How long does an LCD screen last ?

[Al Dykes]

A high-end CRT gets fuzzy and dim anfter a couple of years.

How does a good LCD change with age and what is the limiting factor
for life ?

Does turning an LCD on and off frequently shorten the life. I have in
mind a laptop that has power saving set to trun the screen off after
15 minutes of activity.


Comments ?

 

[Bob Myers]

A high-end CRT gets fuzzy and dim anfter a couple of years.

Well, OK. Actually, what you are seeing here is the aging of the
cathode and phosphor screen, which combine to reduce the
brightness and change the operating characteristics of the
tube. (I.e., it doesn't HAVE to get "fuzzy," but the charactertistics
of the electron gun change such that the focus voltages provided
by the initial adjustment are no longer correct.) It's rarely worth
the trouble, though, to go through a complete readjustment of
the monitor given the main impact of all this aging, which is
to make the cathode a less efficient producer of electrons, and
the phosphor less efficient at converting them into light. This
typically means a lifetime for the CRT of something on the order
of 15-20,000 hours of operation. IF you left the thing running
all the time, that would be a couple of years (one year is
about 8800 hours), but then if you only use the thing 8 hours
a day, five days a week, this might be a useful life of 10 years
or so.

How does a good LCD change with age and what is the limiting factor
for life ?

The LCD itself - i.e., the inner workings of the panel, the action
of the liquid-crystal molecules themselves - doesn't really have
much in the way of aging mechanisms that degrade its performance,
assuming that there's no significant net DC voltage applied across
the material for extended periods of time (and this is true for
any drive scheme you're going to come across these days). So
it's tempting to say that the life of the LC itself, assuming that the
panel doesn't get mechanically stressed (cracked, etc.), or operated
under extreme temperatures is practically unlimited. However, what
you DO run into within the panel itself is the eventual failure of one
or more drivers (which will generally take out an entire row or
column of pixels), or (much less likely, but there's way more of them)
the failure of a transistor within the TFT "active matrix" array on the
glass (which knocks out a subpixel). Driver failures are generally
considered to fail the panel, whereas with subpixel losses, it's a
question of how many you can tolerate before you consider the
panel as "failed." In any event, once the "infant mortality" sorts of
fails are past (those failures that occur very early in the life of a
given component, generally due to something wrong in the
construction of that component that wasn't caught when it was
made), the cumulative MTBF of the panel is very likely at least
in the 50-100,000 hours range, meaning that it's very likely to
give you at least a 10 year useful life and possibly quite a bit
more. This again assumes no unusual stresses, either mechanical,
thermal, or electrical.

More likely than a panel failure by far is failure of one or more of
the backlight tubes, and/or the power supplies (you've likely got
at least two to deal with in an LCD monitor, namely the main
supply and the inverter which provides high voltage to drive the
backlights). Backlights of the CCFL type (cold-cathode
fluorescent, which most notebook and monitor panels will be
using) also DO age, similar to the way a CRT ages, meaning
that their light output goes down with time - so you may reach
a point where everything is working, but you find the display
unacceptably dim.

The bottom line, with all of these factors considered, is that
you currently see LCD monitors generally quoting an MTBF in
the low-to-mid tens-of-thousands of hours range, with the
limiting factors being the backlight tubes themselves, followed
by the inverter, power supply, and finally either panel or the
interface board. As with any electronics, mechanical and
thermal stresses outside the normal operating conditions will
shorten the life considerably.

Does turning an LCD on and off frequently shorten the life. I have in
mind a laptop that has power saving set to trun the screen off after
15 minutes of activity.

As with any electronics, turning the unit off for power-savings
reasons involves a tradeoff. You're trading straight aging
factors for the degradation that results from the on/off
transients. It's really hard to answer this question generally, as
a lot depends on just how often the unit is being power-cycled
vs. how long it would be left on an idle (i.e., burning time on the
"aging" clock, but not really doing anything useful). Turning off
after 15 minutes of inactivity isn't necessarily a bad thing, as long
as it's not being immediately turned back on each and every
time that happens, all day long. That's then the worst possible
case, as you get all of the aging AND the on/off transients.

Bob M.

 

[Gumby]

.......
More likely than a panel failure by far is failure of one or more of
the backlight tubes, and/or the power supplies (you've likely got
at least two to deal with in an LCD monitor, namely the main
supply and the inverter which provides high voltage to drive the
backlights). Backlights of the CCFL type (cold-cathode
fluorescent, which most notebook and monitor panels will be
using) also DO age, similar to the way a CRT ages, meaning
that their light output goes down with time - so you may reach
a point where everything is working, but you find the display
unacceptably dim.

The bottom line, with all of these factors considered, is that
you currently see LCD monitors generally quoting an MTBF in
the low-to-mid tens-of-thousands of hours range, with the
limiting factors being the backlight tubes themselves, followed
by the inverter, power supply, and finally either panel or the
interface board. As with any electronics, mechanical and
thermal stresses outside the normal operating conditions will
shorten the life considerably.
...

