Nichia Develops 60 Lumen Per Watt White LED

[Stereodude]

What, the mercury is going to leach out through the glass? :skepo:

Fluorescent tubes have had mercury in them since the beginning. I haven't noticed a bunch of mad hatters running around.

 

[LunarMist]

What, the mercury is going to leach out through the glass? :skepo:

Fluorescent tubes have had mercury in them since the beginning. I haven't noticed a bunch of mad hatters running around.

Sure, and it is behind a cover. Unless something on the range explodes and flies 2-3 feet in the air, it's not a concern. Besides, for decades people had thermometers stuck up the ass with far more mercury than that. :frowner: At work we ordered a few hundred grams of mercury at a time from the early 80's until the early 90's.

 

[CougTek]

At work we ordered a few hundred grams of mercury at a time from the early 80's until the early 90's.

Did you like the taste? I heard it's bitter.

 

[time]

I've usually found no-name CFLs to be poor in every aspect: reliability, longevity, color rendition and brightness. This test supports that prejudice.

I've seen US tests that put GE and Osram head and shoulders above the others in terms of reliability at least. The linked test confirms my own observations that you're also likely to be dudded in brightness, with the better brands being up to 50% brighter!

A decent CFL should be equivalent to an incandescent with 5 times the power, yet popular recommendations have claimed as little as 2 to 3 times. Now you know why.

 

[Fushigi]

I bought Lights of America brand CFL bulbs a few years ago from either Menards or Sam's Club. They were absolute crap & didn't last as long as a regular bulb let alone anything close to what a CFL should do. I won't ever buy them again, which is too bad as they are offering bulbs I would consider buying, like LED floodlights.

 

[Howell]

What, the mercury is going to leach out through the glass? :skepo:

Fluorescent tubes have had mercury in them since the beginning. I haven't noticed a bunch of mad hatters running around.

I would not expect it to no. But the operating environment combined with the hazardous substance leads me toward a different solution. IMO, the risk/reward is less than desired.

 

[Stereodude]

I would not expect it to no. But the operating environment combined with the hazardous substance leads me toward a different solution. IMO, the risk/reward is less than desired.

FWIW, I agree that a fridge is not a good location for a CFL. My opinion is due to temperature and usage patterns, not because of the mercury though.

 

[LunarMist]

I checked my ~3 YO fridge and there is very little space for a CFL. The bulb is a 40W appliance type.

 

[sechs]

FWIW, I agree that a fridge is not a good location for a CFL. My opinion is due to temperature and usage patterns, not because of the mercury though.

And at what temperature and usage patterns would it be good location?

 

[sechs]

I checked my ~3 YO fridge and there is very little space for a CFL. The bulb is a 40W appliance type.

I'm sure that there's plenty of space if you pick the right lamp. As the ballasts can be bulky, that can really be a hinderance.

It's bizarre that refrigerators still use those little appliance bulbs.

 

[LunarMist]

I'm sure that there's plenty of space if you pick the right lamp. As the ballasts can be bulky, that can really be a hinderance.

It's bizarre that refrigerators still use those little appliance bulbs.

Why? They work just fine for the 10-30 seconds that the fridge is usually open. A string of built-in LEDs might be a good idea as they would light a larger area more evenly.

 

[Howell]

FWIW, I agree that a fridge is not a good location for a CFL. My opinion is due to temperature and usage patterns, not because of the mercury though.

And here, I was mostly talking about the range hood. : D
I agree with you about the fridge.

 

[jtr1962]

A refrigerator is perfect for LEDs. LEDs love cold, and aren't affected by the on-off cycling like CFLs.

 

[jtr1962]

I've usually found no-name CFLs to be poor in every aspect: reliability, longevity, color rendition and brightness. This test supports that prejudice.

I've seen US tests that put GE and Osram head and shoulders above the others in terms of reliability at least. The linked test confirms my own observations that you're also likely to be dudded in brightness, with the better brands being up to 50% brighter!

A decent CFL should be equivalent to an incandescent with 5 times the power, yet popular recommendations have claimed as little as 2 to 3 times. Now you know why.

No surprise here. The old adage you get what you pay for is especially true with CFLs.

CFLs fail mostly because you really can't design decent ballast electronics for the price these things are selling for. Back when CFLs cost $20 and up they seemed to be more reliable. But people complained about the price, so they made them less expensive. We all know how well that went.

