Philips 75w equivalent LED lamp, 1100 lumens

[PhotonWrangler]

The latest version of the Philips EnduraLED remote-phosphor lamp just showed up at our local "orange" home improvement store. This is the lamp with the three yellow plastic lenses. I picked one up and I'll be comparing it to the 800 lumen 60-watt equivalent version of this lamp. Paid $40 for it.

[purduephotog]

The latest version of the Philips EnduraLED remote-phosphor lamp just showed up at our local "orange" home improvement store. This is the lamp with the three yellow plastic lenses. I picked one up and I'll be comparing it to the 800 lumen 60-watt equivalent version of this lamp. Paid $40 for it.

I've been waiting to hear if it has 18 LEDs in it or 24. With the larger heat sinking area I could also see them upping the current by about 20%.

[netprince]

Subscribed, I'm curious about your findings.

[PhotonWrangler]

Subscribed, I'm curious about your findings.

I've managed to pry up one of the yellow phosphor lenses and peeked underneath it. They're still using six rebel LEDs per lens for a total of 18. The layout appears to be identical to their previous 60-watt equivalent shown in this teardown report. I'm guessing they're pushing the LEDs a little harder. I plugged it in with the lens partially pried up and those blue LEDs are intensely bright!

More details to come.

[Steve K]

any signs that the heatsinking was increased/improved in order to accomodate the increased light output?

Steve K.

[PhotonWrangler]

The body of the lamp is longer and it's noticeably heavier, so I believe there is more heatsinking.

[blasterman]

Prior ones got too hot to touch as I recall, so this is a good thing.

Still curious why they're using so many rebels at low current -vs- fewer ones at higher current and what the configuration solves from an engineering standpoint. Can't gain that much efficiency I would think, but there has to be a reason. Obviously using so many Rebels doesn't increase the price - it is a Phillips bulb BTW :-)

[PhotonWrangler]

I'm thinking that using multiple smaller LEDs helps to tailor the beam pattern as well as guarding against catastrophic failure of a single large LED. In this sense it's engineered like an LED traffic light - a single LED failure won't render the device useless.

[blasterman]

Mmm.....maybe. Still, my experience with power LEDs is that they simply don't die unless the power supply goes bad or they are given pretty severe mechanical shock, and in this case it would nerf all the other LEDs. I don't think redundancy is the issue here.

Maybe it has something to do with power. Bunches of LEDs would allow for a much higher forward voltage and smaller total current loads. That makes for a simplier and less complex driver, but would require the expense of more LEDs, which obviously aren't an issue because it's nit like Philips is buying Crees.

Curious if PhotonWranger can make any sense of this with the bulb open. Anyway to determine how the arrays of LEDs are wired; number in series, etc.

[PhotonWrangler]

Curious if PhotonWranger can make any sense of this with the bulb open. Anyway to determine how the arrays of LEDs are wired; number in series, etc.

I haven't been able to pry the lens all the way off yet for fear of breaking the hinges off. However if you look at the 6th photo in EDN's teardown report you'll see that each cluster of 6 LEDs is wired in series. Beyond that I don't know if the three clusters are in series or parallel with each other. If I can manage to get it apart without destroying it I'll let you know.

[patrickhuang]

The body of the lamp is longer and it's noticeably heavier, so I believe there is more heatsinking.

Nice information!
How is the temperature of heat sink and remote cover ?
are PCB boards still ceramic ?
I think maybe the Rebel were over drive to get more light output.

[PhotonWrangler]

Here are a couple of quick comparison photos that I took tonight. The 17W bulb is visibly brighter than the 12W one. Temperature of the lenses on both bulbs was subjectively the same (my IR thermometer is broken) but the neck of the 17W bulb was slightly cooler than the 12W, indicating a significant improvement in thermal management.

[Vesper]

Thanks for this info and the nice pics. I was staring at these today at the local orange wondering about them. Sounds like they're dialing in the heat sinking - good news.

[patrickhuang]

Here are a couple of quick comparison photos that I took tonight. The 17W bulb is visibly brighter than the 12W one. Temperature of the lenses on both bulbs was subjectively the same (my IR thermometer is broken) but the neck of the 17W bulb was slightly cooler than the 12W, indicating a significant improvement in thermal management.

Could you please take a photo about the LEDs?

[PhotonWrangler]

Could you please take a photo about the LEDs?

If I can get the lens popped off I will take some pictures of the LEDs inside. So far I've only been able to pry it up a little bit for a quick peek. They look just like the Rebels shown in the EDN teardown report mentioned earlier in this thread.

[ratsbew]

I think you have to pop off the top cover.

[PhotonWrangler]

I think you have to pop off the top cover.

I've been trying. It's either glued shut or it's a pretty tight fit with the top latches on the lenses.

[purduephotog]

Here are a couple of quick comparison photos that I took tonight. The 17W bulb is visibly brighter than the 12W one. Temperature of the lenses on both bulbs was subjectively the same (my IR thermometer is broken) but the neck of the 17W bulb was slightly cooler than the 12W, indicating a significant improvement in thermal management.

Or that the heat isn't getting out fast enough and instead rising around the dies themselves.

I'm assuming you had this on about steady state 1hour ish? (Sorry, please don't read active dispute tone in this note- just stating the old issues).

