Solar outdoor Led light

[rocky79]

Hello,

I am designing a Solar outdoor light and I want your input on the design aspect. I realize you guys are expert in this field.

The outdoor light should be reliable and lives up to it's promise of meeting the following criteria:

It must run for 10 hours and maintain constant light output for at least 8 hours throughout the night.

LED:
I am going to use one cree MX-3 LEDs per light running at 0.68 watts/0.2Amps
for 10 hours that's 2000mah.

Battery:

I am trying to select a quality battery for the system that should last at least 2 years.
I found the Rayovac hybrid rechargeable 2000mah ( realistically I want more mah but that should do it for now). The manufacturer claims it can be recharged 1000 times. So that's 1000/365 =2.7 years. The question is does that 1000 times assuming your draining the battery all the way down or to a certain percentage from it's fully charged state?

Charging system:
I will have to you use a Nimh charging chip..that's work in progress

Solar power:
I selected a 3w monocrystalline panel. Assuming 6 hours of light and panel outputting half it's claimed wattage so 1.5 watts.
6hours x1.5 watts= 9 watts-hours

Does anyone you have a better recommendation for a battery system or LED selection?

Thank you

 

[SemiMan]

I predict the light will be out often.

Unless you are in Arizona, your panel is likely too small unless you have a power point tracking switch mode controller.

Your battery needs to be 5-10x the size if you want it to come on every night.

 

[deadrx7conv]

I have a lawn light that has a 1-watt LED. Your design is workable. Your battery choice isn't. Build it around a single 18650 battery and your .7w:3w LED to solar panel ratio should work fine. The other option is 3-4 AA batteries.

I do recommend that you buy one of these to see/test if your end-design will work for your requirements.
http://www.homedepot.com/Outdoors/h...splay?langId=-1&storeId=10051&catalogId=10053
The older version of the HomeDepot light above uses 3x 1.2v rechargeable AA NiMh batteries. The newer version uses a single lipofe or LiFePo? 3.2v battery. Read the box carefully as I prefer the 3xAA better than the single LiPoFe? battery. My Duracell/Energizer NiMh batteries have made it thru year 2. I don't ever bother using the batteries that the lawn lights came with(always upgrade since the generic nimh/nicad batteries are pathetic). So, your 3 year battery life is definitely there with quality batteries. My AA batteries are the pre-charged low-self-discharge and not the ones with mega mah rating.

Lowes has a similar lights too:
http://www.lowes.com/pd_329359-59179-SS5C-P3-BK-1_0__?productId=3316670&Ntt=led+solar&Ns=p_product_price|1&pl=1&currentURL=%2Fpl__0__s%3FNs%3Dp_product_price|1%26Ntt%3Dled%2Bsolar
http://www.lowes.com/pd_329358-59179-SS6C-P2-BK-1_0__?productId=3316668&Ntt=led+solar&Ns=p_product_price|1&pl=1&currentURL=%2Fpl__0__s%3FNs%3Dp_product_price|1%26Ntt%3Dled%2Bsolar
http://www.lowes.com/pd_329355-59179-SS4C-P2-BK-1_0__?productId=3316666&Ntt=led+solar&Ns=p_product_price|1&pl=1&currentURL=%2Fpl__0__s%3FNs%3Dp_product_price|1%26Ntt%3Dled%2Bsolar

Once you test how "1" of the above lights works for you, you'll be able to size your battery/panel a little better.

The Homedepot light uses a simple LED star mount. Its 20mm. I've upgraded a couple of the 1-watt Homedepots lights with pre-mounted XPG's for double the light output. I'm considering dropping the power to 3/4 watt to have better run times. The MX-3 LED should perform equivalently.

Also your solar panel needs to be mounted toward the sun. For example, all my panels aim about 45 degrees pointing a tad west of compass south. This means that the lights work great on the left side of the house, work somewhat OK in the front and back, and work horribly on the right side. You'll have to make up for 'location' using bigger batteries or solar panels in some areas, and smaller in prime areas. You can also re-aim the solar panel monthly to optimize its charging.

