Fixed 12v lighting with XR-E

[Veto]

Last night I took the head off my D-Mini and used it as a candle and it was very bright and usable room lighting. Got me to thinking and doing a little bit of research. So now I'm thinking of doing some garage lighting via 12v (solar panels to come in a future iteration).

Would it be possible to drive a single XR-E star via a BuckToot @ 350ma from a 12v lead-acid battery? ($6.30 + $14.99 per emitter)

Because I don't know, could I run 2 or 3 XR-E's from the same bucktoot to reduce the 'non-photon emitting' cost in the bucktoot? I simply don't know how to tell if that's feasible or what additional electronics would be needed.

 

[crewcabrob]

Hi,

I'm thinking of the same thing for the garage. I have a huge garage, 1000+ sf, and I hate to turn on the 600w florecents every time I go out there.

Every time I take my new fenix out there with me, I think about modding 3-6 new Crees to provide illumination. Likewise, I would run it off of a solar cell and a 12v battery. Heck, I'm even thinking about running landscaping lighting off of the same type of arrangment. I hat those cheap solar/led looking lights people stick in thier yards. Any one come up with a mod for that?

Thanks for starting the thread, I hope some one can help is out. I thought about buying the $13 crees flashlights from DealExtreem and just modding the body and wiring to accomplish what I need to have done.

Do you think 6 of those little lights could run off a solar cell about 1 foot square charging a 12v battery for ever?

Thanks,

Rob

 

[Veto]

Rob,

I -think- (20+ years ago that I last tried this) that you'll need around 4v/350ma and that will will be in the neighborhood of 1.5watts to drive it. Here is my math for someone to correct:

4v * 0.350ma = 1.4w

Now a 12"x12" panel might come in around 5w @ 16v (12v panel) which would be around 300ma or to close for comfort probably. Again, here is my math for someone to correct:

5w / 16v = 312.5ma

So I -think- you would be at a power deficit which would eventually end up with a dead battery and no photons if you ran it 12hr on battery and 12hr of charging...but take all of this with a grain of salt until someone much brighter comes along and corrects my homework.

Hmm, to continue my bad math exercise: Given 5 hrs of good light in the US you might get a upside of 312mah * 5 = 1.5ah and with your 350ma draw you could expect to run your emitter for around 1.5ah = 1500mah/350 = 4.28hr (4h15m).

More bad math homework:

312ma * 5hr = ~1.5ah from panel to battery
1500mah / 350ma = ~4.28hrs

So your 5w 12x12 panel in 5hrs of sunlight might gather enough to give your 4:15 of runtime. Not something you would want to have on all the time, but you could bank it (store it in the battery) and use a switch when needed.

...Awaiting my homework grade...

-veto

 

[2xTrinity]

I'm thinking of the same thing for the garage. I have a huge garage, 1000+ sf, and I hate to turn on the 600w florecents every time I go out there.

Is the idea of the LEDs is so that it would act as a "night light", or basically low level illumination to see your way around by so as not to cycle the fluorescent as much? That's a pretty good idea. I wonder if it woudl be possible to rig something like that to be connected to a motion-sensor controller -- we have one like that in our garage with a 150W halogen floodlight for when people are unloading from the car, replacing that thing with an LED would use a lot less power and get the same job done (enough light to see by -- the halogen is overkill). In one room we have a T8 fixture, and lightswitch is on the opposite side from the door, so I just put a little photocell controlled nightlight with some 5mm LEDs up in an outlet in the ceiling -- lights up the room enoguh to see by when the fluorescent is off.

If those are Magnetic T12 fluorescent in there, you could probably save a lot of money and improve the lighting in the room (read: better color rendering, no flickering) by going to 320W worth of Electronic T8 fluorescent with efficient reflectors. (5 two-tube shoplight fixtures)

 

[crewcabrob]

Thanks Guys!

Yep, all I need is the night light to grab beers out of the fridge. The kids leave thier bikes, rollerblades and scooters in the middle of the floor and I hate the noise and flickering of the huge shop lights.

I tried to put a motions sensor on the light switch for teh garage, but the lights were over the operating range of the switch and the whole thing got very hot when in operation. I took it out and put it in the storage room where there are two incans haning from the ceeling.