A big factor regarding CFLs is the fact that you're trying to shoehorn a technology into a form factor not really designed for it. Incandescent fixtures are designed to hold in heat due to the nature of their light source. No surprise then that designing CFLs which last is an expensive proposition. It's a major engineering challenge. You're taking something which wants to be long and with a separate ballast, curling the tube into a spiral, and then putting the ballast in close proximity. And we're surprised they don't last? Frankly, I'm surprised they last as long as they do. There's a reason most CFLs are rated between 6,000 and 10,000 hours, while linear tubes are anywhere from 20,000 to 35,000 hours. Even in the best case, and if everything is made well, lifetime is going to be compromised. Truth is if you want to use fluorescents optimally, you need a purpose-designed fixture which takes either bi-pin CFLs or linear tubes. You can get CFLs which work well if you're willing to spend $15. However, I figure if a fixture takes a few lamps, then the quality CFLs are setting you back around what a replacement fixture will cost, and you'll be spending that much again every time you need to replace lamps. Might as well just drop the $$$ on a decent fluorescent fixture from the get go. In the long run you'll be happier. Any yes, I practice what I preach. Right now my house has 17 linear tube fixtures (mostly T8s, all electronic ballasts except the one in the garage), one circline fixture, and CFLs in the few remaining Edison-based lamps (18 sockets if I counted correctly-that includes the outdoor lamps). Also one incandescent fixture which I converted to LED (picture below). Unfortunately, no viable replacements yet for the small-base candelabra chandelier lamps but LED should take care of those fairly soon.

 

[sechs]

A refrigerator is perfect for LEDs. LEDs love cold, and aren't affected by the on-off cycling like CFLs.

Exactly. And most LEDs should outlast the appliance.

 

[time]

My daughter's sewing machine uses twin LED lamps. As you say, they'll last the life of the appliance.

 

[udaman]

http://bsl.lacity.org/LED1.htm

^1k down, 140k to go. 1 of the streets on the Sept 9 map is right near where I live...will have to go take a look.

http://www.ledsmagazine.com/news/6/9/24

Every year it is estimated that more than 425 million 60-Watt incandescent light bulbs are sold in the United States alone, representing approximately 50% of the domestic incandescent light bulb market. According to DOE, an LED replacement for this could save 34 terawatt-hours of electricity in one year, enough to power the lights of 17.4 million U.S. households. It would also avoid 5.6 million metric tons of carbon emissions annually.

lol, if you have bad eyes, you *need* 100w replacement or higher...60w just won't do it.

 

[time]

lol, if you have bad eyes, you *need* 100w replacement or higher...60w just won't do it.

You're mistaken Uda. It's all about contrast and color rendition. Ideally, you don't want your light source to be visible at all - like daylight indoors, in other words. The lack of glare that ensures means effective contrast is very high.

If the light source is visible, the iris in your eye simply closes up the pupil. It's exactly the same as trying to shoot into the light with a camera, which I'm sure you can relate to. Having said that, the human eye can handle low light levels a lot better than most cameras.

 

[P5-133XL]

You're mistaken Uda. It's all about contrast and color rendition. Ideally, you don't want your light source to be visible at all - like daylight indoors, in other words. The lack of glare that ensures means effective contrast is very high.

If the light source is visible, the iris in your eye simply closes up the pupil. It's exactly the same as trying to shoot into the light with a camera, which I'm sure you can relate to. Having said that, the human eye can handle low light levels a lot better than most cameras.

As I age, my eyes seem to be going down hill, rapidly: Me not pleased about this. Glasses now make a big difference when reading small print and reading computer screens, but are not sufficient in all things. Specifically, I have discovered that the amount of light matters a lot. I can read much smaller text in bright sunlight than I can with any artificial light. Further even with artificial light and glasses, the brighter the better.

This is particularly annoying when reading things like serial numbers. Unless the item can be taken outside, most S/N's are impossible for me to read now even with with glasses and very bright artificial light. I am now carrying a magnifying glass, in my standard tool kit, just to read S/N's.

 

[time]

I have the same problem, Mark, and the same concerns.

What I also noticed the other day was how much clearer printed text was when I was near a window, than under artificial light. To my surprise, the light meter told me that light levels on the page were actually quite modest.

Could it be that the pupil is more closed in daylight, and that that somehow enhances edge detection? Any photography/medical fans know?

 

[jtr1962]

Could it be that the pupil is more closed in daylight, and that that somehow enhances edge detection? Any photography/medical fans know?

More blue in the light spectrum does indeed close the pupil, and as with a camera, that enhances depth of field. Here's a good explanation of that along with the other benefits of using light containing more blue than the usual household incandescent sources. Another explanation.

