1W "S" bin : IntegratingSphere results

McGizmo

Flashaholic
Joined
May 1, 2002
Messages
17,290
Location
Maui
Hi Guys,

Zach from LEDsupply sent me a SX1J Star for comparison in my IS to some of the other LED's. First, a table on some other LED's:

LED-samples-efficacy.jpg


Unfortunately, I didn't keep the Excel Spread Sheet used or I could have just added the S bin sample to the rest. :shrug:

So what I measured on the S sample follows:

If (mA)****** Vf********** Lumens

100 ---------- 3.0-------------- 17
200 ---------- 3.2-------------- 33
300 ---------- 3.3-------------- 46
350 ---------- 3.3-------------- 51
400 ---------- 3.4-------------- 56
500 ---------- 3.5-------------- 65
600 ---------- 3.5-------------- 74
700 ---------- 3.6-------------- 81
800 ---------- 3.6-------------- 88
900 ---------- 3.7-------------- 94
1000 --------- 3.8-------------- 98

It is really important to keep in mind that these measurements are reasonable in terms or relative comparisons among the samples tested but as to how representative any one of these LED's is of its brother bin members, well, your guess is as good as mine. :) This particular S bin held its own against the measured competiton but did not win in any category; nor did it loose!
 

Kiessling

Flashaholic
Joined
Nov 26, 2002
Messages
16,140
Location
Old World
very interesting! thanx for the work.
if some of those would show up again, I'd be really interested ... :D :D ... the resurrection of the old LuxI :)
bernie
 

Klaus

Flashlight Enthusiast
Joined
Sep 6, 2001
Messages
1,998
Location
Germany
zapper said:
So the "S" bin out performed the "T" bin !?

:wow:

Either Lumileds measurement for the binning is screwed or Dons is off I think :confused:

Bernie - who said this SX1J was a LuxI ? Edit: OKOK - I re-read the title - but running a 1W LuxI at 1000 ma or almost 4W is a bit adventurous then me thinks :whistle:

While even with LuxI vs LuxIII the binning regarding the flux should be coherent and a "S" outperforming a "T" bin by 10-20% over all currents is indeed strange and confusing - at least to me ;)

Klaus
 
Last edited:

cy

Flashaholic
Joined
Dec 20, 2003
Messages
8,186
Location
USA
nice chart!

Sbin one watt is rated at 350 milliamps

Tbin three watt is rated at 700 milliamps

so when you drive Sbin at 700 milliamps, output naturally goes up.

had no problems driving one watt at 700 milliamps
 

Cypher

Enlightened
Joined
Nov 14, 2004
Messages
270
Location
CO
Lux III "U" bin beat K2 "U". Seems like a step backward if that is true but I guess driving a Lux III so hard will shorten lifespan but the K2 is made for the higher current and temp? Isn't the K2 still supposed to be more efficent even at lower current? Are the bins roughly equivalent between LuxIII and K2?

Daegan
 

joema

Flashlight Enthusiast
Joined
Aug 14, 2005
Messages
1,189
Location
Nashville, TN
Cypher said:
Lux III "U" bin beat K2 "U". Seems like a step backward if that is true...
Considering how we've hyped the K2, that is indeed interesting.

While the K2 looks good in these tests, neither its max output NOR efficiency beats the U-bin Lux III, at ANY drive current.

The Lux III max specified drive current is 1000 mA, and the max spec'd K2 drive current is apparently 1500 mA (from prelim datasheet), BUT the output visually wouldn't be much greater -- just 16%, which I doubt you could even see.

As you said, I wonder if the bin #s are equivalent, or if the K2 is pre-production and will be improved later, or if the K2 superiority is just marketing myth.
 

wasBlinded

Flashlight Enthusiast
Joined
Dec 14, 2004
Messages
1,222
Location
Oklahoma
Cypher said:
Lux III "U" bin beat K2 "U". Seems like a step backward if that is true but I guess driving a Lux III so hard will shorten lifespan but the K2 is made for the higher current and temp? Isn't the K2 still supposed to be more efficent even at lower current? Are the bins roughly equivalent between LuxIII and K2?