I need the shop lights for washing the vehicles in teh winter time, tinkering in the garage and restocking the "cooler". (if you know what I mean) But I would love to have lights on all the time out there, espscially if it is free light from solar.

Veto,

OK, I follow your math, which means it is right, or I'm really clueless. No opinion on it either way at this point, which might mean I am clueless. I'm thinking a larger cell might be needed to run them 24 hours a day. I'm guessing at least 16 on the batteries and 8 on the pannel itself while also charging the batteries. I can see already that my best bet is to add both a motion and a photo sensor. Either that or cover the whole roof with solar panels.

I have been looking on the internet for solar panels that are under $80. I need to cap my project at $200 or so. I'm still looking at either 3 or 6 Cree emitters running off of a motorcycle size 12v battery. What I was thinking about was the drop-in reflector and regulated 3w CREEfrom Dealextream. I have the battery and cable, and can fassion my own mounting hardware from junk in the garage. All I need to do is get a panel ($?), the emitters ([email protected]) and any regulation and charging of my circut. Now add the photoelectic sensor and the motion sensor. Think it is doable?

I'm guessing that my panel will need to be about 15x36"?

Rob

 

[Veto]

Rob,

Be assured...something is wrong with my math...just awaiting someone to correct/grade it!

You might consider what I'm looking at: Harbor Freight 45watt kit and I'm trying to do some math to see how many XR-E's it can sustain @ 350ma a pop.

That's what prompted my initlal post, I'm trying to figure out how to drive the XR-E's. A constant current driver seems to be the best way to go, but I might be able to drive them with just a inline resistor at the LED off a run from the battery which would be cheaper, but something seems dicey about driving a XR-E star via just a resistor and a wire to a lead-acid battery (or charge controller).

Again, I'm not sure so I'm looking for information/direction.

 

[crewcabrob]

Veto,

I went out and looked at the kit, (HF 45 watt). That would more than drive my needs for lighting the garage. But, I would also get driven out of my neighborhood by the neighbors. :thumbsdow I think they may see that as an eyesore. I hate to admit it, but my house is in one of "those" neighborhoods where being "green" is a sign of weakness. Muffy and Chas might spill thier martinis driving past my house. :naughty:

Three of the 15 watt panels would be overkill for me, I really only need 1 of them, but the rest of the kit looks exacatly like what I want.

When you get some real feedback, I will be watching. I love the idea too, but after you start adding it all up, it seems cheaper to just pay the electric company.

Rob

 

[Veto]

True about paying the electric company, but we had a pretty serious storm last night, which started the entire thought process, and I sure would like to have at least the ability to drive some lights when the power company is having electron routing issues.

 

[Veto]

Some additional math for review:

Using the Harbor Freight 45w (3x15w panels) solar kit:
15w @ 16v = ~0.9a * 3 panels = ~2.7a

What can we expect to "bank" in a normal 5hr solar day?
2.7a * 5hrs = ~13.5ah

What does it take to run an XR-E @ 350ma for 8hrs?
350ma * 8hrs = 2800mah = 2.8ah

What can we run out of our expected intake?
13.5ah / 2.8ah = 4.8 lights for 8hrs or 4 + insurance.

 

[winny]

If those are Magnetic T12 fluorescent in there, you could probably save a lot of money and improve the lighting in the room (read: better color rendering, no flickering) by going to 320W worth of Electronic T8 fluorescent with efficient reflectors. (5 two-tube shoplight fixtures)

That, or use something even better and post-eighties like T5's while you are at it.

 

[scottv]

I'm working on something similar.
An all night light using solar panels, gel cell and as many Cree XR-Es that I can sustain.
Started with a 12V MR16 triple Cree bulb but my odd small panels can't quite keep up. A single Cree with no optics spreads quite a bit of light so I'm trying that next.

Last year I successfully used a 21 LED MR16. Runs from dusk until dawn every night and gathers enough charge during the day. Just modifying that setup to work with the Cree. The 21 LED bulb is very blue, clumpy and not so bright.

 

[crewcabrob]

Hi,

I'm starting to see that this project is going to be a labor of love. Money wise, it isn't making much sense.

