A Partial Teardown of a Philips LED Bulb

Nov 2013

I never liked compact fluorescent bulbs in principle because they contain mercury and because of lingering fears that the bulbs emit ultraviolet which can bleach dyes and possibly contribute to cataracts. LED bulbs don't use mercury, and some experimentation suggests that they do not emit ultraviolet or even violet light. LED bulbs are now at a price where their alleged lifespan and documented efficiency make them a better buy than CFL bulbs to replace incandescent bulbs, at least for normal purposes.

The Philips AmbientLED "Alien Head" bulb. Not the subject of this teardown. Buy more of these bulbs so Philips will continue making them.

My first LED bulbs were the three-lobed Philips "Alien Head" 60 Watt equivalents (see picture at right). I find them to be attractive with a cool new-technology look to them but apparently normal consumers don't like them. I've seen stock sitting in the local Home Depot with dust on the packaging; yes, really, they are not selling. That's a shame as what research I've seen online suggests that they're technically good bulbs. I find the Alien Head bulbs so attractive that I haven't been able to bring myself to do a teardown of any of them for fear of damage.

I bought a couple of the newer Philips 60 W equivalents, model 9290002268. Like the Alien Head bulbs these are 2700 K and use Luxeon Royal Blue LEDs to cause a remote phosphor to fluoresce; the same technology in a redesigned body. The remote phosphor is now in the shape of a yellow egg-yolk dome but is hidden from view by a frosted glass envelope; the bulb has a similar profile as the Alien Head bulbs but looks "normal" enough to not scare off consumers.

A Happy Accident

The only problem I had with these newer bulbs is the packaging. Like the Alien Head bulbs these are packaged in that annoying super-tough moulded PET plastic shell whose halves are welded together to make it impossible to remove the bulb by hand. Unlike the Alien Head bulbs, which are metal and plastic, these bulbs are plastic and thin glass — in a word, fragile. You can guess what happened, right?

In my excitement I must have gotten a little too energetic when cutting open one of these packages and bending the plastic out of the way to get at the bulb. The tough plastic packaging snapped back and hit the glass part, punching a hole into it! This was irritating but not a setback for me as I was going to break one of the bulbs anyway for access to the remote phosphor dome and LEDs. This experience simply led to the experiment taking place sooner than planned. If you don't want a broken bulb, take away this warning: Cut open the packaging completely with scissors so that no force is needed to remove the bulb. Then don't drop it.

As an aside, I wonder if a glass-envelope LED bulb is more likely to shatter than an incandescent if it falls with the glass facing down. An incandescent doesn't have much mass but an LED bulb has a lot of mass in its electronics and heat sink. If a glass-envelope LED bulb is dropped on its head then that glass would seem to have to handle an impact that has much more energy than an incandescent bulb would without shattering! Food for thought?

Blue Light Hazard

Words of warning: If you don't know about the Blue Light Hazard then, please, do some reading before operating a disassembled LED bulb. Without the remote phosphor dome in place this bulb emits a powerful, intensely blue light: The Luxeon specifications indicate a typical peak at 447.5 nm. This is in the important region for the Blue Light Hazard. It seems that a Royal Blue LED could be the poster child for the Blue Light Hazard — so BE CAREFUL!

Partial Teardown

This is only a partial teardown because I'm not interested in the potted driver electronics, only the light emitting parts: The LEDs and the remote phosphor dome. I'd also like to be able to use the bulb once I'm done with this experiment.

Fig. 1. An undamaged bulb and one with its glass envelope removed.

Figure 1 shows an undamaged bulb on the left. The bulb to its right has had the glass envelope removed after it was broken. The cement holding the glass to the plastic is slightly flexible which allows one to tease out the glass fragments with gentle prying. Care was taken to ensure that all of the fragments were removed to avoid accidental cuts in the future.

I didn't test whether or not heat would soften the cement but considering that the heat sink has been observed at over 80°C in operation, I wouldn't expect heat to soften the cement at all.

Judging from appearances, I suspect an unknown Philips designer had some fun with the concept of this bulb before being forced to put the glass envelope over the yellow dome, ruining the visuals of a good hard-boiled egg joke!

Fig. 2. The naked Royal Blue LED array and the remote phosphor dome attached to its plastic carrier.

Figure 2 shows the bulb after the white plastic carrier holding the remote phosphor dome has been removed. The dome is cemented to the carrier which is held in place against the heat sink by three plastic tabs on its bottom which mate with the three slots visible close to the edge on the top of the heat sink. Cement applied to hold the glass envelope to the plastic body may also stray to the edge of the plastic carrier and adhere to it.

I don't know if I broke one of the three tabs (unseen beneath the carrier) while prying out the glass shards or whether or not it had already been broken by careless assembly. Certainly I wasn't aggressive with the tool I was using to pry out the glass fragments; it never pays to be aggressive with razor-sharp pieces of glass just waiting for a chance to slice open a finger!

The egg-yolk dome is made of a flexible plastic. If one were to remove the cement holding it to the plastic mount then one would be able to flex the whole thing quite easily. The rim on the dome is inserted into the four hold-downs on the plastic carrier and cemented down.

Fig. 3. Close-up of the array of eight Luxeon Royal Blue LEDs which drive the remote phosphor dome. Note the damage to the insulation on both wires from the dome carrier pressing on them (see text).

To the left, Figure 3, shows a closer look at the Luxeon Royal Blue LED array. There are eight LEDs mounted on a board that is bonded to the heat sink and connected to DC power from the electronics beneath the heat sink.

A close look at both wire leads shows a potential problem: The insulation has been crushed by the phosphor dome's plastic carrier! The carrier does have a channel molded into it for each wire but they were designed for each wire to be butted against the curved indentation seen in the LED board. In fact, the uncrushed part of each wire shows where the edge of those channels is! Was this an assembly error or did the carrier tend to rattle and someone realized that using the wires to wedge it would be a cheap way to stop it?

The heat sink appears to be electrically isolated from both DC leads and from the screw-in AC connection, so a fuse or other protection may not trip unless both DC leads short to the heat sink.

Measurments show that the voltage on the terminals was 22.6 V while the current drawn was 410 mA. That's 9.3 Watts drawn by the LED array.

Primitive Spectrum Testing

Concerned about the spectral emission of a bulb? One can use a DVD as a cheap diffraction grating to view the spectrum, at least the visible part. If one is looking at a CFL bulb one will see mercury emission lines and whatever lines the phosphor emits. If the lens in one's eyes has not aged too much then one can see the 405 nm violet line from mercury in CFL bulbs. The LED bulbs I've tested haven't had this line, the high-energy part of the spectrum being limited to the blue light coming from the royal-blue LEDs used to cause the phosphors to fluoresce.

Have some fun with a DVD spectroscope and identify lamp technology such as what is used for your local street lamps. This gives a scientific answer rather than a guess based on the lamp's colour, as the emission and absorption lines depend on the materials used: High pressure sodium has a different spectrum than mercury vapour.

There are plenty of web sites online discussion the use of a DVD as a cheap difraction grating. Lots of DIY projects. Have fun!