an outdoor LED lighting project

idleprocess

Flashaholic
Joined
Feb 29, 2004
Messages
7,197
Location
decamped
Wow. I haven't posted to CPF in years.



I've not been so interested in flashlights, but since I acquired a house I'm quite interested in fixed lighting - particularly LED lighting since it's getting more powerful and the lumens/watt and lumens/$ figures keep getting better. I guess I was inspired by the local 7-11's swapping all their outdoor floro and metal halide lighting for LED modules (I can't figure out who they're using - wish I could source some of those for a reasonable price).



I've never liked the driveway lighting for my house - it came with a dual incandescent flood fixture that produced two distinct spots



I happened across 3-up Rebel MCPCB's sometime in the third quarter of 2010 and immediately realized their potential for lighting bright enough to be quite useful singly or in groups. I obtained 6 while I waffled on how to power them - switching power supply plus resistors is cheaper, but less efficient than a fixed-current driver and requires significantly more fabrication.



I then came into possession of a pair of handy constant-current drivers and the plan was ready for execution.



I fabbed up some fixtures:

fixturemockup.jpg


Triple-emitter is AAd to the holder, which is AAed to the mounting plate (2.00 x 0.125 aluminum bar); the holder is really a tiny aluminum sample container with a glass lid that wasnt terribly easy to frost with sandpaper (not pretty, but these were destined to be out of sight)



I hung some metallic raceway:

eaves.jpg


Out of sight from more than five paces, so don't give me any lip about them not being lined up; the other wire is pre-existing



After some frustration with popped solder joints, ripping a few solder pads off the MCPCB's, and the distinct pleasure of soldering something attached to the ceiling while standing on a ladder, I scaled back the design to 4 fixtures. I'm quite pleased with the results:

outdoors1.jpg




outdoors2.jpg


Cellphone pics are all I have. Sorry.



Each fixture should be sipping about 3 watts for 12 watts net. Figure in some driver inefficiency and I'm looking at maybe 15 watts total power consumption. Fewer absolute lumens than even one of the halogen lamps it's replacing, but the light is uniform and where I want it rather than two spots on each half of the driveway. Color is a tad cool, but not too far off from some of the CFL's in the neighborhood.



Learned a few lessons:
  • My soldering skills need some work. Some serious work. The bigger the blob, the better the job is OK for through-hole components on a breadboard with generous clearances, not so great on surface-mount and other devices requiring far more attention to detail with a bit more heat sensitivity than a 2W flameproof resistor
  • Strain-relief is important for solder joints, especially when your solder pad is little more than a trace on a single-sided PCB - I ruined 2 modules and a third is decidedly ... marginal (but working - nothing some RTE or liquid electrical tape won't fix whenever I decide to fix it ForReal)
  • A dual-gang box is a lot smaller when you're on a ladder trying to stuff 2 power supples and a bunch of wires with slack connected with wire nuts into it than when you're test-fitting it at a bench
  • Arctic Alumina has a tight hold, but it has limits
  • Measure twice, three times, a fifth time, then screw something down

Not entirely sure how much it cost me in the end. Emitters were $17 each, drivers were $20 each, and the holders were $9/20. I didnt track all the other sundries (wire, aluminum bar, wire clamps, raceway, J-box, way too much solder, RTV, and too many trips to Home Depot). Obviously more than just getting a better off-the-shelf fixture from Home Depot...



Ill probably do some more of these again elsewhere, but will redesign the fixtures a bit. I imagine Ill just use thermal grease between the elements and just screw the assemblies together. Ill also find a way to strain-relieve the wire - maybe with a grommet in the holder or cable clamp outside.



I'm also interested in the Cree XP-G 3-up MCPCB's that LEDSupply recently released; they had previously only had XP-E in cool white, which I did not feel was acceptable for outdoor lighting.



I'm sure that some will point out that drivers and LED's are available on dodgy websites like DX for cheap. Trouble is that those sources are not known for high quality, timely delivery, nor could I count on their drivers being CE/UL recognized.



Anyway - I appreciate any feedback.
 

fmzip

Newly Enlightened
Joined
Nov 23, 2011
Messages
32
I like your idea.

