What Happened To The 100,000-Hour LED Bulbs? | Hackaday

Early adopters of LED lighting will remember that the box has a service life of 50,000 hours or even 100,000 hours. But during a recent trip to a hardware store, I found that the longest lifespan in the advertisement was 25,000 hours. Others only claim 7,500 or 15,000 hours. Yes, these are Cree and GE brand bulbs.

So, what happened to those 100,000-hour household LED bulbs? Was the initial estimate too optimistic? Is it all marketing hype? Or, do we not know enough about the aging of LEDs to predict the true life of the bulb?

I tested these issues. Join me after a break to learn about the background knowledge of bulb cartels in the incandescent bulb era (not kidding, the cartel controls the life of the bulb), and the destruction of some modern LED bulbs to understand why the lifespan is much lower than the original LED Alternative wave.

Any discussion about bulb life will be incomplete without mentioning the Phoebus Cartel, an international organization established in 1924 by the world's leading bulb manufacturers to manipulate the bulb market. As Markus Krajewski in 

, The cartel allocated regions to member companies, restricted production, and stipulated that the lamp life should be shortened by 1,000 hours. The previous light bulb has burned for more than 1500 to 2500 hours. It is said that in order to improve quality, efficiency and light output, the new 1,000-hour limit also brings more bulb sales. The archived documents show that a lot of research is spent on designing bulbs that can last their designated 1,000 hours. It's not just home lighting that has suffered: The flashlight bulbs that can use three sets of batteries have been reduced to two sets, and a proposal to limit their lifespan to one set has been proposed. Again, the increase in brightness is touted as the reason. However, the final step to halve the life of the bulb can only increase brightness by 11%-16% while doubling sales. This is about selling more bulbs and making more money.

Cartel uses the strong strength of GE's patent portfolio to enforce production quotas and lamp life through a fine system. The bulbs produced by each manufacturer were tested, and the bulb life was significantly shorter or longer than 1,000 hours. Before the end of World War II, Phoebus has been exerting influence on the market. Cartels are often regarded as one of the earliest examples of obsolete plans: designing products that artificially shorten their lifespan. Documentary in 2010 

 Explore the history of cartels and outdated examples in the plan. I want to know, do the conspirators think of bulbs that are said to last 100,000 hours? Even 7,500?

A few lonely incandescent light bulbs are hidden on a lower shelf on the lighting island of the hardware store, waiting for some Ludit consumers. Picking up the box, I read the rated life: 1,000 hours.

What does this 1,000-hour service life mean? This is the average rated life (ARL) of the bulb-the length of time that 50% of the initial sample of the bulb fails (abbreviated as B

). The meaning of "fault" depends on the type of bulb. We will discuss this in more depth later. Definition of B

Reveals a common misconception that the bulb will be used within its rated life. In fact, although half of them don't tell you any information about the distribution of failures in the average lifetime, only half of them can last this long.

Manufacturers use these ARL values ​​to predict how many years the light bulb can last based on the number of hours (usually 3 hours) of light bulbs used per day. Compared with incandescent lamps, LED bulbs suffer less wear and tear during the power cycle, so conversion is just a division: service life = ARL / (3 * 365). For example, based on this calculation, half of a set of 100,000-hour bulbs will still be in use after 91 years. But this simple indicator does not tell the whole story. The failure mechanism of LED bulbs is complex and fundamentally different from the well-known incandescent bulbs. To learn more, we need to clarify the inner workings of the bulb.

Before leaving the store, I threw a few light bulbs in the shopping cart so that I could see the contents with my own eyes.

LED bulbs are more than just LEDs. In fact, the socket in our house is quite dirty AC power. The LED needs a clean constant current DC power supply, so the circuit inside the bulb must rectify and filter the input AC, and then limit the current flowing to the LED package. In order to understand how to do this, I analyzed three different types of A19 bulbs: one from GE "Basic" and "Classic" series (7,500 and 15,000 hours), and the other is Cree model with 25,000 hours. Service life.

The GE bulb has a plastic dome covered with a circular aluminum PCB on which eight LED packages and driver electronics are mounted. The driver includes a MB10F bridge rectifier, an electrolytic capacitor rated at 105°C and an SM2082D linear constant current driver. There are 3 resistors on the PCB: one deflates the capacitor when the bulb is off, and the other two set the SM2082D current to 54 mA. In fact, the circuit looks like it was taken directly from the SM2082D data sheet.

Seven of the 3.5 x 2.8 mm LED packages showed a forward voltage drop of 18 V when driven at 50 mA, indicating that they contained six LED chips in series. The voltage drop of one LED on the board is 9 V, so there are only three LED chips. All LEDs (45 dice in total) are connected in series to reduce approximately 135V.

When they say classics, they are serious. This bulb is housed in a glass case like an incandescent lamp, and like those old bulbs, the glass can be easily removed with a round head hammer. Instead of tungsten wire, the aluminum PCB is folded into an obelisk. Sixteen 3.5 x 2.8 mm LED packages are connected in series on the board, and each package displays a forward voltage of approximately 9 V at 50 mA. Therefore, this version has 48 LED chips, and the basic bulb is 45, except that they have twice the number of packages, which is very useful for maintaining the heat dissipation of the LED.

Another difference of this longer-life bulb is that the driver electronics are not thermally coupled to the LED. They are hidden on a separate PCB in the screw base. This prevents the remaining components from heating along with the LED. On the driver PCB is a bridge rectifier, which is also an electrolytic capacitor rated at 105

C, and SOIC-8 IC. Interestingly, this bulb also contains a metal oxide varistor for transient suppression. Although I am not sure what the driver IC labeled "BYSACT" is, the lack of any inductive components on the PCB indicates that this is another linear power supply.

The Cree bulb has a diffuse plastic dome similar to the GE Basic model. Inside, the larger aluminum PCB houses (16) 3.5 x 2.8 mm LED packages. Each LED drops about 8.5 V at 50 mA, so it contains 3 chips; like the GE Classic bulb, this bulb uses a total of 48 LED dice. The LED wiring consists of eight parts of two parallel LEDs, so the total voltage drop is approximately 68V. The LEDPCB is coupled to a thick aluminum heat sink with silicone thermally conductive compound.

Like the GE Classic bulb, the power electronics are located on a separate PCB, thermally separated from the LED. The driver IC is a SOT23-5 package with almost no mark "SaAOC", but the presence of a transformer and a robust Schottky diode indicates that this is a switch mode power supply. The filter capacitor on the output of the switch is made of aluminum electrolysis, with a rated temperature of 130°C.