Daegan

Keep in mind that a "U bin" is a premium bin for the LuxIII, while the K2 that was tested might only be an average (or worse) performer out of the population of K2s.

It does seem, at least so far, that the K2 as we have seen it is not going to be all that impressive from the CPF user point of view.
 

McGizmo

Flashaholic
Joined
May 1, 2002
Messages
17,290
Location
Maui
Hi guys,

I would like to add some comments/ qualifications as most of you are not likely familiar with my test set up.

I have calibrated my system best I can with a calibration lamp of known and documented spectral and radiometric output but the lamp was not designed to be used with the IS I have so there is no doubt error in the absolute values my system measures. I recently got some sample Nichia LED's which were provided with measured flux values and one that I introduced to my IS in much the same manner as these other LED's measured 15% greater on my system than the number provided by Nichia. Since a 15% variation is often still within a single bin range, I figure my system is close enough for my needs. :shrug:

As already noted, bin clasification assigned by Lumileds is based on a drive current level that is not the same between the 1W, Lux III and the K2. Since color tint is also being assigned the LED at the measured current and color tint can shift due to drive current, I can see why Lumileds does not use a standard drive current for all of the parts and in fact, on the K2, they have parts binned in two drive current categories due to the wide possible range of drive currents the K2's might be operated at.

The K2 sample above was certainly an early engineering sample and not a production sample. I hear Beta2 samples are now starting to get into the field and I have hear tell that they are better than the first samples. One can hope this is the case!!

Since this information is based on a single sample of LED in each case, it may well be more of a disservice than a useful tool unless you allow for significant variation fromthe results!! For instance, the first K2 I tested, at 1400 mA was delivering 13 lumens/watt!!! Not sure if it was more a hand warmer than flashlight at that level! :green:

I do suspect that there is some utility in this table in consideration of a particular LED and how its efficacy changes due to drive current changes. This is based on the assumption that the physics of the actual LED design will be more consistant across the population than the flux output would, in and of itself. My assumption may be faulty in which case the information provided above becomesalmost completely useless!! :)
 

jtr1962

Flashaholic
Joined
Nov 22, 2003
Messages
7,506
Location
Flushing, NY
44 lm/W at rated current isn't bad at all. BTW, are S-bin stars starting to become fairly common? Anyone ever get a T-bin 1 watter?
 

TrueBlue

Flashlight Enthusiast
Joined
Mar 5, 2004
Messages
2,373
Location
Central CA
To pull this thread more off topic.

IMO…the Lumileds K2 is in the early production stages and needs to mature. The K2 has the potential to be driven at higher currents for more light. From what a couple of people, me included, have observed the K2 is not a quantum leap Lumileds would like you to believe in making more lights better.

Thank goodness I do not work at Lumileds. I'd be diverting development into the Lux V emitter. I've had a chance to play with a good number of Lux V emitters. The Lux V premium bin emitters are bright. If the company could make consistently brighter Lux V emitters that can be made or proven they are more reliable than the company's specs I'd be 100% sold on the Lux V. An estimated life of 500 hours is not inspiring. For a lot of us 500 hours is more than a lifetime of lighting. But if they could get the Lux V to even double the 500 hour estimated life I'd be happy.

The Lux III emitter has its place too. It is a proven, durable, design. If Lumileds needed a torture bench I'd give the emitters to CPF members! We've tested, tortured and abused the emitters more Lumileds could ever do! The Lux III U-bin emitter, with the right tint around a W0 or X1 (and my preference of a V1) is a great emitter to work with. I've actually thought that the U-bin is better than a K2 emitter. If Lumileds could, once again, concentrate on getting consistent U-bin type emitters out I'd be very happy. I'm not turning over my UX1J emitters to replace the K2 emitters yet. And if anyone has UX1J emitters I'd like to have them!

I'd be concentrating on building up the reliability of an already known but little used emitter…the Lux V.

