Nichia Develops 60 Lumen Per Watt White LED

[jtr1962]

I thought this article might be of some interest since I've noticed some others besides myself have an interest in white LEDs. This is a fascinating development since it puts the efficiency of white LEDs at about the same point as compact fluorescent bulbs and T-12 linear fluorescent tubes. The T-8 tubes still do somewhat better, however, at 80 to 100 lm/W, but white LEDs should surpass this benchmark within a few more years at the most. Current production white LEDs, at 20 to 30 lm/W, are already the most efficient means of generating small amounts of light. Very small fluorescent tubes are far less efficient than larger ones, and require far more complex driver circuits than white LEDs. Incandescent bulbs of any size are inefficient(<20 lm/W), but the small ones can be as low as 1 lm/W. Moreover, dimming bulbs to save power causes them to be even less efficient whereas white LEDs actually are more efficient at smaller currents. For example, decreasing the drive current from 20 mA to 5 mA results in an increase in efficiency of about 40%(power is reduced by 75% but light output is only reduced by 65%).

 

[CougTek]

Thanks for spreading the news. I would really like to replace all my lamps by LED-based ones. Be it just for the better color reproduction they allow compaired to the reddish light of incandescent bulbs.

Unfortunately, Nichia doesn't plan to market its new LED before 2005. Two years away.

 

[Handruin]

If I remember correctly, florescent lighting always played havoc on video recording because of the frequency and color temperature. (Professional studio recording)

Do you think LED lighting would help the film and video industry for more accurate light reproduction? I'd be curious to see how the color temperature is compared to halogen and florescent lighting.

 

[jtr1962]

I would really like to replace all my lamps by LED-based ones. Be it just for the better color reproduction they allow compaired to the reddish light of incandescent bulbs.

Amen to that. One of the first things I did when my sister moved out and I got her room was to change the fixture to a fluorescent one. I hate incandescent light, and besides that, I like any area where I live to be brightly lit. Prior to removing the bulb fixture I had 2 200W bulbs which made the room too hot and still left quite a bit to be desired as far as lighting quality goes. Up until recently, I was using the standard cool whites(4100K), but I recently tried daylight(6500K) for my bedroom and like it much better. I even got a 19W daylight compact fluorescent bulb for my desk lamp. My brother said it looks just like a big white LED. Until LED becomes common place, you might want to try fluorescent, especially daylight. The light is very similar to white LEDs, and I think the color reproduction is just as good. The newer T-8 fixtures use electronic ballasts with none of the flicker problems associated with older magnetic ballasts. I have sunlight(5000K) for my workroom, and here again it's a big improvement over the cool whites.

 

[jtr1962]

If I remember correctly, florescent lighting always played havoc on video recording because of the frequency and color temperature. (Professional studio recording)

Do you think LED lighting would help the film and video industry for more accurate light reproduction? I'd be curious to see how the color temperature is compared to halogen and florescent lighting.

The older fluorescents ran on magnetic ballasts and flickered twice every line cycle, or 120 times per second. Besides that, the color reproduction was awful, with far too much green. Newer tubes, whether T-12(1.5 inch) or the newer T-8(1 inch), have much better color rendering properties. The daylight tubes in my bedroom have a CRI of 85, and for the sunlight ones in my workroom(GE Chroma 50s) it is 90. The sun is 100. Colors look quite good under these lights. As to the flickering problem, it is non-existent with newer T-8 tubes running on electronic ballasts. The typical operating frequency of these ballasts is about 50KHz. The tube lights continuously because the driving frequency is far faster than the tube's time constant, so it doesn't have a chance to turn on and off as it does at 60 Hz.

LEDs run on pure DC, so the light is continuous. The CRI of white LEDs can be over 90, and in fact one of the express goals of the Solid State Lighting Initiative is to have LEDs with an efficiency over 150 lm/W and a CRI of 90+ by the year 2012. As far as color temperature goes, here is the general breakdown for conventional lighting: incandescent(2600-2800K), halogen(3000-3200K), warm white fluorescent(3000K), cool white fluorescent(4000K-4200K), natural or sunlight fluorescent(5000-5500K), daylight fluorescent(6500K). For most current production white LEDs the color temperature varies from around 5500K to 8000K, with the design goal usually being somewhere in the 5500K-6500K range, which in my opinion is just about perfect for white light. The only problem is that color temperature for most mass-produced white LEDs is highly variable even within a given batch. In time I'm sure this problem will be solved, however, and there will also be a wider selection of color temperatures. In fact, Nichia recently developed a warm white LED(3000K-3200K) to satisfy incandescent freaks. I'm guessing perhaps any lighting system designed for LEDs in the future may even let the user select whatever color temperature they prefer so as to avoid having to mass produce LEDs in a plethora of color temperatures. This can easily be done by having mostly pure white (6500K), and then lighting a smaller percentage of yellow or red LEDs for lower color temperatures.