It's not a big deal, but what conclusion can we draw from the design of these three bulbs? It helps to consider how they usually fail and what factors affect their life.

Since the LED bulb contains many parts, it is natural to ask which may be the cause of the malfunction. US Department of Energy (DoE)

Supporting the research and development of LED technology, its website contains a lot of data about LED lighting systems. its

Contains data on the failure rate of 5,400 outdoor lights in more than 34 million hours of operation. Interestingly, the LED itself only accounts for 10% of failures. On the other hand, the driver circuit takes up almost 60% of the time. The remaining failures are caused by shell problems and may not be suitable for indoor bulbs. This data shows that, at least for catastrophic failures (the bulb stops emitting light), extending the life span means improving the power supply.

The life of the bulb (or power supply) must not exceed the life of any of its components. Among the components found inside the bulb, there are two life limits: semiconductors and electrolytic capacitors. The failure rate of these two components is an important function of temperature. Based on the Arrhenius equation, the typical model for this effect predicts that the life will double for every 10 degrees Celsius decrease in temperature at least within a limited range.

The two longer-life bulbs use twice as many packages as GE Basic bulbs, and carry roughly the same number of LED dice, thereby reducing the thermal resistance of their respective heat sinks and possibly lowering their temperature. These bulbs also mount the faulty driver electronics on a PCB separate from the LED to keep it cool. Finally, 25,000 hours of Cree bulbs used electrolytic capacitors rated at 130 

C instead of the upper limit of 105°C in the other two. For similar operating temperatures, this may increase the life expectancy of the capacitor five times. Each of these measures may result in delayed catastrophic failure of the bulb, resulting in a longer rated life.

However, as far as the LED itself is concerned, life estimation is more than predicting catastrophic failure.

Just like the soldiers in Douglas MacArthur's famous route, the old LED will not disappear, it will only disappear. We all know what the failure of an incandescent lamp looks like: for one second, it burns brightly; next, it is not (every time, you will hear a popping sound, followed by a faint jingle, which is the release of Out of the filament richochets). In addition to the power supply, LEDs usually do not fail to a large extent. On the contrary, they gradually lose brightness as they age. In the lighting industry, this is called lumen depreciation and is a separate failure mode of catastrophic failures that we usually consider.

Facts have proved that incandescent bulbs can also suffer from lumen reduction. By the end of its 1,000-hour lifespan, production usually drops by 10-15%, but no one notices. When using LEDs, the effect will be worse, and as the equipment ages, the output will continue to drop. At some point, even if the LED has not "burned out", it no longer produces enough light to meet its original purpose. Research shows that most users will not notice a 30% drop in light levels. Therefore, the industry has defined L

The time when the output drops to 70% of its initial level is used as the end point for measuring the life of the LED bulb. According to the estimation method, this indicator is usually expressed as B

-L

, That is, the point in time when 50% of the initial sample of the bulb retains 70% of its rated output.

As phosphor-based white LEDs age, other things will happen: they will change color. U.S. Department of Energy report

The four color shifts observed in LED lights (blue, yellow, red, and green) are defined, although yellow shifts dominate in high-power white LEDs. Since the temperature of the phosphor may be 30 C – 50°C higher than the temperature of the LED junction, cracking, delamination and thermal effects of the phosphor will cause the light output to gradually turn yellow. Modeling and predicting the color shift of LEDs is a difficult task because all the mechanisms are not yet fully understood. As a result, no standards have been established for accelerating testing or predicting color stability over time.

Ultimately, these effects may adversely affect the function of the bulb like a catastrophic failure. Given that the decline in lumens and color shifts will cause LEDs to fail in time, it may not make sense for manufacturers to design bulbs with long service lives. The reduced lifetime ratings we see on current bulbs may simply reflect a better understanding of the actual performance of existing LED technology over time.

I have witnessed the drop in luminous flux and color cast. In June 2010, I replaced twelve 65W incandescent PAR30 floodlight bulbs in the kitchen with equivalent LEDs. At the same time, I replaced the three lights in another room with the same LED bulbs. These three bulbs are much less used, so when preparing this article, I picked up a bulb from each position and placed it side by side to see if I could tell the difference in output. The recessed luminaires in the two rooms are the same, so I want the bulbs to be exposed to similar temperatures when turned on: any difference should only be due to aging effects. The results are shocking. Since these two bulbs are in different rooms, I never saw them side by side, so I didn't notice how serious the depreciation of lumens and color changes were. Of course, I know they are dimmer and yellower than when they were installed, but I don't know how bad it is.

The life span of these bulbs is 30,000 hours. I estimate the total usage time is 15,000-20,000 hours. In 8.5 years of these services, one of them failed completely. Instead of replacing it (or replacing all) with a new bulb that did not match the color of the old bulb, I left the socket empty.

At the hardware store, I noticed that the new 9-watt BR30 LED bulbs cost $5 each. The PAR30 I bought in 2010 was priced at US$45 and consumes 11 watts. Quick calculations show that, compared with the replaced incandescent lamp, the old bulb can save more than three times its own electricity bill, and it emits much less carbon into the atmosphere. They may continue to burn for 15,000 hours, but after weighing the degraded output and the cost of replacing them with a brighter, more efficient version, I will return to the store.

I studied some technical issues in LED lighting. Of course, the lifespan of LED bulbs is much more than that-color temperature and color rendering index (CRI) should be considerations in any purchasing decision. Many larger issues are also involved, including economic and sustainability issues. JB MacKinnon talked about some of these issues in a 2016 article, 

, In "The New Yorker".

Of course, it makes sense to switch from incandescent bulbs to more efficient lighting, but maybe we don’t need 100,000-hour bulbs at all in the beginning. Compared with the rapid development of lighting technology, even a 7,500-hour bulb has a long service life. Does it make sense to buy expensive long-life bulbs today when better, cheaper, and more efficient bulbs may appear in the near future?

The longest surviving incandescent lamp, called

It is a dim carbon filament bulb that has been burning continuously since 1901-more than 1 million hours. In its current state, it emits as much light as a modern 4-watt incandescent lamp. If we understand the progress of the next 117 years, is it necessary to pay a premium for such "million-hour bulbs" at the turn of the 20th century?

The new $5 BR30 LED bulb I just installed in the kitchen is bright and clear: tests with a lumen show that the illuminance is more than 60% higher. In addition, they will save more electricity costs compared to old, inefficient companies

They replaced the light bulb.

Those bastards...

Check with you the documentary "Lightbulb Novel".

