home  wiki



To come up with several designs (for powering nodes), who's principle goal is efficiency, and another design who's sole constraint is cost.


S O L A R~~P O W E R E D~~A P ' S


Editor Slave: Chooken
PART 1: Members List

PART 2: Collection Of Emails Reposts Relating to Progress + Ideas

PART 3: Mailing List

PART 4: Idea flinging etc - Be Creative and constructive


PART 1: Members List


If you are very interested in the possible creation and technical designs of a Solar Powered AP then please list your Name/Emai/IRCnick in the table below. I will endeavour to email the people who enlist with important information regarding the progress of other members thoughts/designs/tests.


Will Lanigan | External linkchooken@jibbed.cjb.net | Chooken

Robert Foster | External linkwireless@mountainvisions.com.au | ACG

Matt Pearce | External linkmattpearce@optushome.com.au

Ben Anderson | External linka_neb@optushome.com.au

Robert Farrar | External linkrobert@secnet.com.au

Clae | External linkclae13@yahoo.com

Andy Freeman | External linkwireless@kawasaki.kz


PART 2: Collection Of Emails Reposts Relating to Progress + Ideas


From: "Ben Anderson" <a_neb@optushome.com.au>

You're making the assumption that the panel -> battery -> use is 100%
efficient... More typically it's like 75%. Which puts you right on the
line. If you're using the cheap cells, be aware that the cells themselves
degrade over time, some of the cheaper, crappier cells can lose half their
conversion efficiency in only a couple of years.

http://www.oatleyelectronics.com/solar.html might be another interesting
source of panels

And where are you getting 80Ah deep cycles for 120 bucks? Sounds dodgy,
though if there real, i want some *grin*

I'd suggest you seriously consider doubling, or tripling the 10W.
Say 4W for the AP, You're regulating 13.8v down to 9, so to get 4 watts of 9
volts, you're going to burn say 1/3 of the power in the regulator, so now to
get 4w, we're up to 6w input. By 24h, that's 144Wh used. Add a fan, say
150 to make rounding easy. Going by your 5 hours of sunlight, which isn't
really that unreasonable for a non-tracking system, you now need 30W of
sunlight input total to get close to break even!! 300 bucks at my 'cheaper'
price. 120 of your DSE cells -- ie just under 500 bucks of cells (and a
whole pile of work to put it together). And I still haven't taken the 19
hours that we've only got 75% energy efficiency of the battery yet!! So now
we're talking 40W to break even. 50 or 60 would be a not unreasonable
safety margin -- remember the solar cell -> regulator to battery link also
has loss. Not only that, the battery has to "catch up" if there's a dark
day -- in winter there's some days where there's not enough energy going
into the cell to activate the solar regulator, so you get zero power that

Trust me, you don't want to under-spec the solar cell (I've experience in
powering a remote weather station using deep cycle/solar). And for
price/performance, the 60W bp solar for 600 or so are probably not
unreasonably priced anymore.

For half reasonable batteries (ie, near the 75% efficiency) I'd suggest
you'd be looking at more around the 200 buck/~80Ah

So system performance, we're probably talking about 8-900 for
panels/batteries, and another hundred would be easy to spend on regulators &
assorted bits and pieces to glue it together.

For a thousand bucks, you can buy a lot of electricity, and usually it's
cheaper to pay an electrician to install a powerpoint on a roof, and pay the
10c/kwh or whatever it is thesedays (ie at 150w/day, that's what,
1.5c/day... 5 dollars 50 a year. So throw the owner a coupla hundred bucks
to cover power, throw 4-5 hundred for a "difficult" electricians job
(probably usually a lot cheaper than this, as there'll be power nearby
anyway). System performance -- more reliable, cheaper...
Solar's really only useful if you need to stick it on a mountain where
there's no power.


From: "Will Lanigan" <chooken@jibbed.cjb.net>

hehehe don't use mega mega good panels such as the BP ones, Solarex is
cheaper - but still very exp.
Dick Smith had 1V, 250mA panels for sale that are very small ($4.10 each
something). 16 of these provides 16V @ 4A peak, with an 80Ah deep cycle
battery ($120ish) - Cover with a sheet of glass for extra protection
free - always glass lying around)
The houseing for the AP etc is pretty cheap also, you could even use an
old filing cabinet and just water seal it?

