We've moved!
213 E Ermina Ave, Spokane WA, 99207


IMPORTANT!! -- Open hours have been temporarily suspended to prevent the spread of COVID-19.

Thanks for supporting your
community 501(c)(3) makerspace

Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
garage LED lighting project
09-13-2013, 01:03 AM,
garage LED lighting project
I'm remodeling my garage and turning it into a shop. The insulation and drywall has been put in, and I have taken delivery of a brand new Haas mini mill:

[Image: drywall1s.jpg]

[Image: drywall2s.jpg]

[Image: brianmills.jpg]

The lighting array I intend to install will consist of thirty custom made fixtures, each containing six Nichia 219's for 180 leds in total, powered by three 95% efficient meanwell 320w drivers, hooked to a dimmer.

The appeal of the Nichia NVSL219AT led is that it is 92 CRI, so it reproduces colors very accurately, including red, which leds typically render very poorly, and also the particular version I have is a quite neutral 4500k color temperature. Typically, high CRI led's are only offered at lower color temperatures such as 2900k to 3500k, which is a much yellower light, similar to a low wattage incandescent bulb. The Nichia 219 is quite special indeed. It's maximum recommended operating current is 1500mA, and it will approach 3.7 volts at that current.

Each of the 320w Meanwell drivers will power 60 leds, in 12s5p configuration, with each led seeing roughly 1340mA. The fixtures will feature a frosted diffuser that has 70% optical efficiency. The overall output of the luminaries after diffuser losses will approach 40,000 lumens and should achieve a lux of over 640 at full brightness. Total system power will be approximately 930 watts. (180x4.7 /.95)

The fixtures themselves will be made of rectangular aluminum tubing, 1"x3"x.125" in size.

[Image: yic0.jpg]

[Image: 69ie.jpg]

[Image: qc2m.jpg]

To fit the diffusers to the tubing I will need to mill a couple of small channels along the 1" side. The angled portion at the base of the diffuser will wedge into these channels, and I will use screws to secure the diffuser to the tubing.

[Image: 5ghx.jpg]

[Image: gmuw.jpg]

I have done some number crunching to get a rough idea of the thermal performance of these fixtures: the Nichia 219A has a thermal resistance of 7 °C/W typical; the MCPCB stars upon which the leds sit add 3-5 °C/W, achieving 12 °C/W worst case in combination. The profiles provide 19 in² of vertical surface per led plus some horizontal surface and also I will leave the ends somewhat open, but at 19 in² of vertical surface the heatsink will have a TR of 5°C/W, so the total additive TR is 17 °C/W. If the led consumes 5 watts, 4 of which are heat, the total temperature rise over ambient will be 68°C. The hottest it's going to get in my garage is 30°C, so the maximum Tj temp will be 98°C. I doubt I will actually be driving these suckers at full power very often, so they would typically be much cooler. At 600mA per led (1.6W of heat) there will be 360 lux in the shop space and the led junction temperature in 30°C ambient would be 57.2°C.

The maximum allowable Tj of the led is 150°C so this fixture should keep it well within operating specs. I will have to take some actual temperature measurements of the setup to more precisely determine the Tj of the led, but I am pretty sure these fixtures will allow for adequate cooling in the worst case operating scenario.

What I have left to do is model up some "end caps" and mounting brackets for their attachment to the aluminum extrusions, then do all of the machining and assembly, then all of the wiring and installation of the fixtures themselves in the shop.

I have actually been idle for a couple months now and need to get this project going again!
09-13-2013, 05:22 AM,
RE: garage LED lighting project
Well dude thanks to you for sharing that pics in according to me it has made of aluminium mostly people have used it in their homes and buildings especially for roof purposes.........So please share me more pics about it buddy????????You can do it buddy?SmileSmile
09-13-2013, 08:02 AM,
RE: garage LED lighting project
Looks great!

awesome Haas too!

"If you didn't build it, you will never own it." - Barton Dring
09-13-2013, 09:33 AM,
RE: garage LED lighting project
You could improve the efficiency of the fixtures with even a basic reflector.

Also it seems that using light sensors you could shut down some of the leds to decrease power consumption when the doors are open and providing light. A lot of commercial spaces plan for that but then cut the sensors because they are too expensive. Then you end up with an over lit space that is consuming too much power. An example of that would be WSU's student rec building. They designed it with daylighting in mind, but they cut the sensors. During the day you will find all of the lighting circuits on consuming power because that s what people do. Automation could give them a consistent light level all day.

Sorry for geeking out a bit. I was a commercial lighting designer at one time.

09-14-2013, 12:49 AM, (This post was last modified: 09-14-2013, 01:06 AM by Brian_H.)
RE: garage LED lighting project
(09-13-2013, 09:33 AM)Photonic Wrote: You could improve the efficiency of the fixtures with even a basic reflector.

Please elaborate.