 

[CougTek]

I found that page when searching for T-8 fluorescent tubes. This is all very informative (just like your post, JTR), but there's a part I have a hard time to understand :

"Full spectrum" lamps are defined as fluorescent lamps having a CRI of 90 or better and a color temperature of 5000K, designed to provide a completely balanced pattern of visible wavelengths of light.

Full spectrum lamps produce the most accurate colors; however, they are not "natural" in that they do not replicate natural sunlight or the normal conditions in which people view things.

A light with a CRI of 90 should be quite close to what sunlight is and therefore, should replicate the normal conditions in which people view things. No? How can a light with a Color Rendering Index of 90 be that different from sunlight, which has a CRI of 100?

Mist, darkness, uncertainty and confusion here...


P.S. Thinking about it and re-reading the sentence, isn't it the 5000K color temperature that f**ked up our perception of the light, rather than its CRI?

 

[jtr1962]

A light with a CRI of 90 should be quite close to what sunlight is and therefore, should replicate the normal conditions in which people view things. No? How can a light with a Color Rendering Index of 90 be that different from sunlight, which has a CRI of 100?

Mist, darkness, uncertainty and confusion here...


P.S. Thinking about it and re-reading the sentence, isn't it the 5000K color temperature that f**ked up our perception of the light, rather than its CRI?

The reasons the light is not exactly the same are fourfold, possibly fivefold. First of all, the CRI of 90 means that the light is still deficient in some wavelengths. Second, the sun is a point source as opposed to a diffuse one. Third, ambient sunlight can be as intense as 130,000 lux, or lumens per meter², whereas few rooms are lit more brightly than a few hundred lux. Fourth, the sky acts as a secondary light source with a somewhat higher color temperature(i.e. bluer) than the sun itself. A possible fifth reason is that sunlight contains UV, which although harmful and invisible, may cause certain objects in nature to fluores(spelling?) and reemit shades of visible light, thus making them appear different than they would under a artificial source with little UV. All of these factors combine to make it virtually impossible to duplicate outdoor viewing conditions indoors, although we can come very close. The full spectrum tubes are effective enough to be used to cure seasonal affective disorder.

There are a few errors in that article you linked to. For starters, you can get both T-8 and T-12 tubes in many color temperatures and CRIs even though the article gives the impression otherwise. All of my tubes are standard T-12s, although I may get a T-8 fixture for my bedroom soon because the (nonvisible) flicker of the current tubes combines with the refresh rate of my monitor and ends up giving me a headache. Retrofitting a T-12 electronic ballast would be just as good, but I haven't been able to find any. When I light the room with my desklamp, which has an electronic ballast, I don't get headaches as quickly. Of course, staring at words, even on paper in sunlight, will give you eyestrain or headaches sooner or later. Second, you can most certainly still purchase T-12 tubes even though they are largely being phased out in commercial buildings. The article gives the impression otherwise.

As to how color temperature relates to our color perception, consider that incandescent sources have a CRI of 100, meaning that they have the same spectrum as a black body radiating light at whatever temperature they are operating at. The sun also has a CRI of 100, but at a much higher color temperature(5000K vs. 2700 to 3000 K). Therefore, incandescents tend to bring out the red in objects more than a cooler(high color temperature) light source would. As such, they are far less natural light sources than the sun despite their high CRI, even though much of the general public may disagree with me(probably because humans can adapt quickly to anything, and even prefer the poor thing over a better one once they get used to it). Therefore, I tend to think it is the overall combination of color temperature and CRI that makes light appear "natural" to us. I also tend to think that the general public will in time come to prefer a cooler light source since it is closer to natural sunlight. I'm still working on getting my sister to switch to fluorescents. Maybe I'll have better luck when white LEDs are in common use since they are point sources like light bulbs.

 

[LiamC]

jtr,

flourescent light sucks. Period. I prefer sunlight in our house, but if not, incandescent. I like halogen most of all, but can't afford to rewire the house yet. Note meaning to offend you, just expressing a different opinion.

In our living room, we have a circular T-12 flourescent. I despise cool white (with a passion) but my wife likes it. I tried a warm flouro (more like incandescent light) which I liked but my wife hated. Recently I went for a daylight flouro and we both hate it.

Now I am getting on - 41 in a month or so in fact and I have noticed that in the last few years, I need brighter lighting to feel comfortable at night and my ability to distinguish objects in low light conditions (contrast??) has fallen off noticeably - I was only thinking on this last night with a mental note to slow down whilst driving at night. What I am trying to get at is that flourescent lighting (at night) seems to "colour" the room in such a way that I feel like have covered my eyes with the finest of gauze - not noticeable and see-through - but slightly distorting/disorienting. Incandescent of halogen light does not affect me so. Wish I new why this was so...