 

[hellokitty[hk]]

New version Sanyo Eneloops are nearly identical to the Rayovac LSD's except that they're rated at 1500 cycles; this includes partial recharges. I haven't read any specifics about what Sanyo's partial recharge ratings, but they said that it includes both full and partial recharges to realistically simulate normal use. Also, it's 1500 charges to 80% original capacity.
Sanyo also has their Eneloop lites rates at 2000 cycles, but their down to 1000mAh; not worth it IMO.

Your calculations are wrong.
A single NiMH's nominal voltage is 1.2v, and the eneloops are rated at 2000mAh, but you need 3.4v and 2000mAh or nearly 7 watt/hours for an entire night assuming high efficiency and a full charge.
I would think that you'll want at least three batteries.

Good luck this sounds like fun.

 

[SemiMan]

The poster said it "should be reliable". A 3W panel will not be reliable and the battery is grossly undersized ... both even for Arizona. If you can eliminate using it reliably from November to February you can reduce your requirements quite a bit. 3 NiMh .... guessing you are direct driving your LEDs so I will assume they are running at less than 1W most of the time.

A 3W panel is only 3W under absolutely ideal conditions. Most of the time it is much less. That has been taken into account by using 1.5W, but reality is the situation is likely worse. That 3W is under ideal loading conditions. Since it is likely ran either direct into the batteries or "maybe" using a linear regulator (hope so), then best case you may only get 2W out of that panel ... but let's say 2.25W or 75%.

In the winter, with very very few exceptions, there is no where in the world where you get 6 hours of equivalent sunlight. It is more likely properly pointed you are getting somewhere between 2 1/2 and 4.

Best case, you just may be getting 9 watt hours into your batteries, but you may also be getting 4 or 5.

Of course, NiMh batteries are not 100% efficient in charging so you have to account for those losses as well (LiFeP04 have very low losses though).

Those hours of equivalent peak sun are averages. Some days you may get 50% more. Some days you may only get 10-20% of that amount ..... and that could happen for a week on end (unless you are in the desert).

Batteries also have degraded storage capacity at low temperatures which should be taken into account.

If I knew where this was located I could tell you fairly accurately what you needed but best guess is the same. Double your panel, and make the battery 5-10 times bigger and the light will stay on most of the time.

 

[rocky79]

Thank you all for your reply. Please find my calculation table link below. I am planning to use A123'S batteries 2 cells AHR32113 Cylindrical Cells in Parallel. ( The hard part is getting hold couple of those batteries for a prototype.

https://spreadsheets.google.com/pub...QlJEQ1FwVWI5ZGc&single=true&gid=0&output=html

You're input is appreciated. Thanks

 

[deadrx7conv]

If you can't get the cells, why even bother?

Your location?

I do alright with my lawn and spot lights in RI/Mass area. The only issue I had this winter was snow, and not the lack of solar power or battery capacity. I was too lazy to wipe the snow off the panel and lost light for a few nights here and there. I also lost a few to my plow/snowblower, snow weight, and snow movement while melting. As the snow melted(glacier movement), it'd knock over, snap or crack the mounting pole for either the panel or the light.

What will the lights be mounted on/in? intended purpose? your electronic background?
Number of these 2/3rd watt lights that you're building? mounting location?

A protected 18650 would be a great choice. It'll shut off the battery if voltage is too high or too low.

You also need to research LED drivers depending on your battery voltage available and voltage needed by your LED.

There is not much between DIY failure or success. Many lean on the failure side caused by budget or jumping into something that they have no idea about. You jumped from 1.2v 2000mah store bought cheap hybrid battery to an unavailable pricey LiFePo4 cell. And, you haven't mentioned much about photo-sensor light switching, timers, drivers, charge controllers, housings, heat sinking, weatherproofing....

Don't forget that you're losing power wasted by the driver, sensor, and charge controller...

BTW, the spreadsheet is useless. Formulas don't work in the real world as well as many think. You'll need some trial and error for your location, build quality/efficiency, mounting location, and is why I recommended test lights from your local hardware store. Buy some batteries, a solar panel, a charge controller, a sensor/timer switch, an LED, mounting/waterproofing, lens/reflector, and your time.... could cost you more than a $40 test light form Homedepot. And, you won't know if your finished project even works.