A reflector in combination with a diffuser would probably be beneficial in a design where the emitters were spaced several inches away from the diffuser, such as in a recessed can design; beyond that scenario I don't think there is much need for the combination in fixed lighting, as LED's are already directional light sources, having a directivity 2θ ½ of 120° or so. Since my diffuser wraps 180° around the LED's, all of the light hits the diffuser. A reflector would introduce losses to this system.

70% efficiency is not too shabby for a diffuser. Without the diffusion there could be some undesirable glare. The diffusion also prevents shadows, which are the enemy in a working environment. The ceilings and walls will be painted white to help the overall efficiency of the system as well as the diffusion. I could also paint the floor but I don't think I will. I only hope that the glare will be reduced enough with the diffuser I have chosen, as the LED's are not going to be recessed in the ceiling or fixtures.

Another option would have been to point the LED's up at the painted ceiling, but with the amount of dust in a shop I did not want to deal with cleaning them all the time.

(09-13-2013, 09:33 AM)Photonic Wrote: Also it seems that using light sensors you could shut down some of the leds to decrease power consumption when the doors are open and providing light.

The array will be controlled by a dimmer and variable from about 0% to 100%. I will probably not have the garage doors open while actually machining anything as my shop is in a residential neighborhood and I wouldn't want to disturb the neighbors with my air compressor and milling machine noise, even though where I live there are no ordinances regarding this.

(09-13-2013, 09:33 AM)Photonic Wrote: Sorry for geeking out a bit. I was a commercial lighting designer at one time.

You would have had to produce some ray traces and rant about angle of incidence or something to geek out.
09-17-2013, 06:16 AM,
RE: garage LED lighting project
(09-14-2013, 12:49 AM)Brian_H Wrote: You would have had to produce some ray traces and rant about angle of incidence or something to geek out.

I love this conversation!

When can we see the finished project?

Brian, did you find any good sources of information on led lighting while designing your project?

30% seems like a lot of light to lose. I wonder if there is a more efficient alternative...

Geoff, you mentioned including sensors in a setup like Brian's, but also said that WSU found them prohibitively expensive. Has the cost come down a lot since then? How hard would it be to design such a setup to maintain a constant light level in changing conditions?
09-20-2013, 02:54 AM, (This post was last modified: 09-20-2013, 03:21 AM by Brian_H.)
RE: garage LED lighting project
Hi David.

I want to get this project done soon... I am nearing the completion of the modeling process so hopefully it won't be long before I start machining it.

As for learning things, this project has taught me a bit more about the heatsinking requirements of fixed led lighting, specifically how to determine the led Tj temperature, and about various diffuser materials. I also looked into existing led lighting products to see if it would be considerably cheaper to do it myself. It is, and DIY led lighting is currently the bleeding edge of the technology.

I have some basic knowledge to build upon, as I have been pretty much obsessed with flashlights and leds since about 2011, though I haven't built much... except for a spotlight which features a 600 watt 28v airplane landing bulb. It's pretty good for walking the dog and getting the mail and stuff. Big Grin

(09-17-2013, 06:16 AM)david tremaine Wrote: 30% seems like a lot of light to lose. I wonder if there is a more efficient alternative...

There are some diffuser materials that have manufacturer claims of higher efficiency. "Makrolon Lumen XT" is one, which is available in several different ratios of diffusion to transmission. There is also "plexiglas diffuse," which claims "good diffusion" and 88-92% transmission on most types, and one which has 60% transmission and "very high diffusion." Then there is "Diakon Frost," which is claimed to achieve nearly 100% diffusion at 80% light transmission or so. I will undoubtedly get ahold of these plastics in the future and test them out, but in the mean time I went with a readily available diffuser profile from "Klus," specifically the square "GIP" profile. Somewhere I saw that it is 70% transmissive, but I don't know how good the diffusion is compared to the aforementioned materials. I would rate it as having moderate diffusion; hopefully it keeps the glare down.

Your average metal reflector has substantial losses as well. A quality rhodium coated reflector will be between 70% and 80% reflective, though "protected aluminum" coatings can be around 90% reflective but are not as durable or as stable over time as rhodium. Silver coatings are the most reflective but are not stable in air and tarnish immediately. Then there will be slight losses through the glass lens, between 1% and 6% or so depending on the glass and whether or not it has anti reflective coating applied.
09-20-2013, 09:00 AM,
RE: garage LED lighting project
Sweet Jesus! (sorry, god)

Never seen a flashlight with a shoulder strap before. Who needs a gun to defend their property when they have the power to instantly blind their assailants? Seriously, that thing is like a sawed-off photon shotgun!

Have you tested its run time on high? Or does it overheat too quickly? I could imagine even with the fans it'd have a hard time staying cool.

This led lighting business gets complicated quick, apparently. I was curious, Brian (and anyone else), if in your research, you've come across particularly good information resources - websites, books, local vendors, that might have helped you come to the understanding of the subject you have now?

