Butchering an LED bulb
When one of the bulbs in our kitchen began to flicker, I saw a chance to sate my curiosity as to what was inside.
After “Made in China”, the glyph that resembles the head of a trident (or maybe half of a Greek letter theta turned on its side) is a factory mark. It would tell someone in the know precisely which facility in China produced the bulb. Alas, the Strictly confidential!
Philips Factory Marks document available as a PDF via the Factory Identification Symbols link on the LAMPTECH Lamp Markings page was last updated in 1986. Thanks to a different PDF file hosted by the same site, I know that the first two characters in the string 5DH
indicate that the bulb was manufactured in April 2015. The significance of the H
is, to me, unknown.
Some hacksawing got the globe off. Unfortunately,it didn’t occur to me to snap a straight-on photo of the LED array until after I’d pressed it face-down against my bench top and spent a few minutes enthusiastically sawing on the base, closer to where the screw-threaded connector was attached.
As you can see, more than half of the twenty-three LEDs inside the bulb have been visibly damaged, losing their yellow phosphor layers. Oops.
Getting the screw-threaded socket connector off was simply a matter of twisting and squeezing. With that done, the PCB was visible, but it wouldn’t come out. The board-to-board connection to the LED array was via 2-pins which slid surprisingly loosey-goosey into a pair of what looked a lot like female crimp-on “DuPont pin” connectors. They’re visible sticking straight up and out through two of the center of the LED array in the preceding photo.
At this point, having used pliers to snap off all of the bits of the bulb base that consisted purely of plastic, I realized that, beneath the remainder, was a one-piece shell of aluminum. Switching blades to one intended for cutting metal and exercising what I thought would be sufficient caution, I cut away enough aluminum to remove the PCB.
See how I managed to saw through one or more traces on the surface-mount side? That put paid to any thought I might have had of trying to diagnose the cause of the flicker. Maybe next time.