Solar help: Keep a 12V battery charged?

December 8, 2008 12:28 PM

I have some remote monitoring equipment out in the woods that runs on a single 12 volt deep cycle car battery. How do I keep my 12 volt battery charged using a solar panel?

I don't know anything about solar panels and such devices. The battery keeps the equipment running for about 15 days (once the voltage drops below about 11, the equipment stops running). Unfortunately, the equipment is in the middle of a forest, with the top of the tree canopy at about 75 feet. In the winter time, the leaves have fallen, so there's a bit more light reaching the ground. The equipment is on the ground. I may have some problem with branches from above breaking, then falling and striking the panel.
What exactly do I need to keep this battery charged, but not over charged, and not allow the battery to discharge a night while the panel is not getting any light?
Thanks for any thoughts, and keep if simple since I'm a complete beginning with solar and electricity issues.

posted by jldindc to technology (12 comments total)

Google shows a bunch of alternatives using "solar battery charger" as the search term.

Unless you just want to homebrew, I suggest calling one of the many vendors, describing the installation and load requirements and solving the problem with a credit card. You could see the problem solved in a day or two.

The simplicity of this question generates a lot of subtle answers. There are thermal issues, charging issues (depending on the nature of the battery chemistry), load issues, panel sizing issues... How long do you need it to work? What if you get a cloudy month? How often can you visit it? How much expensive gear do you want to leave out in the woods? What temperature range are you dealing with?

I've seen simple units with just a panel, blocking diode, and the battery. They aren't well engineered, but they are homebrew. If I were doing it, I'd use the resources of the charger vendors. Fastest way to a solution.

Good luck!
posted by FauxScot at 1:02 PM on December 8, 2008

Another good google phrase would be "charge controller". You want a charge controller that sits between the power source (solar panel) and battery, and regulates the current flow to the battery based on its state of charge. It's not too expensive for a basic one, just look for something with good reviews.
posted by knave at 2:16 PM on December 8, 2008

You need what's called a charge controller.

http://store.solar-electric.com/sg-4.html

This one will handle 4.5A. Be sure to get one that's rated to handle the current your equipment draws.
posted by lalas at 2:18 PM on December 8, 2008

It sounds like your biggest challenge will come in the spring, when the trees leaf out and there's no sun reaching the forest floor. It's nearly impossible to place your solar panel above the tree canopy, as it would require very long utility pole or similar — you can't mount the solar panel to a tree since there are leaves on the limbs above.
posted by exphysicist345 at 3:47 PM on December 8, 2008

Just some rough calculations: assume your battery is rated at 100 amp-hours. 15 days * 24 hours = 360 hours. Divide this into 100 amp-hours and it means you are drawing about 0.28 amps. 0.28 amps * 12 volts = 3.33 watts. That is a substantial amount of power for an electronic monitoring device.

Instead of the complications of a solar charger, you might look into redesigning your equipment to use less power. For example, using a switching voltage regulator instead of a linear regulator for the electronics. And using low power electronics that can be shut down when not taking a sample. For example if you need one sample per second and taking a sample requires 100 milliseconds, you only need to draw power for 10% of the time. This effectively increases you battery time by a factor of 10 to 150 days.

Another alternative is to supplement your single battery with 2 or 3 more batteries connected in parallel. Four batteries would last 60 days and are a lot cheaper and simpler than a solar panel -- although if you have to backpack these in, you better have a few friends.
posted by JackFlash at 5:27 PM on December 8, 2008

Cheap and cheerful solution here. Hook up a 5watt or 10watt solar panel directly to the battery (in paralell).
Assuming a normal-ish car battery, you are sucking out only a little bit of juice, 5watts will probably cover it (where I am, you get an average of 6 hours a day equivalent full sunlight, so 30watts from a 5watt panel each day).
Ideally, swing it up to the top of a tree or in a clearing to get plenty of sun.
The controllers mentioned above will prolong the life of your battery if you are pumping in more energy than the battery can take, with a low wattage panel this is unlikely, so they offer no benefit. If you could visit once a season, you could probably have indefinite power if you took in a fully charged battery and replaced the existing one, figuring the small panel would largely 'top-up' the on-site battery in between times. Take the older battery home and recharge it for next season.
In my real world experience, you will get 5yrs+ per battery if you treat it this way.
Note, my real world experience is in occasionally used batteries for camping, but I reckon this is something that is easy to over-engineer, and if two batteries and a small panel work for a few years - good enough
posted by bystander at 5:47 AM on December 9, 2008

you get an average of 6 hours a day equivalent full sunlight, so 30watts from a 5watt panel each day ...

