LED bulbs reveal different design approaches

-April 21, 2011

A recent article examined light patterns from common lighting bulbs and took apart a Philips LED bulb to understand how its use of a secondary phosphor affects its light pattern (Reference 1). This article continues that discussion of the Philips 12.5W LED bulb, which relies on a secondary phosphor to spread its light in an almost-360° pattern (Figure 1). Figure 2 shows the top view of the bulb looking down into the wiring and connectors that take the power to the three LED boards.

LED bulbs reveal different design approaches figures 1-2

LED bulbs reveal different design approaches figure 3This bulb uses connectors rather than relying on low-cost labor for hand-soldering, which manufacturers of CFLs (compact fluorescent lamps) use. For example, Figure 3 shows a view of a PCB (printed-circuit board) from a hand-soldered CFL. The big blobs of solder inspire little confidence in manufacturing quality and probably have played a role in shorter-than-expected lifetimes for some CFLs. Philips’ decision to rely on connectors makes for a more repeatable, reliable assembly process and should pay off in long-term bulb reliability.

My praise for the use of high-quality connectors resulted in some virtually raised eyebrows in the online comments when I posted these photos online. For example, one reader asked how a connector can be more reliable than a solder joint (Reference 2). Because these solder joints are on the top side of the board, they are necessarily hand-soldered, and it’s difficult, though not impossible, to achieve a repeatable, high-quality hand-soldered joint in a cramped space in a high-volume manufacturing line. Connectors, on the other hand, are simple to use and consistent in their performance.

LED bulbs reveal different design approaches figures 4-5

Figure 4
shows the LED PCB and connector, as well as the thermal-interface material that helps pull the heat away from the LEDs into the relatively massive heat sink. Popping off the bulb’s power socket exposes the power-management circuitry (Figure 5) encapsulated in a rubbery potting compound (Figure 6). I couldn’t get a good photo of the dimming LED-driver IC, but the lettering looked something like CY8OLED, which I believe is a 12W Cypress CY8CLED dimmable LED driver (Figure 7 and Reference 3). When I connected it to a TRIAC (triode-alternating-current) dimmer, the bulb dimmed beautifully from almost 0 to the full 100%.

LED bulbs reveal different design approaches figures 6-7

My deconstruction of LED bulbs has in the past been utilitarian, relying heavily on hammers, wrenches, and safety glasses (Figure 8). Shortly after I performed the Philips bulb tear-down, Peter Di Maso, marketing manager for lighting-power products at Texas Instruments, told me that he’d had good luck with baking bulbs in the oven at 200°F for 30 minutes before pulling them apart. I tried this approach on my next tear-down. This time, the victim was a Lemnis Pharox 300, a 6W, dimmable LED bulb capable of 360 lumens at 2900K.

LED bulbs reveal different design approaches figure 8LED bulbs reveal different design approaches figure 9

Before popping the snow-cone-type bulb into the oven, I turned it on to see its light pattern (Figure 9), which turned out to be fuller than the usual 180° you see with a snow-cone-type design. How does the manufacturer achieve that trick? The bulb emitted a noticeable hum, however; I could hear it from 10 feet away.

LED bulbs reveal different design approaches figure 10I had to leave the bulb in the oven for an hour. Wearing leather gloves, I twisted the bulb quite a bit before pulling the white top straight out to reveal an array of six LEDs: Two red LEDs flank four white LEDs, visible in their off-state as yellow (Figure 10). This instance marks the first time I’ve seen “color-tuning” red LEDs in a relatively inexpensive replacement-type LED bulb.

Why would you want to add red to a white light? In general, we associate warmer-colored lights with incandescent lights because most interior color schemes target use with incandescents. However, LEDs are generally more expensive and less efficacious at higher color temperatures. Adding a couple of red LEDs to warm up the white light can make for a more attractive light. George Kelly, an Avnet “illumineer” who spoke at a panel on LED lighting at last month’s APEC (Applied Power Electronics Conference), believes that our preference for warm colors dates back to prehistoric times when firelight was the only option for light at night. Blue light is more prevalent during the day when the sun is high, whereas redder, warmer light is a signal that the day is winding down and it’s time to relax.

LED bulbs reveal different design approaches figure 11At this point in the tear-down, the socket was still on the bulb, and I was able to fire up the bulb to see how the red LEDs behaved: They were on whenever the white LEDs were on, dimming and growing brighter in the same way that the white devices were. This scenario contrasts with what I observed in the tear-down of the Cree TrueWhite LED module, which also uses red LEDs inside the white-LED module (Reference 4). However, those red LEDs come on only at full-on power to the module because white LEDs generally shift away from the warmer colors at higher drive currents. The Pharox uses always-on red LEDs to achieve a warmer white but without the added intelligence—and circuitry—of the TrueWhite approach.

LED bulbs reveal different design approaches figure 12Once I got the bulb cover off, I could easily see how the Pharox achieves a larger light pattern than most snow-cone designs (Figure 11). The LED’s array is slightly elevated. As a result, it projects down more than it would if aligned with the edge of the LED’s heat-sink base. The elevation is slight, but the bulb cover, acting as a diffuser, adds a slight reflection, so the light pattern is deeper than that of other snow-cone designs.

Continuing to dismantle the bulb, I removed the base, exposing the power-control electronics (Figure 12). This experience marks the first time that I’ve seen an LED bulb that did not encase its electronic components in a rubbery potting compound, and this lack of compound may be the reason for the audible hum: The potting compound would have either prevented the vibration or muffled the noise from a discrete component. This problem may well have been an isolated one, however.

See this video comparing the light patterns and dimming responses for the Philips 12.5W LED bulb to incandescent and CFL bulbs:

  1. Conner, Margery, “Remote phosphor expands reach of LED light,” EDN, March 17, 2011, pg 26.
  2. Conner, Margery, “Remote phosphors: Philips LED bulb tear-down, part 2,” EDN, Feb 15, 2011.
  3. CY8CLEDAC02 A55 Highline 12 W Dimmable LED Driver Reference Design Guide,” Cypress Semiconductor.
  4. Conner, Margery, “Cree’s LMR4 LED light module: What’s inside and how TrueWhite works,” EDN, Sept 14, 2010.

Also see:

Loading comments...

Write a Comment

To comment please Log In