After $200 on the solar panel, I still have to buy emitters and regulators, I'm guessing I will have spent at least $300. My current electric bill runs only about $112 every month all year long, and that is with 4500sf of house and another 1000sf for the garage. And I'm all electric except for the water heater and tow gas fireplaces. There are 5 of us at our house. Yeah, the house was built right.

By my estimations, running a couple cheap off the shelf LED pucks will cost me a few pennies a month.

While I think most of us here have a little "fireman" in us, I'm not sure I can really build this with good conscience. I like the idea of having a contingency light source for storms and such but that is also why I bought a bunch of flashlights. True, I could use the existing infastructure of the solar panel and battery to run other things in case of a total outage, but my power has never been off for more than an hour or so. Even then, I have a generator to run the fridge and some other items.

Maybe I am looking at this wrong. A new Fenix l2d-ce on low can run for 50+ hours. 9 lumens is a lot in total darkness and more than enough if placed directly over the beer fridge in the garage. A cheap solar battery charger should be able to charge at least two batteries in two solar days. What I could do is just mount a remote box on the wall for easy access, set up the remote switch, and let her run, I could change the batteries in the remote box I place on the wall when they get tired. This may be the way to go for me.

I will keep watching, I still love the idea of dedicated solar circiuts powering all sorts of things inside and out.

Rob

 

[sortafast]

You can always bypass the drivers and just go strait with resistors. Be a heck of a lot cheaper, although not recommended by Cree. Also this would allow you to run the LED's at a lot lower current so it would save battery juice. If you just want to use them for a night light style illumination, you could get it down below 100mA and still have a fair amount of light. You could also go with blue LEDs as your eyes pick up blues alot better at night and it wont wreck your night vision. I have plans to build some blue leds into some fixtures for night lights as my house gets crazy dark inside at night.
Also in regards to the battery, I don't know too much about motorcycle batteries, but you will want to make sure that its a deep cycle battery as you would kill a regular batt in no time with the constant charge and discharge cycles. But yeah, I totally agree that its not super cost effective to do the whole solar thing, at this point in time and technology anyway. Its more about piece of mind and the green factor. If money was not object and I was guaranteed return on investment when we sell our house, I would totally spring for solar powered lighting. But such guarantees are hard to come by.

 

[Brlux]

The problem with your math is you are counting the amp hours for 1 LED emitter and comparing it to the 12V battery amp hours. The battery is 12V and the emitter is less than 4V. If you used one emitter with a buck converter, the converter should only draw about 120ma from the 12V battery to be able to drive an LED at 350ma. Your math does work for driving 3 LED's in series with a current limiting resistor.

I have been running a custom solar powered indore LED lighting system for several years now. I have done a few threads on my system and some sources for cheep solar panels.

I have been thinking of creating a microcontroler based solar charge controller specifically designed for driving LED lighting. It would probably have several PWM channels and would scale down output as the batteries got lower and probably put time limits on the highest output levels. I have many other ideas that would be cool and useful, but I have to see how good I can improve my programing skills. Is seems that many could benefit from such a controller.

 

[Veto]

Thanks Brlux, that's actually what I was hoping! That means I can drive ~3x the number of XR-E's I thought I could off of 45w of solar. So my math was wrong here:

350ma * 8hrs = ~2800mah @ 4v = ~933mah @ 12v

Therefore:

13.5ah (@12v) / 0.933ah (@12v) = ~14.46 XR-E's @ 350ma for 8hrs each.

 

[crewcabrob]

Thanks for the great thread! I missed the discussion on your solar/LED project.

I went to Ebay and looked up some of the panels you mentioned. Funny how I also missed them in my search???

When I look at the numbers to really run solar anything, it looks like you have to spend a lot to get serious about running anything other than LEDs. Kind of sad in a way.

I will keep dreaming right now. I hope that some day I am running my house (new house) on some alternative fuel. I watched a great program on hydrogen the other day and it's use in fuel cells. That may be the ticket for storing the energy for a home.

I hope others keep the thread alive, I really don't want to give up hope.

Thanks,

Rob

 

[2xTrinity]

That, or use something even better and post-eighties like T5's while you are at it.