(I will be running regulators etc. to prevent the panel from putting a
full 16V in to the battery... the deep cycle batteries should however be ideally charged at around 13-13.5V to ensure they get full charge - 12V devices
will take up to 13.8V before they get shitty so there isn't a problem with this. I will also be running a regulator to the AP which only allows 9V @ 300mA or something along those lines)

The panel will be connected directly to the battery terminals, and the AP
will also have a feed directly to the battery terminals. There will be a
regulator of some description on each positive feed so as not to allow
excess power to travel through which could cause undesirable effects.

I am doing all calculations assuming the panels only put out 2A for... 5
hours per day? I think that is a fair estimate for quality sunlight. That
is 10A charge to the battery per day. I am coincidently performing my calculations for an airport too :P
I will call it 300mA just to make it easier. 300mA will pull 7.2A from the
battery per 24hr cycle. The battery is recieving 10A from only 1/2 the potential power at a reduced number of hours, most typically there would
be more power generated from the panels and also more sunlight hours than i have performed my calculations on.
I have done my calculations on what could be the 'worst case scenario' and
we still have 2.8A per day to spare. The battery alone could support the AP for the length of an 11 day eclipse... which i dont think will ever happen :P

So yeah, i think the power consumption is fine. Even if we move to a 500mA
AP then that is only 12A per day... this could be solved by throwing another couple of these small panels on to boost the amperage to the battery.

Expected Costs:
AP = whatever you want to use?
BOX = scrap metal - whatever u can get your hands on $30?
PANELS =$65.76
GLASS = free
ASSEMBLY = time :)

TOTAL (no AP) = $217.76

I will email you the file with some of my calculations off list ok :)
(file is a little to large to mass email :P)

From: "Julian Featherston" <julian@concentrate.com.au>

Heres a post I sent to the list bout 2 months ago re: a remote solar setup Send me a copy of your plan so far, and give us a yell if u've got electronic questions.

I cant help feeling it will be cheaper and easier to get 240V to the places we need, there may be some effort involved but it beats the major effort of solar.

The problem is the AP heavy power consumption some as high as 500mA @9-12V (remember if u supply a 9-12V AP with 12V the AP will drop it to 9V internally but it is still pulling the same amperage and thus wasting power)

The Apple Airport looks good for low power - I used that as a figure
(~290mA @ 12V)



Hi All
I couldn't help being a bit sceptical of cheap solar powered AP's (wind looks good on tall buildings / hilltops). So here's my attempt to cost a solar power supply for 12V 290mA.
All power would be drawn from cheap deep cycle batteries so one would charge to Max~13.6V. Average charge would be ~13V say we use cheap Vreg for 12V supply, average drain would be 13V x .29A = 3.77W/hr

It's easier to work in days so
3.77W/hr x 24hr = 90.48W/hr per day
charge/discharge loss ~16%, say battery power is used half the time, so battery loss is equivalent to 8%.
90.48 plus 8% loss = 98.347W/hr per day

Efficient solar battery charger/regulators are expensive and very difficult to substitute, but suitable ~75% efficiency models can be sourced for ~$100.
98.347 plus 25% loss = 131.129W/hr per day.

The people from Murdoch reckon a cell will average in Aus.
winter about twice its rated power output in watt hours per day.
131.129 x .5 = 65.564W

So one would need
60 or 80 Watt rated cell $659 or $799
Charger/regulator $100
Deep cycle batt $120
Output Vreg & bits $15
Mounting etc. $10

60W solution $904
80W solution $1044

Plus: AP; Ants; Cable and mounting etc. whole node could easily be done for under $2000 I'd reckon the 60W would be fine in better than average locations, the 80W probably mandatory for sheltered locations. This is cheaper than I'd thought but involves pretty big cells (60W - 502x1105mm) (80W - 502x1456mm). I'd guess the cells in this shot would be 25W each (wonder if he's tops up the charge occasionally?) [*Note* Each cell is 40 watts.]

