47
\$\begingroup\$

I have a 15 year old ReplayTV DVR. I got it used, I'm not sure when, but probably about a decade ago. It's turned on almost all the time, and it has a blue power LED on the front.

The LED FAQ says that many LEDs have a rated lifetime of 50,000 hours, which implies that it should only last about 5.5 years when used 24x7. Green Efficient Homes says that "an individual LED may well last 100,000 hours", so that bumps it up to 11 years.

These sites seem to be about LED light bulbs, not the individual LEDs that are used as indicators on electronic devices.

Why has this little guy been able to keep shining continuously for so long?

\$\endgroup\$
  • 44
    \$\begingroup\$ FWIW I have a clock with a 7-segment LED display which I built myself about 35 years ago. It has been in constant use ever since, and still works fine. The display is a bit dimmer than it used to be, but still perfectly readable unless direct sunlight is shining on it. \$\endgroup\$ – alephzero Oct 23 '18 at 20:52
  • 4
    \$\begingroup\$ @Passerby More likely they're betting that most customers will not use the product that long, or will not save the warranty information. And most LEDs are used as components by manufacturers, not individual consumers. \$\endgroup\$ – Barmar Oct 23 '18 at 21:32
  • 5
    \$\begingroup\$ In addition to everything else, if the TV PWMs that LED, it might have only been on for a small fraction of those 15 years, even if your human eyeball thought it was seeing light the whole time. \$\endgroup\$ – Jay Kominek Oct 23 '18 at 22:45
  • 1
    \$\begingroup\$ My first instinct was (as many others have stated in various ways) indicator lights, clocks, etc. have LEDs that last "forever" for a variety of reasons - e.g., not subject to high levels of power/heat like LED lighting fixtures, not inherently prone to failure the way incandescent and fluourescent bulbs fail, etc. And then I remembered that I have two dishwashers ~ 18 years old with failing LED indicators in their control panels - controls still work just fine but various indicator lights, most of which are only on during the wash cycle, have failed. All of which may be planned... \$\endgroup\$ – manassehkatz-Reinstate Monica Oct 24 '18 at 1:03
  • 3
    \$\begingroup\$ @manassehkatz A number of answers and comments lead me to consider LEDs to be the "cockroaches" of the electronics world. They'll last long after any other devices have worn out or gone extinct. \$\endgroup\$ – Barmar Oct 25 '18 at 15:44

10 Answers 10

51
\$\begingroup\$

Properly designed, built and used, today's LEDs have incredibly long lives and the wearout mechanisms are not catastrophic in nature, so instead of using incandescent lamp MTTF statistics, a luminosity percentage (70%) is often used to define the lifetime- this doesn't mean that the LED burns out, it means that the light out put is only 70% of what it was when brand new. Because the change is slow, you normally don't notice it.

For white light applications this makes sense because after a period of time the light works but may be too dim to be useful in the application such as a flashlight or reading lamp.

FWIW incandescent lamps are prone to something similar in that the filament would slowly evaporate leaving an dark coating on the glass shell, dimming the light. They are able to compensate a little bit because the filament glows a bit brighter due to the increased resistance (until it breaks).