Is the movie in German or English

I had a very bad experience when using a certain brand of 100W LED bulb (I had 12 failures in 15 installations, and some installations more than once). I gave up 100W LED bulbs, which is too bad because they can save the most money. I might replace them with more and smaller LED bulbs.

When Home Depot started selling Cree products, they proposed a "return" policy for waste bulbs. I have brought in scammers the last few times, and this has made me more and more annoyed (although I have receipts etc.).

The 100W led bulbs we get from Home Depot can usually be used for a long time for a year, but the few failures I have seen are in the ceiling fan, the bulb is installed in a socket with a glass cover (effectively spreading heat across the bulb Around the radiator near the end and cooking electronics. There is no doubt that other well-ventilated bulbs in the socket last longer than the bulbs in the ceiling fan.

Interestingly, I have been observing the LED bulb failure in one of my ceiling fans, and I think it is caused by vibration. I never thought that this might be a hot issue on a global scale.

After replacing the bathroom lighting fixture, I reversed it (bulb facing up) so that the new LED bulb can run cooler and soften the light when it bounces off the ceiling.

So far, very satisfied with the mod.

As someone in the lighting industry, the first question I want to know is whether any lights in the same room as the fan have dimming functions. If so, someone is not running your dimming control line correctly (in any case, the possibility is greater). It causes interference in some way. The second thing I want to do is to stop buying Cree and GE. Despite having a good reputation throughout my career, I found that both companies are seriously lacking in LED quality control. Try to use TCP lights or topaz lights. You will get better results. Curiously, Cree produces their own LED diodes. Although they appear to be of high quality and many manufacturers use them in their fixtures, Cree does not seem to extend this quality to heat sinks and drivers.

I am an electrician and I also prefer TCP.

I'm a bit thinking, buy a higher power bulb and cut it to produce a lower output, which should equal a longer life. Compared with this bulb, it will effectively use a larger heat sink and heavier components than the equivalent bulb. The problem is solved, all of which only cost a resistor and some glue.

I suspect that replacing industrial soldered surface mount components with some household soldering heads will not actually extend the life of the circuit board substantially. As the author pointed out, it is entirely possible that you can save the life of the bulb for several years, but it is really not worth it.

Is that really worth the trouble? I don't think so.

That is feasible, except that many of them are potted or have a thick conformal coating, which would be a huge hassle to remove without damaging anything.

At the American Lighting Group, we have manufactured a bulb (especially BH4) with a service life of 190,000 hours. This is not spam, please visit our website www.uslightinggroup.com

How did you test? You haven't opened it in 21 years.

Hmm... do you have LM80 data to support this? I was thinking "No." This would mean nearly 32,000 hours of testing.

Is this the same company?

My name is wim. I have been in the leadership field for more than 23 years, and the content of the statement is impossible. Even if a rapid spike appears on the website, it indicates that the company is selling thin air. It's not that they are lying about everything they claim. Maximum lumens/watt life expectancy 190.000 h?? A huge Transformer that has been replaced by a miraculous weird part? And such examples are too numerous to list. For everyone with a little bit of head and knowledge about leadership, check it out. They said that if their own LEDs can work under perfect conditions, they can handle 100.000 hours... As far as I know, 100,000 is not 190.000, perfect conditions, radiator, drive, power grid, power failure, Fluctuations and temperatures need to be 100% perfect at all times, and their miraculous led chips can last up to 100.000 hours, maybe when you buy these led lights, you will get an extra 90.000 hours from the gods of led, they will give You brought 90.000 to nowhere to be hit and were found to enter the bulb for hours. Is the largest lumen output on the market theirs? With 160lm/w. I want to say that 160lm/w has been on the market for nearly 3 years, and the maximum value of commercial lm/w is now 220lm/w. Looking at the picture and its statement, it is clear that this company is another company because they are so many, because these companies are destroying many leadership fields and are doing well and doing well. product. Yes, when it looks good, you can be sure.........................

Another way to shorten the life of ceiling fans is the voltage limiter. They may disappear in the latest models, but I have a lot of fans that cannot run CFL because of them. It is a box inside the fan designed to reduce the consumption of incandescent lamps. When working at a university with thousands of students, I had to lay off some of them.

Did they really do it? Incandescent bulbs that reduce poop are even more efficient?

Heat is the killer. I have been obsessed with these low-power bulbs from 4chan since June 2016, and have never burned them since. They take about half a minute to brighten after opening.

I noticed that all bulbs use aluminum(i)um "radiator". But what are these radiators coupled with?

Simply thermally connecting the LED to aluminum does not solve the heat dissipation problem-aluminum needs to be connected to a heat sink or other room temperature heat sink. Aluminum PCB can only balance the temperature between all LEDs. Unless there is a place where heat is released, this will reduce the temperature of the PCB, otherwise there will be no benefit.

The small squares of the aluminum PCB are cute, but they don't seem to be thermally connected to anything.

But it is. It is coupled with the air in the bulb and coupled with the outside air through the bulb. Believe it or not, engineers are involved in the design of these bulbs, and they can calculate like this (or at least try to use them). Moreover, aluminum PCBs can operate at higher temperatures, which is of no benefit to LEDs, but it does affect the entire heat dissipation problem. Do you think that if FR4 can do the job, they will use aluminum PCB?

Jim is correct.

Do you want to know the difference between a computer CPU with hot synchronization and a computer CPU without hot synchronization?

A *no*...you will be lucky to make it through the startup screen.

One may run for several hours with no load.

Same as Peltier device.

These are made of semiconductors! ! ! ! High-end materials are made of silicon carbide, which requires very high temperatures (for example, the hot surface igniter of a furnace). There may be other materials now.

The difference between passively cooling any semiconductor and inadequate cooling is the difference between success and failure.

No need to deal with the store, just call CREE. Every time I call them, I send me a new mail for free.

All electronic products seem to have "baby ethics" issues. I have many LED bulbs that fail within a few days/weeks, and other bulbs only failed last year.

This is the so-called bathtub curve. Compared with the middle of life, things are more likely to fail in the early or late stages.

This is true for mature products, but the LED "bulb replacement" market does not yet exist. As a relatively new product, responsible companies are still observing how to effectively make them (cheaper!) and remain decent, while competing with "overseas" manufacturers, in some cases, as long as they don’t really care Quality can be fiddled with shelf space in the store. Since the cost of processing failed returns is higher than the cost of the store's initial purchase, they will not be shelved for long. Once unreliable brands are eliminated, the low end of the curve should rise a bit.

