Z-Power P7 Series - World’s Highest Brightness of 900 lm at 10-watt

[Juctuc]

http://www.led-professional.com/content/view/930/29/

900lm at 10watt,,,well well...no PDF and no information about availability. But i think CREE will come out with they own led about in1-2 weeks.

 

[Lightingguy321]

The SSC P7 at this point is not available to the consumer market and is exclusively provided to surefire only. It was a new LED created in conjunction of both Surefire and SSC.

 

[mds82]

I would love to be able to get my hands on a few of those .... cant wait till the multi core chips become available to consumers.

 

[evan9162]

Some comments:

The package is HUGE. 12mm = about 1/2" in diameter. This would be a tight fit even in a Mag reflector.

All 4 dies are in parallel. I think this has been covered before, but you're going to need to drive this thing with > 2A. At 10W, each die is using 2.5W - which is about 700mA/die (700mA * 3.5V ~= 2.5W). So for full power, you'll need a power supply capable of 2.8A at 3.5ish volts.

Look at all those bond wires. 9 total under the dome. If you thought the bond wires on the P4 were fragile, you now have 3x the chances to break one accidentally.

The apparent size of this light source is going to be HUGE. It'll be bigger than a Lux V by far. The dies are spaced out from each other, AND the phosphor deposition is far larger than the dies themselves. This will be quite difficult to focus into a tight spot - this will be a good floody device though.

The giant package size and huge die spacing are likely compromises to ensure proper thermal performance.

I'll be fun to get ahold of these to try them out.

 

[SemiMan]

Re: Z-Power P7 Series - World's Highest Brightness of 900 lm at 10-watt

Who wants to question the likelihood of Seoul delivering 90 lumens/watt at 2.5watts? Considering the R2 only delivers about 80 lumens/watt at 2.5 watts (700mA) it seems unlikely they could deliver these in any sort of volume where they actually hit those specs. Of course when they heat up in a real light fixture, that 900 suddenly becomes 800......

 

[jirik_cz]

Datasheet can be found here http://zled.com/en/product/prd/zpowerLEDp7.asp

This image is a snapshot from the datasheet


900 lumens is only maximum value, not typical...

Some quick calculations 700-900lm at 10-11,8W -> 60-90lm/W (2800mA)

 

[matrixshaman]

Re: Z-Power P7 Series - World's Highest Brightness of 900 lm at 10-watt

Think residential lighting - it looks like it's about 50% more efficient than flourescent and after seeing a big article today on mercury warnings for flourescent bulb replacements I think this would be the logical answer. Assuming production can be ramped up enough to keep cost reasonable I think if they put these into a 120 volt bulb replacement setup they will sell like crazy if promoted properly. The will of course even outlast flourescent if they have a life anything like other LED's. I predicted about 8 or 9 years ago that within about 10 years LED's would be replacing household lights - looks like it has a good chance of happening or at least getting a good start soon.

 

[IMSabbel]

Some comments:
The apparent size of this light source is going to be HUGE. It'll be bigger than a Lux V by far. The dies are spaced out from each other, AND the phosphor deposition is far larger than the dies themselves. This will be quite difficult to focus into a tight spot - this will be a good floody device though.

I'll be fun to get ahold of these to try them out.

That may be, but i see the other side: at 900lumens, this is an adequate replacement for a fixed lighting point.
No messing around with multi-emitters needed anymore.

Edit: should have read the thread to the end...

also: Think about economey of scale: How cheap are normal light bulbs, even though they have a fragile thungsten coil in a delicate glass housing?
Or how we can create computer chips with 100s of million transitors that sell for $30.
As soon as you sell those leds in 10s of millions instead of tens of thousands, they _will_ be cheap. (Just compare to the price drop in compact fluorescents, which are vastly more complex devices).

 

[jtr1962]

As soon as you sell those leds in 10s of millions instead of tens of thousands, they _will_ be cheap. (Just compare to the price drop in compact fluorescents, which are vastly more complex devices).

