CFL Color Temperature

[TKM]

19:38 GMT, [email protected]
|> wrote:
|>
|>> I have not yet seen color bulbs in CFL form, but it should not be
|>> hard to make them. I'm sure they eventually will.
|>
|> I'm not sure there is a big enough market for color CFLs.
|
| FWIW, I saw colored CFLs at Wal-Mart while shopping earlier this
| evening. They were available in red, orange, yellow, green, blue,
| and "black"/UV. These were all 13 W (60 W equivalent) spiral CFLs by
| Feit Electric. The red, orange, green, and blue ones were marketed as
| "party bulbs"; the yellow one was actually marked as an anti-insect
| lamp. I took a close look at the red and orange ones, and I *think*
| they used colored glass, as opposed to a coating on the outside of
| clear glass. They sold for about $4 or $5 each.
|
| Wal-Mart sells another brand of anti-insect CFL (GE or Philips - I
| can't remember) that appears to have a translucent yellow plastic
| covering over a normal CFL.

Now all I need are some colored HID lights. Maybe they would know to
use different HID technologies for the different colors. Unfortunately
there is probably no market for 400 watt HID colored bulbs. They could
be used on giant Christmas trees, though.

Venture Lighting has sold metal halides with different saturated colors
using arc tube chemistry for some time. They're shy about listing the lamps
though and only show them as special order products. There's an example
photo at:
http://www.venturelighting.com/Applications/ArchitecturalStory.html

Terry McGowan

 

[[email protected]]

|
|>Well, there is some physics driving the high CCT of CFLs. As
|>you know, in normal fluorescent lamps the mercury produces a
|>bit of blue and green light that is mixed with the slightly
|>yellow light from the phosphor to make white.
|>
|>It turns out that the current density in most CFLs is so
|>high that they generate a lot of blue light. So much in
|>fact that many CFLs use only a two-component rare earth
|>phosphor mix instead of the normal three-component mix.
|>There is just too much blue to have any more coming from the
|>phosphor. (If the phosphor does have all three components,
|>then the blue component is reduced in small diameter lamps.)
|
| Somehow I see either 3 or 4 phosphor bands/features in the spectra of
| CFLs and "triphosphor" T8 fluorescents, Philips "Ultralume", etc. Could
| two of these spectral features be from the same phosphor?
|
| What I notice, in order from longer wavelength to shorter:
|
| 1. The strong orange-red linelike extremely narrow band around 611 nm,
| along with some very weak similarly narrow bands nearby from yellow to
| deep red,
|
| 2. a small group of wider but still very narrow bands in the green, with
| the dominant feature maybe typically around 542 nm,
|
| 3. a dimmer, moderately narrow band in the green-blue/blue-green
| with its brightest part around 485-490 nm but extending into the
| blue-green around 500 nm, and
|
| 4. a wider still blue band, that with CFLs I usually only see when
| nominal CCT is at least 3500K, mainly from 440-475 or 440-480 nm or so.

Is there any kind of "catalog" of phosphor material and their spectral
lines that would identify what is being used.

My own interest is in find a way to make a CFL that could fill out the
spectrum fully enough to be visually seen much like incandescent. But
I don't know if there is enough phosphor material choices to do that.

I've considered the same for LED. I've seen LEDs listed for sale with
specific wavelengths and have seen at least 22 different wavelengths.
That might be quite a balancing act to get "flat white" out of it all,
especially with different rates of output decay over time.

 

[Victor Roberts]

FWIW, I saw colored CFLs at Wal-Mart while shopping earlier this
evening. They were available in red, orange, yellow, green, blue,
and "black"/UV. These were all 13 W (60 W equivalent) spiral CFLs by
Feit Electric. The red, orange, green, and blue ones were marketed as
"party bulbs"; the yellow one was actually marked as an anti-insect
lamp. I took a close look at the red and orange ones, and I *think*
they used colored glass, as opposed to a coating on the outside of
clear glass. They sold for about $4 or $5 each.

I hope they don't use white phosphor and colored glass. What
a waste of energy.

--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.

 

[Victor Roberts]

Venture Lighting has sold metal halides with different saturated colors
using arc tube chemistry for some time. They're shy about listing the lamps
though and only show them as special order products. There's an example
photo at:
http://www.venturelighting.com/Applications/ArchitecturalStory.html

Terry McGowan

And the special color metal halide lamps that GE designed
for the Statue of Liberty.

--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.

 

[Victor Roberts]

Well, there is some physics driving the high CCT of CFLs. As
you know, in normal fluorescent lamps the mercury produces a
bit of blue and green light that is mixed with the slightly
yellow light from the phosphor to make white.