White LEDs are getting pretty bright now, why don't LCD manufacturers
use LEDs for the backlight instead of the tubes? That would have much
higher lifespan(100000+ hrs for LEDS), better ruggedness (no brittle
glass tubes) and avoid the toxic materials in those tubes (probably mercury
in them?)
Mark

 

[J. Clarke]

White LEDs are getting pretty bright now, why don't LCD manufacturers
use LEDs for the backlight instead of the tubes?

It's not the LCD manufacturer, it's the manufacturer of the monitor that
decides. There are some with LEDs--they're fairly common on PDAs for
example.> That would have much

higher lifespan(100000+ hrs for LEDS),

Not necessarily. White LEDs use a phosphor similar to fluorescents and I
believe that that phosphor ages.

better ruggedness (no brittle
glass tubes)

The "brittleness" of the glass tube (engineered glass can be tied in knots
or made to stop bullets) is secondary to the fragility of the LCD panel
itself.

and avoid the toxic materials in those tubes (probably
mercury in them?)

Very little. And LEDs have their own toxic materials--arsenic instead of
mercury.

 

[Bob Myers]

White LEDs are getting pretty bright now, why don't LCD manufacturers
use LEDs for the backlight instead of the tubes? That would have much
higher lifespan(100000+ hrs for LEDS), better ruggedness (no brittle
glass tubes) and avoid the toxic materials in those tubes (probably mercury
in them?)

Two major reasons: cost and power/thermal issues. LEDs
aren't as efficient (yet) as CCFLs, and they're definitely more
costly. They've started to show up more in smaller panels
(cameras, cell phones, PDAs, that sort of thing) already, and
will make it to TVs at the high end very soon (and will then
be working their way down the market to monitors and
notebook PCs).

But expect the TV/monitor/etc. LED backlight systems to
use RGB LEDs, not white. Having the separate colors
is generally more efficient, and gives a FAR better color
gamut. (Better, in fact, than any display that has been in
mainstream use to date!)

Bob M.

 

[Bob Myers]

Very little. And LEDs have their own toxic materials--arsenic instead of
mercury.

That's not really much of a concern, though, when comparing
the environmental impact of LEDs vs. CCFLs; in the LEDs,
the quantity of the toxic element or compound is extremely
small, and it's very tightly bound. CCFLs, on the other hand,
contain free mercury.

Bob M.

 

[J. Clarke]

That's not really much of a concern, though, when comparing
the environmental impact of LEDs vs. CCFLs; in the LEDs,
the quantity of the toxic element or compound is extremely
small, and it's very tightly bound. CCFLs, on the other hand,
contain free mercury.

Which is less of a concern than organic compounds of mercury. Being "very
tightly bound" and being "nontoxic" are not the same thing.

 

[J. Clarke]

Two major reasons: cost and power/thermal issues. LEDs
aren't as efficient (yet) as CCFLs, and they're definitely more
costly. They've started to show up more in smaller panels
(cameras, cell phones, PDAs, that sort of thing) already, and
will make it to TVs at the high end very soon (and will then
be working their way down the market to monitors and
notebook PCs).

But expect the TV/monitor/etc. LED backlight systems to
use RGB LEDs, not white. Having the separate colors
is generally more efficient, and gives a FAR better color
gamut. (Better, in fact, than any display that has been in
mainstream use to date!)

I can see where a system that uses RGB LEDs as pixels would have a good
color gamut, but I'm a little bit puzzled as to how that would give as
different color gamut when used as a backlight for an LCD than a white
phosphor would. As far as being "more efficient", the white phosphor LEDs
seem to give off more light per unit of power consumed than colored LEDs.

 

[Bob Myers]

Which is less of a concern than organic compounds of mercury. Being "very
tightly bound" and being "nontoxic" are not the same thing.

Nor did I say they were; but toxic compounds are
not as much of a concern IF they are, and reliably will
be, kept away from possible leaching into groundwater,
etc., whether it's through physical containment (i.e.,
the material in question is reliably sealed away) or
through being chemically bound so tightly that it's
extremely unlikely that any hazard will exist. That's
why the lead in CRTs, while it hasn't been an
ignorable hazard, has not been the concern that one
might initially think. It's contained within the glass, and
is about as likely to be a health hazard under typical
conditions as drinking from your grandmother's prized
lead crystal goblets would be.