[Theatre Booth Guy]

This thread gave me an idea to try one of these with the yellow covers removed as an anti SAD light. I had read in a few places that blue light is supposed to work best. Tonight, I got the bulb prepped and tomorrow, I'll plug it in and see if it makes me any happier :-)

[purduephotog]

This thread gave me an idea to try one of these with the yellow covers removed as an anti SAD light. I had read in a few places that blue light is supposed to work best. Tonight, I got the bulb prepped and tomorrow, I'll plug it in and see if it makes me any happier :-)

DO NOT look at the dies. They can cause problems with your vision- so use it in a bounced configuration.

[Kinnza]

I think the point of using so many LEDs at low current is about thermal management. On incans replacement, the main problem is to achieve junction temps low enough given the limitation on heatsink surface area. By running each LED at lower current, you can run them on a hotter heatsink while keeping Tj low enough to guarantee a lifetime over 25Kh. Aditionally, it achieves higher efficiency, thus reducing the total watts required to get the target light output and more important with the limited heatsink area, the total thermal load.

[blasterman]

I know that spreading lower powered LEDs round -vs- fewer higher powered ones in theory keeps the on-die temp lower, but I'm still skeptical of this. Your heatsink efficiency, if run long enough, is ultimatley a factor of energy in -vs- radiating effiency. Using a dozen lower powered LEDs doesn't allow you to cheat because the total energy being handled by the sink is the same. The heatsink will heat up to the exact same temp and present each LED with the same surface heat.

However, that assumes were talking about heatsinks with a decent base thickness. If the heatsink is too thin then heat generated by the LED's 'pools' up in proximity on the heatsink because it can't move away fast enough. In this respect using more LEDs and lower wattages might have a thermal coefficient via the limited convection of one of these bulbs to buy you a few degrees. Given these bulbs run really, really hot that might be critical.

Given the increasing linearity of high powered LED efficiency I'm not sure if the improved efficiency would amount to much, but not sure how much a difference we're talking about. Possible they are using many smaller LEDs because they are that much cheaper.

[Kinnza]

Its clear that the heatsink will heat the same at the same heat load, if its well designed. But in thermal management, there is a near fixed figure which not depends of the heatsink temperature, the thermal resistance of the LED package. From solder to junction, the temperature raise is linear with the power of the LED.

So for a given heatsink temperature, you get lower junction temps using more lower power LEDs than fewer more powerful ones. The Rebel has a thermal resistance solder to junction of ~10 ºC/W. If you use one at 700mA, it has a typical power of 2.24W (0.7A*3.2V), thus junction temp is about 22.4ºC over heatsink. But if you use 2 Rebels running at 350mA, each one runs at just 10.5ºC over heatsink temp.

In this lamp, Rebels runs yet below 350mA, so the junction temps are below 10ºC over heatsink temp, allowing heatsink to run really hot without compromising lamp lifetime (at least, considering just LEDs reliability).

It is true that lately linearity of emission at each current is way better than in the past, but it not translates directly to similar efficiency, as efficiency is a function of total power, not just current. As current-voltage curve is flatter at lower currents, the gain on efficiency is not just achieved by the better efficacy emitting light at lower currents, but by the reduced voltage. For example, in the Rebel, still if the emission would be completely linear from 350 to 700mA, you get a gain on efficiency of 7% just due the lower voltage. As emission is not completely linear, the whole gain on efficiency is very noticeable, from 10 to 20% (it depends of the thermal management, as different power affect temperature, and temp affect efficiency) depending of the exact current compared.

I guess by using Rebels below 300mA as Phillips designed this lamp, they get an efficiency gain of minimun 10% as compared to higher currents (not 700mA, but lower), allowing to use less total watts and run junction temps low enough to get very good lifetime (if driver is designed accordingly).

It would be very interesting to know the heatsink temperature of the lamp after several hours of operation. We could make some guesses about expected degradation with time knowing it, and confirm if they were in the limit of heat load when decided touse so many LEDs. Maybe they just used so many because it is cheap for them and they want this lamp works as the flagship of LED lamps into the residential market, by using as little watts as possible to get the target output and with excelent reliability.

True 75W incan replacement with 17W and true long lifetime is the way most consumers accept the technology: 40$ is not excessive to try, for example on a place were the lamp is working many time but with frequent switching on/off. I dont see that price excessive at all, although of course it should drop with time to a price competitive with CFLs.

[patrickhuang]

.......

True 75W incan replacement with 17W and true long lifetime is the way most consumers accept the technology: 40$ is not excessive to try, for example on a place were the lamp is working many time but with frequent switching on/off. I dont see that price excessive at all, although of course it should drop with time to a price competitive with CFLs.

Additionally, lm per w of LED chip is upgrading every season, maybe Rebel in 75w bulb(2012) is better than Rebel in 60w bulb(2010).

[SemiMan]

I would add to Kinnza's excellent explanation that more LEDs at lower currents may also be a matter of economics. Whether Lumileds is a Philips company or not, someone still needs to pay for those LEDs either in direct costs or lost sales.

It may be cheaper to run 18 lower binned parts at lower currents to achieve effiency X, then to run fewer higher binned parts at higher currents to get the same effiency X. Premium bins command a premium which may be beyond the benefit of using less.

Semiman

[Theatre Booth Guy]

DO NOT look at the dies. They can cause problems with your vision- so use it in a bounced configuration.

Yup, they are bright. I'm concerned about the content of UV with the yellow covers removed.

[don.gwapo]

Yup, they are bright. I'm concerned about the content of UV with the yellow covers removed.

+1 coz i'm using a led bulb too with the white frosted cap removed. I'm using 40w equivalent @ 510 lumens with 18 leds in it.