Solar Panel / Battery setup

I need the help of the experts here on the forums. Let's say I have a device (to keep things CPF style, let's call it a light) that has a 12VDC current draw spec'd at 277mA. I'd like to have a solar panel to run it during the day and charge a battery that would be capable of running it at night. It probably won't be used all night, but it needs to be capable.

What would you recommend as far as batteries and electronics to make this happen? I'm assuming the solar panel will need some sort of voltage regulator to keep things on the level - both for battery charging and operation of the "light" during the day.

Nicad batteries work nicely for simple charging from a solar panel. They can handle C/10 trickle charging much better than nimh, and don't require voltage regulation like SLAs.

My first choice would be a 12v nicad battery, paired with a 12v (nominal) solar panel. The panel's output at 12v should be roughly a tenth of the battery's AHr rating (i.e. if you need a 4 Ahr battery, the panel should supply 400mA at 12v).
ooops... since you want to power the load at the same time, you'd have to add the load's current. For our example, that means the panel would have to source 400mA plus 277mA, or roughly 700mA at 12v.

The battery should be sized to have the required capacity, plus some margin. Assuming that it'll have to power the load through a 12 hour night, it'll need a capacity of 0.277A x 12hr, or 3.3Ahr. Make it 4Ahr, just to provide a bit of spare power.

Without knowing more about the load, I can't make any recommendations about regulation.

I would suggest adding a circuit to disconnect the load when the battery voltage gets low. Failure to do so will destroy cells in the battery pretty quickly.

have fun!

Steve K.

Steve - Thanks for the info. For the nicad's - should I just build a battery pack with 10 x 1.2V cells then? Panasonic's got some 4000mAh D size "High Temperature Trickel Charge" cells - you think those would be OK? They list standard charging at 400mA for 16 hours. Would the difference in length of sunlight during winter/summer days be a problem for battery life?

A little info on the load: It will be a piece of electronics that's meant to fit on heavy equipment. I'm assuming it has a pretty good regulation system of its own - seeing as heavy equipment voltage might not be the cleanest in the world. Most solar panels have varying voltage depending on load and light amount, so I imagine you couldn't get by without some form of regulation.

Ten 4Ah D nicad cells sound appropriate. Reduced sunlight in the winter will certainly be a concern, as well as the issue of nicads being harder to charge in cold weather (the internal resistance is higher). I'd still encourage the use of a low-voltage cutoff to protect the battery.

What exactly is the load?
If it's designed to handle the power noise and spikes that are found on a vehicle, then it'll do okay with the nicads.
The nicads will act as a load on the solar panel, and limit the voltage quite nicely. I've done this sort of thing myself with a 7.2v nicad battery, and it's hooked up to a 12v solar panel.
If you have any doubts, you can throw a high power zener across the panel to act as a shunt regulator (pick something with a suitably high zener voltage... maybe 15v or so. You don't want the zener to discharge the battery).

Steve K.

I think you normally wouldn't use fancy electronics for charge control for something this small. Just have enough panels to charge the batteries at > 12 volts, and have a diode to prevent discharging.

You have to take into account the inefficiency of solar panels and the limited availability of sunlight because of solar angle, clouds, rain, nighttime, etc. Solar panel power ratings are in terms of peak power, which is the power you get at maximum sunlight (noon, no clouds). Average sunlight will always be less than 50% of peak sunlight because it's nighttime half of the time. In reality, 20% is more typical and it can be as bad as 10% depending on your location. So to run your 277 mA device 24 hours a day, (i.e. about 4 watts), you may need a 40 watt solar panel, which is several square feet. A panel like that will cost a few hundred $$ and weigh quite a bit. Are you sure you want to deal with this?

Did anyone mention that nicd cells have a charge inefficiency of about 30%?

Looking a little bit further into this, I find that similar applications and controllers are partial toward lead acid batteries.

There are a few off-the-shelf charging controllers that seem to accomplish what I need. The good folks at PowerStream have a nice little controller that looks like it might fit the bill:
http://www.powerstream.com/pv-control.htm

[ QUOTE ]
flashlightlens.com said:
Looking a little bit further into this, I find that similar applications and controllers are partial toward lead acid batteries.

There are a few off-the-shelf charging controllers that seem to accomplish what I need. The good folks at PowerStream have a nice little controller that looks like it might fit the bill:
http://www.powerstream.com/pv-control.htm

[/ QUOTE ]

That's pretty steep, lad. Low end charge controllers should be a quarter or third that, $30 to $40, 'all over'. Try a google search on 'charge controller' and enjoy the read (and savings...).

Doug Owen

The controller isn't really the problem. The solar panel itself will consume most of the cost, if you really want such a large one.

Another word of advice: Solar panels made out of silicon have a pretty high dark current. They're essentially diodes working in reverse. If the panel get no or low levels of light, the battery will discharge over the panel. If you want to keep a Nicad connected to the panel all the time, I'll strongly suggest to get some regulation circuit. It's much easier with CdTe solar panels, which really do not let any current flow backwards. The catch: CdTe solar panels are more expensive than silicon ones.

