Help bringing light to the less developed world

[wingnutLP]

Hi,

I recently moved to Sierra Leone in West Africa, in fact I drove here from the UK but that is another story! I have always been convinced that LED is the lighting technology to help bring light to parts of the world where it has been previously unavailable, it is bright , rugged and uses very little power so it is an obvious choice for off grid use.

There are a range of cheap REALLY badly made flashlights here; some like traditional flashlights and some area lights usually featuring arrays of LED's. They are badly made, flimsy, don't last long as they are fragile and use disposable batteries which is terrible for the environment.

I think that we can do better and I think that with a good enough spec you could produce something that could really change the way a lot of people live their lives.

I see the perfect light as:

- Cheap

- Solar or renewable powered I don't like the wind up things but the rip cord ones don't seem bad, solar is probably the obvious choice.

- Featuring either an in built battery pack that is rechargeable or possibly the ability to put in one or more standard rechargable NIMH batteries. Maybe it could have the option of running of say regular AA (or other) disposables or from an AA rechargable cell that the renewable power source charges. The rechargable would have to be capable of being topped up constantly from a solar cell and would also need to be capable of thousands of recharge cycles.

- Able to act as an area light attached to the roof of a hut so that the family can cook etc after dark. this probably means having the light on a flying lead with the switch mounted on the light itself or the battery pack.

- In a perfect world the system would be modular so you could add to it or replace bits as necessary.

- Several brightness levels but not necessarily regulated, I am not sure what this would do to the cost though.

- The battery and light would need to be capable of being run for say 6 hours per day each day but it wouldn't matter if this was from a full charge to completely empty.

The way I could see the product being set up is with a small solar cell on the roof with a lead running to the battery pack and a further lead or leads powering the LED's. It would need to be cheap enough that it was affordable to people with very little money. Although it needs to be cheap, at the moment peoples only options for lighting rooms are very expensive generally. A generator or a big solar pannel and car batteries combined either with very innefficient incan bulbs or very expensive ennergy saving ones. Cost for such setups range between $50 and $1000's depending on how much light you need.

I am doing some research at the moment but everyones thoughts on what the spec or components for such a light would look like would be really good.

Specifically; if anyone can suggest what cheap wide angle LED or LED's would be appropriate and also what battery capacity or type would be suitable then that would be great.

Thanks for any help

 

[Marduke]

Someone already attempted to make a cheap light for off-the-grid use for developing countries.

http://www.bogolight.com/

 

[sparkysko]

Aside from bogolight, I'd consider something like this:

http://www.boaterbarn.com/powerfilm_aa_charger_2_4_30349_prd1.htm

and then just hang a light with a diffuser from the ceiling.

One of the problems with doing solar and rechargeable, is that you have two options for Nimh/nicad. You can either charge at a very slow 'trickle charge' rate with no extra electronics needed (Requiring 2 days to recharge a battery) or you can fast charge with a bigger panel, but you need a charge controller.

Perhaps 3-4 x D cell nimhs @ 10 amp hour being trickle charged by a 500ma-1 amp panel, and then try to keep your power consumption down to about 1/2 watt for the LED that is using a diffuser.

 

[wingnutLP]

The sun night/bogolight is one of the ones that I have come across but it doesn't seem right to me. It is too complex and too expensive IMO.

The use of a hard solar cell is definitely good as they are cheap but it needs to be moved in and out to charge and use. A modular approach with the solar cell outside all of the time facing the correct direction would be better I think. With a flashlight like this you need to put it outside, where do you put it and how do you ensure that it stays in the sun and doesn't get stolen?

I also suspect that it is brighter than it needs to be. You can get by with the light from a zebralight attached to the roof on low.

It sounds like NiMh is the way forward, they don't need to be LSD unless that gives a significant improvement in overall lifespan of the battery.

Is it possible to trickle charge in one day using a slightly larger pannel or would you immediately need electronics?

 

[sparkysko]

Is it possible to trickle charge in one day using a slightly larger pannel or would you immediately need electronics?

No, you can't trickle charge in one day. Let's assume 1/10C as a conservative trickle charge rate. The peak sunlight can't exceed 1/10C, and as the day goes on, the light output and energy received is far less. If you perfectly matched it, you can probably trickle charge in 2 days, but not less, because the suns intensity is not constant for 10 hours, and if you sized it to charge in 10 hours, you would overcharge the cells during peak sunlight.

