Solar Charger

[tjetson]

For my school project next year (the teacher has advised us to think about our designs and such now), I want to build a weather station type thing. One requirement is that our projects must be powered by renewable energy. I thought it would be easiest to use a solar panel.

So, I decided that I would use a solar panel and a pair of Ni-MH batteries. I would need some type of sensor to detect when the solar panel wasn't providing enough current to run the circuit (ie it's too dark) and switch over to the Ni-MH batteries. I would use some form of voltage step-up to increase the batteries from 2.4 V up to 5 volts. I would also want the batteries to be charged whenever the circuit was currently being powered by solar.

So, my question is, how would I go about detecting the current? Also, I understand that I need a solar regulator for the solar panel. Could I just use a 9 volt panel and a 7805 + caps?


EDIT: For the solar charging, would either of these be appropriate?
http://www.futurlec.com/Maxim/MAX713.shtml
http://www.futurlec.com/Maxim/MAX712.shtml

 

[MFB]

The US based Nuts & Volts magazine has been running a series entitled “Experimenting with Alternative Energy” by John Gavlik that’s currently reached Part 5. Solar panels and battery charging have been covered in extensive detail, including basic theory, schematics and even examples PICAXE code. I think this is a resource well worth you downloading from www.nutsvolts.com.

 

[eclectic]

tjetson.

See here:
http://www.picaxeforum.co.uk/showthread.php?t=13234&highlight=garden+light

then do a search using the terms

garden
solar
light

Lots of info on the foum.

e

 

[hippy]

A common way to combine two or more power sources is to use diode mixing. Current will then only be drawn from the supply which has the highest voltage. As long as the solar unit is putting out a higher voltage than the batteries it will not draw any current from the batteries.

 

[Andrew Cowan]

And if the solar cells are unable to supply a high enough voltage (ie it is cloudy), the voltage falls and the batteries kick in to help.

A

 

[tjetson]

The US based Nuts & Volts magazine has been running a series entitled “Experimenting with Alternative Energy” by John Gavlik that’s currently reached Part 5. Solar panels and battery charging have been covered in extensive detail, including basic theory, schematics and even examples PICAXE code. I think this is a resource well worth you downloading from www.nutsvolts.com.

For the reference of others, I would like to put forward this link:
http://www.learnonline.com/experimenter kit.htm

At the bottom of the page are the 5 parts you mentioned for download. I couldn't find them on the actual Nuts and Volts website.

 

[manuka]

You're in luck-weather stations don't need much energy! Did your teacher say NEXT year? You could still do it THIS year! Even 50-100 milliWatts average output from a small PV may be enough. Grab some solar garden lamps perhaps & explore their circuitry (see below), & dumbfound your teacher before the summer break kicks in. Stan

 

[lbenson]

Note that for anything solar- or battery-powered, the 7805 is quite power-hungry, and needs a lot of headroom. The LM2931 seems to be a favored replacement (up to 100ma).

 

[BeanieBots]

This can be very simple if you can get the right battery & panel voltages to start with.
Don't think about switching between the two supplies. ALWAYS run from the batteries. Any type of switching will either use power to switch or cost a diode volt-drop which is just a waste.

Simply run direct from the battery.
If your panel is the correct voltage, you can connect it to the battery via a diode to prevent back flow when dark. Current will either flow out of the battery when the panel output is low or into the battery when the panel output is higher than the load. It really is that simple.

Do not use a 7805 regulator. As mentioned by lbenson, it is power thirsty itself and is VERY inefficient at converting voltage which is true for ANY linear type regulator.

 

[tjetson]

You're in luck-weather stations don't need much energy! Did your teacher say NEXT year? You could still do it THIS year! Even 50-100 milliWatts average output from a small PV may be enough. Grab some solar garden lamps perhaps & explore their circuitry (see below), & dumbfound your teacher before the summer break kicks in. Stan

Well, he did say next year. This is for a sort of secondary class I take once a week (a VET course, for those who know what that is), which unfortunately ended around a month ago. But, I figure if I can put together a circuit diagram or something over the holidays then that'd be great (class will probably start again in February 2010).



Also, I liked the idea of a Ni-MH battery so that I could still run the circuit on a dark day. Perhaps I should give a list of some of the components so far:

-Picaxe (probably 18X)
-DS18B20 temp sensor
-ModernDevice's LCD controller board
-Humidity sensor
-Perhaps an air pressure sensor
-Some form of wind speed detection

@manuka: I like your circuit there, but I can't seem to find the IC specified there. Are there any simple substitutes?

 

[tjetson]

This can be very simple if you can get the right battery & panel voltages to start with.
Don't think about switching between the two supplies. ALWAYS run from the batteries. Any type of switching will either use power to switch or cost a diode volt-drop which is just a waste.

Simply run direct from the battery.
If your panel is the correct voltage, you can connect it to the battery via a diode to prevent back flow when dark. Current will either flow out of the battery when the panel output is low or into the battery when the panel output is higher than the load. It really is that simple.

Do not use a 7805 regulator. As mentioned by lbenson, it is power thirsty itself and is VERY inefficient at converting voltage which is true for ANY linear type regulator.

