• Please review our updated Terms and Rules here

Electronics pros here? How to add a solar panel to a computer?

kyodai

Experienced Member
Joined
Jan 26, 2011
Messages
467
Location
Kerpen, Germany
OK, I am not that deep into electronic circuits and stuff. I can replace a capacitor, repair some broken contacts, but i never designed any circuit from scratch.

So to make it short - my idea was to add a solar panel to one of my palmtop PCs. I think something like a tidalwave clone is a good candidate because their main power comes from 2 AA-batteries in a series circuit, so basically you need 3 Volt. I don't know how many ampere it consumes, but it probably depends on the program you run and the processor speed selected. Usually these have about 10 hours of battery life in normal use, so I'd rate the typical AA-battery (a good one) at 2000 mAh, so my guess is it draws around 200 mili Ampere in standard use, probably more at peak. Watt = Volts x Ampere (If my school time wasn't a total waste of time), so a 3 Volt 0.6 Watt Solar cell should be sufficient for normal use, might not be enough for "peak" power consumption.

Well so i looked around a bit for solar cells that could fit.

Radio Shack has some that seem quite the size i like. The lid of the Tidalwave is 4.5 inch x 9.7 inch (11.5 x 24.5 cm), so the solar cell should be roughly that size, preferably a little bit smaller.

The one i looked at is this:

http://www.amazon.com/RadioShack®-...1456004820&sr=8-30&keywords=radio+shack+solar

From the size i could fit 3 of them on the lid if i calculate correctly. 1 Watt peak power at 6 Volt each, that would be 3 Watt - well probably much less unless it's a very sunny and i point the lid directly into the sun. But i think 3 Watt already sounds like it would be sufficient.

OK they are 6 Volt (Well at peak i guess), so the goof in me says like "hey you need a resistor. Wanna reduce the volatge by 3 Volts and we assume the Palmtop could use up to 500 mili Ampere at peak (Not sure how realistic that is though), then we'd need a resistor that is 3V / 0.5 A = 6 Ohm

OK, now some of you are hopefully already laughing. Well I have a feeling that there might be a better way to reduce the volatge? Especially since the power the solar cell produces varies incredibly - i think on a cloudy day we would likely get only 1/10th of the theoretical 3 watt. I know there are things like step-down transformers so you can use a 120 Volt device on a 240 Volt poer socket. But for 6 volt to 3 volt?


I know, a good solution would be to use rechargable batteries, buy a 2 dollar solar cell charging circuit and be done, but hey, making it easy wouldn't be so much fun, right? So at first I'd really like to evaluate the possibility of running the device completely from a solar cell without a battery buffering it. How do these 2 dollar solar calculators do that without risking supplying too much voltage or power during peak times?


Imho the perfect solution would have a little switch so i can select if the solar panel charges the battery (Here comes our cheap charging circuit from china again) or directly powers the palmtop. The switch also makes sense when you use non-rechargable batteries since trying to recharge alkalines can lead to them blowing up, leaking or other nasty problems, right?
 
Last edited:
Given the variable nature of solar energy, you're going to need something to store the energy in. Rechargeable batteries would be my first choice, but maybe a supercap will do. Secondly, you're going to need something to manage the variable nature again. A buck/boost charger will do the trick; that is, it reduces over-voltage or raises under-voltage conditions to what you're aiming for.

Here's a TI PDF that talks about that If you do a web search for "buck/boost solar battery charger" or the like, you'll get lots of hits.
 
First figure out how much power you actually need. Then determine if a solar cell that makes 30% more power than that is reasonable for your intentions.
 
I've done this with a raspberry pi which has been running for a couple of years now.

First thing is as Chuck said you need an energy storage device. Solar power is notoriously unreliable and even a small cloud floating over will drop the wattage produced by quite a lot. My choice is a 12V SLA battery as they are cheap and pretty long lived. As for sizing of the battery that depends on how much load you are planning to run on the system. My rPi draws about 2W at 5V so thats 2/5th of an amp. I have a 5AH battery so I can get around 12.5hrs of running with no sunlight (I don't run it at night). This calculation is up to you; if you want it to run all night and charge during the day you'll need something bigger. Don't run your palmtop directly from the panel. It will cut in and out like a yoyo.