Imagine a bell curve for the amount of sun energy hitting your panel throughout the day. If the peak is 1/10C or less, then there's no way an average day can total 1C unless you were in Alaska or something where the sun doesn't set in the summer.

http://www.thecaptainsmemos.com/wp-content/uploads/2009/06/BellCurve.jpg

To get around this problem, in my house, I use a very large NIMH battery that can be trickle charged at a very high rate (still takes several days to fully charge my battery bank). I use effectively a 12 volt, 14 amp hour nimh battery that I charge at 0.5 amps.

1C = draining/charging battery in 1 hour. 1/10C = draining or charging battery in 10 hours. With a AA, 1/10C would be ~200mA, with a high capacity D cell, 1/10C would be about 1 amp.

Therefore, if you increase the storage capacity of your battery banks amperage, you can also charge at a correspondingly higher rate.

That is the approach that I would consider taking, if you don't want to deal with charging electronics. The extra capacity also acts as a buffer during days of less sunshine.

Lets say you were to build a simple 1 cell light (with a boost driver). One AA cell charged at 1/10C I would expect to get on average 1.2 amp/hours a day (1.4 watts), if you were to use one D cell, I would expect to get on average 6 amp/hours a day (7.2 watts), of course you'd need to size up the panel to match, but these numbers would be without exceeding the capabilities of the battery as far as 'dumb charging'.

 

[Linger]

How cheap? How much work will you (or whom-ever) do?
Standard led's are cheap, dealextreme $4 for a xpeQ4. Multi-level boost drivers are also very cheap ($3), simply interupt power supply (clicky switch) to change modes). The emiter can sit in a copper end-cap - heatsinking plus physical protection. The emitter with-out any lens / reflector is an excellent flood (like zebralight). - add a pop-bottle-cap that fits over the copper end-cap and holds a lens / or reflecor and you've converted your flood to a spot light.
Another great option is a p60 drop-in - the pill includes reflector, emiter, driver, all in a metal protective case with decent heatdisipating properties - the pill drops nicely into the copper endcap as well, extra protection thermal mass and modularity.

The internals of your rechargable light system, 1 hour assembly time + $14 + batteries + solar panel (2 travel consumer panels)

any powerled emitter on star ($3-8)
wire leads soldered on
Copper end-cap with three holes drilled out ($1)
star into end-cap, endcap onto heatsink, screw the lot together ($1)
leads to driver ($3-4)
quick disconnects to battery pack (0.50)
battery pack, 4AA nicd/nihm 2s2p, has quick-disconnects
overcharge / over discharge / short-circuit board protects pack (controller circuit from batteryspace.com, customizable)
battery disconnects and is plugged into solar array. Or remains connected and driver/emitter isolated during charging with push-button switch.

 

[sparkysko]

I would avoid doing 2S2P. NIMH in parallel becomes tricky because of resistance differences causing current to favor one cell over another. I have a 10S7P pack and once I start doing appreciable current charge levels, it gets ridiculously hot.

 

[HarryN]

I remember seeing a presentation about an organization that was working in India about a similar need. In their case, they were focused on indoor lighting, as the typical local indoor lighting was a sort of "fat fueled lamp".

In that case, the village had one central solar panel set, which charged up 12V SLA batteries (still one of the cheapest and most robust setups today). This makes a lot of sense, as it builds "community" as well as takes advantage of economy of scale, and important part of any real program.

To "bring home" the power, each family had its own battery that they carted home. Local women ran the business and "sold" a charge IIRC, as part of the concept was to empower women, especially women whose husbands had died.

Once you are home with a 12V SLA, then of course your options are pretty open. You can of course make up an electronics "driver" for indoor light, but perhaps given the locations and need for simple reliability, a simple RGBW led string in series with a couple of ohms resistance is more practical - and perhaps more interesting as a conversation piece. You really don't gain much efficiency going to a driver over a resistor once you hit 3 - 4 junctions with a 12 volt setup.

 

[yellow]

"less developed" like:
- difficult to obtain batteries?
- having money to purchase batteries?
then battery powered lights are crap

those wind-up lights, when the movement is of "normal" quality, are quite good + easy to learn to handle them.
Enough light for 99 % of chores one can think of.
"Endlessly" running, imho way better than solar powered
(note: I am typing on haldheld lights)


PS: NOTHING can beat a gas/petrol lamp for area lighting, especially in that environment
Only con is the "open" flame and its dangers

 

[wingnutLP]

Thanks for the pointers.

If the price point and quality are feasable then I would be looking to manufacture a few prototypes and then pitch them to NGO's after field testing...

I heraa what you are saying about not using batteries but they are available most places. the problem is that people use disposables and they are as you say not really a good alternatibe to kerosine or even candles.

HarryN I like the idea, I will try to see if I can find that paper as it sounds really interesting.

