Is the Luxeon Lottery Rigged?

Well yes, by a reasonable definition it is. What I am wondering is how rigged. This thread is intended to be an informal survey to get an idea.
Background: Some consumers [OEMs] of LEDs consider certain bins to be premium for their application. There is only one authorized distributor of Luxeons in the USA which is Future electronics. Consumers with adequate volume and/or negociating skills can arrange for some degree of bin selection in their purchases. My question is this: In accommodating those customers, are *only* the dregs left for the small volume purchaser or does one have a chance to getting a desirable bin? To do this survey, we must define what is a desirable bin. Also, to keep this from getting complicated, let's limit this to only white Lux III Luxeons. Since high vs low Vf bin is not a clear indicator lets disregard Vf binning. For flux binning, for practical purposed only "S" and "T" exist and all would agree that "T" is the premium of the two. While there will be less agreement on color bin, I am going to declare V0, W0, and X0 to be the premium bins. If you have ordered white Lux IIIs directly from Future, please post what you have received for bins. Please state the number of different orders you have made and list all of the bins you have received.
I'll go first:
2 orders, all were SV1K.
TIA in advance for your participation.

oo2eoo had twoh's for sale, i am guessing those came from future.

[ QUOTE ]
ufokillerz said:
oo2eoo had twoh's for sale, i am guessing those came from future.

[/ QUOTE ]
No guesses please. I am looking for input from those who have received white Lux IIIs directly from Future without benefit of a negociated bin selection.

I ordered 30 and they were SL1's.

[ QUOTE ]
flashlightlens.com said:
I ordered 30 and they were SL1's.

[/ QUOTE ]
Could you check on that? SL1 is not a valid bin number.
Hey guys! Surely there are more than two of us here buying directly from Future.

I cannot speak for Luxeon/Future, but from other experiences, I will tell you that as more customers get on board, the quality that most see goes down hill. I was an OEM for Magenetic peripherals for many years. We were a very early adopter, and we measured quality based upon the number of defects in the media of the devices we purchased. We didn't have a purchase spec, because early on, everything we saw we could live with.

Early on we got complete random distribution. Most were very good, some were not so good, but were never saw anything really bad. In fact I actually ended up with a 400mb device with ZERO media flaws(that was one out of several thousand units of that product we purchased).

As other customers came on board with purchase specs (and the purchase specs were usually designed to subtly weed out the product with lousy media, once you understood a little about the source of defects it was possible to design a spec that didn't actually call for a very limited number of defects, but exploited the nature of the defects so that only the best media would meet the spec). As a result we saw the average number of flaws per device more than double. The good stuff was being selected out by virtue of the purchase spec, and we were getting the leftovers, although our costs didn't go down.

So the answer is that it certainly happens, although it is not explicity by design on the part of the Future, it is explicity by design on the part of Future's customers with purchase specifications.

I remember a statement from Future that they did not make specific Luxeon bins available to anyone - you got what you got.

Yet many manufacturers (aka SureFire) seem to be getting consistently better bins than smaller buyers without the leverage to force Future to ship only specific product. Perhaps those consistently getting premium bins just tightly specified what they wanted and had the volume to force the issue.

I heard from a person who does business with Future that they will quote specific bins as being available IF you were not considered a distributer. This means that you have to call them each day to ask what's available at that time. At least that was my understanding of the conversation.

If you are a distributer, they would not quote specific bins, but simply sell them in lots.

I have the un-substantiated feeling that the lottery is loaded in another way. I suspect that the LED lumen output tends to be near the boundaries of the bins, as in most T are probably closer to 67 with very few near the upper boundries of 87. Likewise, I suspect that most S bins hover around the 51 mark, not the 67. I'd love to be able to run a few hundred through an integrating sphere to see if that's true.

Daniel

Gadget,

Why would you suspect that the S bins are mostly near the low end? This would make the distribution bi-modal, which doesn't seem to fit what we've seen with luxeons.

My guess would be that the overall L3 population is a standard distribution, centered somewhere in the middle of the s-bin range. It is artifically bounded at the low end of the S bin due to product labeling reasons (sudden dropoff to 0 at the 51 lm mark).