SOURCE: http://acre.murdoch.edu.au/refiles/pv/text.html
"The daily energy output from PV panels will vary depending on the orientation, location, daily weather and season. On average, in summer, a panel will produce about five times its rated power output in watt hours per day, and in winter about two times that amount. For example, in summer a 50 watt panel will produce an average of 250 watt-hours of energy, and in winter about 100 watt-hours. These figures are indicative only, and professional assistance should be sought for more precise calculations.

Trackers are used to keep PV panels directly facing the sun, thereby increasing the output from the panels. Trackers can nearly double the output of an array (see figure 7). Careful analysis is required to determine whether the increased cost and mechanical complexity of using a tracker is cost effective in particular circumstances. A variety of trackers, which will take about 10 panels, are manufactured in Australia."



PART 3: Mailing List

External linkWGsolar@wireless.org.au
External linkSubscribe


PART 4: Idea flinging etc - Be Creative and Constructive


yeah throw some ideas here :)
saves putting them all on the list which is now becoming quite cumbersome considering the volume of messages we now have regarding solar power

In other news my inital emails have attracted a lot of attention from the group, a lot of attention has been directed at flaws in my plans. Good!
Bloody oath we need more of this, i'm glad i turned a few heads with my inital plan, now lets all work together and come up with several designs, one who's principle goal is efficiency, and another design who's soul constraint is cost.

Thanks to whoever edited the pages and formatted the links :D
I'm thinking it was Robert Foster - ta
Don't go unnoticed, if you make some changes feel free to post in the section (Part 3) and boast or whatever you want.

Small wind turbines may be worth investigating too, I'll see what I can dig up. Could provide supplemental charging to solar panels - works at night and on cloudy days. May mean less expenditure on batteries. Can be homebrewed. And as the APs will be nice and high, there'll be plenty of wind. Blades would need to be transparent to 2.4g and/or mounted away from antennas, eg higher or lower on the same mast.
Disadvantages are moving parts are less reliable than PV, and wind loading on masts would need to be allowed for.

Seperate idea: if we are remote adminning these APs it might be worth having a bright light viewable from a distance which could be remotely switched on, as a basic "i'm here!" signal. I would also love to have some remote monitoring for the batteries and power system.

On the other hand, KISS. more devices = more power = more panels = bulko more $$ The 500 mA (6 watts@ 12V) WL400 looks ideal.

Paul van den Bergen:
I came across a subsidy scheme for Ti-based solar panels a while ago... thought you might be interested... trying to find details, will post more when I do...

Chooken: Thanks for the feedback and new innovative ideas Clae, i will also try to do some investigation in to this also. On another note, Paul, sounds nice - i believe there was a government subsidy for getting your whole house wired for solar (i think the most it cost the user was $8000 - govt. payed rest? or something like that)... what we could possibly look at doing is getting a house done on solar then perhaps dismantling the many solar panels and diving them amongst many nodes - this could be a fesible solution in the future :P
dodgy but feasible hehehe

Clae: The subsidy is 50% of the panel cost up to a limit. System must be specced and installed by a licensed installer. Must be primary place of residence. Further details avaliable from www.ata.org.au

Andy: Should the additions here be cronological? Anyway... I have a one of those WL400 access points that Clae mentioned. It is currently up a pole but will be at gound level in the near future. I also have a great Nokia AP that I think might be better suited as it does bridging and DHCP out of the box and doesnt require the additional servers on the ether like the WL400 does.

TysonClugg: The [Solar page seems to be what would be ideally suited to a mailing list - ask Steve to create it for you. If you can, see if the info that is in Solar can be transferred to here.

Clae: External linkhere's a link to a low-powered AP howto, using HostAP and NoCatAuth

Back to WorkingGroups

Version 5 (current) modified Tue, 03 Jul 2007 23:12:22 +1000 by graybeard
[EditText] [Spelling] [Current] [Raw] [Code] [Diff] [Subscribe] [VersionHistory] [Revert] [Delete] [RecentChanges]
> home> about> events> files> members> maps> wiki board   > home   > categories   > search   > changes   > formatting   > extras> site map


 Remember me.

> forgotten password?
> register?
currently 0 users online
Node Statistics