Philips Award Winning LED Bulb Review

[Arilou]

I read what you wrote on the feedback on the economist page and I am sorry, but what you wrote is predominantly crap.

1) With the exception of cheap Christmas lights, no LED lights flicker. Actually for that matter, no modern CFL or linear fluorescent flickers either.

I guess you haven't tried the latest from Cree, they do noticably flicker.

 

[idleprocess]

I guess you haven't tried the latest from Cree, they do noticably flicker.

The example I'm looking at in my hallway certainly doesn't.

 

[Arilou]

The example I'm looking at in my hallway certainly doesn't.

I don't know what you have in your hallway, but I'm talking about these:

http://www.candlepowerforums.com/vb/showthread.php?356710-Cree-A19-9-5w-60w-800lm-2700K-for-13-97

 

[idleprocess]

I don't know what you have in your hallway, but I'm talking about these:

http://www.candlepowerforums.com/vb/showthread.php?356710-Cree-A19-9-5w-60w-800lm-2700K-for-13-97

I have the 5000K 60W-equivalent and it doesn't flicker.

 

[Arilou]

I have the 5000K 60W-equivalent and it doesn't flicker.

The 2700K 60W-equivalent sure does. It looks okay when things are still, but any motion looks like you're watching a movie. Cree way underspeced the power supply components. The 6W (40W-equivalent) isn't as bad, but it flickers too.

 

[idleprocess]

The 2700K 60W-equivalent sure does. It looks okay when things are still, but any motion looks like you're watching a movie. Cree way underspeced the power supply components. The 6W (40W-equivalent) isn't as bad, but it flickers too.

I seriously doubt the innards are any different between the two other than diffferent LEDs.

 

[Arilou]

I seriously doubt the innards are any different between the two other than diffferent LEDs.

That would explain why the 9.5W bulb flickers more. The capacitors are insufficient for the higher current.

Or maybe they just had a bad batch, but it's a new product that has not been in production long.

 

[SemiMan]

I personally do not have experience with the Cree bulbs, but have many others, Philips, LG, Feit, some crappy Luminous, TCP and some offshore ones. I can't say that I have noticed an issue with flicker with any of them.

That said, I did do a quick web search and came across Mike Holt's web page, http://forums.mikeholt.com/showthread.php?t=152320. He did find detectable flicker with the new Cree bulb as well as the low cost Philips non-dimmable. He is not able to quantify it, so I am not sure how serious it is, but if you are finding issues as well, perhaps they have cut the design a bit tight.

I believe I am a bit sensitive to flicker as I could notice the rainbow issue with the early DLP projectors, but have not had issue with any CFL or linear fluorescent with an electronic ballast.

I have not seen the Cree bulb at Home Depot in Canada yet, but may try to pick one up in a few weeks when I go down to Philly for Light Fair. It is pretty easy to measure flicker.

Semiman

 

[wws944]

I have noticed flickering with a number of my LED bulbs. However the reason is that they are on dimmer circuits. Playing with the dimmer a bit usually allows them to settle into a non-flickering state. Note that my dimmers are non-C/L Lutron Divas. So maybe moving to LED-specific dimmers, or just a good old on/off switch, would help.

 

[idleprocess]

That would explain why the 9.5W bulb flickers more. The capacitors are insufficient for the higher current.

Or maybe they just had a bad batch, but it's a new product that has not been in production long.

The only difference between 2700K and 5000K parts of the same equivalent wattage is going to be the LEDs used - of which the difference will almost certainly be just the phosphor composition. Electrically, they should be identical since there is no benefit to making the parts any different - especially when they are trying to keep costs down.

Mine is in a very plain fixture with a simple switch - no dimmers and likely a number of other LED / CFL bulbs on the same breaker. Closest I get to perceptible flicker on any LED bulb I've used is whenever I leave the laser printer idle on the same circuit as my desk lamp - causes that LED bulb to blink off for perhaps 50ms every other minute or so. I have tries to detect flicker and can't - and I notice PWM on vehicle LED tail lights all the time. If there is flicker, it's below my ability to detect.

 

[SemiMan]

The 2700K is two strings of LED (20 leds), the 5000K is one string. Electrically I believe they are the same other than that.

 

[idleprocess]

The 2700K is two strings of LED (20 leds), the 5000K is one string. Electrically I believe they are the same other than that.

Oh, interesting. Given the prominent goal of low cost, I really expected them to be identical save for the one SKU selection at time of final assembly - which typically makes for the cheapest way to achieve two part numbers.

I'm trying to envision how they're wired up. Do they use array LED packages, with the 5000K parts containing twice as many LED die as 2700K parts? Or are the 2700K strings run in parallel? I can envision something needlessly complex along the lines of a fast PWM switcher alternating the current between strings, but that's added cost/complexity as an add-on component for the 2700K part or additional cost as an unused option for 5000K parts.

Idle curiosity really since I doubt I'll ever own a 2700K model.