It turns out that the current density in most CFLs is so
high that they generate a lot of blue light. So much in
fact that many CFLs use only a two-component rare earth
phosphor mix instead of the normal three-component mix.
There is just too much blue to have any more coming from the
phosphor. (If the phosphor does have all three components,
then the blue component is reduced in small diameter lamps.)
The 2D is made in a rather large diameter tube, at least
measured by CFL standards. I believe it is T5 for most
wattages, while most lower power CFLs use T4 tubing or even
smaller. The larger diameter discharge in the 2D, along
with the length, reduces the current density, which in turn
reduces the amount of blue light produced by the discharge.
This allows a three-component rare earth phosphor to be used
which may lead to high quality light.

The above looks like it is validated with spectrogram [1.3.3], which corresponds
to a triphosphor fluorescent with a CCT of 2700K on my Amici page.

The 2700K CCT "triphosphor"'s spectrum, looks like it is truly "bi-phosphor" (as
far as I know ions of Terbium and Europium), since it has no blue emissions,
except the mercury blue line @435.8, which is very strong.

All other spectrograms of triphosphors (those with CCT's of 4000K, 6000K, 8000K
and 17000K, [1.3.4]/[1.3.5]/[1.3.8/[1.3.9]), contain additional continuous
emissions in the blue area around the blue mercury line, betraying the presence
of a "third" phosphor component which is not present in the 2700K CFL
fluorescent.

My question now is why was the 2700K CFL called "triphosphor" if its phosphors
were two-component only. The previous looks like a chronological inconsistency,
unless the "triphosphor" technology for higher CCT fluorescents was already
known at the time of production of 2700K CFLs.

Any opinions on the above?

Would you accept the answer that it is tri-phosphor with 0%
of the blue component?

The better answer is that people are familiar with the term
"tri-phosphor" and since the bi-phosphor uses two of the
same components, AND only one of the available color
temperatures uses only two of the components, this is the
best way to explain it to the world.

--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.

 

[[email protected]]

| I've noticed the same thing and have long suspected that the width or
| "fatness" of spectral lines - especially as found in fluorescent and HID
| lamps -- has something to do with the color quality and acceptance of light
| by users. That aspect of light and color is not well-captured in our current
| color rendering and chromaticity metrics.
|
| The same thing is likely to affect the perception of color quality for LEDs
| and I've discussed the idea with the group that's looking at new color
| measurement standards for LEDs. Plus the notion came up in the LRO Light
| and Color Symposium in 2006, so there's now a bit of research on it.

A metric that metric be used would be what percentage of spectra is covered
by the top N% of emission "lines" (defining a line perhaps as 1 nm of band).
For a range of N% plot a curve that shows the portion of spectra. A laser
at a single wavelength would be a flat line at virtually 0% for all N%.
Incandescent would quickly rise to nearly 100% spectral coverage with N%
going over 50%, I would think. From studing a variety of these, we might
be able to pick a standard N% emission level to reasonably evaluate the
spectral coverage from for most light sources.

A good name for this metric might be "spectral fill" or "spectral coverage".

This all fits in with my determination a few years ago that it is not the
flicker that affects me (I get headaches after about 20 to 40 minutes work
under a fluorescent light, and so I refuse to have them in the kitchen).

 

[TKM]

On Thu, 14 Feb 2008 02:18:46 +0000 (UTC) Don Klipstein

part:
|
|>Well, there is some physics driving the high CCT of CFLs. As
|>you know, in normal fluorescent lamps the mercury produces a
|>bit of blue and green light that is mixed with the slightly
|>yellow light from the phosphor to make white.
|>
|>It turns out that the current density in most CFLs is so
|>high that they generate a lot of blue light. So much in
|>fact that many CFLs use only a two-component rare earth
|>phosphor mix instead of the normal three-component mix.
|>There is just too much blue to have any more coming from the
|>phosphor. (If the phosphor does have all three components,
|>then the blue component is reduced in small diameter lamps.)
|
| Somehow I see either 3 or 4 phosphor bands/features in the spectra of
| CFLs and "triphosphor" T8 fluorescents, Philips "Ultralume", etc. Could
| two of these spectral features be from the same phosphor?
|
| What I notice, in order from longer wavelength to shorter:
|
| 1. The strong orange-red linelike extremely narrow band around 611 nm,
| along with some very weak similarly narrow bands nearby from yellow to
| deep red,
|
| 2. a small group of wider but still very narrow bands in the green,
with
| the dominant feature maybe typically around 542 nm,
|
| 3. a dimmer, moderately narrow band in the green-blue/blue-green
| with its brightest part around 485-490 nm but extending into the
| blue-green around 500 nm, and
|
| 4. a wider still blue band, that with CFLs I usually only see when
| nominal CCT is at least 3500K, mainly from 440-475 or 440-480 nm or so.

