Alternatively, you can still look towards Li-titanate

Find more questions by tags Do it yourselfElectronicsArduino