\$\endgroup\$
  • 1
    \$\begingroup\$ I'm not talking about LEDs that are parts of bulbs that replace incandescent bulbs. I'm talking about single LEDs used as indicator lights. \$\endgroup\$ – Barmar Oct 23 '18 at 19:45
  • 17
    \$\begingroup\$ Its' the same technology. "white" LEDs as opposed to RGB types are a blue/ near-UV LED that stimulates a phosphor to give off white light, that's about the only difference (aside from the heterojunction materials etc.). If you wear glasses or know somebody, ask them about that annoying blue rainbow effect that they see in all the solid state car headlights- that's the active LED itself. Even if it's an indicator there's a cutoff point where it's too dim to be seen well. \$\endgroup\$ – isdi Oct 23 '18 at 19:52
  • 5
    \$\begingroup\$ "They are able to compensate a little bit because the filament glows a bit brighter due to the increased resistance" - Since incandescent lamps are usually powered by a constant voltage, wouldn't increased resistance result in the lamp consuming less power, and thus being dimmer? \$\endgroup\$ – Tanner Swett Oct 23 '18 at 22:31
  • 2
    \$\begingroup\$ whoops I meant to say localized resistance "hot spots". There are several tradeoffs in filament wire metallurgy which over time erode the filament cross sectional area and heating to tungsten to an even higher temperature at that spot. There's a nice review paper by Mathew Peet "Tungsten Filament Lamps A Case Study". There are of course a bunch of academic papers but they are all behind paywalls. If you google the above title it should pop up. \$\endgroup\$ – isdi Oct 23 '18 at 22:48
  • 1
    \$\begingroup\$ As a side-note, street lights and other "professional" lights often have circuitry to compensate for the aging of LED's They start out at perhaps 70% current, and increase the current over time to still produce the same light output for many years. I have replaced the inside of the ceiling light in the kitchen because I subjectively found it less bright than at installation, and I might have to do the same with the under-cabinet LED's. (Or I might add lights at the front of the top cabinet underside, because the current lights are at the center, and doesn't illumunate the whole table enough \$\endgroup\$ – Lenne Oct 24 '18 at 17:11
24
\$\begingroup\$

Best guesstimate I have for an indicator LED life span is more than 5,000,000 hours.

I don't think anyone would be surprised if a micro controller lasted 10-15 years. More likely one would EXPECT one to last that long. Think about how many pn junctions could fail in a microcontroller. An LED has only one pn junction. An indicator LED will last longer than anyone can live to find out.

Indicator LEDs would likely never fail in your lifetime. The average human life time is now about 78 years or 683,280 hrs. Indicator LEDs can last longer than that.

The power supply (the weakest link in most products) powering the LED in the DVR has lasted 15 years (23,400 hrs).

Reliability is based on not only the LED but the system the LED is used in.

enter image description here

LEDs rarely die. LED light bulbs die because the driver failed.
The rating on an LED lifespan is based on lumen maintenance.


A common specification for high power lighting LEDs is for 70% lumen maintenance (that is, output at 70% of its peak intensity) after 50,000 hours of operation.

The most notable industry standard is the IES LM-80 (LM-80), which is according to the U.S. Department of Energy an “approved method for measuring lumen depreciation of solid-state (LED) light sources, arrays and modules,” (U.S. DOE).

This Lumiled white paper covers the topic in detail:
Evaluating the Lifetime Behavior of LED Systems

I took Lumileds 149,000 hr estimate graph (Fig 5) and extrapolated it to 1,000,000 hrs. Keeping in mind this is a high power LED tested at high temperature. Indicator LEDs last much longer.
enter image description here

Rule of thumb is the life of a semiconductor is reduced by 50% for each 10° C rise in temperature. In the LM-80 testing a junction temperature of over 130° C is used (120 °C ambient is specified). And indicator LED runs at room temperature. So for that well over 5X longer for an indicator. This is where I get my best guesstimate I have is 5,000,000 hours.


That is for lighting LEDs. Indicator LEDs have always been know for very high reliability. AT&T was one of the first to see this.

AT&T has found their telephone indicator LEDs would last for up to 50 years. That's 438,000 hours. 50,000 hrs applies to high power (high temperature) that drop to 70%. They last much longer than 50,000 before failure. Lumileds has tested L70 to 148,000 hours with a 50,000 test. If they were go to the point where it could no longer be seen it likely be 500,000 to 1,000,000 hours. See figure 5 in their white paper. And that was a high power high temperature Luxeon Rebel tested at 85°C, 0.35A (Tjunction ≅98°C).