Yep, I've been saying for a long time now that emitters like this will probably hit the 25 cent price point once made in huge quantities. The raw materials cost less than those in a 25 cent incandescent lamp. It's just a matter of the figuring out how to manufacture these inexpensively. That WILL happen sooner rather than later. Once it does, the incandescent lamp loses it's only real advantage-initial sales price.

 

[SemiMan]

Re: Z-Power P7 Series - World's Highest Brightness of 900 lm at 10-watt

I would guess you could make 4 Rebels for less money than this device.

In terms of volume manufacturing, making a good LED is still a complex process...... a heck of a lot more complex and exacting than a light bulb. It takes a while to properly diffuse a semiconductor using expensive equipment, unlike the lightbulb which is quite simple.

Keep in mind there are many semiconductors that are made in the 10's of millions that are relatively simple that do not cost $0.25.

Semiman

 

[2xTrinity]

Think residential lighting - it looks like it's about 50% more efficient than flourescent and after seeing a big article today on mercury warnings for flourescent bulb replacements I think this would be the logical answer.

I presume you're talking about cheap CFLs -- those have ballast losses, and self-shadowing losses due to the twisted tube. Linear fixtures can already be over 50% more efficient than CFLs.

However, these LEDs have a couple potentially useful applications. Even though they might not be able to focused into super-narrow-angle flashlight beams, they could certainly be packaged to replace household halogen spots and floods -- two applications fluorescents can't easily replace.

 

[Gryloc]

Ooh, more news, meaning the Seoul P7 is getting closer to be in our hands. Cool...

Now, What is the deal with Lumileds? I wonder why Lumileds doesn't release something similar, but either in the form of the LuxV package (or scaled up a bit), or in a larger ceramic substrate like the Rebel and Cree XR-E. Imagine a substrate, say 6mm X 6mm, with 4 TFFC dies bonded in the center, and connected in series with each other. When you look at the big picture of the Soul P7 in the above link, you will notice that they had a problem with the bond wires going to each die. The Seoul P4 uses two bond wires going to the top of the die, while the P7 uses only one each (to keep the bond wires from obstructing the light any further). With the TFFC technology, all electrical contacts are below the die (connected to the top of the die by multiple small vias), and there are not any obstructed views! Die placement can also be closer (if the power density is not a problem). You can then have a smaller package, a nicer beam (for the most part), and the package would only need a small glob of silicone to cover the dies that may be slightly larger than the normal Lumileds dome. Wires are soldered to the top or bottom on pads like the Cree XR-E and sort of like the Rebel, so fragile leads (like on the standard Lumileds Luexeon emitter and the Seoul P4) do not break off, which usually renders the emitter useless.

Does any one else wish for something similar? Maybe if Cree jumps on soon, Lumileds will feel pressured to update their multi-die emitters. With the above substrate, you could also increase the number of dies for increased brightness (considering the power consumption and power density is not too high). 3 X 3 dies would be wonderful, to compete with the nearly lethal OSRAM OSTAR. Lumileds can then compete in the market of forward automotive lighting, and also for light sources for projectors and rear-projection televisions. They can even spread out the dies to make it more suitable for fixed lighting. I guess a handful of Rebels could be used for fixed lighting .

There I go: dreaming... and ranting. Why can a fellow's dream come true? It can be done now, and I bet that demand for such an emitter has to be high (if the awkward Seoul P4 is getting such a warm welcome), so why not try to release products like this? Come on, Lumileds! <sigh> :sigh:

-Tony

 

[skalomax]

They're Here

http://www.cpfmarketplace.com/mp/showthread.php?p=2129735#post2129735

 

[evan9162]

Crap!
They must have changed the construction. HgRyu's photos show 17 (yes, 17!) bond wires under that dome!

