Why is the K2 TFFC shape the way it is?

[tebore]

Why is the K2 TFFC Shaped the way it is?
Why doesn't LumiLED put their new tech in to the LuxIII?

I just shoe horned a K2 TFFC in to a Heatsink made for a LuxIII and really wished they'd just make a TFFC powered LED with a Lux 3 footprint.

 

[Marduke]

Re: Why is the K2 shape the way it is?

Why use a K2? SSC P4 is the same footprint as the Lux III, and will be over twice as bright driven identically.

 

[tebore]

Re: Why is the K2 shape the way it is?

A number of reasons.

The K2 has way way better color/tint uniformity than the P4.
The non-positive slug.
Better tints. Seriously a V0 P4 looks terribly blue compared to a V0 K2.
The K2 is actually as or more efficient at high currents than P4. This is really apparent at drive currents <1A.
Thermals of a K2 is also much better than P4.
The die of the K2 appears smaller than the P4 which means tighter beams when using reflectors.

All in all the K2 is a superior LED to the P4. The only downside I see is some of the K2s have a real lotto when it comes to Vf even within the bin. This is especially the case when I'm going under 350mA the vf doesn't scale downward so well.

When I say K2 I'm talking about TFFC K2. So forget the 2x brightness stuff you're talking about.

 

[Marduke]

Re: Why is the K2 shape the way it is?

A number of reasons.

The K2 has way way better color/tint uniformity than the P4.
The non-positive slug.
Better tints. Seriously a V0 P4 looks terribly blue compared to a V0 K2.
The K2 is actually as or more efficient at high currents than P4. This is really apparent at drive currents <1A.
Thermals of a K2 is also much better than P4.
The die of the K2 appears smaller than the P4 which means tighter beams when using reflectors.

All in all the K2 is a superior LED to the P4. The only downside I see is some of the K2s have a real lotto when it comes to Vf even within the bin. This is especially the case when I'm going under 350mA the vf doesn't scale downward so well.

When I say K2 I'm talking about TFFC K2. So forget the 2x brightness stuff you're talking about.

Doing a direct swap for a Lux III, the SSC is 2x brighter, and is a much more direct swap. It's not terribly difficult to get whatever tint bin you want.

All in all, the K2 is a terrible excuse to use a high current LED in a light with improper heatsinking.

 

[HarryN]

Hi Marduke - I was just wondering if you could help me out here. I am having trouble understanding how the specs of a common P4 are really better than the specs of a common K2 TFFC.

Here is a link to Future (PL's main distributor) when I can buy in 1 each qty a binned K2 rated at 220 Lumens at 1 amp and 300 lumens (typical) at 1.5 amps for under $ 7. That is no muss, no fuss, a pretty decent LED off the shelf from a tier 1 supplier.

http://www.futureelectronics.com/en...tters/white/Pages/1184142-LXK2-PWC4-0220.aspx

I don't think any user of LEDs should avoid heat transfer known good practice, but it is nice to have LEDs out there that can take some abuse - not that anyone on CPF ever abuses an LED. :naughty:

 

[yellow]

deciding between a Luxeon (even these K2s) and a P4 for tint-stability?

funny

why not use an XR-E?
brighter than the better datasheet-binned P4, tint as ordered (while both others its lottery), best thermal package, neutral slug, ...
When using Sandwich Shoppe's reflectors, even that argued-to-death "ring" problem (in spill) is not evident

 

[tebore]

Hi Marduke - I was just wondering if you could help me out here. I am having trouble understanding how the specs of a common P4 are really better than the specs of a common K2 TFFC...

Your post is the only one that seems to understand the benefits of the K2 over the P4. Or you at least took the time to look at the spec sheets before posting. While the others don't seem to understand how to read spec sheets or take the time to look at the spec sheets or understand what I've posted.

The K2's thermals refers to thermal resistance. It is almost half that of the P4. I don't care you well you heatsink if the LED itself can't get the heat to the heatsink. If you noticed Jtr's testing the results for the P4 would point out that it can't seem to get the heat out or the phospher just is not designed to take the output of the die much over 1A. Heck even at 1A I'm questioning the longitivity of the LED.

deciding between a Luxeon (even these K2s) and a P4 for tint-stability?

funny

What is that supposed to mean? Have you even used the K2? The latest U2 bin P4 even when compared to the original series of lux has terrible color uniformity. The P4 has a corona I just can't stand.

The tints on the XR-E while is much nicer than P4. Is still lacking compared to the K2.

And I suppose Retrofitting Lux3s with XR-Es would be easier than using K2s...?
:shakehead
Cree is not the end all of LEDs.

Point is couldn't we agree that if LumiLED's had continued to improve the Lux III with TFFC and supporting technologies we'd all be happier? I know I'd be, heck I wouldn't have to touch a P4 ever again. Well maybe except to get the high CRI versions.

