Small Solar Panel to Power Electronics with 2 AA Batteries

[bondmatt]

My grandfather has a small weather station which is powered by 2 AA batteries. The batteries need to replaced quite often. The weather station archives extreme conditions but this information is lost when the batteries are replaced. I was thinking of purchasing a small 3V thin film solar panel from Amazon. I was planning to just wire it in parallel with the batteries (the batteries are in series). Peak current draw is about 10 mA. I am thinking it would be ideal to charge the batteries and power the weather station whenever possible.

I would just use an AC/DC adapter but the station is basically on the roof nowhere near an outlet (and I think it would just be cool to build this).

Any advice would be greatly appreciated.

Thank you.

[Mike90250]

I can't say it will work for sure, but you would have to use rechargeable batteries. Good luck

[bondmatt]

I was doing some rough calculations and a bit of searching last night. The smallest solar panel is about 20mA (at 3V). NiMH rechargeable batteries should last for 1000 charge cycles (3 years, thats much better than the 1 month the batteries last on their own), ~1.2V each, and 1100 mAh. At 3V the panel should be able to trickle charge the batteries and power the weather station. The batteries will almost certainly overcharge but since conditions will undoubtedly not be ideal at all times I can probably count on less current.

Assuming the panel supplies 20 mAh for 12 hours and the weather station uses 10 mAh the panel will charge the batteries 120 mAh each day. Overnight the batteries should lose 10 mAh for 12 hours or 120 mAh. Realistically I probably will not see 20 mAh but the station probably draws 5 mA or less on average.

It looks like I will need a diode to prevent the battery from discharging through the solar panel at night. A silicon diode would probably have too large of a voltage drop to allow the batteries to charge during the day (0.7V). Any idea where I can get a Germanium diode?

Thank you.

[Mike90250]

You actually need a schotty diode, but if you find Germanium that can handle 50mA, and the Vf and Vr is OK, use it.

Digikey.com Mouser.com avoid radio shaft - as they sell remarked floor sweepings (production line fallout)

[bondmatt]

Thank you.

I really appreciate the suggestions for sourcing components. I dont like making purchases online unless I have some faith I can trust the retailer. I am a mechanical engineering student, I wanted to go into electrical engineering but family persuaded me that there were more opportunities in mechanical. I have come to love the subject but with some time on my hands I am taking an electronics course. I have dabbled with simple electronics before but with some exposure to components like op amps and micro controllers the possibilites have increased exponentially. I will probably be on the look out for more and more odds and ends that I cannot get at radioshack.

[Sunking]

Well you are going to have to manufacture an interface between the radio, batteries, and solar panel. Otherwise if you think replacing alkaline batteries is expensive once every month or so, wait until you you replace rechargeable batteries and cleaning up the mess when they leak all over the insides of your radio you will have to replace once a month from overcharge.

[bondmatt]

Expense was never an issue but inconvenience is so I would like to prevent the batteries from leaking. The solar panel alone at $5 is going to cost far more than years worth of batteries. So I do need to consider overcharge. I did a bit of research. It looks like NiMH would be one of the worst choices if overcharge is a possibility. I just figured I would use NiMH because they are the only batteries I have around the house right now. NiCad would be much cheaper. If I do this I will start by not having the panel charge the batteries at all times (ie manually connect and disconnect it, probably use a switch). I am not over there all that often and usually not for long periods of time so I will probably run the risk of losing the data if the batteries lose sufficient charge. Im sure I can find a charging circuit somewhere and getting something accomplished in the meantime should motivate me to actually make some progress. I think I need some decent data before I go any further in terms of a charging circuit. I measured the current draw while trying to juggle the batteries and the multimeter/its probes while on a ladder. I suspect the 10mA peak was erroneous.

[r0g3r]

If you are a bit handy with elektronics you can make a ideal diode.
I found this circuit on the net and use it a lot in my little solar projects.

http://www.solarbotics.net/library/circuits/misc_switching.html

If the solar panel is 3volt then the output voltage is olso 3 volt.
Ideal for charging 2 NiMh cells.

Good luck

[Steve]

Hi bondmatt:

It looks like NiMH would be one of the worst choices if overcharge is a possibility.

I agree. NiMh cells require a different charger than NiCD or lead acid. If you simply connect unregulated current the cell will not last long.

I personally don't think you'll be satisfied using either types located on your roof. The reason I say that is because I had a set of those solar yard lights. They had a solar cell on top feeding a small charge controller/switcher and an AA size NiCD battery. At night a single blue high power LED would come on.

