5mm White LEDs with hugly different MCD ratings

[VidPro]

I will try and keep this simple

when rating a 5mm LED there is ANGLE 5* 20* 60*
and angle AT a specific MCD rating can be all the difference in the world.
if i have a 60* that rates 6000 it might put out as much overall light at a 5* rated for 20000
(correct any of that if its wrong)

so there is angle, THEN there is "efficiency"
having 2 LEDS of the same Angle binned for less or more MCD.

so now the questions
when binned for output, is it possible that (basically) the light depending on how its DRIVEN would be no different than the others.
like
if they test it at a specific current, or voltage, and the place that it is being meterd happens to read higher.

OR

are the brighter at the supposed same specs, ACTUALLY more efficient, and actually put out more overall light.


The project:
assemble 100 5mm try to get the most light for the least power, without any concideration for how bright it is in ONE spot .
will wizz bang 22000 LEDs VS 12000 leds make a whole room 2x brighter when using the exact same Power overall?

 

[jtr1962]

You can estimate lumens by using the equation lumens=2*pi[1-cos(angle/2)]*mcd/1000. This will tend to underestimate the output for LEDs with viewing angles under about 25° and overestimate it for wider angle LEDs but at least it's a start. For 30° LEDs it seems to be about right. For 20° multiply the result by about 1.25 or 1.3. For a 60° viewing angle multiplying by 0.75 or 0.8 works for me.

In terms of raw lumens, 22000 mcd LEDs with the usual 20° viewing angle will put out about 2.6 to 2.8 lumens. However, a 50° LED rated for about 5500 mcd would put out roughly the same amount of light.

 

[VidPro]

ok so when the angle is Exactly the same, yet the "rated" output is alegedly higher.

is it because it acted different in the testing, or had a spot point that managed to be brighter when it hit the meter.
OR
because at the controlled input (V or ma) it burned brighter just because it was using more power.
OR
because it actually was more efficient, and put out more actual light.


and if its for a LAMP anyways, and your figures are even close to being correct, wouldnt i be better with a 6000mcd at 50* ?
(now you made me think different again)

 

[jtr1962]

If the angle is the same, then the one with a higher rated output should put out more light provided they're both rated at the same current (usually 20 mA). Of course, this assumes the rating is accurate. As for different LEDs with similar ratings and beam angles differing in lumen output, this is possibly but there usually isn't that much of a difference between LEDs of the same size and current rating. After all, the die size is usually about the same, the packages are almost the same, and the placement of the die in the package is almost the same. All of this makes the light distribution about the same for a given beam angle. What this means in that in order to have a higher mcd rating for a given beam angle then the LED must use a brighter chip. That is really the main difference between an 11000 mcd and a 26000 mcd LED if both have a 20° angle. The chip in the 26000 mcd one produces almost 2.5 times as much light, meaning that it is more efficient.

I also want to point out that you always need to look at the current rating which is used to determine the mcd output. Some unscrupulous suppliers might say an LED is 50000 mcd but don't tell you it is at 100 mA. The same LED at the usual 20 mA would be less than 15000 mcd.

Yes, for area light wide angle is always better. I like the Luxeons and Lamina Ceramics as is for area lights. Focused light is very distracting. You may be able to light a given small area brighter with a certain number of 20° LEDs than if you used 60° ones, but the constrast between the dark and light areas is annoying to say the least.

BTW, you can type the ° symbol by pressing down the ALT key and holding it while typing 0176 on the numeric keypad (i.e. the one on the far right of your keyboard), and then releasing it after you finished entering 0176. You can do all sorts of symbols like this. ALT+0178 = ², for example.

 

[VidPro]

"there usually isn't that much of a difference between LEDs of the same size and current rating"

this is exactally my thoughts
the different rated ones, have SEEMED to be just more sensitive at the "rated" voltage. meaning in a flashlight they will get brighter, and may just burn out faster when over current.

it always SEEMED like they have the same general light output (overall light not spot).

they alledge that the testing is done at spec current 20ma
but that doesnt PROVE they ran at the same voltage. meaning the overall power could still be different.

reverse that
IF they specked them at a specific voltage, a more sensitive one, might run at a higher current at that same voltage.
BUT
they would both put out the same light AT the same overall power, exact same volts, exact same amps.

that is my OPINION.

but if they are more EFFICIENT, then for theis Solar project, i would prefer more efficient.

I love the high-powered ones, like the 3W Luxeon, have a 10X bar of them, that puts out some good light, and works at a large variation of voltage.
I am very unlikly to smoke the thing, running unregulated as it is.

but i THOUGHT that it might be possible to have a multitude of 5mm (type) running more "efficientally.

at ebay in cheap quantity, they have 13000 and 26000. the price difference is not a big issue.

my OPINION is that it will be just as total power efficient to use the 13000, but i just dont know for positive.

Leds is leds

 

[jtr1962]

Honestly, with eBay the only way is to buy some of each type and test them. Sometimes suppliers exaggerate their specs, somethings they don't. As for rating LEDs, at 20 mA they usually have more or less the same voltage. Usually this is about 3.2 or 3.3 volts for white LEDs, but it can be as high as 3.7 or as low as 3.0. The extremes only represent a 20% difference in power. However, an LED with a very low Vf can indeed draw a lot more power in an unregulated circuit and thus burn brighter as you say. It will of course also burn out faster since it's running above rated current.

All other things being equal, just go for the highest mcd LEDs if there isn't a huge price difference. It seems on eBay you have 10,000 to 12,000 mcd, 20° whites going for about 20 to 25 cents while 22,000 mcd, 20° ones are sometimes going for about 35 to 40 cents. This is probably about the same cost per lumen, but with the brighter ones you use fewer LEDs for a given lumen output (and also less power).

 

[VidPro]

OK
yes the extra 15-20$ in this situation is nothing, compared to having to hand solder them all

 

[NewBie]

Here, go to page 2, item #3, all are ran at the same 20mA, same degree beam output, but yet you get 11,000mcd to 31,000mcd:

http://www.nichia.co.jp/specification/led_lamp/NSPW500CS.pdf

Personally, if I wanted efficiency, I'd try the upper bins of the Nichia NCCW022 or NCCW023, and under power them, or the go for the CREE XL7090, and do the same with the N bins. Unless you are trying to focus and throw the beam, the LumiLEDs parts really aren't in the game anymore, at least until this summer or late spring...and even then, we'll have to hold our breath and see what Future ends up getting in stock...

 

[VidPro]

thanks, i was using those specs, but they are from the "seller" so to speak.
and a tester usually can cipher the sellers adds into reality , and reality is the only thing that counts

see if there is a voltage diferential even at the same exact current, a more sensitive one could get to 20ma at 3v, and a less sensitive one could get to 20ma at 3.9v
so testing at the current, is not 100% sufficient to test its actual efficiency.
there could be (in so much as) a 1/4 of the voltage diferential.
which would make a total power or WATTAGE difference.

course 1/4 does not explain a 2x brightness of the meterd SPOT.
but add in how the one SPOT is metered, and the binning might not be a big deal.
the binning would still be effective to MATCH like units for banking.

but its already a done deal, i bought the 26,000 things.

and for CPF peoples , to FIND them on e-bay the KEY is to go to the STORE, not the normal auctions.
(now they are gonna be all gone /ubbthreads/images/graemlins/frown.gif )