 

[LiamC]

D'OH!

substitute Not for Note (2nd para) and Incandescent or halogen (not of) at the end of the 4th para.

 

[jtr1962]

jtr,

flourescent light sucks. Period. I prefer sunlight in our house, but if not, incandescent. I like halogen most of all, but can't afford to rewire the house yet. Note meaning to offend you, just expressing a different opinion.

....

What I am trying to get at is that flourescent lighting (at night) seems to "colour" the room in such a way that I feel like have covered my eyes with the finest of gauze - not noticeable and see-through - but slightly distorting/disorienting. Incandescent of halogen light does not affect me so. Wish I new why this was so...

No offense taken. My sister's views are close to yours on this subject, although I'm starting to think that her fairly regular depressed moods are in part caused by the incandescent lighting in her house. Ditto for my brother.

My guess is there are two reasons you're not happy with the lighting in your living room. First of all, since you're using a conventional circular fixture with a magnetic ballast the light flickers 100 times per second(120 times in the USA). You may not be able to see this, but it can certainly affect you in very profound ways, including giving you headaches, and it is definitely more noticeable at the lower line frequency used in your country. For all these reasons I recommend a T-8 fixture with an electronic ballast to anybody putting in a new fixture. Since the tubes are run at a faster frequency than their time constant, the light is continuous just like a bulb or LED.

The second reason is that in all likelyhood the tubes you have purchased, especially mass-produced cool white and warm white, have awful color rendering proporties. Some daylight bulbs, such as the Philips Alto tubes in my bedroom, are somewhat better, with a CRI of 85. Since you said you like sunlight, you'll likely be very happy with full spectrum sunlight tubes(5000 to 5500K) running off an electronic ballast. This will be far better than any other solution. I've noticed a huge improvement in my workroom since installing the sunlights, and as soon as I find an electronic ballast capable of running T-12 tubes I'm replacing the magnetic ballasts in my shop lights. Here is a link to an online store specializing in full spectrum solutions. I wish I had found this sooner since their prices are quite reasonable($9 each for a high efficiency sunlight tube which gives 3600 lumens as opposed to the 2250 lumens my $6 Chroma 50s give). They also have full spectrum T-8s which will work in any standard T-8 electronic ballast fixture. All told you're talking an investment of maybe $50 US(Home Depot prices) for a decent fixture and maybe about the same for 6 tubes(cheaper in quantities of 6+). Well worth it since light affects your well-being in so many ways as I've only recently discovered. Doubtless your wife will love the full-spectrum tubes as well since they appear slightly whiter than cool-whites, with none of the awful yellow characteristics of either warm white tubes or incandescent lighting. A somewhat cheaper way to try full-spectrum light is to purchase full-spectrum compact fluorescent bulbs(which all have electronic ballasts) from the same place. It might be a good idea to do that first in one room before you commit to a large fixture and T-8 tubes for your living room. Some of the better T-8 fixtures are even dimmable. I'm considering a 4-tube dimmable fixture for my bedroom since it'll let me light the room like a ballpark if I choose and have more normal lighting levels the rest of the time.

In summary, not all fluorescent tubes are created equal, and I'll even agree that low-priced, mass produced fluorescent tubes produce lousy light, although a bulb is worse in my opinion. Halogens are somewhat better, but I still dislike the yellow light they put out. If you decide to try some of my suggestions please let me know how they work out. I'm a year younger than you, and although I don't notice any difference in how much light I require I've always liked any room I'm in to be brightly lit except when I'm watching television. For example, my 77 ft² work area has 4 40W tubes(currently 9000 lumens). When these go, I'm getting the 3600 lumen, 40W tubes from the place I linked to above. 14,400 lumens ought to make my workroom really bright and cheery.

 

[LiamC]

Thanks for the advice, I'll hunt up a compact bulb ASAP and go from there.

 

[jtr1962]

Home Depot of all places had electronic ballasts for 40 watt T-12 tubes(I had almost given up looking). A little pricey at $21.97 each but still inexpensive enough to try. I just purchased and installed two today in the shoplights in my workroom. I'll keep the old magnetic ballasts which served me well for 20 years as spares. Too soon to tell if the lack of 60 Hz flickering makes any difference in how quickly I get fatigued. I'll know better after I pull one of my all nighters down there. Some immediately noticeable advantages are that there is less heat in the place where the ballast is bolted to and the tubes are powered independently so if one goes it won't cause the other one not to light. I also believe the tubes are burning a bit brighter but I can't confirm that without a light meter.