1960s an antiquated 110V GaP LED would last over 438,000 hrs. (50 years)

At the end of the 1960s AT&T Bell Laboratories immediately realized that there were many possible applications for LEDs. Indicator lamps. were becoming useful in the telephone business. All such lights used at that time in the USA operated using 110 V. An example is the “Princess” telephone, which was intended to be used in bedrooms – the dial lit up when the phone was picked up from its cradle. The “Princess” was a prestigious telephone and the latest fad in the 1960s but had to be installed near a 110 V outlet. A service call to the local phone company was needed if the bulb burned out. If LEDs were to replace the 110 V light bulbs, the phone line could power the LEDs and a 110 V outlet would no longer be needed. In addition, GaP LEDs had an expected lifetime exceeding 50 years when used in telephones, much longer than 110 V light bulbs...


LED material makes a difference.
From the book Light emitting Diodes by E. Fred Schubert

The reliability of AlGaAs devices is known to be lower than that of AlGaInP devices that do not contain any AlGaAs. High-Al-content AlGaAs layers are subject to oxidation and corrosion, thereby lowering the device lifetime.


If a 110 V GaP lasted 50 years then a 2V LED should last much longer. As the IV characteristics of LED have improved the lifespan has also improved.

Low values of [Vf] are consistently correlated with high device reliability

enter image description here

\$\endgroup\$
  • \$\begingroup\$ As I said in the question, the LED has been on almost constantly since I purchased it around a decade ago. \$\endgroup\$ – Barmar Oct 23 '18 at 20:46
  • \$\begingroup\$ I asked to make a point rather than to elicit an answer. Plus you said: "I got it used, I'm not sure when, but probably about a decade ago." How often was it on the first ten years before you owned it? Again rhetorical. \$\endgroup\$ – Misunderstood Oct 23 '18 at 20:54
  • 6
    \$\begingroup\$ I would be more impressed by the reliability of the power supply powering the LED. \$\endgroup\$ – Misunderstood Oct 23 '18 at 21:01
  • 2
    \$\begingroup\$ Based on my long time working with efficient panel LED's running at 10% of full power, I stated in a company meeting that we could expect them to last maybe 200 years. This shocked everyone except the chief engineer. Your answer is closest to the truth. +1 \$\endgroup\$ – Sparky256 Oct 24 '18 at 3:27
  • 7
    \$\begingroup\$ @Sparky256 200 years or 570 years. Close enough. Less than 3x difference. I still feel my 570 year guesstimate is conservative. By my next lifetime we'll know better. I'm going to wait at least a few hundred years before coming back. \$\endgroup\$ – Misunderstood Oct 24 '18 at 4:22
14
\$\begingroup\$

These sites seem to be about LED light bulbs, not the individual LEDs that are used as indicators on electronic devices.

This is at the essence of the issue.

The first component to fail in an LED lightbulb is usually the power supply, not the LEDs themselves. The failure of the power supply may induce a failure in the LED -- either by delivering too much power, or by generating excessive heat -- but the LEDs themselves are rarely at fault. Since indicator LEDs do not contain their own power supply, they are not affected by these issues. (A failure of the power supply in a device which contains LEDs is usually attributed to the device itself failing, not its LEDs!)

Additionally, LED lightbulbs are a much more demanding application than indicator LEDs. The LEDs in a lightbulb are often run close to their maximum design power, at high temperatures. This causes them to age more rapidly than an indicator LED, which is usually only run at the power required to make it visibly light up.

Indicator LEDs typically have a useful lifespan measured in decades. The brightness of the LED may decrease slightly as it ages, but it is rare for one to "burn out".

\$\endgroup\$
  • \$\begingroup\$ I guess the corollary to my question is "why are LEDs so long-lasting?" \$\endgroup\$ – Barmar Oct 23 '18 at 21:25
  • 3
    \$\begingroup\$ @Barmar en.wikipedia.org/wiki/List_of_LED_failure_modes most of the failure modes that cause slow dimming of the LED are just that: slow. most of the others are caused by one time abuse that either kill the LED right away, or it survives. \$\endgroup\$ – Aaron Oct 23 '18 at 21:39
  • 7
    \$\begingroup\$ @Barmar Why shouldn't they be? The physical process by which LEDs emit light isn't inherently damaging. \$\endgroup\$ – duskwuff -inactive- Oct 23 '18 at 21:45
  • 3
    \$\begingroup\$ @duskwuff I think that is the major point here, OP is used to traditional filament-based bulbs that are inherently damaging (or at the very least incredibly fragile when operating) and has based all his assumptions on lights around this. \$\endgroup\$ – Trotski94 Oct 24 '18 at 7:58
6
\$\begingroup\$

The indicator LED is quite possibly not operated at nominal power.