 

[SemiMan]

Judging by your spreadsheet, I get the impression that perhaps this is a plan for a commercial product? If so, let me know, it may temper my responses. For one, I know there are no off the shelf MPPT controllers that will be suitable at such power levels.

In terms of "field trials", this is one time where calculations are generally more meaningful than just putting it outside and claiming it works. The solar energy variation from year to year in terms of worst case conditions are very large, so you may get lucky, put it outside, and think that it works, however, you may have just been lucky in terms of not having long periods of low sunlight.

Your spreadsheet is very basic ... If I know more about what you are trying to accomplish, I can provide more assistance.

 

[rocky79]

If you can't get the cells, why even bother?

Your location?

I do alright with my lawn and spot lights in RI/Mass area. The only issue I had this winter was snow, and not the lack of solar power or battery capacity. I was too lazy to wipe the snow off the panel and lost light for a few nights here and there. I also lost a few to my plow/snowblower, snow weight, and snow movement while melting. As the snow melted(glacier movement), it'd knock over, snap or crack the mounting pole for either the panel or the light.

What will the lights be mounted on/in? intended purpose? your electronic background?
Number of these 2/3rd watt lights that you're building? mounting location?

A protected 18650 would be a great choice. It'll shut off the battery if voltage is too high or too low.

You also need to research LED drivers depending on your battery voltage available and voltage needed by your LED.

There is not much between DIY failure or success. Many lean on the failure side caused by budget or jumping into something that they have no idea about. You jumped from 1.2v 2000mah store bought cheap hybrid battery to an unavailable pricey LiFePo4 cell. And, you haven't mentioned much about photo-sensor light switching, timers, drivers, charge controllers, housings, heat sinking, weatherproofing....

Don't forget that you're losing power wasted by the driver, sensor, and charge controller...

BTW, the spreadsheet is useless. Formulas don't work in the real world as well as many think. You'll need some trial and error for your location, build quality/efficiency, mounting location, and is why I recommended test lights from your local hardware store. Buy some batteries, a solar panel, a charge controller, a sensor/timer switch, an LED, mounting/waterproofing, lens/reflector, and your time.... could cost you more than a $40 test light form Homedepot. And, you won't know if your finished project even works.

You make good points here. I haven't mentioned all the project details yet however it's going to have a light sensor/led driver/charge controller. As of now i am going to make a prototype and see how things go and then make the necessary adjustments.

As far as experience I am an electrical engineer. I have designed a board for a switching LED driver driving an XPE LEDs and other linear drivers.I know I am loosing power from all the board components. I already mentioned that in the data sheet.

Formulas are useless but you need to start somewhere and I think that the spreadsheet I made gives me a rough estimates of I should expect.

Many lean on the failure side caused by budget or jumping into something that they have no idea about. You jumped from 1.2v 2000mah store bought cheap hybrid battery to an unavailable pricey LiFePo4 cell.

BTW what's wrong with jumping around to find the perfect battery? If i can't find this battery, i will look for a similar one. What's the big deal. Is this a no no? You have some good advices but you're coming off as too strong.

Thanks !

 

[rocky79]

Judging by your spreadsheet, I get the impression that perhaps this is a plan for a commercial product? If so, let me know, it may temper my responses. For one, I know there are no off the shelf MPPT controllers that will be suitable at such power levels.

In terms of "field trials", this is one time where calculations are generally more meaningful than just putting it outside and claiming it works. The solar energy variation from year to year in terms of worst case conditions are very large, so you may get lucky, put it outside, and think that it works, however, you may have just been lucky in terms of not having long periods of low sunlight.

Your spreadsheet is very basic ... If I know more about what you are trying to accomplish, I can provide more assistance.

Hello, thanks for your response. My goal is to make a robust outdoor architectural solar light that last longer than most ones available in the market today. I am aiming at 10 hours run time with at least 8 hours of consistent light output. My intention is to make a working prototype. It may turn into a product but as for now I have many hills to climb.