Is there any kind of "catalog" of phosphor material and their spectral
lines that would identify what is being used.

My own interest is in find a way to make a CFL that could fill out the
spectrum fully enough to be visually seen much like incandescent. But
I don't know if there is enough phosphor material choices to do that.

I've considered the same for LED. I've seen LEDs listed for sale with
specific wavelengths and have seen at least 22 different wavelengths.
That might be quite a balancing act to get "flat white" out of it all,
especially with different rates of output decay over time.

There's very little/none light output decay with either straight LEDs or
LEDs with phosphor; but combining the light from many different colors of
LEDs is being researched at NIST (Washington, DC) in a new facility designed
to do just that as part of an effort to come up with ways to better describe
LED color. I haven't seen a write-up on the facility or any results yet,
though.

The technique that I use to get white light from a mixture of colored
sources is to plot the chromaticity of each source on the x,y chromaticity
diagram. Then just connect all of the dots. If any of the lines or group
of lines surround or go through (it depends upon the number of colors that
you're working with) the white light portion of the diagram, you know that,
with proper mixing, white light will result. This technique does not tell
you anything about color rendering, however.

Terry McGowan

 

[Don Klipstein]

Somehow I see either 3 or 4 phosphor bands/features
Could two of these spectral features be from the same phosphor?

1. The strong orange-red linelike narrow band around 611 nm, along with
some very weak similarly narrow bands nearby from yellow to deep red,

2. a small group of wider but still very narrow bands in the green, with
the dominant feature maybe typically around 542 nm,

3. a dimmer, moderately narrow band in the green-blue/blue-green
with its brightest part around 485-490 nm but extending into the
blue-green around 500 nm, and

4. a wider still blue band, that with CFLs I usually only see when
nominal CCT is at least 3500K, mainly from 440-475 or 440-480 nm or so.

I spent some time with Google, and I think I found my answer. I found a
few bits saying that the green component (with terbium) produces the main
green emission and the dimmer blue-green emission. In addition, I found a
couple saying it produces one of the dim yellow bands - one that appears
to me to be mostly 582-584 nm or maybe a bit wider, wider than the other
little yellow/orange bands.

- Don Klipstein ([email protected])

 

[[email protected]]

FWIW, I saw colored CFLs at Wal-Mart while shopping earlier this
evening. [...] I took a close look at the red and orange ones,
and I *think* they used colored glass, as opposed to a coating
on the outside of clear glass.

I hope they don't use white phosphor and colored glass. What
a waste of energy.

If so, it's probably an absolute waste of energy, but not a relative
waste. In other words, these 13 W colored CFLs are going to be
replacing 40 W or 60 W colored incandescent lamps, so there is already
something like a 65% to 75% energy savings, just as with any other
"incadescent replacement" CFL. It is probably true that if they used
clear (or only slightly tinted) glass and colored phosphor, the lamps
themselves would be more efficacious.

However, I suspect that economies of production also come into play;
it's probably cheaper to buy small volumes of colored glass than it is
to buy small volumes of colored phosphors. (There are people here
that are closer to the lighting industry that know a lot more about this
than I do.) The only big market I can think of for colored phosphors
now would be neon lighting; CRTs are almost extinct, and neon lighting
is a small market, IMHO. Colored glass is popular in tableware and has
not yet been totally replaced by LEDs in strings of holiday lamps, so
I suspect it might still be relatively cheap.

Matt Roberds

 

[Victor Roberts]

There's very little/none light output decay with either straight LEDs or
LEDs with phosphor;

I don't see how you can say this. LED life is now most
often specified as the time at which they have lost 30% of
their initial output. That's worse than most T5 and T8
fluorescent lamps, and would be a considerable challenge for
any color matching system.

--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.

 

[TKM]

I don't see how you can say this. LED life is now most
often specified as the time at which they have lost 30% of
their initial output. That's worse than most T5 and T8
fluorescent lamps, and would be a considerable challenge for
any color matching system.

My apologies. I wasn't clear. I was referring to instantaneous decay, not
light output depreciation over life.

Some LED systems already incorporate feedback circuits to compensate for
color shift and/or light output depreciation, of course.

Terry McGowan