DIY: SOLAR-POWERED LIGHTS 1

'Success? Failure?' Rating: ★★★★☆

Having dabbled with solar-powered lighting for my solar-heated outdoor bath project I was keen to have a crack at something bigger/better. I've decided upon two projects: a miniaturised 'proper' system (i.e. one that uses a small PV panel, a charging controller, a 12-Volt battery and several LED clusters); and an extension of the solar-heated outdoor bath lighting system (hacked up garden lights) for indoor use. First up I wanted a complete system to light our bedroom. This would be the focus of the miniaturised 'proper' system. Here is what I did...

The Solar Panel

I bought a 20-Watt PV panel from Cellpower NZ Ltd for NZ$240 (model number CP-20). It is a mono-crystaline panel (i.e. superior to poly-crystaline and amorphous panels) and is made by a Chinese Company. It outputs a fraction over 1 Amp at a fraction under 20 Volts (i.e. 20 Watts exactly) for a good part of the day. I mounted it on a home-made wooden frame and attached the frame to the gutter board on the almost-due-North facing portion of our roof. The frame is hinged to facilitate changing the angle-of-incline on the panel, but, quite by chance, the flush-mounted angle-of-incline came out to being within a degree or two of our latitude (the optimal angle-of-incline for a fixed panel), so the hinges may not get much use (which is good - we live beside the sea in a very windy place, so I'm keen to minimise the amount of exposure to the elements the frame and panel gets). I also went to some lengths to isolate every place where two different types of metal come in contact with one another (a potential point of corrosion), e.g. steel bolts on aluminium PV frame. All contact points were isolated by using outdoor-grade adhesive tape. I also wrapped the frame in tape where it was in contact with the wood as I'm sure these joints will be full of wet salt in no time - aluminium's good for corrosion resistance, but is it that good? I'm taking no chances.

PV on the roof
PV on the roof
PV up close
PV up close
PV and frame
PV and frame
Frame attachment detail
Frame attachment detail

The Charging Controller

The charging controller is a kit (CirKits SPC2 Solar Power Center Kit) from CirKits.com which I assembled myself. The kit cost US$49.95 + US$10.00 P&P. "A 6 Amp charge controller, manually operated load switch, and low voltage disconnect that is installed between a solar panel, battery, and load. Ideal for remote solar lighting systems. For 12 Volt systems." I found a number of kits (e.g. the Solar Regulator Kit from Oatley Electronics, AU$24.00) and pre-built modules that either did most-but-not-all of what the SPC2 kit did, or all of what the SPC2 kit did, but for a lot more money. Plus the kit was fun to build (Forrest Cook at CirKits was a pleasure to deal with and his kit and instructions were of the highest quality). I housed the completed board in a box from JayCar Electronics.

CirKits SPC2 Solar Power Center Kit
CirKits SPC2 Solar Power Center Kit

The Battery

The battery is a 12-Volt, 18-Amp-hour, Sealed Lead Acid unit from JayCar Electronics. Cost was NZ$54.95. I was assured that the batteries on their shelves were new stock, but I'm not entirely convinced of the validity of this claim (old batteries, especially those that aren't kept charged whilst on the shelf, can become sulfated and thus stuffed). I'm getting mixed messages from my battery (which I'm monitoring closely as the system settles in). One issue that was news to me is the fact that you can't simply attach a Volt meter across the battery terminals and read off the result (well you can, but it won't tell you the real voltage the battery is at). This is because voltage can only be accurately worked out by measuring the voltage drop across a known resistance. So to get a true measurement you need to connect a resistor of a known size across the battery terminals, measure the voltage drop across it and then use Ohm's Law to calculate the battery voltage - easily done in theory, but you'd need a resistor rated to several hundred Watts to do this on a typical 12-Volt battery and most resistors you purchase from an electronics store are rated at less than one Watt. Connecting a Volt meter across the battery terminals does give an indicative result - you're effectively using the internal resistance of the battery to do the measurement (but this internal resistance is unknown and changes with battery temperature and state of charge) - but you need to be aware that the true battery voltage is likely to be lower than the reading (see All About Circuits: Battery ratings for more on this).

Update: After upgrading my battery to a Sonnenschein Dryfit A500 Series Cyclic Battery (A512C/28,0 G6 - 12V/28,0Ah C20) my system now performs perfectly. Lesson learnt: don't buy a cheap battery, nor a battery that has been sitting unattended on a shelf for an unknown amount of time.