 

[Piyono]

This thread has reignited my interest in LED lighting (I had 'plans' at one time, of an elaborate lighting scheme which would illuminate my bedroom entirely wth LEDs of various colors).

I know very little about flourescent lighting, however. T8 v.s. T12, magnetic v.s. electronic -- all Greek to me. Alls I knows is that flourescents tend to make me go stir-crazy faster than filament lamps.

Anyway, I was Googling LEDs and flouros and I came upon this article, which you guys may or may not find interesting:

http://irc.nrc-cnrc.gc.ca/fulltext/nrcc44745/nrcc44745.pdf


Enjoy.


Piyono

 

[jtr1962]

Strangely enough I came upon that very same article when I was learning all about fluorescents a few months back. I've found the whole subject quite a bit, well, illuminating.

I really like your ideas for LED lighting. They might very well be feasible in 5 years as LEDs decrease in cost and increase in efficiency. In the mean time if you enjoy playing with LEDs there are a couple of sellers(all factory direct sellers from Hong Kong) on eBay who have high brightness LEDs, including white, relatively cheap(~$25 to $30 for 100 pieces). I've found the quality to be very good and they have many colors available-white, blue, green, orange, amber, red, aqua, ocean blue, pink, violet, and ultraviolet. I'm currently in the process of buying a few new colors every month and as you can imagine I'm quite excited at getting white LEDs for a quarter a piece shipped. Another thing that excites me about all this is that LEDs have finally gotten bright enough to be a source of general illumination in flashlights and the like rather than just lowly indicator lights. I put 12 white LEDs(~1.2W total power) in my bike headlight and it's just as good as the former 2.5W halogen bulb. In fact, it's better since the blue-white light works better with my night vision than the yellow halogen bulb. Another great use for white LEDs is "imitation" fluorescent lighting of scale models, especially HO passenger coaches. They also make nice simulated HID headlights on car models. And the color LEDs have loads of uses in models as well. As far as I'm concerned grain-of-wheat bulbs are history.

 

[Howell]

Some interesting info here.
http://photoweb.net/pw_tech/floures1.html

 

[Howell]

jtr, what do you think about this place and their product?
http://www.naturalux.com/index.html
Some of the site navigation is broken, but all the links on this page work.
Be sure to check out the "Helpfull Links" section.

Based on similar marketing and technical phrases, you can buy their product from this place.

 

[jtr1962]

It's interesting that something like this is made. The main drawback I see with it is that it cuts out some of the light(and therefore reduces efficiency), and that the filters + cheap cool white bulbs cost more than I'm paying for full-spectrum tubes. I was trying to find out exactly how much they reduced the light by so I could do a direct comparison efficiency-wise. In any case, here's a couple of quick numbers on the subject. A typical cool white T-12 tube gets somewhere around 80 lm/W efficiency. Let's say the filter cuts that to 60 lm/W by cutting out the excess green. Tube($1) plus filter($6.25 in quantities of 24) runs me $7.25. A full-spectrum high efficiency T-12 tube from the link I provided above will run me $9.00 in quantities of 6+. Efficiency is 90 lm/W. I might save a bit by going the filter route but I lose out big time on efficiency. T-8 tubes from the same place run me $7.00. No savings there, and efficiency is 92 lm/W. In fact, I just replaced the fixture in the kitchen with an electronic ballasted T-8 one, and the full spectrum tubes arrived today. Huge difference over the unfiltered T-12 cool whites. At least 60% more light, 25% less power consumption, appearance exactly like daylight. I'm sold on full-spectrum tubes from this place. The filters might be a good idea if they were cheaper(say $1 or $2 each) and one had a huge stock of cool whites.

Three other things bother me a bit about the filters. First off, not all cool whites are created equal. Most of the high performance ones(80 lm/W or more) do indeed have an excess of green and a rather poor CRI(say 62). However, others such as the cool white deluxe that we were using prior to today have good CRI(89), although at the expense of lower efficiency(55 lm/W). The filters might help here, but then again they might distort the already fairly good spectrum horribly, and they would cut the already not so good(for a fluoro tube) efficiency. As such, the filters are best suited for the plain old cool whites. Second, while I'm aware that fluorescent tubes emit UV as part of the down conversion to white light this is the first I've heard that they emit harmful amounts of UVB. Not saying that these people are wrong but it sounds more like a plug for their product. Light bulbs, especially halogens, also emit UV just by virtue of being a heated blackbody. Metal halide lamps emit copious amounts of UV. In all cases the fixture is designed to absorb the UV. I would imagine fluoro tube makers already have something in the glass to absorb any UV that may escape being turned into white light by the phosphors. Third, they don't do a thing about the 120 Hz flickering that is a byproduct of using magnetic ballasts. This problem is probably more responsible than anything for the headaches people experience under flurorescent lighting and the fact that CFLs(which incidentally don't have magnetic ballasts) are taking forever to replace bulbs due to the preconceived notions too much of the population has about fluorescent lighting. In the end only an electronic ballast can eliminate flickering, and once you decide to upgrade your lighting to electronically ballasted T-8s you need to buy all new bulbs anyway so you might as well buy genuine full spectrum ones which only cost maybe $5 more each than the cheap cool whites. So in the end I see this product as more a stop-gap measure than a long term solution. Perhaps the best use is for a company that wishes to do something about it's lighting, has a huge stock of T-12 cool whites, and just can't afford to replace all their fixtures and bulbs. However, since fixture replacement is usually done on a 20 to 25-year cycle I tend to think that if you're within 5 years of the next upgrade you may as well just push it up.