The lifetime of an LED is typically specified at nominal power. As blue LEDs are quite annoying indicator lights, these are often dimmed down to a less annoying brightness (for example, nominal specified current is 20mA, the actual operating current in a specific device is just 3mA). This increases the useful lifetime of the LED significantly.

That is in addition to the other answers. The LED will be dimmer by now than when it was new, but it's still useful as an indicator light.

\$\endgroup\$
  • 1
    \$\begingroup\$ this answer already explains how reduced current and temperature extend the lifetime of a LED and mentions that LEDs get dimmer as they age. \$\endgroup\$ – Dmitry Grigoryev Oct 24 '18 at 11:09
  • 1
    \$\begingroup\$ @Dmitry Grigoryev: Oh. Indeed. I skimmed it, and also sis a full-text search across this page...unfortunately, the relevant part of that answer was contained in a graphic. Thanks for pointing out! \$\endgroup\$ – Klaws Oct 24 '18 at 12:05
  • 2
    \$\begingroup\$ Actually, this is a very astute observation. Blue LEDs are unnecessary bright for indication purposes. The "specified current" (20 mA) is tuned by marketing/process engineering to meet the impressive 50,000 (etc.) life time. So running the LED at 1/10th of nominal will likely increase its lifetime by maybe 20,000X, given that all aging models have exponential behavior. \$\endgroup\$ – Ale..chenski Oct 24 '18 at 21:00
  • 1
    \$\begingroup\$ Running below (way below) rated current also vastly decreases the die temperature, which would add another big factor to MTBF. I would say that this answer offers the best explanation to OP question (togheter with my remark about e^+10 factor). \$\endgroup\$ – Ale..chenski Oct 24 '18 at 21:27
4
\$\begingroup\$

The rating is only a MTBF (mean time between failure)

LED MTBF

For any component or system, the MTBF is the mean time between failures. The MTBF is the elapsed time which is predicted between inherent failures of a component or system during operation.

The MTBF is a figure used in calculations for the reliability of items of equipment. In order to be able to calculate the MTBF of the equipment, it is necessary to know the MTBF of the individual components, e.g. the LED MTBF in this case.

The failure rate for a component, and the MTBF are linked. MTBF can be calculated as the inverse of the failure rate if it is assumed that there is a constant failure rate, which is not unreasonable as a first order assumption.

MTBF = Hours of operation / Number of failures

The MTBF figures are often quoted in the manufacturers data sheets. However the MTBF can be considerably reduced by operating components close to their rated limits. Hostile environments such as high temperature and vibration also reduce the MTBF.

However when run within their limits, the LED lamps have a long lifetime, and do not fail very often.

Source: https://www.radio-electronics.com/info/data/semicond/leds-light-emitting-diodes/lifespan-lifetime-expectancy-mtbf.php

Without a sample population and testing of that specific LED it will be impossible to tell what the lifetime of that LED is. the manufacturer can only guarantee through what they know about that LED (probably through running it through months worth of testing, measuring the degradation) and then extrapolating that over the lifetime. They then guarantee an MTBF figure and if you get a batch of their LED's on average, they will last more than x amount of hours. But they could last more, the MTBF figure is only a lower bound.