Unreliable brands will not disappear because they will constantly change brands, depending on how many batches they can sell before they are found to be bad products.

The reason for sticking with it is because people are always willing to try cheaper options in case they actually work. This allows manufacturers to consciously push a certain number of defective products to the market.

At the same time, I bought a large box of store-branded LEDs two years ago, but I haven't seen one of them go out. Maybe this problem is bothering some brands?

I think it is very possible. I also think that if you have not encountered any failures, then you are lucky. But this seems to be nonsense-you can't get everything you need to pay.

Well, how does an innate logical creature show morality?

I leave the room quietly now...

Screamed silently. Curse it is a short-lived existence. Desire to return to the earth quickly, all the components have started their journey. A hypothesis.

Is there a source for this quotation?

It is said that the only reason parents survive the toddler years of their offspring is because they are at that age. This kid can't kill with his bare hands. They haven't mastered the skills of lethal weapons.

LEDs can last for so long, but the electronic devices that support dimmable bulbs and other things they need are the highlights of the week.

Fortunately, even the weakest bulb can last more than a week. :-)

Conspiracy theories are much more interesting than critical thinking. Doubling the life of the bulb (by reducing the surface power load of the filament) will result in a 10% reduction in luminous efficiency. In other words, during the entire life of the bulb, approximately 10% more energy is required to obtain the same amount of light. During its service life, a 66 W lamp with a service life of 2000 hours will consume 132 kWh of electricity. Assuming the price tag of the bulb is 50 cents and the price of kWh is 10 cents, the total cost is $13.7. You will need two 60 W 1000 hour bulbs to match lifetime and light output, but the total price is now as low as $12.70. Different bulbs and electricity prices can distort the optimal lifespan, but the mandatory 1,000-hour lifespan is a regulation to protect consumers, not a collusion by the industry to defraud consumers. This seems reasonable.

"Incandescent lamps are manufactured in accordance with British Standard BS 161 (IEC 60064). The committee responsible for this work includes representatives from various fields such as government departments, professional institutions, organizations representing large users, electrical contractors and the lighting industry. They They reviewed all the factors involved, such as bulb cost, electricity bill, cost of replacing bulbs, bulb efficiency under different rated lifetimes, etc. In each case, they concluded that the 1000-hour long-term service life should be kept as standard Rated value. In fact, in order to minimize customer complaints, the average design life of a standard filament with a nominal life of 1,000 hours is usually 1,200 hours."

However, although it is cheaper for customers due to higher efficiency, bulb manufacturers do sell twice as many bulbs.

Either way, whether it is the initial energy saving and emission reduction or the bulb cartel hoping to sell more bulbs, I believe they will be happy to accept it because they will sell more bulbs.

Either way, the cartel, which was ultimately called the "government", eventually enacted the same regulations to produce the same effect, thereby making the whole problem meaningless.

Those who point to Phoebes Cartel usually do so in some kind of accusatory criticism, or oppose "capitalism" to "see that it always happens!", or support other conspiracy theories saying "Look, it happened!" ".

"Different lamp prices and electricity prices can distort the optimal lifespan, but the mandatory 1,000-hour lifespan is a regulation to protect consumers, not a conspiracy by the industry to deprive consumers. It seems reasonable. "

Since the first edition of IEC 60064 came out in 1954, this is interesting because conspiracy theories claim that manufacturers have been together for about 20 years or so. I can’t find the original publication date of the British Standard for verification, but is it entirely possible that the standard was developed after the company’s conclusion is proven? I mean, it’s not like a company that will never convince government officials that what they do is good for everyone... Wait…

Don’t ignore the conspiracy until you have all the facts. I personally think that these companies are fully capable of carrying out such vicious conspiracies, and there are quite a few companies involved in such acts (you know, the river in full swing). If you can prove to me that standards have been established since the conspiracy theory was proposed, then I would be more willing to trust you, but until then... I will continue to distrust any companies, because historically they have been used to cover up individuals who make wrong decisions for personal profit.

The crux of the issue is not who has set the standards, but whether the standards are reasonable and beneficial to consumers.

The physics of filament bulbs is well known-you can calculate the efficiency and electricity bill of the bulb yourself-and the story about the extremely efficient permanent bulb is in the same category as the 100 MPG carburetor.

Yes! As you describe, many people seem to like conspiracy.

"Industry conspiracy to defraud consumers"-Please explain why the cars in Asia and Europe have more than doubled the MPG# in the US? Is it because our dollar is based on oil?

Why did the 1980 Honda car cost 50 miles per gallon of gasoline, while the new Prius hybrid car only had 43 miles?

Compare VW.com or Ford.com with: VW.uk & Ford.uk!

Why can't you and I buy these models (especially when buying here)!

Greed and games make me waste time to stop cheering-unnecessary!

Watch the film pores.

Honda in 1980 did not have a 50 mpg (about 42 mph) car, and you are combining a car with 60 horsepower and minimal electronics with a much heavier car (with tons of electronics and two Times the horsepower) for comparison.

By the way, the 2019 Prius gets 50/54 mpg.

Yes, about three years ago, Home Depot, where I live, put a bunch of cheap LED bulbs on the market (the price of each bulb dropped from $5 to $10 to $1 to $3), and they all claimed to use Long lifespan, but many fail an interesting failure mode: They flicker annoyingly at a rate between 4hz and 30hz. In the early stages of failure mode, they would stop flashing after warming up and operate normally, but eventually the flashing lasted longer and longer and became unbearable.

So far, every time he happened, the LED and capacitor were normal. This smells like an unpleasant solder joint or something of similar nature...

I have seen this failure mode in pth and smd discrete LEDs. As you described, this mode will flicker due to the thermal expansion and contraction of the bond wires connecting the LED dies causing intermittent connections.

Two years ago, I moved into a fairly new home. The house itself is only 2 years old, so less than 2 years when the LED bulbs started to malfunction. After blinking for a while, he finally died. I separated a few. They are Cree brands and both failed due to poor solder joints that held the LEDs in place.

I have seen the same thing happen in power transistors, which are caused by intermittent bond wire contact and hot/cold oscillation.

Your parts are completely fake.

I suspect it is now, but in the past (presumably in 1980), a transistor in a transistor factory in the United States (National Semi??) had malfunctioning, which has been reported in the trade media. This part is usually used in the radio output stage (I have encountered it) and it sounds like the power cover is broken. Putting Putting Putting....

This reminds me of the old BC148 transistor. Although it is a low-power device used under stress-free conditions, it may crack and become unreliable due to its unusual structure (there is a photo in Google images) .

