Looks interesting, but for the data disclosed in the news, the 5050 60W 6500lm spec doesn’t seem to be much of a game changer compared to ChangDa’s 5050 SFQ60 (4x60mil), rated 3V20A 5000lm, max 3V40A 16800lm.
Also I contacted a couple of sales, and unsurprisingly there are no samples for sale at the moment, but they implied that the price is roughly above 100CNY(SFQ60 8.5CNY per), the Light Emitting Surface area is almost same as the package size, uses 16 chips (probably 16x43mil chips). Also, the sales refused to disclose its max power, but that’s probably what we’re most concerned about.
Edit: I just realized that the illustration in pic 1 is actually what a single emitter looks like, not 16 emitters. There are 16 chips and the LES is equal to the package size.
Edit2: One of the sales told me they will be used on DJI emergency lighting drones, maybe it’s intended to use smaller package and MCPCB for use on smaller devices? After all, with such a huge LES, even if the output is high the intensity will be still low that it’s hard to tell what its purpose is.
The data they provide, as you pointed out, is nothing to write home about. 6491.6lm (how can they be confident about so many significant figures?) per 25mm^2 is around 260lm/mm^2, which is really bad for something intended to be high intensity. A SBT90.2 manages around 600lm/mm^2, SFT25R does 800+, and Osram W1 does around 1000.
At the same time, if one fixes power density and looks at efficacy: at 260lm/mm^2, a SFT90 makes 2500lm with 118lm/W, better than the 104.7 of the new emitter. A SFT25R makes 585lm at 146lm/W.
Looks like a cool idea, but the implementation needs a lot of work to even compare to existing offerings in performance or price–the much better thermal conductivity of diamond does not seem to come across at all.
Generally larger LES emitters have lower power density, but I’m not aware of any evidence pointing to the issues you mentioned being worse than smaller LEDs.
I don’t know a lot about the San’an emitters, but even in classically known emitters like XHP70.3 HI, more than 8,000lm is achievable in a LES less than 4mmx4mm square. Quite a bit better than 6491lm over 5mmx5mm.
I think one of the main benefits is that with proper heatsinking, the brightness would remain more constant over the life of the LED. In 5 years of continual operation, it would probably be brighter than its xhp70.3 competitor…although arguably better tech should be out by then.
The website mentions use in projectors, I hope they aren’t planning on sticking them into cars, lol …
The first implementation of a new technology rarely is groundbreaking. Give it some time.
They pushed some new tech and likely wanted to be the first on the market with it, so they’d rather launch it with a mediocre implementation than launch it perfectly, but only as second or third company.
We’ll see how the second/third generation of this looks, and if other LED mnufacturers pick it up as well.
The main issue of 5050 and 3535 LEDs right now is heatsinking. This tech could in theory allow a 3535 LED with potentially 5050 class performance, and a 5050 LED with 7070 class performance (overcurrent potential, efficiency). Maybe even higher.
To be more exact: the thermal resistance between chip and substrate/thermal pad. For small LED chips in particular, this is one of the reasons why luminance is currently hardly increasing any further. More than 250-270 cd/mm2 is currently hardly possible, and if it is, then usually with a greenish light color (which is required for very high efficiency)
Interesting point–the longevity of LEDs (particularly under overcurrent) is rarely studied here. Performance and longevity are correlated though: an LED with good heatsinking (low thermal resistance) demonstrates it with high efficacy, which this new LED does not have. It would probably be a while before a decent implementation of this new tech emerges.
These LEDs with the large LES and low intensity should not be used in car headlights–doing so results in a beam that either (1) does not light up very far ahead or (2) blinds lots of oncoming traffic.
They’d have to incorporate it into a projector style headlight, perhaps an indirect “recoil thrower” style.
*the led’s could sit on a flared surface, a few degrees slightly back towards the reflector assembly, this would slightly increase the heatsink capability of the mounting post, and wouldn’t look quite so obnoxious as a thin post coming up from the bottom of an assembly.
I doubt it, it just not related to the science of throwing, it’s just that cars have enough places to house all kinds of cooling structures without the need for these diamond emitters that are so special and costly. Though as always, the website describes that they can be used in cars, streetlights, projectors, etc.
Ofc I have no doubt that luxury cars will use them, but they can only be used for low beam. Especially now that cars are no longer popular with big eyes, according to the current disclosed data, they cannot be used as high beam.
Diamond-based Emitter, in fact, we have studied the thermal conductivity of diamond quite a long time ago. Its thermal conductivity efficiency is indeed very good. However, the thermal resistance bottleneck of LED’s thermal conductivity does not lie in this aspect. When applied in practice, it only improves by 3%.