The LEDs

The final pieces in the puzzle were the LED clusters. For these I again went for a kit, this time from Altenergy. Tom Woods at Altenergy is an extremely helpful guy and his kits and instructions are of the highest quality. I bought enough bits to make up six 'half loaded' LED clusters (including a single power supply sub-kit) and assembled roughly half of the LEDs into three lighting units for use in this project (the rest I'm saving for a future project). I installed 24 LEDs and the power supply onto one board to act as my 'general' light and then built two further boards with 16 LEDs apiece for use as individual reading lamps. I mounted the two reading lamps inside a pair of translucent boxes (such that the light is slightly diffused by the box) and then mounted each box on a home-built wood conduit strip (with a switch at one end and a connector block at the other). The joint that connects the LED box to the wood strip swivels so the light direction can be adjusted. The resulting reading lamps are perfect. I'm less happy with the 24-LED, 'general' light - I had underestimated just how 'directional' LEDs are. Using undiffused LEDs for anything other than spot lighting is probably not a good idea - it's like trying to light your room with a car headlight. Some kind of diffuser would undoubtedly help, but might require more than 24 LEDs to work well (Tom's boards hold up to 40 LEDs each). Having said all this, the 'general' light is making a reasonable stab at the task I've bent it to. One small surprise was just how much heat these LED clusters put out - I'm glad I didn't seal the units into the boxes!

Reading light switch (back)
Reading light switch (back)
Reading light connector block (back)
Reading light connector block (back)
Reading light connector block (front)
Reading light connector block (front)
Reading light LED box
Reading light LED box
Reading light LED box
Reading light LED box
Reading light installed
Reading light installed
General room light (top)
General room light (top)
General room light (front)
General room light (front)
The full set-up (day)
The full set-up (day)
The full set-up (night)
The full set-up (night)

The Wiring

One final point worth mentioning is what I did for wiring. Again this came from JayCar Electronics. Most of the wiring is two-strand, 17AWG copper wire (rated to 15 Amps by the manufacturer, but to only 2.9 Amps (for power transmission) here). The reading lights are internally wired and connected to the 'general' light with thick copper speaker wire. The speaker wire was adequately rated for it's job and much more flexible and easy to work with than the heavy-duty stuff used in the rest of the system - 17AWG (or greater) cable is necessary, but challenging to work with. After the PV photos, above, were taken I wrapped the cabling outside the house in foil tape to give it some UV protection (as I wasn't sure whether or not the cable sheathing had any in-built UV-resistance). I also wired a cigarette lighter socket into the Load circuit (nestling in the light shade of our now-redundant bed light in the photos above) so I can use the excess output of the solar panel to power anything with a cigarette lighter plug (endless free cellphone charging and laptop-in-bed use here I come!).

Success? Failure?

★★★★☆

Pros: Robust and reliable system (once the battery was upgraded); great source of 'focused' lighting.

Cons: Not great for general room lighting.

Enhancement Suggestions: Using a 'collimator' in front of the general room light would greatly improve the even dispersion of the light produced.

Update: Tom at Altenergy posted a link to this page in his 'blog. Cheers, Tom!

Update: I received the following e-mail from a reader requesting more information about the PV frame:

"Hello my name is David O'Neill and I am building a Renewable Energy Trailer for my University studies. I was looking at your website and am very interested in the pv panels. What I am trying to do at the moment is design a frame for the panels to be placed on. These panels then need to be able to adjust to face a certain angle. I am having difficulties with the frame itself? What material to use? etc. These panel will be placed on the top of the trailer so will need to withstand up to 60mph. If you could give me any information or help that would be greatly appricated. Thank you. David O'Neill"

My reply was as follows (in case it is of use to anyone else):

To build your own frame for your PV(s) you've got a couple of choices of materials - the one you choose will depend on your budget and building skills.

You'll need to choose between:

For my PV frame (which I'm guessing you saw on [link to this page]) I used wood. The wood was treated for outdoor use before I bought it. I made the frame such that the PV was snugly held between two horizontal beams. The horizontal beams were bolted to a pair of vertical beams that support the PV from below. The vertical beams line up with holes on the back of my PV and 4 Marine Grade Stainless Steel bolts secure the PV to the vertical beams. Every place where the PV is touched by another material ( i.e. the wooden frame and the metal bolts that hold it to the frame) has been wrapped in outdoor grade plastic sticky tape - I did this to electrically isolated the bolts from the PV and to ensure I don't get any salt working its way down between the PV and the wood (we live beside the sea in a very wind-swept area). The wood frame was then attached to a wooden base using a line of hinges along its bottom edge so that the frame can be tilted relative to the wooden base. As it turned out I didn't need to tilt my frame (our roof pitch was sufficient), so I never got as far as coming up with a mechanism to secure the frame when it was tilted up away from the base (mine just lies flush and is latched at the top to stop it flipping up in the wind). All the wood in the frame and base has been protected from the elements with high-gloss, white, outdoor paint. Over the Christmas period gusts of wind in our area reached 165km/hr (102mph) and my PV frame was totally secure throughout.