Also: Solar panels and their load have to be optimized for each other. Quite often the manufacturers advertise only the open circuit voltage (open contacts) and the short circuit current (no load). Those do NOT reflect the power you can get out of the panel. The optimal voltage/current combination is lower and depends on the efficiency of the panel. The panel works best with this load.

A series Schottky diode will prevent that discharge.

Right. It's just that I have seen crappy designs of garden lamps with just solar panel, photo diode (as light switch), lamp and nicad. The nicads died pretty quickly...though an additional diode would just be pennies.

OTOH, I used to work in research with CdTe solar panels. Those things are much easier to handle, especially with simple circuits, and well worth the higher price. And they typically look nicer than silicon /ubbthreads/images/graemlins/smile.gif.

I suggest that for a 12V application, one should have a 14V solar panel at least. Charging the battery will require slightly higher voltage than the nominal voltage stated. Hopefully the regulator in your gizmo will be able to handle the higher voltage, or else you're going to need a regulator downstream from the battery/solar panel setup.

I haven't seen the answer to these questions yet:

- how reliable, i.e. how many dark days do you need to cover: two days? a week? a month?
- where do you want to set up this system, e.g. Death Valley as opposed to Anchorage?

Since I don't like to charge NiCds without any controller, I'd vote for the SLA version. With charge/discharge controller. But your app seems to be more or less on the limit between a cheap NiCd system and a small SLA system.

Bye
Markus

[ QUOTE ]
markus_i said:
- how reliable, i.e. how many dark days do you need to cover: two days? a week? a month?

[/ QUOTE ]

Well,....That's a good question. Will I get no current at all out of a panel on a cloudy day?

[ QUOTE ]
markus_i said:
- where do you want to set up this system, e.g. Death Valley as opposed to Anchorage?

[/ QUOTE ]

This would be on a piece of equipment that would probably only see service in Utah, Wyoming, maybe Idaho, Nevada, and Oregon. I guess you could say these are all fairly sunny states with the exception of Oregon - closer to the coast. It would be used year-round.

What about these panels?
http://www.uni-solar.com/cons_products_tech.html
Street price on the US-11 is around $100.
10.3W
.62A
16.5V
recommended for a 4Ah/day battery

[quote
This would be on a piece of equipment that would probably only see service in Utah, Wyoming, maybe Idaho, Nevada, and Oregon. I guess you could say these are all fairly sunny states with the exception of Oregon - closer to the coast. It would be used year-round.

What about these panels?
http://www.uni-solar.com/cons_products_tech.html
Street price on the US-11 is around $100.
10.3W
.62A
16.5V
recommended for a 4Ah/day battery

[/ QUOTE ]

Now we're getting to the meat of the matter. Some general rules are an average of five hours of full sun equivlent per day, a bit less in higher latitudes, say four. This is for panels pointing to the sun (but not tracking), that is pointed south at about your latitude (actually a bit less, favoring the shorter winter light). Mounting them flat (as you need to do on boats, RVs and the like) cuts into this some. I'd count on 30 or a few more Watts per day on the panels you suggest. Count on getting say 80% of that back from the battery at night.

Size the battery to take care of worst case dark periods. Might be a week even?

In general, you end up needing more of the expensive panels than you think at first.

Doug Owen

There's another wrench to throw in the mix - The equipment it's mounted on won't always be facing the same direction, so I can't necessarily angle them toward the South.

OK, so we have hardware that you are installing in/on heavy equipment. In general, that means that you have an existing battery source. You also have lots of room for a solar panel. For brevity, I'm calling your "heavy equipment" a tractor.

You might try what I do with my truck. I only drive it occasionally. I don't need to charge the battery 100% from the solar panel, but I want the battery to have plenty of juice when I do drive it once a month. The solar panel I have is just a trickle charger that offsets the drain from the electronics. I should be able to start the truck after several months even if it only gets direct sunlight for a few hours per day.


In your application, you could use the tractors' generator to charge a fairly large battery, and select a solar panel that will keep it from going completely dead between uses. If money is no object, you could set up a pyramid of panels to allow the tractor to be parked at any angle with relation to the sun.

The solar panels I've used do not do well in indirect sunlight. Overcast days drop the available power dramatically. Pointing it so that it directly faces the sun is optimal. Pointing it so that it's off by 30 degrees will drop the voltage a lot. Dust/dirt on the panel will also impact it's power output.

I'd be willing to bet that solar panels are not the answer for your app.

Daniel

paulr said "A series Schottky diode will prevent that discharge." Right, and many panels come with them built in, so refer to the Sandia study and DON'T add one if there is one. Here's a quote from it "blocking diodes were eliminated from the systems as designers found that the daytime losses through the diode exceeded the very low losses associated with night-time reverse current flow through the modules when the diode was removed." You may need a blocking diode to meet NEC requirements, but it doesn't make much difference either way as far as energy input to the battery. See here: http://www.sandia.gov/pv/docs/PDF/IEEE6.PDF
Note that you don't have to meet NEC for a piece of equipment, but give some thought to the safety considerations. A 4ah battery isn't going to be much of a problem, IMHO.

Steelwolf, the typical 12v nominal solar panel is 36 cell @ 21v open circuit. Older 33-cell panels are about 19.5voc. Both are called 12v.