 

[Yoda4561]

Wave a broom handle or other long thin object around under the light, if it has a stroboscopic effect it's "flickering" I can't detect it until I'm moving stuff around or working in a room with lighting like that, and then it's just one of those things that keeps nagging at me unless I figure out what it is so i can ignore it. I tried an ecosmart 60w warm white and decided to return it because of that(and it ran hotter than I was willing to tolerate).

 

[idleprocess]

Last post I intend to make in this thread on the subject of the Cree bulbs ...

Apparently, the 2700K 60W-equivalent model uses parallel strings:

"The LEDs are arranged in a parallel configuration to get as close as possible to the line voltage of the power supply," says Watson. "We can reduce the component count on the driver, making it simpler and lighter, and we need less metal, helping us to replicate the look of an incandescent bulb."

Only found it because I'm researching Cree for a paper.

 

[PhotonWrangler]

Last post I intend to make in this thread on the subject of the Cree bulbs ...

Apparently, the 2700K 60W-equivalent model uses parallel strings:

I think they meant to say series strings, not parallel. A series string would act as a voltage divider.

 

[SemiMan]

2 series strings .... In parallel.

 

[Anders Hoveland]

As for the design of the bulb, some will love it and some will hate it. The yellow plastic lenses on this new bulb are an ugly yellow in contrast to the nice golden yellow lenses on the first generation bulb. Personally, I think that is is cool looking bulb, but the first generation is much more visually appealing to me (the ugly yellow lenses are really quite hideous).

Really? Because I rather like the lemon-yellow colored phosphor covers more than the ugly orange covers. Guess it is just personal preference. I have read some reviews online where people said the three-segmented design looks ugly, one even comparing it to a Hannibal Lecter face mask!

It really does not matter what the bulb itself looks like, but what the light it gives off looks like, and how the colors in the room look.

I've compared this bulb with a BT15 halogen soft-white bulb. Actually both lights look quite similar so this L-Prize bulb mimics the BT15 bulb pretty well. I've been used to standard incandescent so my reference were probably wrong.

Something a little strange I noticed is that my BT15 halogen bulbs seem to have a slightly pinker tint than regular 750 hour incandescents. And the pink tint still does not quite compare to some 1125 hour incandescent bulbs I have, which have a more orangish light. I am guessing all those thick layers of glass in the BT15 bulb might be acting as a slight filter. Another possible explanation may be the halogen filling inside, usually the volatile element iodine which has a purple color and deposits on the inside of the quartz capsule. I have read somewhere else, in fact, that the use of bromine over iodine in halogen lamps can slightly increase the lumen output because the iodine absorbs a small ammount of the light, so this explanation is very plausible.

In fact, I actually bought a xenon-filled BT15 halogen (71 watts). It is rated for 2000 hours, but I never ended up using it in my room because the light was just slightly annoyingly pinkish in comparison to my 71 watt incandescent bulb (750 hours). It was not a big difference, but it was apparent, and I just preferred my old bulb. I know I am probably fussing over nothing here, but lighting is important and I want to optimize it just right.

A paper about the "Brilliant mix" concept by Osram explain it very well. Its a very similar concept than the Cree "True White" and the used by Phillips in the L-Prize lamp.

EQ white is used in the so-called "Brilliant Mix" LED technology, in which the Nd:YAG phoshor is adjusted to emit a bit more toward the greenish side of the spectrum to counterbalance the overall color since there are separate red chip emitters.

No, it is the orange Rebel. 660nm is deep red with very low photometric efficacy (little lumens produced), nobody uses it for human's lighting, as it adds little lm and just improve marginally CRI (if any).

Yes, 660nm does have lower photometric efficacy, which is probably why the L-prize bulb uses 635nm red emitters, because they were focusing on maximizing efficiency. The downside to this approach is that 635nm is not really into the deep red part of the spectrum, and at 635nm there is still much overlap with the orangish Nd:YAG phosphor hump. Probably not the best for rendering red colors or providing full spectrum coverage. From the specifications I have seen, the enhanced white LED lamps that use 660nm seem to be rated about 2 CRI higher than those using 635nm.



No matter how enhanced white LEDs are, many people are still going to find them off because of that narrow deep blue frequency spike. I wish they could use the enhanced red emitter technology, but use electroluminescent blue/blue-green phosphors to get the blue part of the spectrum, instead of simply allowing the LED's harsh blue frequency spike to leak through.

The harshness of blue LED light is due to two things. First, the short wavelength, it is "hard" blue frequency light. This makes it appear a little purplish, but also makes it more difficult for the eye to focus on. Second, the narrow concentration into a frequency spike makes it appear harsh rather than soft. In contrast, the light from blue-green electroluminescent sources have a very soft appearance.

In fact, the deep blue wavelength peak from white LEDs (465nm) is almost indigo (considered to start at 464 nm).

 

[foxtrot824]

Since there is some much talk about the wave lengths I thought I'd most this. We took apart an L prize bulb when they came out and thought I'd share some of the numbers.