Could you post a pic of the can attached to the house?
 

idleprocess

Flashaholic
Joined
Feb 29, 2004
Messages
7,197
Location
decamped
Hmmm... Took a photo, but forgot to post it.

08cba13a.jpg


This thread was posted before the forum crash several months ago, so the replies were lost.
 

idleprocess

Flashaholic
Joined
Feb 29, 2004
Messages
7,197
Location
decamped
Those are cool, but if I were to have many, it'd be costly. I'm looking at the 1.5oz jars OHX1 here. They're sending a sample my way so I'll see if it might work.

The small jars from SciPlus are cheap ($5 for 20 or $0.25 each) and made of aluminum - makes for good thermal conduction to the ultimate heatsink beyond the container.

Odds are the lids for those glass sample jars are made from something other than aluminum like steel. Lack of a uniform flat surface will make interfacing with a heatsink difficult, and odds are there's some rubber coating to aid with sealing the bottle.
 
Last edited:

OLED76

Newly Enlightened
Joined
Jul 25, 2012
Messages
10
I need some input, I'm trying to decied between the Utilitech ES303L42-10 3-Light, 360-degree motion detector area flood light [1929 lumens ] or the Lithonia OFLR 9LC 120 MO BZ similar 3-head design [1,719 lumen ] . My question, does this 200 lumen make that much of a difference. i'm trying to repalce a 300w Mercry vapor light in my driveway/side of house. I have no living on that side of the house so blinding the nieborgs is not an issue. I think I found someone who posted a photo of the Lithonia but I can not find any night photos for the the utilitech. what would be nice is to have over head night photos of each the LED floods and spots so we could compare light output coverage
 

AnAppleSnail

Flashlight Enthusiast
Joined
Aug 21, 2009
Messages
4,200
Location
South Hill, VA
I guess I was inspired by the local 7-11's swapping all their outdoor floro and metal halide lighting for LED modules (I can't figure out who they're using - wish I could source some of those for a reasonable price).

Synergy Lighting. That's a link to their press release. I do not know who will sell you a couple of those units though.






[1929 lumens ] or [1,719 lumen ] . My question, does this 200 lumen make that much of a difference?

You won't be able to tell these fixtures apart by their brightness. It takes a 40% change in lumens to be noticeable in a side-by-side comparison, and that's a very slight difference. You need 4x the lumens to look twice as bright.
 

idleprocess

Flashaholic
Joined
Feb 29, 2004
Messages
7,197
Location
decamped
Synergy Lighting. That's a link to their press release. I do not know who will sell you a couple of those units though.
They look to be a distributor/integrator rather than manufacturer.

I've more or less resigned myself to rolling my own since LED lighting seems to have two binary poles - cheap w/ no frills (and usually no safety ratings, no warranty) or premium w/ all the options selected (and a price to match).

There's a local supplier of most things LED lighting nowadays that stocks far more than they list on their website. I suspect that they could source some modules if I truly had a desire to buy rather than fabricate.
 

TheExpert

Newly Enlightened
Joined
Nov 8, 2011
Messages
194
How much did this project cost? I am looking into setting up some lights like this for my parents quarry, this is exactly what they need.
 

idleprocess

Flashaholic
Joined
Feb 29, 2004
Messages
7,197
Location
decamped
How much did this project cost? I am looking into setting up some lights like this for my parents quarry, this is exactly what they need.

Apparently my reply to this was lost...

LED's were $17 each (now $12/each @ LED Supply) x4
Drivers were $20 each x2
Housings were $9/20 or $0.45 (now $5/20 or $0.25) x4

Don't remember the specifics on the other materials - wire, solder, Al bar, wire straps, J-box. I'll guess $40. My project cost was just under $150.

With better planning, it could be done for less:
  • Wire raceway was unnecessary in this case - a wire strap every 24" would have been adequate
  • Lighter gauge wire would have been sufficient - 18 ga is probably overkill for 350mA @ <24V
  • A single driver might work better than a pair of low-wattage drivers - I gather that Meanwells have excellent performance in addition to lower costs

Shortcomings:
  • No strain relief on the LED's - My original plan for 6 fixtures was undone by ripping solder pads off of the LED's due to lack of strain relief
  • Insufficient heatsinking - At perhaps 8 square inches effective surface area per ~3W fixture, I would not be be surprised if the LED's eventually fail (although they survived the hottest summer on record in 2011).
 