The interesting thing is that the filaments of the triode and rectifier that I have directly heated since the 1920s have been reliable, and they are always on today. Similar technology, but different "needs".

When the filaments are red, they are almost always there, just like in a vacuum tube. When you run them with incandescent lamps, the tungsten will evaporate in a large amount, causing failure within a year.

When building a house five years ago, I installed LEDs in every slot. The bulb burned out in the first half, I took it apart, and every time it was a power failure instead of an LED failure.

As a result, according to my policy, do not use dimmable bulbs unless the socket is located on the dimmer. The additional complexity of the circuit only means more opportunities for failure. Is it true? I don’t know, but it seems logical.

on the contrary. Dimmable bulbs can cope with large voltage changes, because most dimmers work by cutting off the AC waveform. The AC waveform will cause huge voltage ripples, which will kill the input capacitor of non-dimmable bulbs. Non-adjustable bulbs are less adaptable to power quality problems.

All other things being equal, more non-redundant parts of the critical failure path will always result in reduced reliability. If you want to improve reliability, use better parts and/or increase redundancy.

Although if the MTBF of these parts is millions of hours (for example, a simple filter coil), then considering that additional parts will make other parts longer life, the problem of reducing reliability through part counting is actually just picking .

good luck. Most of the latest LED bulbs from good suppliers seem to be dimmable.

I am currently decorating the house, and my lighting will all be LED strips with independent 12VDC power supply. In addition to improving the lighting characteristics (for example, making the light distribution more uniform), it also means that I can easily replace each component as they will degrade, fail or be improved. It can also ensure the installation of control options in the future. Moreover, low-voltage lighting and 120/220V wiring are not subject to the same code constraints.

In other words, for me, there are no more bad LED "bulbs"!

I have a 12V power supply in the cellar and an LED strip (with PWM dimmer) under the kitchen cabinet. I also have some motion detectors and night lights on the dark stairs. If the power distribution voltage is low, you can power all devices that require 12V power, such as routers and modems. Get rid of a bunch of wall warts.

It seems silly to run 3 wires of 120VAC on the #12 Romex* wire to run the LED.

*Previously did it in Rome, New York.

However, the distribution loss is very bad at 12V, and modern wall warts are very efficient. This may mean less confusion, but only for low-power devices.

There are 24 or even 48V LED strips. For "serious" lighting work, I use at least 24V. The voltage of 12V may be bad, this is because there is a voltage drop in the light bar itself. Some time after peeling a few meters, you only have 8-9V left.

I feed the two ends of the cabinet lamp that is 2 meters long. no problem. That tape is fragile.

Double as a communication device.

I think you are on the right track. In order to give full play to the potential of LED lighting, the era of A19 and similar bulb sockets must end. Inserting the power supply and its thermal components into a small light bulb that is combined with the transmitter is a good way to shorten the service life.

I plan to study 48V Power over Ethernet to distribute power for LED lighting.

POE++ makes swapping PSU more efficient.

The power efficiency of the best LEDs now exceeds 50%. As technology further improves efficiency, the heat generated in the lamp will no longer be a problem.

FWIW, obtained high CRI LED light bar. I bought a reel with "DIY Perks" (search here) for his photo/video lighting, and a 2m light bar (on the PWM dimmer) to illuminate 4m x 8m with *excellent* brightness s room. And quality.

Why don't you just use POE lighting. A POE network switch and standard Ethernet cable are required. Ubiquity has some panels, and I know that Williams Lighting has some different lamps. Circuits are usually limited to less than 100 watts, which is more due to data cable limitations.

Because I allow high-power configurations, these are the main room lights. In addition, my setup allows mixing 120VAC power (up to 2 amps per circuit) with 12VDC wiring, just in case I find that I need it in a specific room. It is much more flexible than what I said in the short description.

I have been engaged in LED lighting for 9 years. My experience is:

One kind. Low-wattage LED lights have the longest lifespan...not surprising here, because low-wattage = less heat, which affects the lifespan. However, the lamp purchased from 5 years ago will definitely last longer than the lamp purchased in the past 2 years. Hmm...is the design outdated? In almost all cases, the fault is the driver rather than the LED itself.

b. High-output LEDs have the highest failure rate. This makes sense, because high power/brightness generates heat. However, 90% of the failures are drivers...about 10% are LED failures, which later made the drivers even more confusing.

As for the "warranty"... I said. For example, Meanwell sells its drivers for 5 to 7 years, but I have some sites where the drivers begin to expire after about 3 years (about 40% of them expire within this time frame)... Completely under warranty. However, to get the warranty, the driver must be returned to the warehouse in the United States + pay for the return and re-import to replace the driver. In addition to the time/effort of recording and packaging the return driver, the shipping cost (return + replacement) is almost the same as the cost of the new driver... Therefore, the warranty is almost useless. In fact, it seems that the last person to touch the lamp is the one who must fulfill the warranty requirements. In this neighborhood, most electricians refuse to provide LED lamps, and will only install any lamps provided by others (customers, general contractors, etc.) to get them out of warranty issues (replacement + time/effort to remove/replace lamps) and drive program).

Interestingly. All the LED lamps designed and manufactured by ourselves (low output and high output...20,000 lumens) have a very low failure rate in the past 5 years (1 in every 10,000...with the lid closed), so it seems possible LED lamps with high reliability

Good news.. I just installed a ceiling light with adjustable color temperature.

I replaced the standard 48-inch tube with LED for a longer life.

Don't start with the bastards of the CFL.

I suspect that printing 100,000 hours of numbers on the packaging is cheaper than actually spending time testing.

If you want to increase efficiency, you would rather use standard 48" tubes with electronic ballasts instead of standard 48" tubes. They beat most LED products on the market in terms of luminous efficiency (lm/W), except for color LEDs and some poor-quality white LEDs (in fact, light blue).

The question then becomes whether you want to save money or replace the lamp one year in advance and spend very little time.

Your argument may be beneficial to fluorescent lamps in terms of lumens/wattage... But for fluorescent tubes, their cavity loss is higher, so compared to LEDs, more lumens are required to illuminate an area (for example, in In all directions where the LED emits light, the tube emits light in all directions) is more directional...provides more lumens when needed). Therefore, compare the lumens where needed, and you will find that the LED is more efficient. For example... our fluorescent lamp replacement product consumes a total of 14.7W (LED and driver) and produces slightly more light than the 32W T-8 (assuming the ballast coefficient is 1, so the power consumption exceeds 32W).