\$\endgroup\$
  • 2
    \$\begingroup\$ This answer links to a page that talks about what defines a failure, and references mean time between failure, but fails to define failure itself, which is an important distinction for this question. The typical "failure" mechanism of an LED is not to completely stop emitting light, but to dim beyond a defined percentage of its original output. \$\endgroup\$ – Aaron Oct 23 '18 at 21:25
  • \$\begingroup\$ @Aaron with an indicator light, a failure would probably be defined as failure to indicate or light at all, not the 70% that is commonly used in the industry to define MTBF \$\endgroup\$ – Voltage Spike Oct 23 '18 at 21:32
  • \$\begingroup\$ which is exactly why I think it's important to mention that distinction. The OP was asking why the LED lasted so much longer than the industry suggested lifetime, and it's because the industry suggested failure criterion is different from his own. \$\endgroup\$ – Aaron Oct 23 '18 at 21:38
  • \$\begingroup\$ RE: "a failure would probably be defined as failure to indicate or light at all" so 50,000 hrs falls way short of the actual expected lifetime. I'd say 10x short. I would expect an indicator LED to last well over 500,000 hrs. \$\endgroup\$ – Misunderstood Oct 24 '18 at 0:27
4
\$\begingroup\$

Estimations of life expectancy of electronic components is a tricky business. As one can imagine, manufacturer can't wait 50,000 hours (or something like 2,290,000,000 hours for a RS232 receiver, see page 7 of this essay) to collect statistics of device failures, market doesn't allow this. All estimations are done using "accelerated models".

The idea is to set a batch of devices under extreme conditions (voltage, current, temperature, environmental factors) until some devices fail in a reasonable test time (days or hours), assuming no shift in failure mechanisms. Then, using some theoretical models of failure mechanisms and other ASSUMPTIONS, the failure rate gets extrapolated to normal operating conditions. Usually the theoretical models are of exponential type, so small errors during accelerated testing can result in vast differences in estimations for tails of these exponential functions. So the models are usually on a conservative side, to avoid liabilities from catastrophic premature failures.

For example, if the accelerated model for a LED did include the air humidity and rust/oxidation factor, but your LED was always at room and dry air conditions, it can last much-much longer than the manufacturer's specified MTBF.

\$\endgroup\$
  • \$\begingroup\$ The LM-80 testing standard use 3 different operating conditions 350, 700, and 1000 mA. At 55, 85, and 120 C. So they are tested under normal operating conditions for a minimum of 6,000 hrs. Except 55 C is to low for 1000 mA and outside normal conditions. MTBF does not apply well to LEDs due to the reduction in intensity over time. \$\endgroup\$ – Misunderstood Oct 23 '18 at 23:50
3
\$\begingroup\$

Those times are rated lifetimes and can be very conservative. So it might well be that most or (in case you are lucky) some of them can handle a factor times more, so 15 years sounds not unreasonable.

\$\endgroup\$
2
\$\begingroup\$

If the rated lifetime is warrantied, or even relied on, then the manufacturer may be being conservative and/or limiting itself to what it knows to be true, rather than what it expects to be true. It may have to defend itself in a lawsuit!

Blue LEDs did not become widely available as cheap indicator LEDs until the late 1990s. In 2010 a manufacturer might not have had much more than a decade's experience of the devices, and it would therefore be risky to state a life expectancy of more than ten years because theory and "accelerated ageing" laboratory experimentation can not substitute for two decades of actual use in the field.

There are reasons to hope that disk drives can last ten years in service, LED light bulbs over twenty, solar panels well over thirty. But experience may triumph over hope, and there are presently "known unknowns" concerning the long-term ageing characteristics of such products which only the passage of much time can make known.

\$\endgroup\$
0
\$\begingroup\$

First as the others have said the lifespan of the LED is a MTBF, second, some manufacturers strobe lights at very high speed, so the light is actually off a fair amount of the time, your eyes just can't register it. This is often done to decrease battery consumption, but I suppose could also be used to increase lifespan.

\$\endgroup\$
  • \$\begingroup\$ This might actually be more stressfull to the LED than running it at a constant low current at equal brightness... \$\endgroup\$ – rackandboneman Oct 28 '18 at 21:01
0
\$\begingroup\$

They run a bunch of LEDs for 50,000 hours and none of them burn out. So they say on the package "Guaranteed for at least 50,000 hours". To guarantee a LED for 20 years would require them to run it for 20 years to test it.

Also, there's a probability of an external event, such as a static discharge, causing an LED failure. This rather small probability influences the 50,000 hours vs 100,000 hours rating.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.