TheExpert

Newly Enlightened
Joined
Nov 8, 2011
Messages
194
yes, these will only be security lights. wow, that is pretty darn cheap I might have to do this. the only catch is they want it to be solar powered, so that will be much more expensive im guessing any body know the price of what would be needed for this setup?
 

idleprocess

Flashaholic
Joined
Feb 29, 2004
Messages
7,197
Location
decamped
yes, these will only be security lights. wow, that is pretty darn cheap I might have to do this. the only catch is they want it to be solar powered, so that will be much more expensive im guessing any body know the price of what would be needed for this setup?

Constant-current DC/DC converters are cheap, so you could run it from 12/24V DC from a battery pack simply enough - ie Meanwell sells some 350mA constant-current drivers for ~$5 each.

Solar power for such an installation likely will be appreciably more expensive than the fixtures themselves because of the need for overprovisioning. If you reliably need light 365 nights a year, this will be an expensive endeavour; if an occasional outage due to low sunlight is acceptable to you, it will be a lot cheaper.

I've gathered that on an average day, solar panels will produce about half of their faceplate watt rating times hours of "noon-equivalent" sunlight. As such, you will need to double the apparent panel requirements. If you expect your LED's to consume, say, 100 watt-hours a day, you need to gather some 200 watt-hours a day during the winter ... if your winter sun is 3 hours a day, then you need 66.67 watts of panel (go with a 75W if available). This will give you a good deal of margin during the summer when there is more sunlight.

At a minimum, your battery bank will need to be sized (in watt-hours) to meet demands of the LEDs' routine operating hours plus some margin. Again, using our example of 100 watt-hours expected demand, oversizing of the battery will make it last a good deal longer as well as provide some margin should your panels not collect so much light on some days. I'm not sure what recommended depth-of-discharge for the lead-acid family of batteries is, but I do recall that deep-cycle batteries (designed for regular deep discharges) have discharge-current limitations.

If you demand high reliability from this system, it starts to get expensive. Your PV & battery array sizes increase considerably as you add the ability to compensate for longer periods of low (or no) sunlight - especially at higher latitudes where the difference between summer and winter sunlight is more pronoounced. The system becomes large enough to gather and store sufficient energy in a single day to power the LED's for days on end. As you buy more margin for low sun, your efficiency of utilization goes down since most days you'll only need the design watt-hour capacity. An alternative could be a small wind turbine (since the wind is often blowing when the sun isn't shining), or an automated generator.
 

TheExpert

Newly Enlightened
Joined
Nov 8, 2011
Messages
194
well that is a lot to take in all at once lol, it does not need to be 100% reliable so to speak, and honestly I think my mother just wants it to look good. And we do not run the quarry in the winter so that is a non issue, thank you for the in depth review of that. Ill prolly have to read it again to absorb it all.
 

idleprocess

Flashaholic
Joined
Feb 29, 2004
Messages
7,197
Location
decamped
Solar panels have nominal (peak) ratings in watts, which I refer to as "faceplate" ratings. When new, clean, at noon, on a cloudless day, with bright sunlight, and pointed directly at the sun - they should produce this output. In reality, change any of the parameters and they don't make this number.

If you follow the advice of some solar energy guides - use panel spec sheets, average noon-equivalent sunlight hours, 90% inverter efficiency, then a 75% "fudge factor", you end up over-estimating what real-world calculators such as the NREL's PV Watts calculator shows. For example, a 4kW system in my area, with about 5 hours' a day noon-equivalent sunlight should theoretically produce about 7300 kWH (4kW * 5 * 365). Per PVWatts, I'm going to see closer to 5400 kWH. If you're just using the DC side, you can skip the 10% inverter losses, although you lose energy charging batteries. Note that these numbers are apparently based largely on solar collector projections (not PV panels), which likely included regular cleaning and ideal (static) alignment for the latitude.

Here's an insolation map showing average hours noon-equivalent sunlight in the continental US.
 
Top