You will not have the reflector that commercial fluorescent light fixtures have, even if you paint it white.

@Ostracus

Even if the reflector is installed, the loss of the fixture cavity is still very high, because not all the cavities produced by the fluorescent tube flow downward...you need them most.

In addition, our 14.7W LED (modified) VS 32W T8 comparison is based on exactly the same lamps for comparison between Apple and Apple.

> "In order to have an apple to apple comparison."

If there is no real integrating sphere, it will be very difficult, because each LED will project light at a narrower angle, and their combined beam pattern is not uniform. In other words, if you place the luminometer under the lamp, you will find it brighter or dimmer – but this is still the problem because the fluorescent tube emits much brighter than the floor under the fixture. Depending on the reflector, the propagation angle of light may exceed 180 degrees.

The point is, to light _room_, you also want to shine light on the walls and ceiling to create diffuse light that does not cast shadows. Ambient light brightens the room. Pointing most of the light directly down to appease the photometer is cheating.

@Luke

For a true comparison using an integrating sphere, and only one measurement from below the light, you are absolutely correct. However, you assume that our comparison is based only on light meters under light... this is not correct. In most applications where lighting is used (offices, warehouses, etc.), there is no need to over-illuminate ceilings and walls. The goal is to improve the lighting and energy efficiency of actual work areas, which tend to be downwards from fixtures rather than walls or ceilings. Taking all these factors into account, the result is a significant reduction in energy use (with no additional controls, the minimum power consumption is 70%), and more lighting is provided where people in the room need it, and I am more comfortable with the people in the room. There is no respect.

> "There is no need to over-illuminate ceilings and walls. The goal is to improve the lighting and energy efficiency of the actual work area"

This is a special case where you can save more money by placing the spotlight on the workstation instead of overhead light pipes, but this can create annoying contrast.

If you are in general, take the average room. Or let us take my room as an example. I have all kinds of lights around the room, standard bulbs with a total value of about 200 watts, but they are all spotlights, illuminating a certain area of ​​the table or floor. Then, a 73-watt halogen bulb was suspended from the ceiling. Turn on all the spotlights, and the room still looks dim. Turn off the spotlights and turn on the ceiling lights-the room looks bright. For general lighting, directional lights are not efficient because you actually don’t want to throw them on the floor – you want to throw them everywhere to provide diffuse ambient light that can illuminate the entire space.

My motivation is to save the flickering of the detonator and the warm-up time in the cold room in the basement.

The problem with my tubes is that they must be recycled as hazardous waste. Energy-saving lamps can be put down in any hardware store, but the tubes must be sent to hazardous waste treatment facilities. Each tube costs less than 65 cents, but you have to find a waste management company that takes it away, and then you have to drive there and return. The pain in @ss.

It is ridiculous that modern house wiring (not to mention most of the installed dimmer switches) is not suitable for the LED lighting commonly used today. I am often asked "If I want to dim a new LED bulb, what dimmer switch do I need to buy?" This is a question I cannot answer, because the "compatibility" between the two is just an accidental game . In principle, dimming LEDs has always been an easy task. But when you have to manipulate the main power input in some way to "coax" the light bulb circuit so that it eventually causes less current to flow through the dice, it's not...so bad: -D

I have not seen a good dimmer-LED bulb combination that works well. The sharp TRIAC hacked alternating current cannot be easily converted into an adjustable constant current driver. Usually, the LED flickers and dims only to 80% of full brightness, which is unacceptable for such darker freaks. In the few places in my home where I want dimming control, I still use halogen bulbs.

Many dimmers designed for use in LED bulbs have a bias adjustment screw that you can use to align the bottom of the knob to the exact position.

I have never encountered a problem with running Cree, Philips or IKEA bulbs.

However, what I want to say is that if you want to maintain consistent dimming, you may need to stick to the same type of bulb. Generally, if you switch to another brand or model, you may have a very bright minimum, or there will be no light until the knob is at the 50% position.

The dimmer is the reason I did not replace most incandescent bulbs. My room is full of 100w BR40 (there are 18 in the restaurant and family room), which can glow from a soft night light to a pleasant, full room brightness. I have tried all brands-they either have insufficient dimming or insufficient brightness (or neither). The sunset effect bulbs are close, but even with compatible dimmers, they will fail at both ends of the range. The only option I saw was to replace the fixtures, wires and dimmers with commercial systems, which would be prohibitive prices. BTW, as far as bulb life is concerned, I have six BR40 100w 130V Sylvania bulbs on my old Lutron Skylark dimmer, which I use for several hours a day and have not been replaced since I moved in. The house... 25 years ago! ????

The only truly dimmable bulbs I recently discovered are some old/Edison style GE bulbs (560 lumens) equivalent to 60w, which can actually be used with my incompatible Lutron dual dimmers. They are dimmed to glow and reach full brightness. Now, if only 100w BR40 version is manufactured...????

Make your own LED bulbs, buy transmitters from component supply sites, and get supply of famous brands (avoid outdated Chinese/auction sites, fakes, etc.). Run the transmitter at 1/3 of the current of the bulb manufacturer. Build a good buck/boost driver for them. Use a good radiator. As an added benefit, their power efficiency is twice that of prefabs (check the spec sheet, they may list lumens/watt at low current). For the rest of your life, or until the thunderstorm interrupts, falls, or becomes better, please enjoy efficient and long-lasting light.

I also noticed that as the claimed lifetime of 100,000 hours disappeared, so did the older large LED cast aluminum heat sinks. These may greatly increase the cost.

I recently performed a lighting conversion in an apartment building from CFL to LED. The current consumption of each bulb is reduced from 300-350ma to 125ma. We found that by upgrading the brightness from 3000k to 4000k bulbs, the cost is slightly higher, and we can reduce the current consumption to 75ma. We have scattered more than 250 light bulbs in the building, and according to our estimation, the electricity bill will drop by 2/3rds. The 30,000-hour lamp is guaranteed to be used for at least 3 years.

Does the lumen output remain comparable, or are you just making all lights dimmer and bluer?

3000k and 4000k do not describe "brightness", they will tell you the color temperature. The counter-intuitive lower number is "warm", the bottom that is usually available in the 2700k range is very much like the color of a low-wattage incandescent lamp, while 5000k is more of a daylight color. 4000k is a good compromise between the two. In other words, unlike incandescent lamps, you can get high lumens of 2700k or dim bulbs of 5000k. What you want to show brightness is "lumens", which is actually the amount of light you see.

Usually, you will choose according to the location of use. Personally, I only use "daylight" white light bulbs in the work area, and beautiful low color temperatures where I want to relax.

