Minimum LiPo mAH capacity needed for solar powered project

Other Hardware Home Automation
1

I'm thinking of powering a wireless outdoor remote weather sensor node connected to my Arduino-based weather station via solar panels, and am wondering what capacity LiPo battery I'll need to assure that it stays on all night on even the longest and coldest nights.

I haven't yet decided whether the remote sensor node, which will likely have a DHT22 for humidity & temperature, an LDR for light levels, and perhaps an MQ-2 gas sensor, will be based on a small-form Arduino such as the Pro Mini or Nano, with an RF24+ handling the communication with the base station, or on a NodeMCU module that uses an ESP8266 for both the microcontroller and WiFi (the base station already has both WiFi & RF24+, so either would work).

This means, of course, that its power requirements will vary depending on which MCU/Wireless solution I use, and whether it'll have an MQ-2 sensor, which uses a fair amount of power to heat up the element.

I'm obviously new to solar/LiPo, but from what I've read, it appears that I'll need the following to make this work:

Solar Cell (I've ordered a cheap 6V/1A panel on eBay)
-->Solar LiPo charger circuit
-->LiPo battery (3.7V, mAH tbd)
-->3.7V-5V DC-DC power boost circuit
-->Sensor node (5V input)

Lots of components there! Did I get this right?

Anyway, I'm wondering, what's the lowest capacity LiPo battery I should consider, for the lowest power-consuming version of the remote sensor node (which I assume is the Pro Mini + RF24+ w/o MQ-2 sensor), and for the highest power-consuming version of it (which I assume is the NodeMCU w/MQ-2 sensor)?

FWIW I live in the NE USA, to give an idea of what kind of sunlight we get here, and how cold it gets. In the dead of winter days can be as short as ~10 hrs long and completely overcast, so I'm guessing that the battery, which I assume loses efficiency when it's cold, would have to have enough power to last several days when solar output is this low and it's this cold, as it trickle charges on overcast days, until the next sunny day comes along.

I'm just looking for ballpark estimates, nothing too detailed. Better yet, if someone could direct me to a good site that helps you estimate a project's power requirements and what kind of battery and solar panels you'll need to keep them running 24/7, and the various technologies and boards you need to make this work, that would be really helpful.

2

You need to measure the average current consumed by your entire project in mA and multiply that by the number of hours of darkness, to get mAh. Double that number for safety and buy a battery with that capacity (C).

LiPo batteries are a poor choice. They MUST be charged by a LiPo battery charger, and MUST be protected against over-discharge by another circuit, or they will be very quickly destroyed. Many LiPo battery chargers do not work properly when the battery is simultaneously being used to power a circuit.

NiMH batteries are much safer, much more tolerant of abuse and can be trickle charged indefinitely, at C/20 mA (where C is the battery capacity in mAh).

For advice on constructing low power Arduino projects, see this excellent post.

3

Also mind the solar cells are mostly rated for a direct sun exposure (no clouds), otherwise you may get a fraction of the power claimed (ie. the current).

4

Thanks! I've never done a solar-powered project so LiPo was my initial choice, but it also occurred to me that one of its main advantages, light weight, didn't really benefit a stationary device.

Are most portable cell phone charger/batteries based on NiMH? Since they already put out 5V, there's no need for output conversion, although I assume I'd still need some sort of board to connect the panel to these. Plus, they're pretty cheap. I have several lying around that I could use.

5

Normally, when building a Solar powered project, the solar panel is the last thing you buy, as it has to be sized based on total project power consumption, latitude , elevation and how reliable you want the project to be.
5 consecutive days of cloudy weather is a good starting point for reliability.
Nimh cells are the easiest to use , as they are the most forgiving in respect of overcharge and undercharge.

6

The panel I bought was just over $1 on eBay, almost an impulse buy. At this stage I'm just planning on playing around with components to get a feel for how this works, but I was hoping for some back of the envelope ballpark idea of what kind of battery capacity I'd need. Once I get the hang of it, I'll start to get more serious about capacity needs and all that. Given that I might work with LiPo, though, I realize that I have to be careful with what I connect to it, so nothing explodes or catches fire.

