In my so-far nearly 30 years of writing for EDN, I’ve learned a lot about my ever-evolving audience (i.e., you), at least two aspects of which are directly relevant to this particular writeup:

• You love consuming any content that’s even remotely LED-related, and
• I’ve pretty much given up trying to figure out what topics will especially attract your attention, aside from relying on my own curiosity as a guide to what you might also like.

Take, for example, my recently published teardown of a LED-based desk lamp. Compared to some other teardowns that I’ve done, it was thankfully fairly speedy and straightforward to both implement and document. But, judging from the quantity and detail of the comments already posted on it, I’m guessing it’s still driving a lot of “eyeballs” to the EDN website. A fading-illumination-intensity-over-time LED apparently piqued more than just my curiosity.

The LED light bulb that transformed into a sorta-strobe light

Or take today’s dissection candidate, a conventional LED light bulb that had begun not fading, but instead, flickering. As historical background, I’ll take you back nine years to when I took apart my first LED light bulb, two dimmable examples of which had prematurely failed, due to (I prognosticated at the time) extended exposure to high temperatures caused by poor ventilation of the ceiling-mount enclosures within which they were installed. At the time, a reader named “docterdon” noted that I hadn’t described those sconces, so in the spirit of “a picture paints a thousand words”, here you go to start:

The room switch controlling the lights wasn’t dimmable anyway, so at the time I went ahead and replaced all of them (including the two still-functional ones, which I ended up reusing elsewhere) with CFLs. Two of those ended up prematurely dying too, so once again I swapped them out for LED bulbs, non-dimmable this time (I was admittedly surprised to realize, when recently reviewing past published teardowns, that in the plethora of LED-based illumination sources I’ve taken apart in recent years, a conventional non-dimmable one hadn’t yet gone under my knife). They came eight to a package; here’s the encompassing cardboard label:

Behind it in each initially shrink-wrapped assemblage, four of which my email archives indicate I’d promotion-purchased from VMInnovations back in October 2018 for $9.99 total, were two boxes, each with four bulbs inside (yes, your math is right; that translates to $0.31 per bulb!):

Here’s our flickering victim, as usual, accompanied by a 0.75″ (19.1 mm) diameter U.S. penny for size comparison purposes:

Shifting the bulb slightly to the left, here’s your reminder that my office desk is perpetually bathed by the light coming from—among other things—the front-panel blue (when the computer it’s connected to is powered on, that is; otherwise white) LED of the expansion hub tethered to my Mac mini, whose illumination you’ll see in some of the shots that follow:

Diving inside

Onward; let’s get the globe off. Extended exposure to my wife’s hair dryer didn’t help much with loosening the adhesive; then again, unlike what my heat gun had done in the past, it didn’t deform the globe itself, either. Nevertheless, using several “spudgers” and aided by plenty of “elbow grease” and “colorful language”, I finally wrestled the globe off the base:

Admittedly, in the process, snapping one of the three resistors off the plate, along with scraping the phosphor cap off one of the LEDs:

That large IC you see at center left is the RM9003T, a high-voltage single-channel constant current LED controller from Shaanxi Reactor Microelectronics. That said, from past experience, I strongly suspected that what I was currently seeing wasn’t the full extent of the circuitry; there was likely more behind the plate. There’s only one way to find out for sure:

At this point, my forward progress was stalled until…ah, yes, those power wires running to the cap end need to be disconnected before I can completely remove the plate. Time to dig out my tongue-and-groove, slip-joint (aka, “Channellock”) pliers and wrest it off…

Determining the Achilles heel

That’s more like it:

The markings on the IC on one side of the now-exposed underside PCB:

are barely discernible:

MB6S
1607

It appears to be a miniature surface-mount bridge rectifier, converting (crudely) AC to DC in combination with the 390 kΩ (“394”) resistor next to it and the 200-V 10-µF aluminum electrolytic capacitor on the PCB’s other side:

Speaking of enclosed spaces with insufficient airflow and consequent overheating potential (generated by the multi-LED array on the plate above it), I’m guessing that here’s where the flickering originates. Agree or disagree, readers? Share your thoughts in the comments!

—Brian Dipert is the Editor-in-Chief of the Edge AI and Vision Alliance, and a Senior Analyst at BDTI and Editor-in-Chief of InsideDSP, the company’s online newsletter.

 Related Content

• Teardown: What killed this LED bulb?
• The empty promise of the LED bulb’s lifetime
• Disassembling a LED-based light that’s not acting quite right…right?
• A brief history of the LED
• Freeing a three-way LED light bulb’s insides from their captivity

The post Diagnosing a flickering LED light bulb appeared first on EDN.