What I have seen a LOT out here are cheap torches which have had the 3 LED bulb unit removed and put on a flying lead. The body of teh torch is attached by the door as a light switch and power pack and the bulb is hung of the roof. It is quite effective but resorts to terrible quality alkaline batteries

Thanks again all...

Alex

 

[wingnutLP]

Is there an idiots guide to building your own flashlight anywhere? I have had a look in the stickys but it is all either pretty advanceed or based on modifying what you have rather than starting from scratch...

Thanks

 

[HarryN]

This might not be what you are thinking of exactly, but I have also seen a presentation for a "natural gas" based arrangement for lighting.

Basically, they made a long trench about 3 wide x 2 deep x 10 meters in size. They filled it about 2/3 full of moist waste - garbage, sewage, manure, etc, then topped that with wet straw - about 1/2 meter thick.

On top of the straw was a 200mm dia perforated plastic pipe, which led out of the trench over to the village. The pipe was then further buried with wet straw, and finally covered up with wet dirt and a stone mound.

As decay happened, it provided more than enough "natural gas" to feed a cooking / lighting gas out to the small village. They actually ducted this out to each house (in solid plastic pipe) and it came in through the floor.

The villagers would "turn on" the gas by removing a small rock from on top of the pipe, and then lighting up the incoming gas. "Off" was accomplished with the same stone.

Remarkably, it all seemed to work, although I have no idea how well, and certainly it lacks a certain amount of safety, such as back fire prevention.

Interstingly, backfires can be nearly eliminated by placing a "wet trap" - similar to a normal kitchen "goose kneck" wet trap found in most any developed country bathroom and running the gas through this at each use point. KEEP WATER IN THE TRAP.

Anyway, it is about as simple and cheap as you can get, and even better, the locals can keep making more capacity relatively easily.

A larger version could be made to feed a small generator, but I would have to really think about how to size such a "plant".

 

[HarryN]

Is there an idiots guide to building your own flashlight anywhere? I have had a look in the stickys but it is all either pretty advanceed or based on modifying what you have rather than starting from scratch...

Thanks

It does not have to be advanced, but doing it simple and remote means of course - trade offs.

Body Materials
- what can you find locally made from aluminum
- pipe (round)
- square tube

LEDs or bulbs
- What can you find locally or order
- Lumileds K2s are very robust - which can be handy
- Osram also makes nice LEDs (in Europe) so perhaps more available
- Cree makes nice LEDs - perhaps brighter? Perhaps not as robust to abusive situations
- Most likely, the locals have a source for light bulbs of some kinds

Machining
- Is there a machine shop around?
- If not, it will take some time with a hack saw, but it is possible

Wire
- Everyone has wire

Recipe
- 3 x NiCd or NiMH in series (size depends on availability)
- wire to LED through a switch
- If it is a K2, you are fine
- If it is anything else, put a 2 ohm resistor in there

Reflector
- find an old car and use the headlamp reflector

Good luck

BTW - interesting trip blog.

Harry

 

[wingnutLP]

- 3 x NiCd or NiMH in series (size depends on availability)
- wire to LED through a switch
- If it is a K2, you are fine
- If it is anything else, put a 2 ohm resistor in there


BTW - interesting trip blog.

I will be buying the parts for the test light from the UK and having them shipped to my company there, they will then DHL the parts to me in Freetown so I will have access to whatever parts I need.

Can you explain why you say use 3 batteries is that because 3.6 volts is about right to direct drive most LED's?

Also why the 2 ohm resistor?

If I was to then charge the batteries direct from a solar pannel could I just have a switch so that either the LED is in circuit and the pannel is not or vice versa?

I seem to remeber reading some time ago that if there is no light then solar panels sap power is that still true?

Thanks for the comment on the blog it was aa really amaxing drive down here!

Thanks,

Alex

 

[waddup]

a rechargeable 2xAA size, plastic light with a small solar panel in the side would be ideal. leave it in the sun all day once a week and have useable light all week.

 

[HarryN]

I will be buying the parts for the test light from the UK and having them shipped to my company there, they will then DHL the parts to me in Freetown so I will have access to whatever parts I need.

Can you explain why you say use 3 batteries is that because 3.6 volts is about right to direct drive most LED's?

Also why the 2 ohm resistor?

If I was to then charge the batteries direct from a solar pannel could I just have a switch so that either the LED is in circuit and the pannel is not or vice versa?

I seem to remeber reading some time ago that if there is no light then solar panels sap power is that still true?

Thanks for the comment on the blog it was aa really amaxing drive down here!

