Linear Technology article discussing solar charging, using maximum power tracking - see extracts
The idea dates back to 2002 when I was working on my dissertation on methods of energy scavenging that could be integrated on silicon. On a footnote, I added simple calculations demonstrating that if the surface area of a 'candy bar' mobile phone is covered with solar cell of 5% efficiency (which is pretty rudimentary), including a DC/DC converter, the overall output power is sufficient to augment the battery, providing power for stand-by indefinitely.
And here it is! Or rather a simplified version that is equally effective if not better!
Specifically what you need:
The Motorola F3 - a very cheap (I suspect the cheapest mobile phone there is) phone that is widely available both in US and UK. It is this affordability and basic ruggedness that I had chosen this model. More on this later...
A solar panel - the one I had acquired is available online (more on this later) but any other suitable make can be used provided it is rated 4.5V and above, having the dimension not larger than 100mm x 40mm.
A Schottky diode, 1N5817 or equivalent - a basic component which is not hard to find, you can even use the SMT version as long as you can solder some wires on both ends.
Some wires, I use enameled wire, more on this later...
Some basic soldering + a sharp pen knife + aluminium tape + epoxy + a hand drill + super glue
That's it! Surprised it is so easy? Why not get your hands dirty and build one!
The finish product is shown below.
Step 1: Introducing the solar powered phone.
A sign of its popularity , the phone has its own Wikipedia page!
Some background on this model.
First off, it is VERY basic, you get more ring tones than functions on this phone, supposedly it was designed for developing countries, interestingly, it has an E-INK display that is very readable in bright sunlight and very good reception.
Turn-offs? SMS on this thing is atrocious! Bottom-line, the price far out-weights everything, if it is made of cardboard it would be disposable.
And here's the solar panel, I chance upon this little panel on an web-shop, www.dealextreme.com (I am not affiliated with the site)
The dimension is PERFECT. Fits the phone like peas in a pod.
I don't have any data on the stand-by power consumption, just an estimate based on the number of hours it takes to drain the battery (3.7V, 700mAh). It works out to about 22mW (Typical P = 6mA x 3.7V). Of course this figure jumps significantly to a watt when the phone is in used (as in transmitting). The idea is to provide power in excess of the stand-by power consumption to charge the battery. More on this later.
Step 2: Getting the work done
The drilling of holes on the cover definitely void your warranty!
Grab some sand paper and roughen the surface of the battery cover, this helps with the adhesion.
Spread the epoxy like butter, I use JB cold weld epoxy, it is supposed to be good, cures in 24 hours, I think super glue might work but it wouldn't be as durable. Also note that you have to attached the solar panel in the correct orientation.
Now as you can see, the holes let the epoxy escape allowing better adhesion, acting as 'grips'
That done let's wrecked the phone! :D
The only place that needs 'trimming' is a gap on the phone's naked frame, this allows a snug fit of the diode.
Step 3: Wiring the doide...
A basic lithium ion charger is a constant current charger, with an end-of-charge detector, charging ends when the battery reach its full charge at 4.2V (most common) and the charger goes into trickle charge mode.
A solar cell can be modeled as a current source in parallel with a diode. Higher voltage is achieved by stacking these individual cell in series.
In this case, the charging circuit consist of just a blocking Schottky diode. Advantageous is its low forward biased voltage, which is around 0.3V.
Is it safe to charge the battery in this manner?
The charging current is very small, comparable to a trickle charger, the solar cell act as a weak constant current source. The panel I brought has a open circuit voltage of 5.8V under bright sunlight and a short circuit current of about 80mA, you can measure this with by simply shorting your digital multimeter across the solar panel terminal, this is slightly more than 10% of battery capacity at 700mAh. But remember when charging occurs, it is at a voltage 4.0V and above (3.7V + 0.3V diode drop), the net charging current (minus the current drawn by the phone) is not going to be any more than the measured short circuit current.
In any case, the rule for charging a battery safely is to do it at 0.1C (i.e 70mA) and here, the charging current does not even excess this figure.
A net positive current in excess of the stand-by power will charge the battery. Some experiment I've done shows promising result. It take 3 hours of good sunshine to charge a 'dead' battery with the phone powered off to last 12 hours when the phone is standby without charging.
Again some calculation, working back, given that the phone draws 6mA on standby, 12h x 6mA = 72mAh, each hour of sunshine gets you 24mAh (72mAh/3), on standby the phone consumes 6mAh, thus a net charge of 18mAh will go to the battery. Consider a day with 6 hours of charging (day) and 18 hour without (night), your phone left on standby will run indefinitely.
It is best however, to let solar panel charge the phone ONLY when battery indicator shows one bar below a full charge.
Step 4: Wiring the rest of the gubbins
The hazard is the crafty-bits. Follow the photos and you will get there. About the aluminium tape, these are used for heating ducts, hence likely available in hardware stores. Alternative may be just to stick double sided tape on aluminium foil.
One last warning, the panel does reduce the reception strength, so when you're indoors, signal could be bad. But because it needs sunlight, ideally it would be placed near a window, so it isn't the worst trade-off.
That's it! Done!
Step 5: Last word
Although circumstances differ, on reflection it is very liberating and the idea certainly does work, not cost-effective but that's wasn't the point to moot. Certainly useful for a out and about camper.
How can it be improved?
Interestingly Texas Instrument has a boost converter, TPS61200, that works down to 0.3V providing a regulated output up to 5V. This particular chip is well-sited solar powered applications, anyone interested may well explore further.
I would assume since a solar panel is augmenting the battery, its dimension could be halved (e.g. using a lithium polymer battery commonly used by RC hobbyist), doubling up real estate for additional circuitry.