Thanks for showing us your walkthrough so far. I'm curious to learn how your experience of your workshop space changes when you get these installed. When I did a trial run in mine, replacing fluorescent tubes with leds, I was surprised and delighted to find that, though the overall light level didn't change much, the quality of the light was much, ah, cleaner - that is, I felt like I could focus better on objects at hand. Not sure why this would be - a whiter spectrum, surely, but I wonder, too, if it had to do with the rapid fluctuation of fluorescent bulbs.

Keep us posted!
09-25-2013, 11:59 PM,
RE: garage LED lighting project
The runtime on high is about 15 to 20 minutes. The front of the bulb gets very hot while the back stays comparatively cool. The bezel doesn't melt but it does start to smell strongly like plastic and rubber and is on the verge of plasticizing after about 8 minutes or so on high. The bulb life is probably only a dozen hours or so because the bulb is overvolted, so when it comes time to change the bulb I will put a silicone O-ring between the bulb and the bezel. This should eliminate any worry of melting plastic, as the bulb will not be in direct contact with any.

Most of my information on lighting comes from candle power forums, or at least stems from there initially. Lighting does get complicated when you start to get into optics and the electronics. Also the manufacturing aspects of a given design require plenty of consideration....
03-02-2014, 10:26 PM, (This post was last modified: 03-04-2014, 01:20 PM by Brian_H.)
RE: garage LED lighting project
Boy it's been a long time!

I'm pretty much done fussing with the shop at this point. The air has been plumbed, work benches and shelves have been built and put in place, and all the scraps and garbage are out of there. Now I can focus on my projects....

[Image: shop_jan_1s_zps1efffa05.jpg]

[Image: shop_jan_3s_zpseeb53dce.jpg]

[Image: shop_jan_2s_zps7b93d0e4.jpg]

The LED light project is moving along at a decent clip now. A couple months back I finished modelling all of the LED light bar components, acquired the material to manufacture the components, and all of the necessary fasteners to assemble it all.

I decided that the end caps will be held in place by a single screw which engages with an angle bracket that bolts onto the extrusions. Each led will be held in place by two 4-40 screws with fiber washers underneath to protect against shorts. The plastic diffuser will be held in place at the ends by the end caps, and in the middle by two 10-32 screws going through the diffuser, spaced equidistant from the ends. The grooves in the extrusion coupled with the end caps and screws should keep the diffuser held firmly in place and with no light able to escape at the seams. The end caps have a lip to them so if they were to flex slightly it will be imperceptible because the lip shrouds the end of the extrusion.

[Image: end_brackets_zps5b920b9d.jpg]

[Image: gmuw.jpg]

I have machined some of the parts so far. First I machined the aluminum extrusions. The total machine time on the 30 extrusions was about 13 hours, and this was because I surfaced the little grooves that the diffuser will sit in with a 1/32 4fl ball end mill, and I am limited to a spindle speed of 6,000 rpm. The extrusions themselves are 19 inches long, and my mill only has 16 inches of X axis travel, so I machined half of the extrusion and then rotated it 180 and finished to other half so it blends in the middle. The extrusions are pretty dimensionally consistent so the blends in the grooves are smooth or unnoticeable.

[Image: extmill3_zps260a38fc.jpg]

[Image: ext_in_mill2_zps58c4ba00.jpg]

[Image: ext_fin_zpsaf365abd.jpg]

I milled ten mounting brackets at a time and then parted them off with a slitting saw, so they are complete in one setup, except for the 10-32 hole on the side, for which I made a simple fixture to bolt the brackets onto.

(this piece of stock was short, so only six and a half were cut from it. )
[Image: s1_mounting_brackets_zps4ca55eb7.jpg]

The slitting saw is a real time saver. No backside operation was necessary after the part off.
[Image: slitting_saw_zps78a7e179.jpg]

Here's a bunch of brackets after the first setup and the fixture for the side hole operation:
[Image: mounting_brackets_s1fin_zps2ccd7d13.jpg]

The finished side holes:
[Image: mb_s2_fin_zpse2aa46bb.jpg]

The mounting bracket in place on the extrusion:
[Image: mb_end_zps122adabe.jpg]

Then I ran an end cap through both setups to make sure it fits well:
[Image: end_cap_outside_zps41be0635.jpg]

[Image: end_cap_inside_zps78ffd9ab.jpg]

And here it is all bolted up and with the diffuser in place. It fits good with minimal slop.
[Image: end_cap_mounted_zps9196bb00.jpg]

So now I need to work out the finer details of feeding the wires into the light fixtures. My plan is to use 3/8" flexible conduit and feed it into the square portion of the end caps. I picked up some 3/8 flex conduit from the scrap yard for cheap, and then I went to the hardware store and picked up some 90* conduit clamps. The clamps are too large though so I will have to return them. I should be able to use a grommet of some sort and feed the conduit into that. After I get that figured out I will go buy some more aluminum and make 60 end caps.

Scrap yard conduit:
[Image: flex_conduit_zpsf8448ed5.jpg]

Forum Jump:

Users browsing this thread: 1 Guest(s)