BTW, the verdict still isn't in yet over whether I'm feeling less eyestrain with the new electronic ballasts in my workroom's shop lights. It does seem that way to me so far, but it could be the placebo effect.

 

[jtr1962]

Some interesting info here.
http://photoweb.net/pw_tech/floures1.html

Indeed. What I thought was most interesting here was the graph that shows just how much red incandescents have in their spectrum. To me they always distorted colors but I didn't realize how badly until I saw that graph. I guess fluoro tubes with large gaps in their spectrum do the same.

Seems to me anyway that we'll get the best of both worlds once LEDs improve a bit. The biggest(really only) disadvantage of fluoro tubes for me is that starting them frequently shortens their life, which is why I use the 20 minute rule. If I'm leaving a room more than 20 minutes I turn the light out. If not, it stays on. No such worry with LEDs, and eventually they'll have all the benefits of both fluorescent(full spectrum, high color temperature, high efficiency) and incandescent(point source, no complex ballast).

 

[blakerwry]

I had a compact fourecent go abd on me the other day... started working intermittantly and then it started making a high pitch sound so I just tossed it... I haven't used the thing more than maybe 4 months I don't think... too bad I didn't save the reciept or box or anything or I would have returned it.

When I bought the new one I looked at the lifetime rating on the bulb... 5 years or so... then I saw in fine print that is when the bulbs are only used something like 3-4 hours per day.

I didn't realize that frequent on/off could kill flourecents, does this apply as much to CF bulbs?

 

[jtr1962]

I didn't realize that frequent on/off could kill flourecents, does this apply as much to CF bulbs?

CF bulbs are killed just as quickly by frequent starting as the linear tubes are which is why you don't see them used inside refreigerators, for example. It's inherent in the technology. Each time the tube starts some of the emitting material is blown off the cathodes. This same material is also depleted during operation but much more slowly. When it's gone the tube either won't start, or it'll flicker. Sometimes the ballast will make funny noises as well.

Most CF bulbs(and linear tubes) are rated on a three hours on, 20 minutes off cycle(the off time doesn't count towards the rated life). On this basis most CF bulbs have lifetimes of 6,000 to 10,000 hours, which works out to 5.5 to 9.1 years burning them three hours a night. Starting more frequently can shorten the lifetime dramatically. For example, if the burn time averages six minutes instead of 3 hours the lifetime is only 20% of the rated lifetime. Conversely, if the bulbs are run continuously they might very well exceed their rated lifetime by 50%. Most linear tubes have somewhat longer lifetimes but the same principle applies. Typically, 48" T-8 or T-12 tubes are rated for 20,000 hours. High quality tubes are rated for more. The full-spectrum T-8 tubes that I purchased are rated at 34,000 hours(this makes the comparison I did in my previous post even more skewed in favor of full spectrum tubes as opposed to cool white+filter). Running them about 10 hours/day and one start, which is typical for us, should give us about 20% extra life, or ~40,000 hours. This works out to about one replacement every 11 years.

 

[Prof.Wizard]

This thread has reignited my interest in LED lighting (I had 'plans' at one time, of an elaborate lighting scheme which would illuminate my bedroom entirely wth LEDs of various colors).

But are they luminous enough? Regarding efficiency they are fabulous, but I wanna be able to read Principles of Surgery without my head spinning after a few minutes.

I know very little about flourescent lighting, however. T8 v.s. T12, magnetic v.s. electronic -- all Greek to me. Alls I knows is that flourescents tend to make me go stir-crazy faster than filament lamps.

What's your problem, man?

Anyway, I was Googling LEDs and flouros and I came upon this article, which you guys may or may not find interesting:

http://irc.nrc-cnrc.gc.ca/fulltext/nrcc44745/nrcc44745.pdf


Enjoy.

Nice read of medical interest. Thank you.

