GE VIO Led - 4W testing

[purduephotog]

Evening-

I just received (at work, sigh) a set of GE 4W LEDs to test. I've got to say their product literature does them a great dis-service- they look nothing like what's on the market right now. In fact they look ALOT like real light bulbs- at least the light emitting part.

The LED itself is encased in a glass dome which is the size of the tip of my pinky- think of an absolutely HUGE LED. The dome itself is coated on the inside with the phosphor- so you can wipe the surface clean.

The PCB is aluminum and is covered with a plastic ground guard, except for where the two screw divets are marked which are bare. I believe it is still recommended to use some sort of insulated fastener to separate the two.

Each board is about 30% larger (square) than the CREE stars, and each board comes with a tiny MOLEX connector. There is a very tiny Tb port to measure the temperature of the board- and all of the rated GE specs are given based upon the temperature of the board.... so no having to guess the junction temp.

I have only one converter at the moment that can power these (11V @ 350ma) so I'm running two of them in series. They are only 'warm' colors with either 70 or 85 CRI.

The light when turned on is very uniform and, surprisingly, emits at 180 degrees pretty evenly (obviously this is a visual appearance and not measured as I have already stated I'm not at work).

I ran them on a 1/4" thick chunk on aluminum for 5 minutes- the board temp was 72F and the aluminum chunk was at 55F- no appreciable increase over ambient. Temperature was measured with a non-contact laser/IR thermometer- highest recorded was 90F somewhere on the dome/die area.

When they're fully evaluated at work I'll take some photos here at home with meat and a color checker with my 40D to give an idea.

Pricing was ~$12 for the 4W models if memory serves. They only sell in lots of 10 so if you are interested in a group buy (never done this before) please PM me and I'll collect names and interests and then post a more detailed breakdown. Remember these are NOT DESIGNED FOR FLASHLIGHTS.... I don't think you COULD make one fit in a flashlight. They're designed for lighting replacements... I think 4 of them would make a nice 360 degree light bulb easily.

Jason

 

[purduephotog]

Reserved for pics 1

After far far too long of doing nothing, I have the FIRST photo from these.

Not as under-exposed as I'd like to show the color differences, so I'm going back- but for now....

Pic 2- 1/8000 f22 - shows 'color' 6500K exposure, showing phosphor zones. LEDs lightly cleaned (these are GLASS domes, so no wear or risk there).

 

[jason 77]

Where did you guys these from... is there a website?

 

[Ken_McE]

I'll take some photos here at home with meat and a color checker with my 40D to give an idea.

:popcorn:

 

[PhotonWrangler]

I toured their factory where these are made! I watched while a lab tech was experimenting with different phosphor domes with various phosphor mixes in them. I also saw the phosphor mixing machine. Of course I got a good look at the color rendition and I really like it.

I'm in for a group buy! :popcorn:

 

[brickbat]

Where did you guys these from... is there a website?

Why, certainly:

http://www.lumination.com/literature/Vio_DataSheet_WEB_122308.pdf

Purdue Jason, What flux bin did they ship and what CCT and CRI did you get?

 

[purduephotog]

Why, certainly:

http://www.lumination.com/literature/Vio_DataSheet_WEB_122308.pdf

Purdue Jason, What flux bin did they ship and what CCT and CRI did you get?

Indy! My hometown!

I actually got a sampler- I received the 3500K 85CRI, 4100K 85CRI, and the 70CRI's for the others. I'll have to get the actual specs.

I am investigating these for work... but there's been absolutely nothing about them here and my coworkers an I are *really* impressed with these chips. I'll be buying at least 10 of them to hook up to an xitanium 40W controller to install downstairs in the dining room. You have *got* to see these lights to believe it.

Anyways... back to work on a saturday. Sigh.

-edit: And no, they don't give out samples. 10x minimum order per wattage, bin, color. Very difficult to obtain testing samples to work with since no one does redistributing / ordering besides them.

 

[blasterman]

I received the 3500K 85CRI, 4100K 85CRI

Whoa...sounds sexy! Glad somebody is taking CRI seriously and just trying to impress flashlight collectors.

I have a couple concept projects where the wide illumination arc of these might be very helpfull.

 

[PhotonWrangler]

Yes, the fact that they're using a violet LED opens up a wider range of phosphor blends for them. This gives them a wider palette of color to play with. I beileie it also helps to stabilize the binning, although I'm not 100% positive of that.

 

[bhvm]

Pics pics pics!!

 

[PhotonWrangler]

I wish I had a photo of that phosphor mixing machine. It has plexiglass walls and looks like a little like a cotton candy machine at a carnival, with splatters of phosphor sticking to the walls. :cool:

 

[snarfer]

Those look really cool. I found the http://www.lumination.com/literature/Vio_DataSheet_WEB_122308.pdfdatasheet (pdf).

I just wish they would make them in 92 CRI as well. If they did I would order a couple hundred right now!

 

[purduephotog]

OK, some progress. Still can't find my stupid color checker OR the greycards...

Using the DX 42V driver (320ma) and the DX 24V driver together since there was no 50V driver.

These are 4W LEDs. I believe them to be 4100K (brightest), 3500K (next brightest), and 3000K 85 CRI (extremely dim).

 

[Dave_H]

Looked over the spec sheet, trying to determine how efficient these LEDs
are from a thermal point of view. I'm not looking for luminous efficacy,
which is stated for various conditions, but the actual amount of heat
dissipated for given vf/If.

Thermal application note (from a different vendor) makes assumption that
heat dissipated is vf*If, which must assume that the LED is 0% efficient
at converting dc to light output. It represents worst case in a way but is
not accurate. I don't see how an accurate thermal analysis can be done
this way.

Is LED efficiency some sort of closely-guarded secret?

Dave

 

[purduephotog]

Looked over the spec sheet, trying to determine how efficient these LEDs
are from a thermal point of view. I'm not looking for luminous efficacy,
which is stated for various conditions, but the actual amount of heat
dissipated for given vf/If.

Thermal application note (from a different vendor) makes assumption that
heat dissipated is vf*If, which must assume that the LED is 0% efficient
at converting dc to light output. It represents worst case in a way but is
not accurate. I don't see how an accurate thermal analysis can be done
this way.

Is LED efficiency some sort of closely-guarded secret?

Dave

Yes, I'm sure it is.

I believe GE has had these tested with the DOE, so there may be some information there.

One test, which I'm not equipped to perform, would be to take a chunk of aluminum or copper of a known mass and just allow it to heat up- well insulated with neoprene- so that air losses are minimized. You could calculate the heat flux then.

And by you, I mean not me ;)