New G-E household LED claims 97 CRI

[PhotonWrangler]

I ran across a new G-E LED bulb in a store that goes by the brand "Sun Filled." A quick search shows that it's based on a Seoul violet LED and a Toshiba RGB phosphor blend, and this article in LEDs Magazine shows the specifics. It supposedly produces a remarkably flat spectral distribution without the blue spike.

I'm tempted to try a couple of these around the house. Anyone have one yet?

 

[idleprocess]

Violet die pump is an interesting idea - especially if the near entirety of the spectrum is being down-converted.

Of the two CCTs available I've limited use for 5000 and somewhat recently swapped out several for 90CRI models so I've not got a need at the moment. Perhaps when some vanity fixtures with super-cheap GEs go out.

 

[DaveTheDude]

I just ordered one on Amazon (Saturday night). Price was US$10. I'll post my impressions after it arrives on Monday.

 

[Dave_H]

Interesting that the promo mentions purple LED, which is not violet. I think some plant grow lights use purple (red+blue) compared to separate red and blue for the two different Chlorophylls.

Not great efficacy though as 60W bulb uses 10.5W (but at 840 lumens). Regular good filament bulbs are down to 6.5W commonly; and new Philips bulb down to 4W. Speaking of that, I'll keep an eye out for both OTC (GE at Lowes and Rona etc. here) but looks a bit early for general availability; and pricey.

Dave

 

[idleprocess]

Interesting that the promo mentions purple LED, which is not violet

I saw that too and assumed it was a misprint or obfuscation since violet has a wavelength range associated with it while purple is a trick of perception involving red and blue wavelengths. One could certainly use a red+blue pump but I imagine that still leaves the blue spike.

 

[Dave_H]

I saw that too and assumed it was a misprint or obfuscation since violet has a wavelength range associated with it while purple is a trick of perception involving red and blue wavelengths. One could certainly use a red+blue pump but I imagine that still leaves the blue spike.

Not to be too hard on marketing, but they tend to lose the distinction like this, at times.

Dave

 

[PhotonWrangler]

I saw that too and assumed it was a misprint or obfuscation since violet has a wavelength range associated with it while purple is a trick of perception involving red and blue wavelengths. One could certainly use a red+blue pump but I imagine that still leaves the blue spike.

Many years ago G-E started experimenting with a violet-pumped LED phosphor. The idea was that the shorter wavelength of the pump source opened the doors to a wider range of available phosphors for a better CRI. It looks like Seoul/Toshiba is picking up the ball on this idea now. I always thought it was a solid idea in terms of more natural color rendering, even if it produced a small dip in efficiency. The target markets at the time were those where CRI was more important, such as supermarkets, medical facilities and some architectural settings.

The G-E bulbs were briefly marketed as "Vio-LED." I haven't seen that name in awhile though. I winder if they sold some patent rights or licensing to other manufacturters.

 

[idleprocess]

Many years ago G-E started experimenting with a violet-pumped LED phosphor. The idea was that the shorter wavelength of the pump source opened the doors to a wider range of available phosphors for a better CRI.

I recall a UV- or NUV-pumped design using the same basic phosphors as fluorescent lamps. Suspect it was not a commercial success due to lifespan concerns - the UV likely caused unacceptably fast degradation.

I also recall that early violet die had some lifespan problems of their own, however I've slept a few times since then.

I always thought it was a solid idea in terms of more natural color rendering, even if it produced a small dip in efficiency. The target markets at the time were those where CRI was more important, such as supermarkets, medical facilities and some architectural settings.

Given that system efficiencies of ~100 lm/W can probably be had with this product the efficiency dip is apt to be quite acceptable, particularly with 97 CRI across the entire range of CCTs.

 

[PhotonWrangler]

I also recall that early violet die had some lifespan problems of their own, however I've slept a few times since then.

I recall this as well. It seems that the shorter the wavelength, the more challenging it is to make them last long. They either fail catastrophically or they have poor radiant flux maintenance.

 

[Dave_H]

I guess 88 lumens/watt is not bad for a high-CRI bulb, although it's efficacy of about a decade ago albeit not high CRI. It would be interesting for side-side comparison with other (real or supposed) high-CRI bulbs.

Dave

 

[idleprocess]

I recall this as well. It seems that the shorter the wavelength, the more challenging it is to make them last long. They either fail catastrophically or they have poor radiant flux maintenance.

Unlike UV/NUV, violet die emitting in the visible spectrum should not inherently cause rapid degradation of surrounding materials.

 

[PhotonWrangler]

Unlike UV/NUV, violet die emitting in the visible spectrum should not inherently cause rapid degradation of surrounding materials.

I agree. It seems there's a thin line between wavelengths that are short enough to excite a wide range of phosphors and wavelengths that cause accelerated deterioration.

 

[jtr1962]

I bought one on eBay. I'll give my thoughts on it as soon as it arrives.

Yes, the efficiency is yesteryear at 80 lm/W, even given the fact it's very high CRI, but it's still not horrible. Keep in mind the Philips CRI 95 bulbs I bought are only 119 lm/W but the spectrum isn't as good as these bulbs.

Should be interesting to see how hot it gets. Given the full spectrum, I'm estimating at least 3.5 watts comes out as light, leaving about 7 watts waste heat. By comparison, the Philips 60W high efficiency bulb only has maybe 1.5 watts of waste heat.

Violet emitters are far less efficient than royal blue. The latter are around 75% or 80% in commercial production. I think violet has yet to crack 50%. With further development, no reason a bulb like this can't eventually get to, say, 125 lm/W.

 

[LEDphile]

Violet emitters are far less efficient than royal blue. The latter are around 75% or 80% in commercial production. I think violet has yet to crack 50%. With further development, no reason a bulb like this can't eventually get to, say, 125 lm/W.

For both Cree XE-G and Lumileds Luxeon C, where both royal blue (450nm) and violet (420nm) are available in the same package making comparisons straightforward, the radiometric efficiencies are basically the same, at around 50-60% (depending on binning and drive conditions, of course).

 

[BVH]

Am I correct in that the daylight version is almost twice the price of the soft white? $23+ for one bulb?

 

[jtr1962]

Am I correct in that the daylight version is almost twice the price of the soft white? $23+ for one bulb?

On Amazon it is but I got mine on eBay for $6.50, free shipping. Unfortunately, it was the last one from that seller. There are other sellers you can get it from for about $10.

 

[PhotonWrangler]

I checked the website for the store that I saw them in, and they're not shown on the website yet. Very curious - it's usually the other way around.

 

[PhotonWrangler]

I snagged one of these, specifically the 840 lumen 2700k version. At $16.00 it's a little spendy for an 840 lumen bulb. Here are my initial impressions -

Pros -
The light output is a pleasant white, and it does render some reds and greens a little better than other bulbs I have.
When it's turned off, it takes a second for it to fully extinguish, indicating some adequately sized filtering capacitors.

Cons -
The radiation pattern is the cardioid "ice cream cone" style with an opaque neck. Presumably the neck is housing the driver and heat sink.
The A21 size seems a bit large for the rated output. Maybe this is to provide better heat dissipation?

 

[idleprocess]

The A21 size seems a bit large for the rated output. Maybe this is to provide better heat dissipation?

(Looks at the 60W incandescent equivalent Philips 'alien head' sitting on the desk rated at 12.5W)

Yeah, probably something to do with heat dissipation.

 

[PhotonWrangler]

Usually I can aim a blue or UV LED into an LED's globe and see some fluorescence coming back from the phosphors. Not this time. The globe itself had a little bit of blue fluorescence but absolutely no phosphor glow, which makes me suspect that there's some UV blocking going on.