Nichia NSPW500GS-K1 ?

[Xe54]

Hi:

The data for this one and the DS are here:

http://www.nichia.com/product/led-lamp-phi5.html

The GS-K1 seems almost identical to DS. GS-K1 has slightly tighter CIE coordinates vs. current, though this might just be in the noise of the testing methods.

Can anyone make out any important differences between the GS-K1 and the DS?

 

[Gransee]

Check the lumen maintenance figures.

peter

 

[Xe54]

Check the lumen maintenance figures.

peter

There are none that I can see in the datasheets. Is there another publication which shows lumens vs. time?

 

[kongfuchicken]

The GS-K1 seems to be a triple die led in the same 5mm package.
That'd explain the disappointing 500hours.

 

[paulr]

DS says the same 500 hours except for the subfreezing (-30C) test. I wish there was a way to make these lights unpotted without sacrificing ruggedness. That would allow soldering in a new led when needed...

 

[Gransee]

My understanding is that the GS is designed specifically for flashlights. It has, if I recall, 2 dies in parallel. Nichia reduced the lumen maintance claims to play it safe. Having the dies in close proximity is going to increase heat. Nichia is usually conservative and a multi-die from them is going to out last a multi-die from anyone else. I plan on building it into the AAA when they become available.

I have some protos of course. The beam they produce is bar shaped with rounded corners (elongated oval). Lumens are about 1-3 more than the DS with the AAA driver. Runtime is similiar.

peter

 

[kongfuchicken]

The DS is rated for twice that much as well as every older model from there on...
That said, it's not a huge deal and Nichia is probably being conservative in the sense of minimal guaranteed performance as said before.

I highly doubt I'd be able to harm a LED to the point of non-function, especially when it's been made into a quality light such as the arc models.

Besides, I've got a bunch of Lux5s (rated at 500hours also) still kicking *** and taking names well after 500hours of continuous use. I'm sure they aren't exactly as bright as they once were but I won't be able to tell outside the lab.

 

[paulr]

Oh cool, another round of AAA improvements . I thought I read that the GS has 3 dice rated 20 ma each, so you'll be running within spec at the 45ma or whatever that the AAA now runs at.

 

[Xe54]

My understanding is that the GS is designed specifically for flashlights. It has, if I recall, 2 dies in parallel. Nichia reduced the lumen maintance claims to play it safe. Having the dies in close proximity is going to increase heat. Nichia is usually conservative and a multi-die from them is going to out last a multi-die from anyone else. I plan on building it into the AAA when they become available.

I have some protos of course. The beam they produce is bar shaped with rounded corners (elongated oval). Lumens are about 1-3 more than the DS with the AAA driver. Runtime is similiar.

peter

Where on earth are lumen maintenance data for CS, DS, or GS?

Also, there is no indication in the datasheet for the GS-K1 that it is multi-die, or slated to run on more than 20-30mA:

http://www.nichia.com/specification/led_lamp/NSPW500GS-K1-E.pdf

However, the web page that lists Nichia's 5mm LEDs says "20*3" under the current column:

http://www.nichia.com/product/led-lamp-phi5.html

Strange and confusing. Well I guess a call to Nichia would clear things up.

 

[Brlux]

I spoke with my Nichia sales rep about the 20*3 and she said it was a foot note see the bottom of the page. The comment attached to it makes no sense for being placed in the If column. They are now claiming a 3000 hour expected life and it looks like they rate the life as the point at which it hits 80% of original output.

My rep claimed that the GS is not a multi die part and Nichia has never made a multi die part? I don't know how informed she really is though? If you look down into the front of the part the two bond wires seem to be quite a bit farther apart than the CS parts I have but then again that could be an illusion created by the change in focal point with the slightly longer lens?

I would expect a multi die part to need more than 2 bond wires, perhaps they cut the die to be a larger surface area than their previous parts?

 

[chimo]

............perhaps they cut the die to be a larger surface area than their previous parts?

That would be my guess. I may sacrifice one in the name of curiosity.

Paul

 

[KobeBeef]

Hi, newbie here wave but a proud 2 year owner of an Arc-P AAA.

I was just curious if anyone knew what the meaning of the LED codes (CS, DS, GS, LS) were. Do they stand for something?

Thanks!

 

[chimo]

Hi, newbie here wave but a proud 2 year owner of an Arc-P AAA.

I was just curious if anyone knew what the meaning of the LED codes (CS, DS, GS, LS) were. Do they stand for something?

Thanks!

They denote different LED models. Try clicking the link in the very first post.

 

[KobeBeef]

They denote different LED models. Try clicking the link in the very first post.

Right, I knew that much. I've looked around the Nichia site, but I couldn't find any explanation of what the codes stand for. Are they acronyms? Or abbreviations for something?

I know company model codes are sometimes just arbitrary, but usually they have some kind of meaning. I was just curious if anyone knew.

 

[Brlux]

They are the model numbers for new LED's The newer the model the brighter and more efficient it is. They go in alphabetic order (BS is the oldest GS is the newest).

 

[KobeBeef]

They are the model numbers for new LED's The newer the model the brighter and more efficient it is. They go in alphabetic order (BS is the oldest GS is the newest).

Oh, OK. That makes sense. I guess I got confused with one of the Arc models using "LS" which is out of sequence and presumably not the LED code afterall. So, does the "S" stand for "shiny"?

Thanks for the reply.

 

[Brlux]

I think LS was an abbreviation for Luxeon Star. The ARC LS used a Luxeon LED instead of a Nichia. I can see how the confusion could be made.

 

[Moat]

Having the dies in close proximity is going to increase heat.

Actually, by doubling the die area while retaining the same total current rating as the CS/DS parts (20/30 mA) - the total heat produced would be the same, but spread over twice the area (of two dies instead of one). If anything, that makes me think these GS's will last longer - the ability of the LED's leads/legs to carry heat away likely being the common CS/DS/GS current/thermal "bottleneck" (as they appear the same size/cross-section as the CS's I have to compare to).

Doubling of the die area = half of the current flow across any given area of the chip junction = lower temp at any point in that junction = longer epoxy/phosphor life... maybe (the heat transfer being the crux).

Since LED's have higher luminous efficiency @ lower drive currents, my guess is that's where Nichia found the efficiency gains (and lower Vf, to boot) with the GS, over the DS - think of it as running two DS's in parallel, but at 15mA each instead of a single one running at a less efficient 30mA.

Maybe... :thinking:

 

[chimo]

I just finished disecting a GS. Here's an answer to the 1 or 2 die question.

Paul