High Powered LED Solar Lighting System

While I love making proof of concept solar projects, or solar projects that I might use if the world comes to an end, I prefer to make solar projects that are useful in my day to day life.  (I love my solar FM radio, but I've never had the need to use it in an emergency situation)

This past summer my father has been complaining about installing a lighting system for the back of his yard.  The big issue being that he hates running wire 100 yards just to get three lights to turn on at night.  Plus, being the over achiever that he is, he decided to bury the wires so that he wouldn't run over them with the lawn mower.  This is when I stepped in.

I decided to make a high powered solar lighting system so no wires would ever need to be run to the back of the yard.  Ever.  

Oh, and I'm not talking about a few little wussy LEDs inside of a jar, I'm talking some high powered 3 watt LEDs inside real metal lighting enclosures.  Ones that will probably attract every moth for ten miles.  (And in defense of LEDs in jars, I happen to have 12 of those on the deck of my apartment.  A couple slowly pulsing in and out.)

So for your enjoyment, here is a quick, easy, and inexpensive guide to making a high powered solar lighting system.

(If you like my project, vote for it in the Off The Grid Contest here at instructables, I'm the first entry.  Seriously.  A vote for me will bring you instant karma and a warm fuzzy feeling.  WARNING:  Warm fuzzy feeling may in fact be low levels of radiation.  Seek medical help.)

Step 1: What You'll Need

To be honest, we're pretty much just hacking a cheap lighting system into a solar lighting system.  Most of the parts we will need come preassembled.  We're mostly just wiring things up.  The nice part about that is the fact that we only have to do a little soldering.

Parts:
Cheap Lighting System (With 2 or more metal "heads")
Heavy Duty Outdoors Wire
High Powered LEDs (1 Watt or 3 Watt depending on your needs)
High Powered LED Driver
Heat Sinks
18V or better Solar Panel
12V Solar Charge Controller
12V Battery
Wire
Plastic Container
Washers
Cable Clamp
Terminal Strip

Dark Detecting Circuit:
PNP Transistor (I used a TIP42)
A Prototyping Board
10,000 Ohm Resistor
1N4001 Diode

Tools:
Soldering Iron
Helping Hand
Screw Drivers
Silicon Calk 
Wire Stripper and Cutter

Cost: $75 - 100
Time: 2 - 3 Hours

I bought all the electronics off of eBay on the cheap.  The hardware parts I bought from a local hardware store.  The total cost of this project was in the range of $75 - 100.  Not overly expensive, but much cheaper than any store bought high powered solar lighting system.  Plus a whole lot more powerful.  I also bought a more expensive driver than I needed in case I wanted extra functions, you could easly shave $10 off the project by getting a cheaper one.

The majority of your time will be spent trying to figure out how to run wires into and out of things.  The amount of real "making" work is rather small. 

Comments

[Bubbler]

I have been wracking my brain to work out a system like this. I need some bright light in the driveway and in the back yard. I can easily avoid the weather in both areas by placing the box of tricks under the roof, and the lights on either end of my boat shed, with the solar panel on top of the roof. However, the piddly little garden lights would be useless for my situation. I wonder now if I can become an electrician. At least a hobbyist type. Thanks for this one, it is great.

[optichead]

very interesting I have been doing optics since 92 and wonder if we could talk sometime my name is Mark Zimmerman and I go by the name optichead I need some fine tuning for a new garden system Im at [e-mail]

[bhvm]

Very well done mate.
How is the thing holding up?
I'm worried about the heatsinks on the LEDs (They seem tiny), Are they adequate? or Could we use some interface that enables whole enclosure as an heatsink?

[JoshuaZimmerman]

So far so good. Nothing's blown up yet.

[Ugifer]

This would be a great situation to use an ATtiny 85 - small and cheap, it could measure the voltage from the solar panel to do your dark detecting and also control any fancy functions, like fading or blinking. You could also set it up to give full light for the first few hours of darkness (for show) and then just a little light during the rest of the night to find your way around!

I love the idea of massive, powerful things controlled by little tiny controllers running your home-written firmware, and this is the perfect setup for it - you have a PWM control input, you have +5V and the solar-cell voltage available. All you need is an ATTiny and a breakout board!

[JoshuaZimmerman]

Oooooo thats an AWESOME idea. I just happen to have a couple of hundred ATtiny85 chips around. I'll have to give it a whirl.

Joshua

[mathieulj]

Another option for the dark detection would be simply to use the PWM pin of your controller with a simple pull down resistor and a low power transistor. Depending on the voltage limits of your controller's PWM pin, you could use essentially the same circuit you did here except the transistor would be feeding the PWM pin instead of the LEDs.

As a side note, isnt the 2N3904 you mention a NPN transistor? The way your circuit works, a PNP transistor is needed (as you correctly said so in your text and circuit drawing). It might be good to change the reference to the complementary 2N3906 (PNP) or to simply state the it is a NPN. That way, the less knowledgeable that dont need high power are not misled into using a transistor that would not work in this application (NPN would always be on).

[JoshuaZimmerman]

Doh! You are correct. I did reference the wrong transistor.

Good call with the PWM pin. I'll give that a try.

Thanks!

[manuka]

Joshua: OK on your comments, & I certainly appreciate the regulator's function.  With only a dusk-midnight need you may well want to include a simple timer too. How about  using the likes of a PICAXE micro =>   www.picaxe.orcon.net.nz   ?  These are a dream for pre-teens!

