Philips LED bulb: Tear-down Part I (light patterns)
Philips sent me one of their 12.5W dimmable LED bulbs to evaluate. (This is Philips L-Prize entry, for those following the DOE contest. WRONG, see below.) Tear-down time!
Some of you have mentioned that long posts with lots of pictures are too painful at slower download speeds, so I’m breaking this it into installments because it’s both lengthy and has lots of photos. Here’s a preview of the main points of this tear-down:
- There is a variation in light patterns created by light bulbs, and manufacturers have found some creative solutions to mimic the near-360-degrees of an incandescent light.
- Philips jumped through a lot of hoops minimize hand soldering in its new bulb.
- A look at which IC vendor’s power management chip is in the bulb.
OK, first, some field-testing was in order, because that’s a pretty odd-looking light-bulb.

Granted, CFLs are *really* odd-looking…

…and “sno-cone” LEDs likewise are an acquired taste.

Come to think of it, the first Edison socket light bulbs were probably greeted with head-scratching by a public used to oil lamps and gas light.

My concern with the Philips bulb was what kind of light pattern did it produce? An incandescent bulb is a uniform, nearly global source of light. Intuitively, the sno-cone light will produce a semi-sphere of light, and the CFL – well, I didn’t really know how light was dispersed from a CFL. So my first test was to find a way of comparing the bulbs’ light patterns.

This is the home-made jig: It’s basically a slanted rim – a conical section of very light cardboard - that encircles the light bulb, reflecting its light over the 360 degrees around the bulb. What were the results? Here’s the familiar global light distribution of the incandescent bulb:

The CFL had a fairly universal light distribution also, although slightly less than the incandescent.

The sno-cone LED light was less than half of a semispherical pattern, which you’d expect because the LEDs rest on a flat surface, pointing up with little to no secondary optics to direct the light to the sides or down.

Now take a look at the Philips light: Virtually the same as the incandescent light:

How does the Philips bulb create such a spherical light pattern? We’ll take a closer look at the bulb in the next part of this tear-down (which I hope to get posted later today.)
UPDATE: Both Maury Wright and reader Navigator below in the comments point out that the L-Prize bulb is a different bulb. I checked with Philips and got this prompt reply: “The L Prize lamp that is in consideration is 900 lumens, less than 10 watts and is not commercially available as the DOE is still in the process of considering it.
The Philips EnduraLED is 806 lumens, 12.5 watts and currently commercially viable. We took what we learned with the L Prize lamp to bring a viable 60 watt equivalent to market.”
Back to uploading photos for the next part…
WDB commented:
"WDB commented on Feb 15 2011:
I bought a bunch of these bulbs, I am testing one of them on a dimmer, every night it runs about 5 hours at 100% and then it is dimmed to 40% and runs another 8 hours dimmed. I have killed off lots off dimmable CFls doing this test. I hope these bulbs hold up."
After a year plus this bulb is still going strong.
CEDUP commented:
Osram owns teh former U.S. based Sylvania lighting, if ya want to make claims that Philips is not U.S., even though they have North american Philips lighting based in Jersey. Osram has their Sylvania div in MA. But if you claim they are really German, then no company can do the L-Prize contest, Philips won anyway, since they are ahead of everybody in LED. GE, seems dormant in LED, even though they had a lot to do with the original stuff in the 60's. Philips see's the future in LED, they are perfecting it.
CEDUP commented:
The U.S. Philips factory in Kansas is the largest fluorescent lamp factory in the world. Most T8 and T5 I think sold here are from there. Philips LED Lumileds are the best LED made. Color, quality, lifetime. Philips invents and pefects, then others follow.
jshuler commented:
I have purchased this bulb and it is in use in my daughter's bedroom. It totally blows away the CFL that was in there before -- plenty of beautiful light, dimmable, instant on, no hum. She loves it.
CEDUP commented:
Philips is the leader in lights, biggest and best. they have superior research capacity worldwide. Philips lighting is based in the U.S. as the U.S. div of Philips lighting. Once they perfect stuff they patent and license it, so other mfgs can make it available, Philips always invents the latest and greatest. Always. From inventing the CFL , car halogen, Cd, DVD, low pressure sodium, Xenon car lamps, to the FIRST LED headlight in Audi R8.
Jim commented:
Looking forward to the teardown
Andy T commented:
Sorry Gene. Place of manufacture is irrelevant in this case according to the LPrize rules.
When Philips clicks its heels twice, and says "there's no place like home", they wind up in the Netherlands, not Kansas.
Gene commented:
Note to A.T. Philips has a very large manufacturing plant in Salina, Kansas. That's close to the center of the continental USA. Also the Philips lamps are marked "MADE IN THE USA" vs China on the cfl's.
joejoe commented:
I side with Andy T: please don't break up articles artificially.
WDB commented:
I bought a bunch of these bulbs, I am testing one of them on a dimmer, every night it runs about 5 hours at 100% and then it is dimmed to 40% and runs another 8 hours dimmed. I have killed off lots off dimmable CFls doing this test. I hope these bulbs hod up.
ernest demaray commented:
The full L Prize requirement information is at www.lightingprize.com. But the real prize would be 50% electric to light conversion and near a CRI of 90, which is also difficult, that would be around 175 lumins/W, not the paltry L Prize sec.
Andy T commented:
L-Prize is only open to US entities. Royal Dutch Philips is based in the Netherlands. Methinks 'tis nothing more than a PR ploy, not to mention getting endorsement, at our expense as taxpayers, by our stupid government (and Time Magazine).
Amcfarl commented:
A simple but ingenious way of comparing spread, Margery! But I would like to see the same test repeated in the Z-axis with the Philips as these 3 large metal indents must cause some shadowing. And since Philips/Lumileds apparently don't yet sell and LED with high enough efficacy for the L-prize they must be using special lab samples for this, which they can't make in volume for the Enduraled.
LB commented:
I agree with Andy T, Margery. Keep posting the full content. I hate two or more part series. For those with slow access in today's world, "Let them eat cake!"
Your tear apart articles are really great. As a test engineer, I think your illumination ring test rig is the best setup for bulb testing I seen yet!
Navigator commented:
A correction that this bulb is not a LPrize bulb. This is a 60W bulb that consumes 12.5 watts. The LPrize submission is 10 W( 900 Lm)
B commented:
are there other brands available? aside from Philips...like Sylvania, Lights of America, etc...to compare with Philips'
Larry commented:
That's a very clever and effective test jig.
I wonder how many people like me are going to be put off buying the Philips bulb because the lens appears so opaque? Only your picture convinces me that it really does work well.
GK commented:
Watt are the wattages or equivalent wattages?
Is the rest of the article done?
Thanks for your work.
bert commented:
I really like this article. The information is very helpful to really see how light bulbs perform. Can you provide light bulb pricing information and energy use information.
Andy T commented:
I have news for the whiners. You are going to spend the SAME amount of time downloading all the photos if they are spread across multiple blog entries.
If you can't download at a decent speed, you are not in a legitimate company or engineering effort...or you live in China in which case you have people who can do your teardown for you for a total labor cost of $0.25
I suggest you put them all up next time, Margery. They can always nav away from the page if it's too much for them. For the rest of us, it's never enough in terms of pics and resolution.
It's not like you to be a tease :-)















