5mm LEDs Generally Unsuitable for General Illumination

Posted: November 30, 2008 in Reviewed - Poor Performers
Tags: , , ,

The 5mm LED has been around for a very long time. Containing a single LED die (chip) and no thermal management beyond the lead wires, these devices are low in cost and available in a range of colors, including white. Optical control is very limited. At very low power levels, these devices serve as instrument lamps, indicator lights, and are often clustered to create arrays for signage, effects, and signal lamps. These are the small LED “lamps” found in virtually all low quality and novelty LED products.

For general illumination, where demands for significant lumen output, long service life and consistent color and lumen output are necessary, the 5mm LED is simply inadequate to the task. The large number of low end products on the market using 5mm LEDs marketed with unsupportable performance claims and very poor performance makes sorting the good from the bad is virtually impossible. Because of this, the best approach is to reject all products using this form of LED outright in favor of packaged LED products. Customers perceptions formed from disappointment with the performance of low quality 5mm LED products will hurt the entire LED market, and slow adoption overall.

Packaged LED devices, or “lamps”, mounted to proper metal clad circuit boards and thermally inductive heat sinks will out perform 5mm LED devices in lumen output, lumen maintenance, and service life, by a large margin.

Unfortunately, counterfeit LEDs that are made to look like quality products have begun to infiltrate the market. However, when compared to the volume of poor performing 5mm LEDs in the market, the best bet is to select a product using packaged LEDs.

There are a few applications where the low cost of the 5mm LED is acceptable. For example, novelty lighting, simple direct view indicator lights, graphic displays, and other small, low energy uses are just fine. However, do not expect them to last “forever”, as they will likely fade in output so dramatically, that they will be discarded after a small percentage of their “rated” life has expired.

For flashlights, nightlights, reading lamps, and other similar applications, even though these devices are common, their performance is so poor the products are essentially trinkets with no real value.

For energy savings, the only value products that depend on 5mm LEDs offer is low total wattage. These devices have an efficacy that is no better than incandescent lamps, and far less than fluorescent, regardless of claims made on packaging. The reality is that virtually all of the claims of energy saving are based on the low wattage rating of the products, and not on actual equivalent light output. Virtually all claims of improvement over fluorescent lamps are false, as the 5mm LED is simply not able to produce the same lumen output, and rarely equal life. Claims to equal the light of MR16, PAR and R lamps in output are dubious, and based on initial values only, and are always overstated. As the product is put in use, the rapid degradation of light output will soon degrade their serviceability. Compounding this, is the fact that the majority of products using 5mm LEDs use very low end power supplies of low efficiency, that cannot be dimmed. Finally, the color produced from 5mm LEDs is generally poor, is inconsistent, and degrades to blue over the life of the product.

Packaged LED devices, with proper thermal control, offer a significant performance advantage over 5mm LEDs, incandescent, halogen, and many CFLs, along with far longer life, greater lumen maintenance, superior optical control, and far greater color consistency. For energy efficiency applications, there is no better choice than high quality, high and medium output packaged LED devices.

The graphic below illustrates the range of performance one can expect from 5mm LEDs (yellow area), and high brightness packaged LED devices (blue area). Where typical 5mm LEDs will have lost 50% of their light at around 8,000 hours of life, HB packaged LED products will maintain 70% of their light output at 50,000 hours.

  1. kwillmorth says:

    Clarification: I have received comments noting that there are quality 5mm LEDs that exhibit good life and lumen performance. This is true, to a point.

    First, the intended application must be suited to the use of low intensity, low brightness light sources, simply because 5mm LEDs have no thermal coupling beyond the legs of the device and the board they are mounted to. There are no High Brightness LED die that can live without proper thermal management. This means that, in applications like a novelty candle or decorative feature where the LEDs are viewed directly, high quality 5mm devices may be fine – of one can sort the high quality from the junk.

    Second, there is simply so many poor quality 5mm LEDs and LED products using junk 5mm LEDs, that sorting them out from quality products is impossible for the consumer and specifier. The approach is far simpler to avoid them altogether, as the odds are against finding good quality product in this format. Some offer that there are counterfeit LEDs in all package designs, that this approach is invalid. I suggest that the ease in which 5mm devices are applied in low quality products, coupled with the proliferation of these devices in poor quality, makes them disproportionately more risky. If the SSL market allows this to spread into all form factors without direct action and resistance, then this market is doomed to fail in the price sensitive markets, where poor quality manufacturers prey on consumer ignorance.

    I stand behind my personal feeling that the 5mm LED is suitable only for intermittent use in display and indicator lighting, and not to general illumination. There are far too many superior products of high value and dependable quality for this market to be need anything even the highest quality 5mm (and all other similar through hole products) bring to the table. As the illustration shows, even the best of these devices exhibit far poorer life and lumen maintenance than surface mount LEDs on proper heat sinks.