It also shows that the dies are much more closely clustered together than the PR photo. That will make focusability better, but concentrates the heat more in the center of the heat sink slug. There is still an excessive amount of phosphor compared to the die size.

The P7 will have the same polarity issue as the P4 did - the slug is electrically connected to the positive power lead - so you'll have to isolate the slug of this guy just like the P4.

 

[tebore]

One of the dies look kinda off in HgRyu's pics. In his pics the bottom right die is slightly crooked, I hope later LEDs don't have a problem with alignment.

 

[precisionworks]

I think if they put these into a 120 volt bulb replacement setup they will sell like crazy if promoted properly

They currently make a model called the Acriche, which has a VF of roughly 110v or 220v:

http://www.seoulsemicon.com/en/product/prd/acriche.asp

 

[IMSabbel]

I would guess you could make 4 Rebels for less money than this device.

...
Keep in mind there are many semiconductors that are made in the 10's of millions that are relatively simple that do not cost $0.25.

Semiman

Not if you include the cost to assemble them together to form a single light source.

And you just cannot compare the complexity of semiconductor devices like microprocessors with leds.
Its like comparing spear throwing with intercontinental missiles.
Even creating the source wafers for the chips is more work than doing a led will ever be. (not to mention that 8-12 stacked layers of lithographically created nanostructures).
The fact why you can get a 1Gbyte USB-Stick for less money than a like CREE XR-E Q5 is simply the fact that the R&D budget of a single of the "big boy" Semiconductor companies, for a single year, is bigger than ALL white led R&D _ever_.

 

[SemiMan]

Re: Z-Power P7 Series - World's Highest Brightness of 900 lm at 10-watt

Not if you include the cost to assemble them together to form a single light source.

And you just cannot compare the complexity of semiconductor devices like microprocessors with leds.
Its like comparing spear throwing with intercontinental missiles.
Even creating the source wafers for the chips is more work than doing a led will ever be. (not to mention that 8-12 stacked layers of lithographically created nanostructures).
The fact why you can get a 1Gbyte USB-Stick for less money than a like CREE XR-E Q5 is simply the fact that the R&D budget of a single of the "big boy" Semiconductor companies, for a single year, is bigger than ALL white led R&D _ever_.

I would disagree. The small source size of the Rebel is going to result in very small optics. I will be forced to large optics for the P7. Given the hassle I see of placing the P7, isolating it, etc. I am guessing I could put 4 Rebels down on FR4 fo less money.

Cost in semiconductor manufacturing is more than just the number of diffusion layers and geometry though that certainly has a big influence in terms of up-front capital cost. There is certainly enormous R&D spent on fine geometry and large wafer sizes.

That said, up to this point, even though 12" wafers are available, 8" is mainstream, and 6" is old tech, LEDS are still predominantly manufactured on 2" wafers, with 4" wafers just coming online. You could pick up old 6" equipment for very little, but the reality is that the processes are different and complex in different ways. To make a good LED wafer requires very tight control of the diffusion over the whole LED surface. Combine that with different defect requirements, optical integrity, back-thinning, surface treatments, etc. and even though there is a huge knowledge of semiconductor processing to draw on, a lot of it can not be used. The reason for using approximately a 1mm*1mm die was much in part a trade-off between how much light could be gotten out of a single die and picking a die size that would result in a suitable yield.

Oh, and I was not comparing 1gig USB sticks to LEDs, I was thinking more along the lines of power-semiconductors (mosfets) and microcontrollers that are made in the 10's of millions like LEDS and have a similar cost structures ($1.00 - 2.00). Those are often made on 8" wafers by the way.

 

[Spypro]

Re: Z-Power P7 Series - World's Highest Brightness of 900 lm at 10-watt

Hmmm... I can image 5 of those P7 in a Mag. It would be incredible.. and the size of the heatsink would be incredible too

 

[tebore]

I guess we can start planning super heatsinks made from solid copper and having pins to remove the heat.