Usually, on higher color temperature (more daylight/blue) LEDs, the efficiency will increase, but based on efficiency alone, fill your house with 4000-5000k bulbs, as if you are in a supermarket instead of at home .

Color temperature does not directly tell you how the light is displayed. That depends on the overall brightness of the light.

At lower lighting levels, redder lights should be used to maintain correct color perception. Otherwise, your lights will look too dazzling, and despite the obvious brightness, you actually see worse results. 5000 K bulbs are rarely suitable for lighting in ordinary rooms-they are more like office/workshop lighting, while living areas work best at 2800-3000 K, while outdoor lighting has even lower color temperatures. Because the overall intensity is usually much lower.

Don't forget CRI, color rendering index. This is a comparison between artificial light emitters and daylight. Artificial light emitters have a spike at a certain wavelength and change with the lamp, while daylight has a good curve. Even light with different temperatures will vary according to the location of the spike.

I just dissected a malfunctioning LED bulb and found that the first LED in the light string on the lighting board was malfunctioning, and there was an obvious black spot below the yellow dot.

Before it fails completely, it flashes on and off every few seconds.

The place where I worked before had 400 volt AC fluorescent lights.

When my light bulb fails, I will call the facility, say a light bulb has gone to "disco" and give the location. They know what I mean...

Therefore, since the LED bulbs seem to have constant current drivers, you may just need to bridge the faulty LED and regain the working light.

If these LEDs have this failure mode, it would be even better, such as some old Christmas tree lights and incandescent bulbs connected in series.

Some LED Christmas tree lights also use similar "short circuit shunts."

Is a capacitive dropper used? They do this mainly due to high peak surge currents.

Connect two systems in series: system reliability = Ra * Rb

Two systems in parallel: system reliability = 1-(1-Ra)*(1-Rb)

It becomes complicated. I bought a lamp with CC driver from Lowes, but the LEDs are connected in series and parallel. In other words, 3 in parallel and 4 strings in series. Therefore, if one LED is turned on, the other two LEDs are subject to current, thus speeding up their extinction. If there is a short circuit, no problem, the light is reduced. However, unless they are overvoltage, I have not seen any short circuits, which is extremely unlikely to happen in CC feeder strings.

So far, they have been running for two years without problems. Some commented that "they are almost burned out". I don't know they may be overheated.

My Cree 100W bulbs all failed within one year. They are not cheap either. :-(

for sure

Dice -> death

?

Usually the dice don’t die, but a distorted dice game may end in someone’s death!

B ^)

Both are correct.

The dice are plural. that's it.

>" Again, increased brightness is believed to be the cause. However, the last step to halve the life of the bulb can only increase brightness by 11%-16% while doubling sales. This is about selling more bulbs and earning more Money."

"as long as"? The cost of the bulb is US cents, and the cost of the energy used to light the bulb is US dollars. Therefore, the savings for consumers is still greater than the money lost by having to replace the bulb twice. Win-win.

The real meaning of "cartel" is that the industry is doing its best because this is the lowest point of competition. The competition among sellers is who owns the longest light bulb. It becomes the actual measure of quality. As a result, the light bulbs become dim and wasteful, and the quality varies so much that they are hardly better than gas lighting. The 11% higher brightness also means that the color temperature is increased, and controlling the life (filament temperature) means that the light color, efficiency and rated power between the bulbs become consistent.

The irony is that nowadays, with standardized CRI, color temperature and power rating, the effect is the same: incandescent bulb life is the same among all manufacturers, and there is no cartel. The government eventually did the same job, because it was reasonable to compromise the life of the bulb to achieve other good qualities.

The bulb-shaped screws make the heat sink worse. Especially when it is installed, it points downward.

What is the service life of lamps designed for LEDs? Recently, I bought a cheap LED workshop lamp, which is equipped with a string of LED lights, and the driver electronics are directly mounted on a long metal frame-the whole lamp is a radiator, and I think the temperature above 40°C is not high.

It sounds more like the form factor of a fluorescent lamp. There are hard parts in the ballast and the bulb has a large surface area.

As far as I know, heat is the killer of LEDs (electronics), and part of it is the use of LED lights in the armature used for incandescent lamps. For most of my lamps, I find it interesting to buy cheap bulk LED panels and COB panels from Ali and design my own armature around them and use a separate power supply.

About those idiot lights:

> "Are the initial estimates overly optimistic? Are they all marketing hype? Or do we not know enough about the aging of LEDs to predict the true life of the bulb?"

Yes, yes, yes-the fourth point is that manufacturers use fake numbers to fake lifespan, because there is not always a standard to measure them, or it is not necessary to follow a standard to sell them. Different incomparable indicators are used between LED bulbs. Some measured light output dropped to 70%, some dropped to 50%, and some dropped to complete failure. The color temperature may also change.

The fifth point is that the brightness of early LEDs was much darker-manufacturers madly claimed that the power of LEDs was equal to incandescent bulbs, so they sold people 3-4 watt lamps, claiming that they were equal to 60 watt lamps, due to lack of relevant Common standards of how to measure. A common technique is to declare equivalence through brightness (bright light at a narrow angle will not illuminate the entire room). These less powerful lamps do not heat too much, so they have a longer life.

Key points:

> "The new $5 BR30 LED bulb I just installed in the kitchen is exceptionally bright and crisp: tests with an illuminance meter show that the illuminance is more than 60% higher."

You don't know if they actually emit more or less light, because the lights are directional. If you replace an omnidirectional bulb with a directional bulb, you may still darken the room even if you measure higher brightness directly under the light because there is less indirect light through the ceiling and walls. Except for the light bulb, the room is dim.

Yes, this can explain why kitchen lights use many lamps mounted on a metal plate that doubles as a reflector and radiator.

Even with recessed/directional bulbs, a common technique is to further reduce the beam by 5 degrees to maintain the brightness directly below the lamp, thereby reducing the edges of the lamp. The illuminance meter shows higher brightness under the same power – consumers are satisfied and think they can get more benefits from it.

Regarding the planned phase-out of incandescent lamps: When I was a young man (USA, 1950s), you could buy a set of bulbs for your house, and never buy replacement bulbs. The energy company has a light bulb replacement program, you can take the sundries to a convenient place (such as a pharmacy), and they will provide you with an equivalent new light bulb in exchange. This lasted for at least twelve years, until someone (probably the bulb manufacturer) sued the company. I never understood the economic significance of this exchange, but it is good for us, so...