Mmm- given your freezing Wisconsin winters an additional aspect  may relate to lead acid batteries poor performance when cold/chilled. Best wrap up that poor SLA  to keep it cosy!

Happy to point out this SLA "Gel Cell" source,especially as you're a teacher with (no doubt) a cash strapped budget. Being sealed they present no fume or acid spill concerns of course. Sigh-if only they'd been available when I was younger. Back in my teens I recall rescuing a regular flooded car battery & building it into bench power supply, with resulting sulfuric acid "migration"-argh!

Discarded SLA batteries are usually just sent for scrap metal, & (even with the present value of lead) only bring in about US$1-$2 each. This may be hardly worthwhile for the security firm to bother with, so you could perhaps even make some science resource "pocket money" by offering to do it for them. Play up your educational needs of course.   Hence rescue quite a swag, use the good ones (or pass to capable  students/friends) & take the really weary ones in for scrap $$ yourself. It sure beats collecting aluminum drink cans.

Health & Safety ALERT: 12V 7Ah SLAs are VERY energetic & start a fire if shorted, and even jump start a car, or power an e-bike/scooter etc. I once gave a 12V 7Ah SLA to a 12 yo. who'd looked a budding bright spark. Word filtered back to me however that he'd just used it to "make a bomb" by shorting the terminals with a heavy duty cable, to then relish the resulting acrid smoke & fumes as the battery destructed... I now only pass out the smaller 12V 2Ah types when in doubt about the end user's motives.

Extra:  A simple charge confirmation indicator can be VERY worthwhile too -at the least it confirms connections haven't come adrift or terminals corroded. I've done a bit in the past with FLEDs ( Flashng LEDs) for this in fact (See Instructable => www.instructables.com/id/Single-LED-ammeter-FLED-based/ )  & the attached relates to one a student of mine recently used for a similar (but switched) LED lighting project.

Stan. - semi-retired  "hands on" educator ( Wellington- NZ)

[manuka]

Be darned - I've got this far so may as well also tweak your initial day/night switch! It's possible that a weak but firm pull down may be needed as well in your circuit- refer picture.

[mathieulj]

The addition of a pull down resistor on the transistor's base shouldn't hinder the function of this circuit however it may not be necessary. PV cells loose their voltage during darkness, that is why you need a blocking diode to stop your batteries from discharging through the panel, this lower voltage should in theory be enough to power the transistor.

[manuka]

By the looks of it I'd guess your solar PV is rated at 10 Watts? Thus with an average of 4 hours bright sun daily the battery will receive approx. half a charge each day. However a 3W LED could be on for 10-12 hours each night, & more in winter. A few days of gloom would mean the battery near drained & it's life possibly greatly reduced!!  With such a lighting load you may hence need a larger PV, or more batteries...

Handy hint: Check a local security firm for free SLA ( Sealed Lead Acid) "discards". Here in NZ  such contacts supply more than I can handle - they're ideal for my educational work. Such 12V 7Ah batteries are a global standby standard & firms often renew them at critical sites every couple of years to ensure reliability.. In less demanding applications they'll usually still be good for  5-8 years if not allowed to go flat or discharged too deeply.

FWIW even 20W PVs,c/w inbuilt controller, are often now under US$50 in the right places.

[JoshuaZimmerman]

Oh, and you don't have to worry about the battery every completely draining! Thats what the solar charge controller is for. When the battery hits a certain "low" voltage the charge controller cuts off the "load." I always recommend people spend the $10-20 for one in order to save their batteries from excess wear and tear.

[JoshuaZimmerman]

I completely agree. The panel is a 10 Watt panel, and I have a second panel around as well if my father wants more light.

He mainly wants lighting from dusk until midnight, after that he's asleep and doesn't care how things look.

I'll look into SLA discards locally, thanks for the hint. That would really bring down the cost of the entire setup.

[Hvysteel]

Thank you! I did some landscaping last year and have wanted to try and make a good and reliable solar light set. I had the idea, just not the specifics.

[JoshuaZimmerman]

I promised my Dad I would make this for him last summer. I even had all the parts sitting in my basement for the past year. I'm glad I waited because I learned quite a bit during that time which helped me with this project. I could have done it, but not as nicely.

If you go with some cheap 1W LEDs and a low cost driver and charge controller, the project's price tag is between $50-100. The biggest issues are the cost of the solar panel and the battery. I got my battery on sale for $10, it's normally $20ish. Same with the solar panel, a random online sale last year.

It's not an overly difficult project, but it does take some time to find and order all the parts. I made a third "head" the other evening and it took all of 10 minutes to do and wire into my setup.

[cpotoso]

Very nice, I liked how you used the solar panel itself as a light sensor.

[JoshuaZimmerman]

It a very easy addition, and I hope the parts come today. I'll add that bit into the guide. All it requires is soldering a transistor, a resistor, and three wires together. While not the most efficient, it is cheap and easy.

I've been using the same design when I make small scale garden lights or projects. I've even used it in past instructables, though I can not take credit for the idea. I believe I got it from evilmadscience.com a couple of year ago.

[nerd7473]

this is awesome!!! hey Joshua how long did it take to build this?

[JoshuaZimmerman]

From start to finish about 2 and a half hours. I was also taking photos at the same time, which added a bit into the mix.

A big chunk of my time was figuing out how to make all my wires "fit" into various things. Such as running the wire up into the metal heads, or connecting to the battery. The "hard" work was quick and easy.

If I was to do this a second time I'm guessing I could do 3 heads in around an hour and a half to two hours.