Thanks,

Alex

There are sort of 4 ways to drive an LED

a) Direct Drive
- Just wired in series to the cells
- Relies on resistance in the cells slowing down the current into the LED
- For alkalines it is fine. For a good quality NiMH, it can be a bit over the top when fully charged
- Yes, 3 x alkaline or NiMH are a good starting point, and it is really handy to have a light that can run on primary or recharge type cells (flexible)

b) Resistored
- Similar to direct drive, but adds a little resistance to make sure things stay under control
- I have several lights like this and they work fine (for me)
- 1 ohm might work as well for that setup, depending on the LED you pick
- A perfectionist will say they are not constant brightness for the full battery charge - they are correct, and I am ok with that

c) Boosted current regulation
- Battery voltage is substantially less than the LED needs, so the regulator does the work for you

d) Buck current regulated
- Battery voltage is substantially higher than the LED needs, so the regulator does that work for you

Take a look at the Philips Lumileds site and download the K2 TFFC data sheet. Read it 4 Xs (no, I am not kidding), then read in the LED section under the K2 information thread. If you post K2 questions there, I will happily answer them, as will others.

There is nothing magical about the K2 over for example the Cree parts, but it is an excellent example of a well flushed out and documented part. Once you understand it, you will understand most any LED brand. Pretty much everyone in the industry started with the PL data sheets, and designed their data sheets to have the same info and format, so once you read theirs, the rest are easier. There are also a wide range of optics and reflectors available to use with it, and using existing optics is key to success, as they are a PITA to develop from scratch unless you really know what you are doing.

There are also very detailed applications examples, etc, so you can pretty much just copy and paste.

Future Electronics has quite a range of support products for the K2 for reasonable prices in volume.

As a practical matter, it is not so easy to do things cheaper yourself than to buy them on the open market, so you are pretty much doing this for other reasons, which is perfectly fine. I do that all of the time.

 

[sparkysko]

I have a 3v 200ma solar panel that I toy with. I removed the blocking diode. When it is night time, the solar panel sucks 0.8mA out of the battery at a constant rate. I lose ~2-4% of my daily charge to the reverse current.

The alternative is to suffer the voltage drop from the diode. Depending on the application, it might or might not make sense to use a diode.

 

[wquiles]

I have a 3v 200ma solar panel that I toy with. I removed the blocking diode. When it is night time, the solar panel sucks 0.8mA out of the battery at a constant rate. I lose ~2-4% of my daily charge to the reverse current.

The alternative is to suffer the voltage drop from the diode. Depending on the application, it might or might not make sense to use a diode.

The other alternative is to use a schottky diode. I have a several 60V 2A schottky diodes, and I routinely solder two of them in parallel for my P7 direct drive Mags, so that non-flashaholics can't destroy the relatively expensive P7 if wiring batteries in reverse:

With the two schottky diodes in parallel I measured something like 0.22 volts drop - that might be low enough to be practical in your application and prevent the drainage of your battery.

Will

 

[wingnutLP]

Ok so I am thinking for prototype 1:

- One of these:
http://www.ultraleds.co.uk/05027-pw14-luxeon-star-white-lumens-p-1377.html
or possibly the lower output one

- this solar panel
http://www.selectsolar.co.uk/pics/MPT3.6-150 (100mA at 3.6V).php
- a three battery caddy
- three NiMh aa cells
- a three way swith so the light is eitheron or charging
- a diode

Does that sound like a good start?

 

[John_Galt]

Well, we're looking at the thirw world. It's going to be dusty, dirty, rainy, hot, humid, monsoon-y, etc. So this will have to be relatively robust, if you want to be able to provide long-term help.

For a cheap housing to hold the electronics, switches, batteries, maybe look at those military surplus angle head lights? Relatively robust, water resistant, easily modifiable. If you took some time to modify them, I think that if you were to take out the bulb holders and contacts, you could drill through the top of the body (where the batteries are, I think there is a bulkhead of plastic in the top of these light, I can't remember, tho), and use that as the switch, battery housing, and figure out a way to


Continued later. Gotta go.
**EDIT**

Sorry about leaving you hanging, I was changing class periods.

"[sic] and figure out a way to" cut a small diameter hole in the top to run wiring, to charge the cells, and provide power to the separate light modules elsewhere.

Alternately, pvc pipe could work, as it is cheap, easily cut and modified, and easily sealed with resealable end caps. This would probably work better, because you could purchase pipe that is slightly larger than the diameter of the battery holder, drill a small hole in the one (permanent) end cap, to run wiring, mount the switch on top with epoxy, and either run additional wiring to the batteries to trickle charge them from a solar panel, or allow easy removal for external charging, under regular charging conditions.