It ensures that you will never consider replacing the utility company.

This is an energy utility. Energy costs far exceed the cost of light bulbs.

Of course, they provide you with long-lasting – inefficient:-( –light bulbs. When you waste electricity and buy more electricity, they will like it. Through the exchange program, they can ensure that you never use it "accidentally" A more efficient light bulb.

Great article. Next, I would love to see information about dimming circuits, especially if you are willing to explain why the minimum brightness of LED bulbs is much brighter than dimming incandescent lamps.

After several disappointing purchases, I gave up looking for a service that could provide the smallest luminous LED at night.

Thank you!

This is because incandescent bulbs are a diffuse light source (especially frosted bulbs), while LEDs are point light sources, so when you look at it at night, even a small amount of light will form a hole in your retina.

In order to minimize the light output, please use one of the LED decorative light strings driven by a single AA battery and insert it into a bottle or vase. Replace the battery with a wall wart, and you will have a night light:

If you use high-quality dimmable LED bulbs, some compatible dimmers are usually recommended. There are two Philips floodlights in my hallway, a dimmer is assigned to the box, and I have full dimming capabilities, and I can even see almost invisible glow to full brightness. Most low-cost dimmers are not really compatible with LED bulbs. Recommend a dimmer for LED bulbs for yourself, the dimmer is actually listed on the box with compatible dimmers, you will find that there is a big difference between them. Yes, it will cost you more, but it is worth it.

Obtain a dimmer for the LED and read the manual. Usually there is an adjusting screw in one of their positions to set the minimum chopping voltage of the triac.

The minimum brightness of a dimmable LED is usually very dim on a good quality LED (although it will not be orange/red like a filament, so it will look different), but an incandescent lamp will need more before doing anything Therefore, dimmers used for incandescent lamps or unregulated LED dimmers will set the zero point too high.

One of the disadvantages is that different bulbs will dim in different ways, so if you use different bulbs on the same dimmer, they may be powered on at significantly different settings. Similarly, if you set a dimmer for a bulb and then switch to another model or brand, it may be very different.

We can see amazing low illumination. I work in a group that designs LED drivers. We cited the ability to dim 50,000:1 on some drives, and you can still see the LED glow at 50,000 of its full rating. Incandescent lamps will emit different light with temperature changes. At the lowest temperature, they will emit a bunch of infrared rays, but no visible light can be seen. All the output of the LED is visible (or UV), so if it is rated at one ampere, its visible light will reach 100 microamps. The switching regulator is dimmed by reducing the duty cycle, and due to the jitter of the duty cycle forward pulse length, it is difficult to reduce the duty cycle to 0.01%.

Purkinje's effect means that when you turn down the dimmer, our sensitivity to light will decrease, because color vision will disappear at lower lighting levels. The red disappears first.

Even when dimmed, the LED will continue to emit blue and green, which makes it stand out in the dark. Due to the persistence of visual effects, people can also perceive that the speed of light strobe is less than 1/2000 second, so the PWM dimming method does not work at all. In fact, it is used in many flashlights to extend battery life, because you have to reduce the duty cycle a lot before the apparent brightness starts to drop.

Incandescent filaments have a highly nonlinear response curve. At half power, it emits a dark red light. LEDs have a highly linear response between current or power and light output. The relationship between voltage or PWM ratio and power varies greatly depending on the PSU technology and structure.

This is not rocket science; it's just that you get what you pay for. Consumers want cheap LED bulbs, so manufacturers are scrupulous about power circuits. (Actually this is the case from the first day). I have never seen an LED bulb fail to light up an LED. They are always power failures, and long-term thermal damage to components or PCBs is almost always visible when turned on. Have you ever felt the base of an LED bulb? Crazy hot. As long as the heat flows away from the component, the LED can handle it, but a typical power supply component is not a fan with a large amount of heat for a long time.

Interestingly, I had some "$2 Home Depot with SMUD discount LED bulbs" a few years ago, and I have never encountered any problems. (SMUD is my local power company) They have only 40W equivalent power, but they are more durable than a few 60W and 75W LED bulbs. My real test is the 100W Philips LED bulb owned by the porch light. You can almost feel the heat at its bottom like an incandescent lamp. So far, all night, every night has lasted for nearly a year. Just wait and see.

The cheapest LEDs have no power drivers-no electrolytic capacitors-they have two LED strings in opposite directions and a constant current limiter.

They are stroboscopes, but they are inexpensive and have a long service life. The moment a capacitor is added to eliminate the ripple, the life will drop to the failure rate of the capacitor, and the failure rate is sensitive to the quality of the line power supply and the ripple current (dimmer).

Then, more expensive lamps will add transient voltage suppressors and power factor correction circuits to work with other equipment, but you are talking about $20 per bulb, and at this price, you expect something lower than CRI , Narrow angle, uneven light distribution, low wattage lamp.

Usually they use bridge rectifiers.

Not the cheapest kind. This is because they can take the nominal power of the LED die and print it on the bulb as the rated power of the bulb, even if each LED only runs half the time.

I have an LED light with 4 watts written on it. The actual power consumption is 1W. They put 4 W LED in it.

To be honest, I cannot tolerate any form of LED lighting, which has almost the same spectrum as fluorescent lamps.

The stroboscopic flashes make me dizzy and nauseous.

I have a supply of incandescent bulbs and halogen reflectors for more than 30 years, most of which are on dimmers, so they help reduce power consumption, and when used with dimmer switches, they seem to last for years.

Sorry, but this is almost completely incorrect. It shows that you either have never seen a suitable LED bulb in your life or just didn't realize that you already have it. I once had a similar argument with one of my uncles-he "hated those new LED bulbs", but we had a conversation in my LED-lit living room and he was very surprised when I pointed this out to him .

First, the color spectrum is completely different. Fluorescent lamps usually use three-color narrow-band phosphors (red, green and blue), and each phosphor has a very obvious peak in the emission spectrum. LED uses a broad-spectrum phosphor mixture in which only a strong blue peak appears in the blue light emitted by the die itself. Here is an interesting article about measurement:

This article is so old that their conclusions on brightness and cost-effectiveness are outdated, but the color measurements summarized on the first page are still valid. You can even find the same chart in each LED data sheet. In this test, there is an amazingly colored LED bulb, but I have not seen anything like this elsewhere or recently.

Second, the only LED lights with strobe flashing I have seen in the past decade are cheap Christmas lights (unless you are running LED bulbs on inappropriate dimmers). I am very sensitive to LED flickering-I can see it in some car taillights, LED Christmas lights