So you reckon that I should just do this?

Battery
| |
| | +
| |
Solar Panel
| |
| | +
| |
Picaxe et al

 

[Haku]

The most basic solar setup for low wattage panels is:

The schottky diode prevents the battery from discharging into the solar panel at night / when it's cloudy etc.

For higher wattage panels & ampage batteries you need a charge controller, and a fuse or two wouldn't go amiss either.

 

[tjetson]

The most basic solar setup for low wattage panels is:

The schottky diode prevents the battery from discharging into the solar panel at night / when it's cloudy etc.

For higher wattage panels & ampage batteries you need a charge controller, and a fuse or two wouldn't go amiss either.

Does it have to be a lead battery? Is there a different setup for Ni-MH batteries? Also, does this circuit charge the batteries?

 

[Haku]

It doesn't have to be a lead acid battery for low power projects, but make sure your panel isn't too powerful (voltage & ampage wise) for the batteries so you don't overcharge them. A lot of the solar powered USB chargers use either 4x NiMH cells or a lithium battery, but generally aren't waterproof, or any good.

That circuit will charge the battery when the solar panel is giving out more power than "your device" is taking, otherwise it'll take the power from the battery when the panel can't deliver enough.

To make things easier take a look at the various solar shed lighting kits around, they all come with waterproof solar panels and often charge up 4x AA batteries. If you're intending to use NiMH then splash out on some low self discharge cells.

The really knowledable guys on http://www.solarpowerforum.net/ should be able to answer more technical/detailed questions, but don't be suprised if they wonder why you're not using lead acid

One day I will finish my ~120watt panel, I tabbed all the cells and built the frame but can't work out how to stop it from flexing before I fit the cells - I'm much more at home with a soldering iron than engineering work.

 

[manuka]

tjetson: That 4 legged ZE002/ANA608 LED driver IC (shown in the solar garden lamp pix) seems only available by purchasing the complete lamp. It's become almost standard in most offerings I now see here in NZ, so suggest you just bite the bullet & spend a few $ at your local hardware store on a swag. If unavailable I'll even offer to post you one Trans-Tasman. To tell the truth I'd really like to know more about them myself, as they look to have potential for all manner of 1 x AA driven projects. Google shows nothing decent on either the ZE002 or ANA608 however...

FWIW I'd however say solar powering a weather station will be the least of your design & tweaking concerns. Look at the assorted transducers that need PICAXE interfacing first, perhaps just powering them off a 3 x AA battery pack or solar shed light kit. This forum has had mountains of "my sensor has gone all wonky" postings over even such a standard device as the DS18B20.

You can pretty much take old sol's output & utility as cheaply & readily organised for such a modest need project - especially in sunny Australia. Worrying about solar charging at this initial stage is perhaps akin to concerning yourself with fuel for a F1 racing car- engine & handling design issues will be much more crucial!

 

[tjetson]

I hadn't even thought of just using a solar garden light's circuitry. That's a really good idea. When I saw your circuit, manuka, I was thinking of buying all of the parts; but buying a solar lamp and using that would probably be much easier.

 

[Haku]

The main problem I've found with garden solar lights - at least here in the UK - is that the solar cell is woefully inadequate to run the light for more than about 1-2 hours after it gets dark, indicating that to run anything more than a single LED you'll need better quality cells or lots of them.

Most of the cells on garden lights look like this: http://en.wikipedia.org/wiki/File:Solar_garden_Lights_top.jpg which are a con because only those four really tiny strips (technically cells) actually produce electricty, 0.55v per cell in direct sunlight to produce up to 2.2v for charging a 1.2v AA.
This is more the sort of cell you want to be looking for in a product or on it's own: http://i13.ebayimg.com/05/i/000/ea/3f/0f49_1_sbl.JPG

 

[manuka]

I certainly agree that some PVs are "more epoxy than PV", but for heavens sake we're talking loose change here-just buy ones with better use of the cell surface area!

Here in NZ the likes of "The Warehouse" have showroom walls festooned with so many dirt cheap SLG (Solar Garden Lamp) offerings, most at ~zilch profit margins, that I wonder why they bother.

Pix below of a single NiMH AA solar powered bedside lamp/emergency torch/AA charger (using a rescued ZE002 based SLG engine),all built into a 2 x AA battery box. Total cost ~NZ$4 (~US$3 or ~£2). Under bright NZ sunshine ~30mA was delivered to the NiMH AA, with only ~10mA current drain when the LED operated. Roughly allow hence a 3:1 discharge/charge ratio for 1 hour of sunshine to give 3 hours of night lighting (confirmed OK). Kids loved them ! Stan.

 

[tjetson]

My original idea for the weather station was just a thing you turn on, use and turn off, not leave on to monitor the weather over an hour or whatever. Therefore, it wouldn't use much of a battery or whatever, because it would only be on for a minute or so. I'd have to check with my teacher to see whether I should do that or a device which can continually monitor the weather.

 

[manuka]

Right- the weather changes quite slowly of course, so just occasional monitoring is usually tolerable & the average current needs can be low. Of course, since this is 2009,get a PICAXE to turn on/monitor/log & then turn off. Us humans have got more to do! Stan.