  2. Jeff in Texas says:

    This article, to which I was led by a link to it on EDN.com (http://www.edn.com/blog/1470000147/post/240041824.html) confirms everything I have suspected about the low-end LEDs currently in common use in Christmas lights and the like. They all tout 10,000+ hour lifetimes but I have had so many strings and displays fail, I was driven to the conclusion that it was all a big con. Thanks for the info.

  3. Steve Nordquist says:

    Cite! ‘Your gut’ and that graph show 2 curves with coefficients of different signs (can you not graph on a log^2 so we can see the lines and work out the coefficients? kthxbi) and no reckoning what actually turned your gut.
    Video of you holding a loupe up to a LED, gauging up the leadframe, zeroing in on the merits of the cupola (if any) and maybe taking IR emission patterns (all to ribald commentary) would be precious. Electronics What Not To Wear (to a specsheet) sounds -tasty-.

    Is that some kind of epoxy bilayer? Did I say thermal fail? I meant squirrely fail. Be green and stick to 24ga. wire and 1.5mm, Darfkin PLC.

  4. kwillmorth says:

    The chart is derived from several different graphs generated by the Lighting Research Center (2002), Osram and Philips presentations (LEDs 2006 and 2007 Strategies in Light) and from manufacturer product data sheets (2007-2008). The range and shape of the graphs is a result of this derivative summary and is intended to represent the range of performance one might expect, based on simply overlaying the various graphics over one another. My own tests have not shown any of the 5mm products attaining the life these curves show, they have been much much steeper in loss (shorter usable life).

  5. Steve Nordquist says:

    Fair enough, that’ll save me some infamy and trouble trying to make 1/2 W sockets for 30cd. devices (not to stop those out there ion milling cooling channels into packages) or wishing on composite devices with no test data. Looking at the package thermal conductivity first makes sense.

  6. […] not: Take a look at this chart from the Inside Solid-State Lighting blog of Kevin Willmorth of […]

  7. […] is an interesting article that clearly points out the deficiencies of 5mm DIP style cluster LEDs. 5mm-leds-generally-unsuitable-for-general-illumination/ Leave a Reply Click here to cancel […]

  8. dave99999 says:

    The primary problem is the products drive the LEDs at too high a current instead of using a greater quantity. Thus, while they “could” make high quality products with them, instead they choose them for lowest cost products. They say it’s a 100mW LED but if you want better lifespan, try to stay under 30mW, particularly with the generics.

    Speaking of generics and really I just mean unfamiliar Chinese companies, things are changing. Today you can even get big 9 x die 10W generic LED packages from China for under $1 ea. in volume. Yes you should also drive those at lower than max spec’d current for best life, but it’s getting a bit crazy how much light you can produce for $10 (sans heatsink, driver, etc.)

    • kwillmorth says:

      I stand by 5mm LEDs and others of similar construction not being usable in general illumination. The new trend now is for manufacturers to place plastic case SMT LEDs with no thermal path, mostly on tape light products, claiming long life. They too are not suitable, and have proven to have severe lumen depreciation and failures. Just because they are cool to the touch does not mean the die inside are not being cooked to death. Plastic cased SMT LEDs are just another type of cheap LED product, not a quality solution. You can stick that tape on a heat sink if you wish, but without a thermal connection between that LED die inside and that heat sink, it might as well be a block of wood. Lastly, I am not a fan of the flood of cheap Chinese made generic junk flowing into the market. The vast majority of it is worth less than the pennies they sell for, while their presence tears the market value away from actual quality products. You can certainly fill a product up with little trashy LEDs driven at super low current to protect their fragile design… or spend a bit more on a single reputable COB module generating more light, with higher efficacy, and a thermal path that can be used to build a long lived product for real.

      • Han-Lin says:

        I bought waterproof LED strips from ebay. At their rated voltage of 12V, they felt very warm. If they’re quality should they feel warm?

  9. Han-Lin says:

    What about using 5mm LEDs for bike lights? The wind from riding can benefit the LEDs. 20mA SMD versions seem to have much better lumens per watt. I can’t seem narrow or medium angled ones that also ship to Canada at low quantities yet.

    • kwillmorth says:

      Through hole LEDs simply do no generate enough light to be effective for any forward lighting. They do work for marker and tail light applications where direct viewing is their purpose. There are now many better options, like star board mounted SMT high power LEDs for headlight use.

      • Han-Lin says:

        I guess they’re bright enough for daytime visibility when aimed straight. They should be dim enough not to be blinding. I use high powered ones for lighting the road at night.

Leave a Reply