 

[HarryN]

Hi - Yes, I have read a few data sheets while since becoming a CPF member. I can assure you that you cannot design the two lights in my sig line without reading way too many data sheets. Those are not mag or SF re-works, those are from scratch, including the entire electronics and optics for the "Breeze" light. (won't do that again)

I actually sell equipment to LED die makers (used by most of them) so I started playing with LEDs and flashlights to better understand why they were asking for so many crazy requests. It didn't take long to see their point.

I actually have been trying to accumulate K2 and K2 TFFC data in one thread, so I had started this one a while back:

http://www.candlepowerforums.com/vb/showthread.php?t=224196

The first time I saw the TFFC concept, I had some doubts about it, especially for achieving the lambertian distribution that PL seems to love - I guess PL has some good optics guys. I do wish they offered a TFFC version for higher CRI versions.

I guess you know that you can clip off the 2 unused leads if that helps.

I can see how that Lux III / Lux V change to K2 is a pain if you don't design for it. I was right in the middle of designing the "Breeze" and had heard that the K2 was coming - but could not get a data sheet diagram - had to make a guess. It was close enough mechancially, but I ended up staying with the Lux V for that build as it put out so much more light then anything else back then. (premium bin)

Now, the K2 TFFC is just as bright as the the premium Lux Vs at similar power levels, but easier to focus. Now that is a package I wish they had continued with - bring back the Lux V in a TFFC !!!! I never understood why the Lux V sold rather poorly but the Cree and SS copies seem to be popular with virtually the same focus challenges.

 

[Marduke]

Here is a link to Future (PL's main distributor) when I can buy in 1 each qty a binned K2 rated at 220 Lumens at 1 amp and 300 lumens (typical) at 1.5 amps for under $ 7. That is no muss, no fuss, a pretty decent LED off the shelf from a tier 1 supplier.

At 1A, the P4-U2 can do the same, and is more efficient at any lower current.

At 1A, the XR-E R2 can do 270 lumens, and is still more efficient at any lower current. At 1.5A, it can do 347 lumens.

Either <$7 also.


So, the real question is, if you're going to use a typically non-compatible emitter to begin with (Lux III upgrades to SSC P4, not K2 directly), why use a "premium" K2 when you can get better performance with a premium XR-E??

 

[schiesz]

why use a "premium" K2 when you can get better performance with a premium XR-E??

How many reasons do you want?

The XR-E requires special optics or deep reflectors compared to LEDs with lambertian distributions. If you are directly replacing a Lux, you will probably want to have something with a similar light distribution to use the existing optics. Deep reflectors are fine if you have room for them, but in many DIY, upgrade type situations this is not an option.

So you say the U2 Seoul is the answer here. Maybe so, if your heatsink is electrically isolated. If not, and especially if you are trying to do something like series connect multiple emitters, the positive slug is a pain. I get better performance from my few TFFC K2's than I do with U2 bin Seouls, but that may not always be the case.

I find it easier to work around the different footprint of the TFFC than to electrically isolated the slug of the Seoul, but YMMV.

 

[Marduke]

The positive slug is a horrid design, but 5 seconds and a light coat of Arctic Alumina fixes that.

Or just swap out the Lux III star for the SSC star. Problem averted.

 

[europachris]

What about a swap from non-TFFC K2 to a TFFC K2?

I have a Ray-O-Vac "Extreme" 2xAA adjustable focus light which advertises 100 lumens. I don't think it's even close. It's about as bright as the River Rock 2xAA adjustable focus light with the Rebel 60. It appears that both lights use the same driver circuit (both lights are easily disassembled), and pull about .5A from fresh alkalines.

I like the feel and shape of the light, but the output is lacking. Do I assume the TFFC emitter can be directly swapped for the old one and which bin/tint TFFC is recommended - cool white is fine, just as long as it's not REALLY cool or blue.

Chris

 

[tebore]

The positive slug is a horrid design, but 5 seconds and a light coat of Arctic Alumina fixes that.

Or just swap out the Lux III star for the SSC star. Problem averted.

Hmmm. Lets see I could use an LED with a lower Thermal resistance to start with, and doesn't require me to isolate the slug unless I'm building a massive circuit.

Or I could use an LED that has worse Thermal resistance then make it worse by using a Thermal epoxy as an isolator...:duh2:

If things were stars to start with everything would be honky dory.

In my experience with the U2 P4 I wouldn't go near 1A if you want consistent tint, and longevity. The LED is great when driven 350mA to 800mA when you start pushing 1A you're at the edge of what the LED was designed to do. It's VERY sensitive to heatsinking and the phosphor is not as stable as Cree's or the K2s. You also claim the U2 is more efficient at lower drive currents than the K2. Testing using a bench power supply and a crude light box shows there's a kind of lotto. Some K2s have extremely low Vfs compared to U2s and vice versa. Brightness is also a toss up if you go with the best K2 bin vs U2 bin.