 

[Howell]

Second, while I'm aware that fluorescent tubes emit UV as part of the down conversion to white light this is the first I've heard that they emit harmful amounts of UVB.

*shrug*

http://www.naturalux.com/UVinfo.htm[/url]]UV radiation has long been known to cause a myriad of health problems. Some sources indicate that fluorescent lights emit more UV than the sun. In a 1992 edition of the American Journal of Epidemiology, it was found that fluorescent lights emit "10-30 times" more UVB radiation than the sun does. It is the UVB radiation that has found to be carcinogenic (cancer causing). The study further goes on to report that "Case reports have documented skin sensitivity of patients to fluorescent light, some of which were specific to energy in the ultraviolet B domain, with no reaction to ultraviolet A exposures. In general populations, ultraviolet B is perhaps 1,000 times more effective in producing erythema than ultraviolet A, leading to ultraviolet B sometimes being referred to as the [sunburn] energy range. There are also animal and human data indicating that non-melanoma skin cancer is more clearly related to ultraviolet B exposure than to ultraviolet A".

http://www.naturalux.com/Absorption.htm[/url]]Fluorescent lighting requires the use of ultraviolet (UV) radiation to produce visible light. UV is present in three forms:

· UVA (320 to 400nm)

· UVB (280-320nm)

· UVC (100-280nm)

While all of UVC and a portion of UVB are absorbed by the glass portion of the fluorescent light bulb, some UVB and all UVA rays pass through the glass and into the environment. NaturaLux Filters absorb 100% of the UV rays from 280 to 380 nm. Between 380-390nm, there is a 99-81% level of UVA absorption. Between 390-400nm, 81-50% of UVA is absorbed.

There are UV filter manufacturers that use a yellow dye in their filters to absorb the miniscule amount of UV we do not absorb with our NaturaLux Filters. From a color accuracy standpoint, the problem with using a yellow dye is that it eliminates all of the violet and a good portion of blue from the light source. We prefer to keep your final lighting result as full spectrum in nature as possible.

 

[Howell]

A typical cool white T-12 tube gets somewhere around 80 lm/W efficiency. Let's say the filter cuts that to 60 lm/W by cutting out the excess green. Tube($1) plus filter($6.25 in quantities of 24) runs me $7.25. A full-spectrum high efficiency T-12 tube from the link I provided above will run me $9.00 in quantities of 6+. Efficiency is 90 lm/W. I might save a bit by going the filter route but I lose out big time on efficiency. T-8 tubes from the same place run me $7.00. No savings there, and efficiency is 92 lm/W.

I think the numbers you are looking for are listed here.
http://www.naturalux.com/Properties.htm

I was thinking trying a filter at work or some other place where I could not get the ballast changed.

 

[jtr1962]

I think the numbers you are looking for are listed here.
http://www.naturalux.com/Properties.htm

82% transmission efficiency-not as bad as I thought. This works out to a net efficiency of 65.6 lm/W if you're starting off with an 80 lm/W cool white tube. The efficiency of my GE Chroma 50s are 56 lm/W although I can get full spectrum T-12 tubes with efficiencies of 90 lm/W for only $9 from the same place I purchased the full spectrum T-8s.

I was thinking trying a filter at work or some other place where I could not get the ballast changed.

By all means go for it. I'm curious of the results myself. One thing I thought was interesting that the filters do is to polarize the light. I'm wondering how much of an effect on fatigue this has. My only concern would be if the light level was already marginal and the filter cut it further but most work places are fairly brightly lit. If you find the filter is a big improvement you might even be able to convince your company to convert wholesale to T-8 full spectrum bulbs and electronically ballasted fixtures. :mrgrn: Certainly colors should look better once you're done.

 

[jtr1962]

This thread has reignited my interest in LED lighting (I had 'plans' at one time, of an elaborate lighting scheme which would illuminate my bedroom entirely wth LEDs of various colors).

But are they luminous enough? Regarding efficiency they are fabulous, but I wanna be able to read Principles of Surgery without my head spinning after a few minutes.

They are if you have enough of them. I lit up the inside of my 1 cubic foot thermoelectric freezer with ten white LEDs consuming 0.9W. The light level inside approximates a fairly well lit room. You could certainly read a book placed inside it if you wanted to. Some numbers for efficiency of various colors are on this page. Right now we're at the point where run of the mill white LEDs are more efficient than nearly all light bulbs, and about two to three times as efficient as the typical small bulbs that they replace. Colored LEDs are several times as efficient as bulbs. If you account for the fact that by filtering a bulb to get colored light you lose perhaps 70 to 80% of the light they are more than ten times as efficient at producing colored light. This is why a 116 watt traffic light bulb can be replaced by an LED array consuming maybe 8 to 11 watts. Experimental red and red-orange LEDs with a TIP(truncated inverted pyramid) die have already exceeded 100 lm/W, which is better than old yellow(the sodium vapor street light).