That is not correct. Watts are a rate, not a quantity. You can't multiply watts by hours to get more watts.

A 5-watt panel provides 5 watts maximum under ideal conditions -- high noon and the perfect angle. Under any other conditions it provides less. Let's say 2.5 watts average during six daylight hours. That is one fourth of the day so the 24-hour average is only 0.625 watts -- all the while the equipment is drawing 3.33 watts by my rough calculations above. That is a losing strategy. In a real world application -- cloudy days, trees, seasonal variation of sun angle, only 6 hours of sunlight, etc. you would probably need a panel that puts out at least 10 to 20 times your load -- that is a 30 to 60 watt panel. You are probably going to spend $500 to $1000 by the time you get it installed with charger. It would be much, much cheaper and simpler to just add more batteries if a six-month service interval is tolerable.
posted by JackFlash at 9:18 AM on December 9, 2008

Thanks for all the good comments. After reading them I'll have to:
- go as cheap as possible. The location is likely prone to vandalism. (The equipment is already housed inside a very strong steel contruction/contractor box for this reason.) The other alternative would be to mount the panel so high on a tree as to make it quite hard to steal or break. Then again, this might pose a fun challenge for someone who is intent on solving challenges.
- a charge controller sounds like something I should get.
- unfortunately I can't redesign the equipment using lower power electronics. It's a real-time river gage, and there aren't too many that would work in the situation in which this is placed.
- I actually might try to find some sort of measuring device that tells me how much sun I can get at that particular place, then use those figures to tell me how big of a panel to get.
- I'll have to design something a bit more robust - able to withstand branches falling, etc.
posted by jldindc at 10:17 AM on December 9, 2008

For a river gauge one sample every minute or even 15 minutes should be sufficient. There is no point in the electronics sucking 3 W continuously. You might look into a timer that would shut the electronics down between samples. Your battery alone could last many months that way.
posted by JackFlash at 12:27 PM on December 9, 2008

Thanks JackFlash. That's a very good suggestion. Right now we have the gage giving 15 minute readings. When it takes a reading, it pumps for about 10 seconds. I'll have to figure out if turning off power to the device (and it's cellular modem) is good for it or not. When it's idle, it's not pumping - and I'm assuming using less juice. However, when the device loses power, upon regain of power it pumps for 6 minutes straight (some sort of default setting). The question will be whether that 6 minutes of pumping uses more power than if I were to leave the power on for the 30 minutes between readings. I'm guessing I need some sort of power meter (is that called a multimeter?) to figure out how much juice the pump is using when idle as well as when it's pumping. You think I have that right?

Unfortunately, I inherited the management of this gage, so I'm learning a bit on the mechanics of how it all works.

At the very least, I suppose I could put some sort of timer on the modem so that at least it is not using power when the whole thing is sitting idle.
Thanks for the good comments.
posted by jldindc at 6:56 AM on December 10, 2008

If you're feeling creative, you could build a turbine to power the equipment via the river itself, no?
posted by knave at 11:55 AM on December 10, 2008

Knave: That would be magic if I could. The river at this point is very slow moving, but may just be something to try. My only problem is to place the turbine at a point low enough so that during the summer the low water levels will not leave the turbine dry, but also not allow someone to simply pull it up and take it, or allow floating objects (of which there are many) to rip it away in fast water.
I hadn't though of that idea yet. Thanks !
posted by jldindc at 1:28 PM on December 10, 2008