In a couple of month we'll be thrilled to see K2 emitters start appearing. I'm conservative and thinking the K2 is more hype than what we will end up seeing in real life. It won't matter by then because Lumileds past and present promotions will guarantee they will have the money. Keep your Lux I, III and V emitters. They are not obsolete.

sfemitterheads8ir.jpg
 

Bullzeyebill

Flashaholic
Joined
Feb 21, 2003
Messages
12,164
Location
CA
jtr1962, T bin is a 1W. A lux III one watter. R bin used to be the highest flux 1 Watt led, till the Lux III's entered the field.

Bill
 

hotbeam

Flashlight Enthusiast
Joined
Jan 2, 2003
Messages
1,737
Location
Melb, AU
Don, very interesting data. Do you have a LuxI, eg. QW0J/K/L that you would run the same test on? It would be interesting to see flux readings of an average LuxI (eg P or Q) compared with the high flux "S" you just tested.
 

McGizmo

Flashaholic
Joined
May 1, 2002
Messages
17,290
Location
Maui
HOTBEAM,
I have a number of 1W Luxeons going back to M bin LD's pre EOS. I have some Q and R 1W HD's of EOS. Any that I want to test require mounting on an E-screw and then testing and I just don't personally see the point?!?! :shrug: I am more interested in mounting and testing the new LED's coming out when time permits. Time, now that is more scarce than Y bin LuxV's! :eek:

You know, the more I work with all of the variations out there now, the more I get upset at the LED's being rated with a power distinction! They should be specified by target current and not wattage! Anyone want to buy a 7.5W K2? It sounds powerful but it's a POS compared to a 4W LuxIII U bin that puts out more flux! :green:
 

hotbeam

Flashlight Enthusiast
Joined
Jan 2, 2003
Messages
1,737
Location
Melb, AU
Fair enough Don.

Yes, they should be rated differently to how they are now. I suspect that the "wattage" rating is to compensate for the lack of knowledge of the majority. In their world, more watts equals more brightness. Clearly that is NOT the case.

Manufacturers may also try to out do each other by packaging multiple dies in a single, larger lumens package. However, when you work it all out, their package lumens per watt is something in the order of 20 and that is plainly going backwards.

The only consolation is that if space is an issue, their package can be used to get the light output, even if you have to put 2X the power in! LOL.
 

andrewwynn

Flashlight Enthusiast
Joined
Apr 28, 2004
Messages
3,763
Location
Racine, WI USA
wow, nice thanks for doing this, very helpful information and nice to see some observed data to match with the theory.

It is not surprising to me that the Ubin K2 was bested in both lumen/watt and lumen by the u-bin lux3.. that aligned exactly with both the calculations and the measured data.. as far as 'going backwards'.. no, not at all..

The K2 can handle a TON more heat than the Lux3... not only can it operate safely much hotter.. it has something like 30-40% lower lumen loss per degree.. so when they get cooking, they will hold their lumens better.

Now.. here is why the K2 kicks the U-bin's butt:

The last Ubin's i bought cost me $25 EACH and that was a special discount. The K2s that will output similar amount of light with a bit of a power disadvantage.. $3.45 each.. we need another column.. lumen per DOLLAR! That is where the K2 is going to beat all the other luxeon emitters.

Compare a 1250mA Ubin vs K2..

the K2 is 5.4W and 119L.. the L3 is 5.08W and 125L. an advantage of 14% more lumen per watt to the ubin lux3.

however... now let's take a look at lumen per dollar.

119L/ 3.45 = 35.5 Lumen per dollar.
125L/ $25 = 5 lumen per dollar.

So there you have it.. 7 times the price per lumen with the Ubin lux 3 compared with the U bin K2. This will definitely be the driving force of the K2.

-awr
 

cy

Flashaholic
Joined
Dec 20, 2003
Messages
8,186
Location
USA
no question K2 has a lower cost basis. but it's a little early yet to tell how production samples will fare with vf and tint.

seems the older Ubin and Tbin still offer qualities like correct focus with existing reflectors and better output for now.

like don has already pointed out, trying to base opinions on ONE sample....
 
Top