Despite their efficiency, each LED can only be absorb a small amount of power(say 50 to 100 milliwatts). Therefore, you need ten to twenty to produce a 1 watt LED bulb, and thousands to light a room. Think along the lines that each LED produces maybe 1 to 4 lumens. In time the light per package and overall efficiency will improve. The 2012 goals are 150 lm/W and 1000 lumens/package. Efficiency should slowly go up after that, and may eventually exceed 250 lm/W(this is pretty close to 100% energy to light conversion efficiency). It should be possible to light a room comfortably with only a few watts and brightly with maybe 30 or 40 watts.

 

[Groltz]

Since we are all here, let me pose a question:

Does anyone have an interest in these?

Back when I was still working for the country's worst managed airline I used these often due to their ability to conserve batteries. I used the 3-cell model in a Maglite and a 4-cell model in a Princeton Tec headlamp. (Graveyard shift outdoor work).

Since I don't see getting recalled from layoff in the foreseeable future, I would like to sell the 3 Ledcorp LED bulbs that I have left. They are the generation before the Epieon series...The only difference being that Ledcorp managed to get the Epieons to have a slightly higher output and whiter color temperature. These look identical to the ones on the webpage. These 3 were my spares, and have no use-time on them except to verify that they were good upon receiving them.

I have two 4.5v (3-cell) bulbs and one 6.0v (4-cell) bulb.

If anyone's interested, PM me. 8)

 

[blakerwry]

Those look interesting, but probably not worth the price to me. Just out of cureosity, how did you like the Princeton Tec compared to the mag-lite?

I used princeton Tec flashlights for hicking/fishing/backpacking and camping years before I saw them sold in retail outlets. They simply blow everything else out of the water in my opinion, including Maglite. My experience is with a 4 cell PrincetonTec 40 flashlight and one of the 4 cell headlamps (dont remember the name).

 

[jtr1962]

Looks like there's a new efficiency champ in town. 8) Good enough reason to resurrect this thread in my book. :mrgrn:

65 lm/W using industry standard packages and 74 lm/W using special laboratory packages! Better than compact flurorescent, and starting to approach T-8 linear tube efficiency. Of course, still not cost-effective for house lighting, but even with the current 20 to 30 lm/W efficiency of production LEDs any lighting application that uses a 5W bulb or less is better served with LEDs.

I retrofitted another bike light recently with 20 white LEDs of higher brightness than my last 12-LED retrofit, and coupled it with a more efficient driving circuit. Double the light output of the last retrofit at only 15% more power, and easily twice that of the stock halogen bulb(which consumes 50% more power than the LEDs). And these are just run of the mill LEDs of ~20 lm/W. I'm imagining my headlight 3.5X brighter with some of Cree's new LEDs for the same power consumption(or alternately the same brightness using only 5 or 6 LEDs instead of 20, and consuming less than one-third the power).

 

[Fushigi]

I retrofitted another bike light recently

Is there a guide online for doing this? I don't have a night light on my bike right now (front or rear) and would be interested in buying or building one.

(right now, I just don't ride in the dark)

 

[jtr1962]

Candlepower forums is a good place to look for different types of lighting projects. They mostly modify flashlights but bike lights aren't much different. If you lived nearby I'd be happy to do this for you. Besides the headlight, I also put 5 red LEDs and a blinker circuit on a standard rear reflector. It gets power from my headlight batteries, and is far brighter than those dim, commercially available ones. In fact, at one or two blocks it is very conspicuous. It could probably be seen 10 blocks away but I've never tried it. Whatever headlight you get, just make sure to use rechargeables. I use 4AA NiMH cells. Disposable batteries can end up costing a small fortune if you ride at night a lot as I do(mainly to avoid traffic and air pollution). This seller on eBay is a good place to buy LEDs. I've ordered 6 times from them. Delivery to NYC from Hong Kong takes about 10 days, and you get some nice HK stamps on the package as a bonus. I used their 7000 mcd, 5 mm white LEDs in my headlight. $25 shipped for 100 beats the usual online prices of $1.50+ by a huge margin, and they seem just as good. I've also order red, blue, amber, and UV LEDs from them with equally satisfying results.

If you need any help with any lighting mods please let me know.

 

[i]

If you have a load of cash burning a hole in your pocket, there's always Color Kinetics.

I'd be more enthusiastic about them if they weren't so patent crazy.

