Sunday, March 4, 2012

A Solar Cell Backup Power System - 4

A Practical Starter Project 

In part 2, I presented a simple calculation that showed in order to run my ham radio emergency communications (EMCOMM) would require 11 AH (amps*hours) of battery charge.  In WH, just multiply that times the battery voltage of 12, and we'll call that 132 WH.  There are maps available that tell you how much sunlight to plan for per day, and in our area, it's about 5 hours per day.  That says if I charge a battery at 132WH divided by 5 hours (H), I need a constant output of 26.4 Watts. 

I started keeping an eye out for options along these lines and found a deal on a pair of 20W panels that would make a good battery charger.  I found these at a ham radio swapmeet, not at a store, but Amazon sells the same panels at about $10 each more than I paid.  Two of these in parallel will be capable of 40W, and should be able to charge a battery with ease in the (average) 5 hours I expect.  They should really charge that much in under 4 hours (4*40 is 160 WH).  In fact, they are capable of more, and as far as possible, I will always "round up" my answers when designing a system, to help overcome the unavoidable little inefficiencies that creep in.
With that in mind, let's put a shopping list together, with a little explanation along the way:

The most important part is the battery.  I want to take 11A hours out, and I'd like to charge my battery when it's no more than 50% discharged.  That says I need a minimum of a 22 AH battery and I'd like more, of course.  Something like this 35 AH battery gives me more than another 11 AH margin (in case a cloudy day doesn't allow any recharging time) and isn't much more expensive than a smaller battery. 

If you're going to wire solar panels in parallel, you have to consider that under some conditions, one of them might try to force current into the other - or the battery might try to force current into them.  To prevent that, most writers recommend a diode in series with the panel output.  A diode is the basic semiconductor element, and we use it here because it passes current in only one direction.  While there is more written online about diodes than most people would want to read, the important thing to know about here is that there is a voltage drop across the diode that means some of our precious, hard-won energy from the sun is going to be lost in that diode.  We want a Schottky diode because they have a lower drop than an typical silicon diode.  Since each panel will deliver 1.2 A, a Schottky rated for 2A or more is all we need, and a 1N5822 (pdf datasheet) is fine.  These are 3A diodes that will drop about 1/2V at 3 A and set you back about a 25 cents each. They should go in the positive lead off each panel with the banded end away from the panel. 

We'll need a charge controller.  An MPPT controller is the most efficient, but we're not dealing with an extreme system here.  Since our charge rate is only about 2-2.5 A, anything above that is for margin or future growth.  A 10A MPPT charger seems to go for about the price of the battery, and a 10A PWM charger is about half that.  I will leave that choice to you, but will probably go with the PWM controller myself.  

Finally, we'll need a whole mess of odds and ends.  Probably the most important thing we'll need is a fast blow fuse to put on the battery output.  As mentioned in the comments to part 3, a battery bank can be a destructive device, and while you could argue this isn't needed in so simple a system, it's just a good idea to put something there.  Mount a regular or fast-blow fuse close to the battery to protect it from a short circuit downstream.  (Big systems use a special, rugged fuse called a "Class T" fuse and holder.  They're a bit hard to find in a smaller current, but worth looking into.)  We'll need wire, and for 2 A between the charger and panels, if they're within five feet, #18 wire is fine. Ten feet of #18 would result in a loss of 0.13 V at 2A (~1/4W) - handy calculator here.  It never hurts to use bigger wire, except for the pain in your wallet.  The wire in your walls is probably #14 or #16 and those are fine to use if you have it lying around.  We'll need terminals to connect everything and probably a wire stripper/crimper, if you don't have that.  The only other thing we'll need is a way to connect the radios to the battery, and a cigarette lighter plug adapter is pretty much a standard way of doing that.  For a permanent installation, Anderson PowerPoles are standard connectors in the amateur world, and a connection from a beginner kit like this to the battery is a good idea.

Oh, yeah.  We should also build some sort of portable stand, perhaps a cart, to carry the panels and battery outside to charge.  More to follow, including usable wiring information and mechanical stuff.


  1. If I may freely spend your money, I would suggest the following items to increase your system's usefulness (the following links are descriptive, not prescriptive, and I earn no money from any purchase from the linked sites):

    Sundance Solar's 60-watt 12-volt panel (at a per-panel cost of $189, this panel is about $20 cheaper than three of your linked 20-watt panels)

    Two 50AH deep-cycle batteries (this combination will allow you to run your amateur radio rig for over 48 hours per battery, while your PV system replenishes the other battery)

    Note that my proposed additions to your suggested backup system both increase your radio uptime and give you some flexibility in powering other needed devices.

    1. Not a problem - spending OPM is probably the most addictive thing there is. Just look at congress!

      Since the two smaller panels are here already, I'll keep going with them, but I like the idea of two batteries: charge one and use the other.


  2. Couple of comments.

    1. I don't like sealed lead batteries. Regular lead-acid batteries take a bit of maintenance, but only a bit of distilled water now and then. They last longer than the maintenance free. If you really must get zero maintenance batteries, get gels or AGMs. Though you will need a charge controller - they are VERY sensitive to over charging and shouldn't be connected to the panels directly. For my money, you can't beat 6-volt golf carts. They stand up to deep discharges, are not too heavy. Have decent amp-hour ratings.

    2. When the temperature is anywhere near 60 degrees F, the temperature of the panel will be above 60 degrees F. Above that and the watts produced fall off from the rated power. In Florida in the summer... Also if you really want full power you need to track the sun. That would mean at least moving your stand/cart every couple of hours.

    3. If you do a search on DIY generator, you will find a site with plans and parts for turning a small gas engine (say from a power edger - or from Harbor Freight) and an alternator into a handy dandy 12-volt quick recharger. Even the small ones will put out 50 amps or more. Not as a replacement for the solar, but as an adjunct. Personally I am also looking into a small 5 to 10 HP steam engine for the same thing.

    If you are going to be without power for an extended period of time (after Andrew, most of Miami was without power for a couple of weeks) you are definitely going to want to consider low-power broadcasting.

    1. Zendo Deb, it's not clear from the Amazon link I posted that the battery I was referring to is an AGM battery. Northern Arizona Wind&Sun has a more descriptive page that shows it's a sealed AGM, shippable by UPS, which means it can't leak. I haven't bought any battery yet - still looking at options.

      The rest of it, I'm pretty wired on, thanks.

  3. SG,

    For a small system such as this, making it portable is an excellent idea. If we ever actually settle somewhere, I'm thinking of doing a much larger system that is trailer-built (perhaps on a 20' unit), to be able to use it at rental property without getting into a fight over "improvements to the property" that cannot be removed.