Sad as it is, 90% of the fluourescent fixtures and lamps at Home Depot/Lowes around here are 60-70CRI Cool White T12 driven by flickery magnetic ballasts to produce 65 lumens per watt. The other 10% are 85-90CRI T8 fixtures at 100 lumens per watt, yet only the former is openly labelled as "energy saving"

I'm starting to see that this project is going to be a labor of love. Money wise, it isn't making much sense.

After $200 on the solar panel, I still have to buy emitters and regulators, I'm guessing I will have spent at least $300. My current electric bill runs only about $112 every month all year long, and that is with 4500sf of house and another 1000sf for the garage. And I'm all electric except for the water heater and tow gas fireplaces. There are 5 of us at our house. Yeah, the house was built right.

A very cheap way to drive Cree Emitters is to simply get cheap switching power supplies (the lightweight transistor based ones, not the heavy magnetic wall warts) used to charge cell phones or PDAs. For example, I have a 5V 400mA charger from an obsolete Cell phone driving a Cree that I use as an accent light. It will drive anything with a Vf of less than 5 at a regulated 400mA.

There are some chargers I've seen that go up to 2A that would be good for powering 4 Crees in parallel with current-limiting resistors on each. You can get chargers like this for a few bucks from Wal-Mart and other places, and run them directly off of a switched outlet. You could even hook them up to a UPS for coverage in event of a black-out a lot cheaper than a solar setup.

Even still, the best perspective in terms of dollars is still probably is to upgrade your fluorescent to some sort of electronic tubes, and leave one running when the others are off, or use a separate 20W CFL as a night light bulb for the same amount of power as the Cree emitters. Wiring up the Crees though is definitely fun, and a lot more interesting from a "techie: perspective than plugging in a CFL.

 

[winny]

Sad as it is, 90% of the fluourescent fixtures and lamps at Home Depot/Lowes around here are 60-70CRI Cool White T12 driven by flickery magnetic ballasts to produce 65 lumens per watt. The other 10% are 85-90CRI T8 fixtures at 100 lumens per watt, yet only the former is openly labelled as "energy saving"

It's impossible to get T12 tubes here without ordering them, and single-phosphor tubes stopped selling in the late-eighties/early nineties here in favor for tri-phosphor tubes as they are way more economic and better in just about all ways. Still, T8's are more popular for domestic use than T5, for no good reason (for the end costumer that is).

 

[Mash]

A very cheap way to drive Cree Emitters is to simply get cheap switching power supplies (the lightweight transistor based ones, not the heavy magnetic wall warts) used to charge cell phones or PDAs. For example, I have a 5V 400mA charger from an obsolete Cell phone driving a Cree that I use as an accent light. It will drive anything with a Vf of less than 5 at a regulated 400mA.

There are some chargers I've seen that go up to 2A that would be good for powering 4 Crees in parallel with current-limiting resistors on each. You can get chargers like this for a few bucks from Wal-Mart and other places, and run them directly off of a switched outlet. You could even hook them up to a UPS for coverage in event of a black-out a lot cheaper than a solar setup.

Can you please elaborate a little on this?

1- how do u tell which kind of cell phone charger you have? I have lots of Nokia chargers sitting around with spec of 5v and 400mA, and some different ones also, so maybe i can start experimenting! Do you connect the LED directly to the charger or do you need resistors etc?

2-My understanding of the technicalities is limited, but im trying to learn, so isnt the 5v output too much for the Crees which are in the 3.5v range?

3- How many Crees can you run with one of these chargers? Am i correct that if you run them in parrallel, you distribute the 400mA between them, eg with 4 LEDs each gets 100mA?
Thanks in advance for your info!

 

[Veto]

Okay, I've located a 12vdc 1a wallwart to play with and will be ordering 3x XR-E's from Dealextreme either tonight or tomorrow. It also seems I'm going to need some resistors and from playing with this resistor calculator (plugged in 12v source, 4 volt drop across LED and target of 500ma) it seems I would just need a single 1.0 ohm resistor @ <1watt capacity inline before the first of the 3 XR-E's that'll be connected in series.

That seems not to jive with what I've seen with people using a resistor between each emitter in other pics, have I figured something wrong?

+12v --> 1.0 ohm --> XR-E --> XR-E --> XR-E --> -12v

Could it be this simple to drive the 3 XR-E's @ ~500ma?