I've noticed myself that things appear much clearer under daylight (and as of yet I don't really have any vision problems beyond being near-sighted), or even my CRI 90+, 5000K fluorescent tubes, than they do under more usual sources of residential illumination (i.e. incandescents and warm-white CFLs). I've been saying for years how awful the typical light sources that people use in homes are, and the research is finally backing me up. And in addition to being the wrong spectrum, most homes are grossly underlit. A 60W incandescent bulb in an average-sized room doesn't cut it unless you're just trying to see where you're walking. To do anything productive you need a couple of times that.

As for the LED streetlights, it's long past time we finally got something better than those disgusting sodium lights. Besides being fugly, the spectrum from those actually pretty much kills peripheral vision, making them decidedly less safe to drive under than a whiter light source.

 

[time]

I don't have access to 90+ CRI fluorescent lamps, but my personal experiences with high color temperature CFLs such as Philips "Cool Daylight" have resulted in excessive glare. I assumed this is why many people prefer low color temperature "Warm White" lamps. My own preference is 4000K around the home, but sometimes I've had to settle for 2700K.

Could it be that higher color temperature lamps become more tolerable when color rendition is improved? I've noticed that I'm very sensitive to color rendition quality. If fluorescents rendered color as well as is claimed, surely it would be difficult to discern their light from daylight?

 

[jtr1962]

I don't have access to 90+ CRI fluorescent lamps, but my personal experiences with high color temperature CFLs such as Philips "Cool Daylight" have resulted in excessive glare. I assumed this is why many people prefer low color temperature "Warm White" lamps.

The preference for low CCT lamps is hardly universal. It seems to be more a cultural thing. In Japan 5000K in residences is common, for example. And some surveys point to 3500K-4500K being most preferred. The thing is many people just go with what they are used to. Incandescent lamps were similar to the gas lights they replaced. So basically people have been used to this type of light for over a century. And we really didn't have any alternative until recently (except linear tubes but few people seem to want them in residences). Now that we have CFLs (and soon LEDs) in a variety of color temperatures, some are finding that warm white isn't really so great.

Glare is more a function of having the light source visible than CCT. Best lighting from a glare standpoint are linear tubes on ceilings. The effect mimics skylight, in that it's diffuse and just wraps around everything.

Could it be that higher color temperature lamps become more tolerable when color rendition is improved? I've noticed that I'm very sensitive to color rendition quality. If fluorescents rendered color as well as is claimed, surely it would be difficult to discern their light from daylight?

Any light source is more tolerable if color rendition is improved, especially high CCT lamps. And even though some fluorescents might have CRIs in the 90s, the spectrum still looks different than daylight. Problem is CRI isn't the greatest way to compare color quality. It measures how eight color patches appear relative to how they appear under a blackbody at the same CCT. Fluorescent lamp makers can get a good match on most of those patches, getting a high CRI score, and yet still render colors like skin tones poorly. In tests with LEDs CRI hasn't necessarily correlated with how pleasing people find a light source. Some combinations of red, green, and blue LEDs had very low CRIs, in the 20s, and yet they were preferred over other sources. And some sources have a CRI of 100 just by virtue of being a blackbody, and yet they render colors very poorly (i.e. candlelight). Right now alternative measures, such as the color quality scale, are being developed to overcome this deficiency.

 

[ddrueding]

Any light source is more tolerable if color rendition is improved, especially high CCT lamps. And even though some fluorescents might have CRIs in the 90s, the spectrum still looks different than daylight. Problem is CRI isn't the greatest way to compare color quality. It measures how eight color patches appear relative to how they appear under a blackbody at the same CCT. Fluorescent lamp makers can get a good match on most of those patches, getting a high CRI score, and yet still render colors like skin tones poorly. In tests with LEDs CRI hasn't necessarily correlated with how pleasing people find a light source. Some combinations of red, green, and blue LEDs had very low CRIs, in the 20s, and yet they were preferred over other sources. And some sources have a CRI of 100 just by virtue of being a blackbody, and yet they render colors very poorly (i.e. candlelight). Right now alternative measures, such as the color quality scale, are being developed to overcome this deficiency.

I just read a chunk of that article. It and your post confirm something that I have felt (high CRI doesn't explicitly make a bulb good). Can you point me in the direction of some quality linear tubes? I don't know the type, they are in my office fixtures (built 5 years ago, 4' long). I'm looking for 6500K with the best color quality available.

 

[jtr1962]

I just read a chunk of that article. It and your post confirm something that I have felt (high CRI doesn't explicitly make a bulb good). Can you point me in the direction of some quality linear tubes? I don't know the type, they are in my office fixtures (built 5 years ago, 4' long). I'm looking for 6500K with the best color quality available.