7

Its impossible to answer you question without knowing what the battery has to power, and what the power consumption rate is.
If you are trying to power some kind of weather node then you need to measure its power consumption.

8

I was hoping for some back of the envelope ballpark idea of what kind of battery capacity I'd need.

See the first sentence of reply #1.

9

I have built a half dozen remote sensors for temperature monitoring, they report once every 2-5 minutes depending on which one and I power them all right off of 3 AA batteries. I have not bothered with solar yet as they are lasting upwards of 100 days right now. One other thing you should figure in is how often they will take the measurements and use the radio. If you do many optimizations to your board, the radio will quickly become the most power intensive part(maybe that air sensor as well, I have no experience there)

I am starting to think about going solar on some of harder to get to sensors as well. I have a few of the same cheap panels you are likely talking about coming. What I have in mind is a 6v lead acid battery in the 4-9AH range. That is over kill for my project, but as I understand it the lead acid batteries are the most forgiving with Solar and the price point is not that bad. You will be looking at something like $10-$15 on ebay for the battery.

10

3xAA/AAA NiMh is increasingly looking like my best option, with a small solar panel for recharging and a DC-DC 5V voltage booster to supply 5V to the board. Although, I might play around with LiPo as I've ordered a small 1200mAh LiPo and solar-LiPo charging circuit. In fact I'm now thinking of designing an even simpler remote sensor node around an ATTiny85 board, which is even smaller, uses even less power, works with the RF24+ and should be more than enough for a couple of sensors. Either of these power solutions should probably be enough for this.

11

Either of these power solutions should probably be enough for this.

So, you haven't measured the actual power consumption, but remain hopeful.

12

jremington:
So, you haven't measured the actual power consumption, but remain hopeful.

Now was that really necessary? It's an enthusiast's forum, not an official meeting of the IEEE or the Samsung battery group, and no one here is building the next Space Shuttle. I realize you've been here forever, but, sheesh, try to keep the gratuitous sarcasm to yourself.

13

Unfortunately you dont seem to want any help.
Jremington is giving you the exact information you need to do to get any kind of Solar powered project running properly, but you seem to ignore the advice given.
Ill try again.
To build a Solar powered project the FIRST thing you need to know is the daily power consumption of the project.
Then you work backwards.
Knowing the power consumption, you can then size the battery, and having sized the battery, you can then work out the Solar panel size.
You dont get the Solar panel and battery first.

14

I disagree. Except for the air sensor, it is easy to get the consumption down so much, that almost all​ batteries are overpowered.
My suggestion is to drop any voltage conversion and make the project run on the voltage range of the chosen battery.

15

it is easy to get the consumption down so much, that almost all​ batteries are overpowered.

Very useful tidbit of advice, thanks.

16

mauried:
Unfortunately you dont seem to want any help.
Jremington is giving you the exact information you need to do to get any kind of Solar powered project running properly, but you seem to ignore the advice given.
Ill try again.
To build a Solar powered project the FIRST thing you need to know is the daily power consumption of the project.
Then you work backwards.
Knowing the power consumption, you can then size the battery, and having sized the battery, you can then work out the Solar panel size.
You dont get the Solar panel and battery first.

Actually, I do, it's just that when someone feels the need to be sarcastic or condescending instead of just explaining themselves, I pretty much write off anything else they have to say at that point, even if it's technically superlative. There are enough tech experts in the world, here and elsewhere, that I don't need to bother with A-R P-A types clearly overcompensating for whatever who don't play well with others. Not so much here, but I've more than enough online forum experience to know the type, and ignore them.

Btw, any real engineer can give back of the envelope, ballpark answers to general questions. Obviously you don't base a circuit on that, but as a starting point, you don't need to get too fancy.