 

[jtr1962]

Those seem cool but also $$$, and not as fun as making it yourself. The driving circuit for my headlight was built from spare parts not worth more than $1. I modified this circuit to drive 20 LEDs(in two series strings of 10) off of 5V(4 AA NiMH cells) instead of 1.5V. I used a 100 uH inductor from my spare parts, replaced Q2 with a ZTX449, and R2 with a 1.2K resistor. I also added a Schottkey diode and 100 uF capacitor to filter the output so it's smooth DC instead of pulses. Measured efficiency is about 88%. I might be able to get this into the mid 90s with a few tweaks. Original circuit is shown below:

 

[Howell]

Thanks for the updates, Joe.

 

[jtr1962]

Another interesting development on the LED front. Osram reaches 96 lm/W in 617 nm. This technology will be employed in the next generation of Osram LEDs. While it is not a white LED (617 nm is red-amber), it is an interesting develpment nonetheless as 96 lm/W in red-orange color represents a wall-plug efficiency of about 34%, meaning 34% of the power used by the LED comes out as light. This is about 6 times the efficiency of an incandescent light bulb, and the first time an LED of such efficiency will make it out of the lab. Cree also plans to put into production in the relatively near future its series of XThin™ blue LED chips which have a light output of 18 mW. This translates into a wall-plug efficiency of about 28%.

 

[sechs]

Only 34% efficient? Aren't fluorescents something like 20%?

How efficient is your average LED?

 

[jtr1962]

Only 34% efficient? Aren't fluorescents something like 20%?

How efficient is your average LED?

The best T-8 and T-5 fluorescents have an efficiency of around 100 lm/W, which translates into about 35% efficiency dependng upon the particular spectrum the tube is putting out.

Most run of the mill high-brightness LEDs these days have efficiencies in the range of 15 to 40 lm/W, depending upon color. When translated into percentage this must first be corrected for the eye's response to determine the actual light output in mW. For example, yellow-green light is over 600 lm/W because the eye is most sensitive to this color. Deep red and blue light causes a lesser eye response so a blue LED may be dimmer than a green one in terms of lumens and actual brightness but may still be converting a greater percentage of its power to light. Anyway, typical LEDs are about 10 to 15% efficient, which is about twice as efficient as an incandescent bulb. However, since they output 100% of the light at the desired color they can be as much as 10 times more efficient than a light bulb with a color filter which absorbs most of the light.

34% may not sound impressive but it is, and some experts say we'll eventually surpass 80% or even 90% efficiency. I don't think we even have a conventional light source now which exceeds 40% efficiency, and that includes the whole family of discharge lamps (sodium, mercury, metal halide).

 

[sechs]

My impression was the LEDs were much closer to 50% efficiency than they are. Except for specific situations concerning colour and space, it seems like fluorescents and others are better choices.

 

[jtr1962]

Some red-orange LEDs in the lab have approached 50% efficiency. Within a few years you'll have production LEDs close to that or perhaps even better. It's a rapidly evolving technology whereas fluorescent and especially incandescent likely won't get much more efficient than they already are.

LEDs are currently best suited for small lighting needs. In fact, even in their current state of development they obsolete incandescents for white lighting needs under about 1 watt, and for most colored lighting needs such as traffic lights. Their much longer lifetime and in most cases superior efficiency more than offsets the additional cost, and in many cases their robustness is an added bonus. LEDs just don't break when dropped.

For general lighting needs, LEDs just aren't there yet. While they can replace incandescent and offer twice the efficiency, current white LEDs are too inconsistent from batch to batch in their quality of light, and they are still far too expensive. When they are uniform from batch to batch and the cost drops to about $0.01 per lumen or less they will begin replacing incandescents for general lighting needs. By then (~2007) efficiency should be over 75 lm/W, which is better than compact fluorescents, and lifetime will be 100,000 hours or more, meaning an LED replacement burning for 6 hours per day will last 45 years. They also won't suffer the drawback of shortened lifetime when turned on and off frequently the way fluorescents do. It will still be a while (probably a decade or more) before LEDs become efficient and cheap enough for it to make economical sense to start using them to replace fluorescents for general lighting needs. The targets frequently cited for that are 150 lm/W efficiency and a cost of $0.001 per lumen (this is about what cheap fluorescent tubes now cost).

 

[sechs]

I honestly don't see LEDs coming into common use for some time. It's tough enough to convince people that compact fluorescents are good replacements for incandescents

 

[Fushigi]

I honestly don't see LEDs coming into common use for some time. It's tough enough to convince people that compact fluorescents are good replacements for incandescents

That's because CFs cost more. If the price was the same, or even just a slight premium, people would buy. When the CF bulb costs several $ more each than the incandescents, what's the payback in terms of cost savings? Years?

Personally, I try to use CFs where appropriate, but my house also hase several fixtures where 'ornamental' lights are used; there's no CF equivalents for those. At least none that look nice.