 

[idleprocess]

Since there is some much talk about the wave lengths I thought I'd most this. We took apart an L prize bulb when they came out and thought I'd share some of the numbers.

A little dispersion there on the bare emitters ... what about the spectrum on the whole thing?

EDIT: Ah, I see the varying drive currents now - nevermind.

 

[landroni]

I read what you wrote on the feedback on the economist page and I am sorry, but what you wrote is predominantly crap.

I beg to differ.

1) With the exception of cheap Christmas lights, no LED lights flicker. Actually for that matter, no modern CFL or linear fluorescent flickers either.

As Arilou points out, LEDs flicker. This flickering is also documented here, along with a method to detect it called the "phantom array" effect. This is essentially the "Wave a broom handle" technique that Yoda4561 describes in his post. Personally I just take a camera with good optics and record a video of the bulb in an otherwise completely dark room. Replay the recording in slow motion, and notice the strong flickering effect.

If you don't take my word for it, feel free to buy a "cold daylight" LED bulb from here or a Megaman LG1014dv2 LED bulb, and perform the above-mentioned experiments. They flicker like hell.

As for (modern) fluorescents, the effect is not "flickering" per se. It's a Neurological/Optical Effect of Fluorescent Lighting. You need to have above average sensitivity to notice that, though.

2) You must hate sunlight ... it has a ton of blue light in it.

Far from it; I relish in sunlight. It's not the absolute amount of light that matters, but its relative proportions.

Sunlight and fluorescent bulbs are very different technologies, in that sunlight comes with a nice continuous spectrum while fluorescents generate spikes of very narrow bands of the spectrum. Even if sunlight contains a lot of the blue band of the spectrum, it is nicely diluted because you have so many other wavelengths. The resulting spectrum is approximately uniform.

Fluorescents are very different beasts. Their spectrum is highly discontinuous and they essentially generate spikes in intensity: one at UV, one at blue, one at red, etc. If you have above-average sensitivity to any of those specific wavelengths, like I am, then your screwed. But even if you're within the average, it doesn't mean you're immune to it: you are just less likely to realize it.

Unfortunately LEDs are not always much better than fluorescents (be it screens or lamps). Even though their spectrum is more continuous than that of fluorescents, as Anders Hoveland explains in his post, LEDs often come with a "harsh [..] narrow deep blue frequency spike".

But if you're so confident in this fluorescent technology, then I invite you to use "cold daylight" fluorescent bulbs 24/7, at home, in the office, in the bathroom, in the basement, and most importantly on your desk while working; be sure to remove all covers, too, so that you nicely get your eyes exposed to these intense sources of light, and to keep the desk lamp at less than 1m from your eyes. If possible, try to stay exposed to them for a week or so with no or little exposure to sunlight. I'd be very surprised if you didn't get eye strain, headaches, decrease in eyesight or the like. Repeat the experiment with incandescents, and I'll wager it would be much nicer.

3) I don't buy for a minute your test with yellow sunglasses and computers reducing eye strain. Those glasses work outside in bright sunlight where there is more than enough blue light to still make it through the glasses and cause the iris to properly close. In an office situation in front of a computer, the surest way of getting eye strain would be to wear yellow glasses. This will eliminate blue from the spectrum causing your pupils to over dilate and consequently kill your depth of field leading to excessive eye strain as when you move your head things will be continuously out of focus. Manufacturers have even added blue lit trim to monitors to compensate for this real issue. Without adequate blue light, your pupils will be wide open which impacts ability to read fine text, see fine details, and adapt to diverse focal distance situations .. i.e. looking at a monitor and other things around your office.

We definitely don't have the same sources on this one. Less blue light improves contrast, hence reduces eye strain. Also yellow-tinted glasses work essentially as sunglasses, thus reducing luminosity; again, less eyestrain. This works when working in front of computers, and if you don't believe it you may as well test it by yourself with a pair of such glasses. They don't help with overhead neon lamps, though, as you always have direct eye contact with the light source.

I am calling bs ..... based on real science, not hand waving. Keep in mind, we evolved in a situation where we were exposed to the sun (super rich in blue) up to 16 hours a day and 12 on average. Sorry, but a few hours of a bit of blue is not going to cause major health issues and likely is saving your eyes and causing less eye strain. But hey, if you want to do reading and other fine work in halogen/incandescent lighting, they are your eyes to wreck. Getting rid of the blue during the day will also play havoc with your circadian rhythm.

Some people have above-average sensitivity to certain wavelengths. This is documented here in the context of autism. When you have such sensitivity and work 8 hours a day on small, dark-blue laptop screen, surrounded by neon and fluorescent bulbs at work and at home, you're bound to hit trouble. But if you personally do not have such issues, please be so kind not to call the issues that I have "crap".


And to get back to my original question: Does the L-Prize flicker, and does it come with excess blue light? Could anyone who got a hold of this bulb let us know?