The batteries began to fail after about 4 months. Some went out while others dimmed. All were nonfunctional after 6 months. I replaced all 10 batteries which cost about 1/2 of a new light set. Again they lasted about 6 months until complete failure. I salvaged the solar cells and discarded the rest of them as garbage.
The lights were located out in the yard receiving sunlight for 3/4 of the day. I can only imagine the heat of a roof would be harder on a similar setup.

If you must use a battery on your roof I'd use a separate battery holder, one you can replace if it leaks goo. That way you can safeguard your equipment. Another option may be to run a low voltage wire across your roof and locate the battery in a more convenient place.

[Mike90250]

If you are a bit handy with elektronics you can make a ideal diode.
I found this circuit on the net and use it a lot in my little solar projects.

http://www.solarbotics.net/library/circuits/misc_switching.html

If the solar panel is 3volt then the output voltage is olso 3 volt.
Ideal for charging 2 NiMh cells.

Good luck

Nice link - thanks for shareing it. That's the same sort of thing that many charge controller circuits use - less voltage drop, less heat generated, and more power delivered.

[bondmatt]

Thanks everyone for the advice. I should clarify what the weather station really is. It consists of a temperature/humidity sensor, a rainfall sensor and a wind speed sensor connected by phonelines. The batteries are in the temperature sensor which is in a somewhat sheltered location. It can be reached by standing on the porch railing (made of 2x4s).

Its too bad, I think I have several of those solar powered garden lights that do not work anymore. At the very least I will have to open one up to see what kind of circuitry is in there. If I can get 6 months of continuous operation of the weather station I will be happy. I should receive the solar panel within a week.

[Mike90250]

I wonder if it uses ALL 4 of the phone wires in the line.....

[bondmatt]

Where the lines connect to the rainfall and wind speed sensors the connections are more or less permanent. Those sensors are also located in much more inconvenient locations than the temperature sensor (the wind speed sensor is on an aluminum pole strapped to the chimney). Still, a good suggestion without any details.

I pulled apart a solar powered garden light today. There is no circuitry to protect the battery from overcharging. There was no sign of significant battery leakage (NiCad batteries). However, it is a very small solar panel. The solar panel was still working. One battery still produced ~1.2V the other produced almost no voltage. The output of the panel was about 2V in low natural light. The batteries have the following label: "1.2V 600 mAh standard charge: 60 mA for 14h" It would seem to me that a small panel would not produce 60 mA for 14 hours each day, if 60mA at all. If overcharge killed the battery then the batteries must not fully discharge overnight. Or it could be that the number of charge cycles coupled with temperature changes killed the one battery. I will try to test the panel tomorrow in bright light. There is a diode in the garden light to prevent discharge through the solar panel (in addition to the LED). I will also try to check the voltage drop of that diode.

[irfanarghi]

Where the lines connect to the rainfall and wind speed sensors the connections are more or less permanent. Those sensors are also located in much more inconvenient locations than the temperature sensor (the wind speed sensor is on an aluminum pole strapped to the chimney). Still, a good suggestion without any details.

I pulled apart a solar powered garden light today. There is no circuitry to protect the battery from overcharging. There was no sign of significant battery leakage (NiCad batteries). However, it is a very small solar panel. The solar panel was still working. One battery still produced ~1.2V the other produced almost no voltage. The output of the panel was about 2V in low natural light. The batteries have the following label: "1.2V 600 mAh standard charge: 60 mA for 14h" It would seem to me that a small panel would not produce 60 mA for 14 hours each day, if 60mA at all. If overcharge killed the battery then the batteries must not fully discharge overnight. Or it could be that the number of charge cycles coupled with temperature changes killed the one battery. I will try to test the panel tomorrow in bright light. There is a diode in the garden light to prevent discharge through the solar panel (in addition to the LED). I will also try to check the voltage drop of that diode.

Hi, I see your problem here is about how to recharge the battery using solar panel. my suggestion is, u can use cellular phone battery such as Li-Ion battery or Li-Poly battery. they are usually having voltage around 3-5 volt with capacity around 1000mAh.
to be able to charge them using solar panel u need at least three of your solar panel connected in series so that you can get 6 volt from the solar panels in the system.

[Mike90250]

please be careful using batteries with improper charge controllers, or one sunny afternoon, your little garden lamps may catch fire from overcharge.

Mixing Liion batteries with nicad chargers could be very exciteing !