solar power

Andy technil at riseup.net
Sun Feb 10 23:32:15 CET 2008


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Sorry for the long e-mail, hopefully it will help folks work out
possible applications of cell phones/solar. I also added it to:

http://wiki.openmoko.org/wiki/Hardware:Neo1973:Alternate_Cases:Solar_power#Design_options

Trickle PV charger integrated into a back case may be feasible. Chances
are it would not significantly lengthen daily operations, unless it was
left in the sun for extended periods of time, and the other problem is
that while it sits in the sun you can't really easily use it as a phone
w/o a headset. In addition, It's unclear how hot the phone might get
(from absorbing the other bands IR etc.. while sitting in the sun), and
whether that might be a problem for the phone electronics or the battery
in some hotter climates.

To fully power and charge the device, plugging in a small PV module to
the USB port is an option. If you integrate the PV module with a Li-ion
battery, charging circuit, and a USB hub, It could make sense for
extended computing operations outdoors or places without power. For
people roaming about the wilderness/bush this might be an option. If it
weren't for the GPS, I might even ask the person why they would bring
the phone with them, particularly if there were no cells/wifi hotspots
around.

For the sake of a high end usage calculation, figure 2.5A @ 5V nominal
for this application (12.5 Watt nominal) depending on what peripherals
you want to power (500mA @ 5V on 2 usb ports, and charging the hub
battery, powering the hub itself, and 500mA @ 5V for the phone running
and charging).

> arbitrary chosen polycrystal
> solar cell 100mm x 100mm: 0.47V, max current (short circuit) 2,6A
> (www.conrad.de, #112135-99)

So we could figure 1 string of 12 cells in series to give us the 5V and
2.5A. arranging them in a 4x3 pattern for 12 total we have a size of
300mm x 400mm for powering and charging everything directly from the
sun. This assumes perfect alignment of the panel with the sun and a
clear sky. The panel will only be able to provide this power and
charging for a normalized number of solar-hours/day (4 or 5 where I
live). Minimally this panel, battery, and hub could probably fold to
something around 320mm by 120mm by 38 mm, making its folded footprint
still about 5 times as large as the phone in area and over twice the
thickness.

Taking into account the rules of thumb for the energy available (out of
the original ~1kW/m^2) in various scenarios:.. the available energy is
further reduced.

full sun, panel square to sun: 100%
full sun, panel at 45 degree angle to sun: 71%
light overcast: 60-80%
heavy overcast: 20-30%
inside double pane window, both window & module square to sun: 84%
inside double pane window, both window & module at 45 degree angle to
sun: 64%
indoor office light at desktop: 00.4%
indoor light store display: 01.3%
indoor light home: 0.2%

If you want to charge the device by laying it or a panel down on the
ground in the sun, understand that you are going to lose about 30% of
the available energy just in not having it positioned perfectly. Of
course, the part of the earth's surface you are on is rotating away from
or towards the sun, changing the angle of the incident radiation. So
even if you position the phone or panel perfectly, it won't be "perfect"
all through the day, unless the light is already being scattered by
cloudy or overcast situations. Indoors is simply a no go situation, even
for trickle charging.

After taking the 30% of the available energy away from our panel due to
imperfect conditions, it leaves us with 825mA @ 5V.

To just trickle charge the phone at the 100mA @ 5V through the usb in
imperfect conditions, we can reduce the area of the cells by about a
factor of 8. 120000mm^2 then shrinks to 15000mm^2 or about twice the
phone footprint if you assume it is a 62mm x 120mm rectangle. You would
still need 12 cells in series, each with an area of 1250mm^2. 30mm 40mm
might be a workable size cell, as about 6 would cover the phone, and
another six could fold out.. this would definitely add some bulk.

To go even further, we could interface directly with the battery in the
case and provide different currents at a slightly lower voltage, meaning
fewer cells in series, larger area, and we might cut out some of the in-
efficiencies of the on-board charger. The fully charged neo battery has
a 4.2 V open circuit voltage, and a nominal voltage of 3.7 volts. This
suggests at an absolute minimum 10 cells @ 0.47V to be able to reliably
charge anything (4.7V). The footprint of these polycrystalline cells is
still larger than the phone footprint.

If we just said, OK, we just have the non-curvy back of the phone
portion as area to use, about 50mm x 80mm, and we are going to use 10
cells for 4.7V to directly charge the battery, 4000mm^2 of cell area
divided by 10 gives us a 400mm^2, which points to a rectangular cell
size of 8mm x 50mm. this gives us 1/25th of the original cell (100mm x
100mm) area and correspondingly about 1/25th of the current. this is
about 104mA under perfect conditions. If we derate it by 30% for non
perfect conditions, alignment, etc. we are looking at a charge current
of just 63mA @ 4.7 V for 10 cells in series.

So if you can get the phone to draw nothing when it is off, and charge
for the full length of the normalized solar day (say 6 hrs.) with the
trickle charger in the last example you would get (optimistically) on
average, somewhere between 378mAh and 624mAh per day, or about 30 to 50
percent of a battery charge a day. In climates with lower normalized
solar hours, you get less, In climates with more you'd get more. I
should stress again that these normalized solar hours are NOT the number
of hours of sunlight you get per day, you need to look them up on a
chart for your region, and then leave your phone outside all day!

What is obvious is that -->Power Management is Important!!!<-- and that
you probably won't solve power management problems by sticking some PV
cells on the back of your phone. However, the better your power
management and device efficiency, the farther the cells you could put on
the back of your phone go in meeting some sort of auxiliary charging
function.

other stuff:

Voltage of a cell is related to the underlying physics of the cell and
its manufacturing process,(crystalline, polycrystalline, amorphous,
printed etc.., its particular junction type, and dopants used to make
the junction). (This is measured Volts open circuit) more cells in
series, more voltage. Cells and modules then have different responses to
the effect of ambient temperature on the voltage of the cells, depending
on their manufacturing process. Usually there is a slight voltage drop
at higher temperatures with all types, but is often more pronounced with
Amorphous.

Current of a cell is proportional to the surface area of the cell that
faces the light source. (This is measured Amps short-circuit current)
Larger area, more current. Cells linked together in series string should
be of the same or extremely close surface areas. The smallest surface
area limits current and can cause overheating and damage within the module.

modules that have to deal with shading should have bypass diodes.
modules should be accompanied by some sort of charge control that has a
blocking diode to prevent current leakage from the battery in periods of
darkness.

Last but not least, the cell and/or module efficiency of terrestrial
cells should be between a low of 9% and high of 17% conversion of
incident radiation to electrical current.

Hopefully I haven't made any embarrassing mistakes. Feel free to correct
me if you find anything, thanks. HTH

Andy
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFHr3tvTHCOtwancvwRAn/4AJ41jC+MGBZwaRUFckwW2ueS429xwACghSln
IHANRxAaQALgNmdGGJpY9Sg=
=MEur
-----END PGP SIGNATURE-----




More information about the community mailing list