U-Bin vs. J-Bin - Light Meter Readings

I grabbed my CPF LionHeart, Serial No. 050/125, and with a well-spent Asian-made R123 lithium-ion battery still inside the short LionCub battery tube, took a light meter reading of 1810 lux from one meter away. It has a TYAJ inside, and I estimate well over a hundred hours of use on it.

Straight out to the workshop, desolder two wires, pull off the TYAJ star, squish a little half-glop of Ceramique onto the heatsink surface, press on the new UX1J star, resolder the two wires, and with the same light, same bezel, same reflector, and same R123 li-ion battery, it now meters 2420 lux, or a 33.7% increase in brightness. So yeah, a U-bin is indeed brighter in this one specific test comparing these two particular Luxeon examples. But...

...the color was greener(!) due to the X1 tint Bin. Not warm, not "less-white", but greener - definitely - Yeesh...

...and I still extend a big, hearty Thanks to You-know-who-you-are, for you have helped all of us to further our collective knowledge base here at CPF.

Just FYI and of course, as always, YMMV.../ubbthreads/images/graemlins/thumbsup.gif

Thats a nice gain Mr Bee...I'm hoping that this will be uniform across transformation from T -> U bin... I might change the star in my LH with UX1J ...

Now...if you can only get some UYOJ/UYAJ stars /ubbthreads/images/graemlins/wink.gif...

Btw...changing the star in LH ... is it as easy as desoldering 2 wires, take off the star, put the new one in then solder the wires back on?.... I've seen somekinda small tube coming from the same hole as the -ve wire and it seems that this small tube is touching the side of the star...any idea what is this "little" tube for? ...

bajaiman,

The tube is a grounding stake. No need to remove it for any reason and yes, replacing the Luxeon star is as easy as I exactly described. In fact one of the steps could likely be eliminated since you might not need to add more thermal paste. I only did it because when I pulled the star off, more of the goop stuck to the bottom of the star than stayed on the heatsink surface and it was less messy to just add fresh paste. /ubbthreads/images/graemlins/thumbsup.gif

What was the current of the two Luxeons? You can't assume they were getting the same current just because they were in the same Vf bin and had the same battery.

WOW! 33%?!

Can't wait until UXOJ/H or UYOJ/H become available. /ubbthreads/images/graemlins/grin.gif

Jon

A true light meter response is heavily weighted in the green, just like the human eye, at 555nm. Good light meters have settings on them for different color light sources to correct for errors that result from different light sources. Most decent light meters have settings for Tungsten/Daylight, Fluorescent, Sodium, and Mercury lamps. This is to correct for errors that result from the light meter sensor. This allows the light meter to give you a true response for the human eye. Some meters will give you only one setting, but will give correction factors in the manual. What is proper for white LEDs, I have no idea.

My single-sample test was also done with the same battery, but without soldering, with the T-bin and U-bin leds in different LionHearts.

This test supports Chrlie's results. Using my trusted old Minolta Flash Meter III which has been dead accurate to 1/10 of an f-stop for over 30 years, and with the LH's UCL centered on and touching the light meter's incident-light attachment (which resembles half of a small ping-pong ball), I measured a 0.4 f-stop gain (which is about 40%) of the LH/UX1J over the LH/original at all brightness levels, confirmed by the naked eye judging a white-ceiling-bounce test. Runtime (measured only on high setting) was shortened by 20%, and temperature was slightly warmer.

In use, the LH/UX1J has slightly better throw, but what is most apparent to the naked eye is the brighter side spill.

Contrary to Charlie's experience (and thanks to CY) my LH/UX1J is noticeably whiter than my original LHs, and has become my new standard for subjective flashlight whiteness.

U-Bin vs. J-Bin ... in the subject makes me confused little..

only that i opened it.. i kame to know that its U vs T Bin /ubbthreads/images/graemlins/smile.gif

and the tint is X1 vs YA

MrBulk, what was the current to the 2 leds? Were they exactly the same??

I have not measured current since I already did the swap and am carrying it as-is now. I do know there can be a variance in actual Vf vs. Bin-rated Vf and thus there could be a difference in the "J" Vf Bin of the two Luxeons, but to me the ultimate and final arbiter is seeing a brighter light. Not just to the meter, but to the naked eye.

However, as Newbie's post seems to suggest, the presence of more green would indeed make the output appear brighter, and if my LM631 MeterMan light meter is also indeed weighted to gauge brightness according to human eye response, then perhaps the U-bins are NOT that much brighter than the T's...?

(sheesh, since I can't get U-bins in any appreciable quantities anyway, it might be that for once, Newbie is actually inadvertently Helping me here! Woo hoo!) /ubbthreads/images/graemlins/grinser2.gif/ubbthreads/images/graemlins/grinser2.gif/ubbthreads/images/graemlins/grinser2.gif

Perhaps someone can take the specs, as provided by Lumilleds of the worst case T bin and best case U bin and compute the potential best case lux % increase if one were to go from a T to U bin. Constrained by the accuracy of Charlie's light meter and method of testing, we now know what he measured in comparing a 100+ hour "old" T bin with a new U bin and to some degree of certainty, we know the difference betwen these two particular LEDs; based on some margin of error as suggested above. Beyond that, certainly YMMV holds true here and if Lumiled's binning system is to be trusted and/or believed, we can look to their parameters for anticipation of improvements.