Second thing is you need a regulator. 12V solar regulators come up on eBay quite cheaply; they have a buck boost converter in them to automatically regulate and charges a 12v battery. Mine also has overcharge protection which is a bonus. Choose one that meets the power requirements of the PANEL - i.e. if you buy a 10W panel then make sure it can handle > 10W.

Third thing is the panel itself. Choose one that is DOUBLE you power requirements if you only plan to run during the day. For example a 10W panel will only produce 5W most of the time UNLESS the sun is perfectly overhead on a hot clear day. Even double is sometimes not enough depending on panel quality. If you plan to charge and run during the day and also run at night you will need to quadruple your requirement as during the day the panel also has to recharge the now almost dead battery.

Forth thing is a regulator for the device you intend to run. I use a simple USB car charger for the rPi as it has a small inbuilt switchmode supply which is already tied to a USB socket. This gives me +5V with reasonable efficiency.

That's a very brief rundown. If you have specific questions let me know.
 
At first thanks everyone for your valuable feedback and helpful explanations.

I will look into these buck-boost chrgers and see if i can come up with something halfway affordable. I think at first i will experiment with alligator clips and see how reliable the cell power production is before i start cutting the lid open and install everything.

I found these step-down modules which are affordable and small enough to be used in the palmtop, so maybe I'll give tehse a try:

http://www.ebay.de/itm/5pcs-Mini-36...eeb084e&pid=100005&rk=1&rkt=6&sd=331744431127

Input is 4.75 Volt min, so maybe I'll go with 9-12 Volt solar cells to be sure to get the minimum voltage even on poor light conditions. Since the output voltage is adjustable i could try a bit with either directly generating 3 Volt or a bit more for a seperate charging circuit (I guess to charge a 3 volt battery assembly you probably need 4 volt or so)...
 
OK, so a short update on this project. I found some gear (on ebay of course) that was cheap enough to give it a try. So here is what i ordered:

- 1.5 Watt Solar Cell @ 12 Volt 115 x 85 mm polycristalline (ordered 6 of them)
- Mini 360 DC/DC stepdown converter (Input: 4.75V - 23V, Output 1.0 - 17V, rated at 1.8A)
- Boost Buck (Input 3.8-32V, Output 1.25 - 35V, rated 3A max)
- Miniature Boost Buck (Input 4.75 - 24V, Output 0.92 - 15V, max 3A)
- Digital Multimeter (Generic El Cheapo)


That should be enough to get some basic experiments going.

the Solar cells arrived today (Already!) - they are a super tight fit, i plan on using 3 of them which would cover the back lid perfectly. Actually i will need to do some sanding to fit them, that's why i ordered 6 of them. So i pick the 3 where the tolerance of the edges allows most sanding and if i blow some i have spares. In Theory that would be 4.5 Watt which is actually a lot. I found old data for this palmtop which claims it will draw a max of 265 mA with hi speed and serial on. With serial port off and lower speed i think i can set it so a peak would be around 150 mA, maybe a bit more if i use the PCMCIA card i think. In idle it should be around 65 mA.

So We have 4500 mWatt at 12 Volt maximum, that is a whooping 375 mA if i calculate right. So at the brightest day in the desert with perfectly aligned cells we could get 2.5 times the power we need (If the advertised 1.5 Watt were not just wishful thinking). In realistic conditions we can probably already be glad if we get the 150 mA at best i think.

To stay realistic - with unalligned (not pointed to the sun) solar cells in low light conditions we will probably rather have somethign around 20 mA i fear. Call me pessimistic, but the chinese advertisement was "Excellent weak light effect" - I'd be surprised if we even get 10% in weak light conditions.

The rest of the stuff will probably take weeks or more to arrive and unlikely that i will experiment much until my wife is on vacation - end of April. Stay tuned for more from the most optimistic solar power project ever.
 
The problem with solar devices is that they do require serious surface area to get a decent return.

12V ones are available as auto solar battery chargers, but as noted, let a dark cloud get in the way, and you're relying on the attached battery. Solar USB power banks have all sorts of freaky power ratings (e.g. 30 A-h at 5V), but that's only after you've charged the internal lithium-ion cells, which, under the best conditions (day in the tropics) takes about 13 hours.
 
Back
Top