I have been observing the LED lighting transition in buildings and homes for several years now and realized that there are a lot of subjects that pertain to engineering and in particular analog engineering. The LED Interior Lighting Craze, Part 1 was an introduction to the industry as it currently exists. This second part focuses on reliability. I found some interesting information on LED reliability for the components as well as the lights as a whole.
LED failure is not a straightforward subject. Unlike a traditional candescent bulb, an LED doesn’t flat out die and stop emitting light according to the reports I read. The LED actually starts changing in lighting terms to a point where the light is not adequate. Therefore the LED is alive, it’s just a bit sick.
LED reliability is not limited to the performance of the LED in terms of providing acceptable light. Other failure mechanisms occur such as failure of the LED drive circuit. Also, LEDs are often matrixed in strings of individual lights. Various strings can fail while others continue to work. So LEDs die in phases like the leaves on the trees in autumn.
My look into LED reliability focused on three reports listed in the references. There are also hyperlinks to the reports in the text. The following paragraphs describe the reports as well as their origins and relevance to the various forms of failures. As with any internet search, these reports are not the governing industry findings per se. They are the top ranking Google search results. Therefore take them with a grain of salt. As with most of my subjects, I’m starting a conversation that readers are free to weigh in on. I’m just the starting fluid. Feel free to provide the fuel.
OSRAM provided an application note on the actual LED lifetime. An application note from an LED company might invoke thoughts of the IRS auditing itself or congress voting on its own raise however the report is quite interesting. The report doesn’t really go much into the deep techno terms such as luminaire lifetime and instead refers to just plain lifetime based on the percentage of light output. For this curve, percentage is a dimensionless quantity. The report is from 2013 which is a bit dated in terms of the rapidly advancing technology in LED lighting. Still, the report offered an insight that the Planet Analog audience might find interesting, the design environment has a lot to do with failure. Mostly this is due to junction temperature as a result of LED drive current as well as thermal resistance and heat dissipation.
The other two reports were not from manufacturers. The Department of Energy report had information on subjects such as failures due to fixtures and wire bonds as well as speculation on power supply effects on performance degradation. There was however little data to back this up with only LED light measurement data being presented. Once again there is reference that current is the major culprit when it comes to the LED failing.
The expected LED life based on drive current indicated by each colored line and target LED junction temperatures. (Source: Philips Lumileds)
The final reference was from a RTI, a research firm that provided results “based upon work supported by the Department of Energy under Award Number DE-EE0005124”. This report provided information on the degradation of the LED lumens and chromaticity. As for the engineering aspects, the power supply and its components were listed as well. Failure modes focused on electrolytic capacitors and MOSFET transistors as well as the LED drivers themselves. What I found interesting was slide 29 (see figure) which indicated that the circuits making the input line purty, failed more often than the circuits actually doing the work of creating the LED drive current.
As you can see, the Filter and Conditioning failed 71% of the time whereas the Shaping and PFC failed 69% of the time while the switched mode control only failed 56% of the time. My fellow power electronics engineers can take a bow (depending of course on what kind of shaping and PFC took place). Then again was it the components that failed or the design of the components (as power electronics engineers usually design the whole system)?
In summary of reliability as a value proposition, are those lights really going to save money on installations and last 15 years? I’ll present this statement from Reference 2 and let you decide if manufacturers are standing behind their products, “There is often a huge gap between the warranted life of a product and the expected life of the LED source in that product.”
Of course I had to once again include “The big lie about LED lighting ,” by Scott Elder of Linear Technology. As evidenced by this article, consumer lighting is not living up to lifetime expectations. Planet Analog engineers like myself have warped expectations that a cell phone should actually be able to answer a call. WE might also expect a circuit we buy to last as long as one we build. So maybe we should ignore the outsourcing that brings these cheap products to our shores in the first place and accept the fact that profit rules over expectations. Wait, my phone is ringing……but the screen won’t let me answer it. Someone fetch me an Uzi, this phone is in need of attention.
“Ut-ah, I know what you’re thinkin’, punk….”
LG might stand for life’s good however I’m wonderin’ if this phone feels lucky. Did I fire six or only five shots? In all of the melee, I don’t recall myself, but seein’ how…… Are you squinting?
- “Reliability and Lifetime of LEDs”, OSRAM application note, December 2013.
- “LED Luminaire Reliability” U. S. Department of Energy (no date given however a reference to 2006 is enclosed and a web page age of April 2013 was noted)
- “Solid-State Lighting Devices: Are the Claims of Higher Reliability True?”, Illumination Engineering Society –Raleigh Section, March 10, 2015, J. Lynn Davis, RTI International
- “The big lie about LED lighting,” December 31, 2014 By Scott Elder
- “The LED Interior Lighting Craze, Part 1”, Planet Analog website, Scott Deuty, 10/14/2016