If we look at the T-bin alone, I would agree that the distribution of the T-bins is probably greatest at the boundary (67 lm mark), and drops off quickly to basically 0 at the high end. While some may think that they can label an average T bin as being in the middle of the lumen range, I would bet that less than 1/100 actually output in the middle of the T-bin range.

Again, all guesses and unsubstantiated gut feelings.

Well if we are sharing unsubstantiated gut feelings, here are mine. BTW, for HD white Lux IIIs the lower limit of the "S" bin is 60 lm, not 51.7 lm [as is the case with all other emitters]. The datasheet "typical" is 65 lm. Like evan9162 I think the HD Lux III flux distribution approximates a standard distribution trucated at the lower limit of 60 lm. Where I differ is that I think that the peak of that distribution is below the 60 lm cutoff. Those dice that fall below the 60 lm cutoff in up in other products [mostly Lux I].

OK, another datapoint. This is from a CPF member who emailed the info to me rather than post here.

One purchase: all SX1M

As an individual I do know I asked for LXHL-PW09 TWOK and received LXHL-PW09 TWOK from Future, qty 10.

I do know if you are in the million devices per year category, you can purchase directly from LumiLEDs.

See datasheet: http://www.lumileds.com/pdfs/protected/DS25A.PDF for windows, especially for colored emitters.

There is a premium for these parts, example:
White LXHL-PW01-00M WA, X1, W0, X0

Now, I don't see this part listed on the Future site, but I did find this one, LXHL-PW01-M4G Not sure what that part is.

But, if you look up LXHL-BB01-00F, which is a narrow sort blue, you will find that on their site.

There are also brightness binned part numbers.

For customers with enough volume, the customer can ask for a three bin wide selection, but Future will not guarantee availability.

I argued for years with Future and LumiLEDs to at least tell us which bins they sent, as they used to not do this. It's printed right on the reel, so it wasn't hard to implement.

I've only ordered a few pieces, and they wern't marked other than the part number.

I'm bought a small quantity (10) of Star LUX I's and I was told that my flux requirements could be met but I would have to accept the full range of color productions. I received 5 PYOL and 5 PX1L. Obviously they had been removed from their original aluminum tree's purposely to get the variation.

CNC DAN, look on the invoice. The bin numbers are labeled somehwere on there. I forget what it's called. Unless you're talking about a long time ago. I ordered some 5W's emitters and it showed the bins.

[ QUOTE ]
evan9162 said:
Gadget,

Why would you suspect that the S bins are mostly near the low end? This would make the distribution bi-modal, which doesn't seem to fit what we've seen with luxeons.


[/ QUOTE ]


I suspect that because of marketing. They could easily make the bins in 10 lumen steps or even mark each one with the exact value. What they do instead is use a range that makes a 67.2 lumen T bin lux 3 as valuable as an 87.4 lumen T bin.

The typical human response is to think that your T bin might be 87.4 lumens. After all, it's a lottery right? So the pervieved value of a T bin is greater than an S bin that may be as little as 1 lumen dimmer.

If there were a random distribution it would be advantageous to mark them in
finer grades so as to maximize the value of the brighter ones. The same holds true if the average was towards the high side of the scale. If I had 50% of my luxeons over 80 lumens I'd make that a special category and charge a premium.

Of course, this is all just speculation based on on a cynical point of view.

Daniel

Hi all,

No speculation in this post..

I have placed 2 total orders with Future, all low quantities.

The first order was 3 LD 1W's and 4 HD 1W's. Nothing special, PWAK's and NV1J's respectively.

My latest order was:
2 - 3W HD - SX1L
1 - 1W LD - PX1L
4 - 1W HD - QYAJ
5 - NX05 optics.

pb

[ QUOTE ]
Doug S said:
Could you check on that? SL1 is not a valid bin number.

[/ QUOTE ]

It was S something something........ Can you tell I don't buy these too often? There was a "1" and an "S" and I swear there was an "L" somewhere in the mix.......

It could have been SX1L or SV1L (assuming they were lux3s)

...or the best-performing 1W luxeon ever seen!