Personally, I have only installed a few of the U bin LED's and they were all subject to the variations in host optics as well as drivers to add confusion to their own contributions; not in isolation. My take ranged from a jaw drop to a stiffled yawn. /ubbthreads/images/graemlins/blush.gif (no worries, the yawn is still on my bench /ubbthreads/images/graemlins/nana.gif)

Rudi, you have the very first lionheart that was converted to Ubin. there's a long thread somewhere about it.

thought it was pretty white too. but white is in eye of beholder and I seldom promise tints anymore. got into too much trouble /ubbthreads/images/graemlins/icon15.gif

Hi Rudi

Is it possible for you to measure the Lux at 1 meter ? I think this distance is a little more informative, at least for me.

Thanks.

I have a KI light that has a TW0H or I can subsitute a UX1J emitter. The TWOH put out a respectable 930 lux at one meter. The light with reflector has an intense hot spot with a decent spill light. With the UH1J emitter The KI light lux reading jumped to 1350 lux. That is a very noticeable increase of light. 45 percent more light for 50mA more current and a UX1J emitter. The KI light is brighter and the beam is a nice frosty white.

Harry, I'm sorry but I don't have a lux meter. My method is a comparative estimation of total light output.

I compared a TW0H to a TWOJ some time ago and if I recall correctly saw something like 1900 lux from the H and 2400 lux from the J. With subsequent builds using the H's I had, they all seemed like they were efficient on the Vf but low end on the lux. /ubbthreads/images/graemlins/icon3.gif

If someone wants to go to the trouble of taking two fresh and out of the tray LED's of known bins and compare them using the same drive current and same optic then we will have two data points which relate to those two particular LED's. /ubbthreads/images/graemlins/thumbsup.gif /ubbthreads/images/graemlins/rolleyes.gif

On the other hand, we can go with the specs as purported by Lumileds and presumably used in the bin sorting.

Argh!! The bane of the bin, the hunt for white walls, the inequity of inequalities in these damn LED's! /ubbthreads/images/graemlins/smile.gif

I've put a UX1J in my LH and the meter reading went up from 3500-3600 lux to a 4300-4400 lux using the same battery (18650 cell)… that's a good 22% gain.

I'm kinda afraid to try using the H bin as when I supplied 3.7V to the H bin I got a reading something like 1.2A current draw /ubbthreads/images/graemlins/huh.gif from the benchtop power supply. This is like only with 3.7V …. At 4.2V (freshly charged battery) I'd think it'll kill the poor luxeon real quick /ubbthreads/images/graemlins/icon15.gif

[ QUOTE ]
McGizmo said:
Perhaps someone can take the specs, as provided by Lumilleds of the worst case T bin and best case U bin and compute the potential best case lux % increase if one were to go from a T to U bin. Constrained by the accuracy of Charlie's light meter and method of testing, we now know what he measured in comparing a 100+ hour "old" T bin with a new U bin and to some degree of certainty, we know the difference betwen these two particular LEDs; based on some margin of error as suggested above. Beyond that, certainly YMMV holds true here and if Lumiled's binning system is to be trusted and/or believed, we can look to their parameters for anticipation of improvements.

Personally, I have only installed a few of the U bin LED's and they were all subject to the variations in host optics as well as drivers to add confusion to their own contributions; not in isolation. My take ranged from a jaw drop to a stiffled yawn. /ubbthreads/images/graemlins/blush.gif (no worries, the yawn is still on my bench /ubbthreads/images/graemlins/nana.gif)

[/ QUOTE ]


Also consider the Vf bin J. One J can have up to 0.24 Volts lower Vf than another. In an unregulated light such as a LionHeart, this can also cause a considerable difference in output, since this would cause very different amounts of current to flow through the emitter.

So far, I have only seen warmer tinted U bins discussed on the forum, but I might have missed the other threads on this. It is interesting because I am wondering if the comment Newbie made has two aspects which help make them a U bin
- Higher sensitivity to green, both the meters and eyes
- Bluer versions would have slightly higher stokes losses in the conversion to yellowish colors. (meaning warmer bins would tend to be slightly brighter I think) Of course, it is late and I might not be thinking clearly.

The best way to compare a T bin to a U bin is to use a constant current power supply and use luxeons of the same color bin.

As far as using T bins with H Vfs, it's been my experience that H Vf T bins have been less than stellar performers. I don't think that you really start to see the T bins shine until you get to a J Vf.

I think that is pretty widely accepted that the warmer tints score higher on a lux meter, so it's really tough comparing a relatively cool bin to a relatively warm bin, especially one sample that was a greenie.

One of the regular bins that I stock is the TWOJ. I recently also had the opportunity to mess with some UWOK bin Lux IIIs. The feedback that I've gotten seems to indicate that the typical increase in performance is about 10 or 12% or so. Sometimes more and sometimes less.

I like the fact that the U bins are hitting the streets now. It shows signs of forward movement in the technology; however, as far as I've seen, a U bin doesn't really make a light do anything that it wouldn't do with a T bin.

Just my two cents worth.