These look promising to me. These look even better if you want to spend the money.

The clue that a fluorescent lamp has good color rendering (beyond the CRI ratings) is when it has much lower lumen output than similar lamps. That's the case here-these lamps are all ~2000 lumens compared to about 2800-3050 lumens for the usual 48" T8 tube. It's not that these tubes convert less energy to light. Rather, it's that more of this energy goes into wavelengths your eye is less sensitive to improve color rendition.

 

[jtr1962]

I just realized the first tube I linked to is 6000K and the second 5000K but you're looking for 6500K. I came across these, but $32 per tube? Ouch!

Just in case you're not familiar with the nomenclature on fluorescent tubes it's a three digit number like this:

965

The last two digits are the CCT in hundreds (65 = 6500K). The first digit is the CRI range:

6=60 to 69
7=70 to 79
8=80 to 89
9=90 to 99

There is no 10 because 100 CRI isn't possible. 98 is about the best we can do, and for all practical purposes it's indistinguishable from 100.

 

[jtr1962]

In case anyone might want the Philips TL 950s I linked to 2 posts above, I found them cheaper here. $6.41 each for a case of 25 versus $7.50 in the original link. Not a bad price for such a high quality tube IMO. If I needed tubes I'd consider buying some myself. If anyone local to me is interested in buying a case, I might be willing to buy a few tubes out of the case just to satisfy my curiosity.

 

[jtr1962]

Philips also makes CRI 98, 6500K tubes according to this. Their TL-D 90 family actually has a more continuous spectrum than the TL90 series, and about 15% higher output, despite both having the same CRI. Only problem is the TL-D 90 are only sold in Europe as far as I know.

 

[jtr1962]

And some surveys point to 3500K-4500K being most preferred.

I mentioned this earlier regarding CCT preferences, and here is a survey from CPF which confirms that (I added the CCT corresponding to each bin):

Note the strong preference for 3700K to 4300K, although large numbers of people still like CCTs in the 5000s. There is also an uptick in the 6350K - 7000K area by those preferring something closer to daylight. One thing funny is how unpopular the really warm bins are, and yet typical household lighting is even warmer than those (and lighting designers always seem to think this type of light is so wonderful). Go figure. It seems once people learn a little about the subject, they mostly gravitate to 4000K or higher.

 

[Stereodude]

Yeah, but there's a huge contingent looking for the look of their incandescent flashlights in there likely throwing off the results. The 5A bin is nice outside, but horrible inside IMHO.

 

[ddrueding]

Have a case of those 960s you linked in the shopping cart...probably tomorrow.

 

[jtr1962]

Yeah, but there's a huge contingent looking for the look of their incandescent flashlights in there likely throwing off the results. The 5A bin is nice outside, but horrible inside IMHO.

No arguments from me there. I'm not big on the warmer tints myself inside.

 

[jtr1962]

Have a case of those 960s you linked in the shopping cart...probably tomorrow.

Please let me know your impressions. I may eventually order a case myself.

 

[ddrueding]

Ordered, along with some 2x4' surface mount fixtures for my workbench.

 

[MaxBurn]

For some reason I have one of those weird cars (newer even) that don't shut off the map lights if you leave them on although the dome does, whatever. Killed my battery twice so I swapped the interior lights out for LED's. Several times I have messed up sense then and found the map lights on when returning to the car, battery fine. Simple solution that even puts out more light, I'm happy.

 

[LunarMist]

Can't they fix it under warranty?

 

[MaxBurn]

Can't they fix it under warranty?

Fix implies that something is broken where here nothing is other than my apparent inability to remember turning off lights. Obviously my wanting the auto off to apply to the map lights would be nice but I see that as a feature, not a fault. Anyway I thought it was worthy to mention that the power savings of converting to LED's in this case was a huge benefit to me.

 

[Stereodude]

A lot of cars turn off the interior lights if you leave them on after say 30 minutes. I thought that was a standard feature since the mid 90's on most cars.

 

[ddrueding]

In the Audi and my VW, if you turn on the lights after the car is off and the keys removed, they stay on. If you turned them on while the car was running and then remove the keys, they will turn off after a delay. Both will also let you turn on the radio and other stuff without the key in and they will never turn off automatically.

 

[MaxBurn]

My last couple Fords 95' and 04' did turn off lights automatically. My 00' Land Rover didn't and I drained its battery once too and now I find my 08' Subaru doesn't either. Definitely a weird hit or miss thing, but nice to know you can just pop in some fairly inexpensive LED's and solve the problem.