One thing you can't argue with is the tint when it comes to P4 vs K2. The K2's tint uniformity is superior to the P4, PL has a patent on this. I'm even tempted to say the Tint variation within each bin is better too, but I haven't got enough data to prove it.

I'd like to see TFFC tech in a Lux5 and Lux3 lay out. There were reason why the Lux5 wasn't successful. Mostly poor thermals and poor lumen maintenance. The poor thermals probably contributed to poor lumen maintenance. However newer TFFC tech seems to have address this problem. If they had a 2S2P set up with specs calling for a max of 1.4A drive it'd probably survive longer than the old 500hours.
I can see how they'd have a problem as it was 4 dies crammed in a small area without increasing the heat extraction surface and the Thermal resistance was really high compared to today's LEDs.

For the sake of discussion lets assume that the use of Cree LEDs are not suitable. Because sometimes for one reason or another it's not.

Based on the logical responses I've seen in this thread I'm not alone in wishing PL would release new LED tech in the older footprints.

As for my shoehorning, I got it to work extra legs gone and extra encapsulation gone. If you didn't know any better you'd think It was TFFC Lux 3. When I got it to look like a Lux 3 and it still worked is why I wondered why they didn't just make it the same shape as the lux 3 I guess you can get more LEDs out using a square mold than a circular one.

 

[gunga]

tebore, do you have any threads on your k2 work? I do a lot of seoul mods but am not super happy with the current tints. I'd like a 5A type tint.

I did purchase a quantity of high CRI seouls but would like more output on some lights...

Which K2's did you purchase? Were there any neutral (4000K is) ones you liked?

 

[tebore]

tebore, do you have any threads on your k2 work? I do a lot of seoul mods but am not super happy with the current tints. I'd like a 5A type tint.

I did purchase a quantity of high CRI seouls but would like more output on some lights...

Which K2's did you purchase? Were there any neutral (4000K is) ones you liked?

I didn't take pics of any of my K2 work but it's basically retrofitting lighting once powered by LuxIIIs. Ranging from a fixed lighting desk lamp to a few of my HDS lights.

The warmest bin I've gotten hold of lately is from Fred the TV0D bin. It's what I'd expect from a V0 tint. Besides the High CRI seouls I'm not happy with the tint and tint uniformity from the cool white P4s.

I'd like to get a hold of some warm or neutral white TFFCs. The spec sheets shows 80lm and 95lm @350mA respectively for the best bins. CRI of 75 slightly worse than Cree but again the advantages outweight it. I've tried some 6500K range K2s and they look more white less blue than equivalent P4s. To me the tints on K2s are more pleasing to work under.

When we're talking retrofits the K2s work really well with LuxIII optics and reflectors you get about 90% of the beam you had with the LuxIII where with the P4 you get much more spill and a bigger hotspot. I also notice that I'm able to get more light out of K2s than P4s at drive currents over 700mA. Not scientific but I had them side by side on P4 CPU heatsinks at I believe 800mA and the K2 looked brighter. The only advantage during my rough run was when I was going under 300mA specifically when I was targeting 100mA. The K2 seemed to show a higher VF than the P4 when I was at 100mA.

I didn't have time to get beam shots of my work but I guess when my schedule clear up I'll get a shot of a V0 K2 vs V0 P4 from my HDS.

 

[Cuso]

"Shoehorning" is the term???

I love this LED period, and I too wish the footprint was similar to LUX, until then shoehorn away..

 

[gunga]

Any ideas on where to get nuetral tffcs?

Are they a suitable retro-fit for classic lux 1 seoul mods?
How bad is the shoehorning?

 

[Black Rose]

Do K2 TFFC LEDs work well with smooth reflectors?

I have a Nuwai ALX-33K22L light I got from The Source that has a warm regular K2 in it and I'm thinking of swapping it out for a K2 TFFC.

 

[HarryN]

AFAIK (from the spec sheets) the K2 TFFC is only available in cool white. The K2 neutral and warm white are not TFFC - at least not yet.

As far as the shape - I think this was actually customer driven. There are a number of features on the K2 package designed to make alignment and fixing the location easier. There might be more to it than that as well.

I think it is unlikely that PL will come out with a 4 die package again like the Lux V in the near term. The roadmap presentations do not show such a package, and I actually directly asked this question to their CTO in a public meeting about a year ago. I didn't particularly like his answer, but I do respect a company that will stick to a plan.

I am pretty comfortable with Cree or PL packages - both are tier 1 companies with a competitive drive to make their products better every year. I do have more experience with the PL product lines, but each package has its advantages. LEDEngin has done a very nice packaging of the Cree die - perhaps better than Cree itself. I am happy enough with these suppliers to not bother testing SS - mostly a bandwidth challenge - you just cannot do everything.

 

[zzonbi]

"Any ideas on where to get nuetral tffcs?"

try led1.de

guess the idea was to generate more revenue, but the old round package seems more logical. a great opportunity for modders to cut corners though.