Similarly, the arsenic that was mentioned as being a
concern for LEDs is both chemically bound up, and
very unlikely to leach out just for that reason, plus
it's embedded in a metal and epoxy package. The
mercury in CCFLs (or other fluorescents, for that
matter), is a LOT more likely to get out into the
environment and cause problems. The mercury content
in backlight tubes, specifically, has been rather strictly
limited by various regulatory agencies (the limits imposed
by the European Union, under directive 2002/95/EC,
are the most commonly referenced). However, no
requirement for mercury-free backlighting has been
imposed yet, because at this time the various national
and international agencies are more interested in
encouraging the move the LCD monitors, for energy-
efficiency reasons. And it's understood that the
mercury-free alternatives simply aren't economically
feasible at this time. Once solid-state backlights are
more cost-competitive, that situation may change.

Bob M.

 

[Bob Myers]

I can see where a system that uses RGB LEDs as pixels would have a good
color gamut, but I'm a little bit puzzled as to how that would give as
different color gamut when used as a backlight for an LCD than a white
phosphor would.

Having a wide color gamut depends on having primaries
that are highly saturated - in other words, the colors are
as spectrally pure as possible. Ideally, the RGB primaries
would be purely monochromatic; completely saturated,
single-wavelength sources. But at present, lasers are about
the only way to get to something like that.

Filtering a white source generally gives a much smaller
gamut, since the white source (especially an incandescent
source, but also a "white" phosphor one) emits over a wide
range of wavelengths, and no practical color filters can
limit the light coming through the display to just a single
wavelength, or even a narrow band around the desired
wavelength. This results in primaries that aren't especially
pure, and therefore a smaller gamut. (Take a look at
where typical CRT phosphors, such as the EBU set, fall
on the CIE color space chart.)

LED sources are much better, because they DO emit in
such a narrow range anyway (generally not a single
wavelength, but there's a narrow range, generally with
one wavelength dominating), and that's made even purer
with the right set of color filters. So you wind up with a
much better gamut than is possible with a white source.

As far as being "more efficient", the white phosphor LEDs
seem to give off more light per unit of power consumed than colored LEDs.

If the goal were just to make white light, you'd be correct
(at least for now). But as noted above, the goal for a
color display is actually to put the most light possible into
a narrow band of wavelengths around the desired
primaries. RGB LEDs which match the color filters well
will always be better in that regard than starting with
a white source, since the color filters will in effect "throw
away" most of the light from a white source.

Bob M.

 

[J. Clarke]

Having a wide color gamut depends on having primaries
that are highly saturated - in other words, the colors are
as spectrally pure as possible. Ideally, the RGB primaries
would be purely monochromatic; completely saturated,
single-wavelength sources. But at present, lasers are about
the only way to get to something like that.

First, I fail to see how making the white light for the backlight from
"completely saturated, single wavelength sources" instead of a phosphor
improves the color gamut of LCD displays, and second, the difference
between an LED and a laser is a little bit of trimming and some mirroring.
You don't need a laser to get monochromatic light, you need a laser to have
_coherent_ light.

Filtering a white source generally gives a much smaller
gamut, since the white source (especially an incandescent
source, but also a "white" phosphor one) emits over a wide
range of wavelengths, and no practical color filters can
limit the light coming through the display to just a single
wavelength, or even a narrow band around the desired
wavelength. This results in primaries that aren't especially
pure, and therefore a smaller gamut. (Take a look at
where typical CRT phosphors, such as the EBU set, fall
on the CIE color space chart.)

So how does backlighting with colored LEDs improve this?

LED sources are much better, because they DO emit in
such a narrow range anyway (generally not a single
wavelength, but there's a narrow range, generally with
one wavelength dominating), and that's made even purer
with the right set of color filters. So you wind up with a
much better gamut than is possible with a white source.

Again, how? You have explained how it is possible to obtain broader gamut
if you are using the LEDs as the pixels, but we were discussing
backlighting. Or have you changed the topic without making it clear that
that was your intent?

If the goal were just to make white light, you'd be correct
(at least for now). But as noted above, the goal for a
color display is actually to put the most light possible into
a narrow band of wavelengths around the desired
primaries. RGB LEDs which match the color filters well
will always be better in that regard than starting with
a white source, since the color filters will in effect "throw
away" most of the light from a white source.

Well that's nice. Next time you change the subject would you make it clear
that you are doing so?

 

[Bob Myers]

First, I fail to see how making the white light for the backlight from
"completely saturated, single wavelength sources" instead of a phosphor
improves the color gamut of LCD displays,

I'm not sure how I can make it any clearer without being
able to show you the spectral curves for the two types
of sources. You DO understand that the LEDs are much
"purer" sources of light of a given color than a white phosphor.
right? (White phosphors are actually generally a mix of
multiple different materials, and have emission peaks all over
the place. If they were truly broadband "white" emitters, as
an incandescent bulb is, the situation would be even worse.)

Take a look, for instance, at the emission spectra of various
"white" LEDs in:

http://www.lumileds.com/pdfs/techpaperspres/IEEEJSTQE-HighPowerPCLEDs.PDF

and then compare these with the spectra of individual RGB
emitters, as in:

http://www.lumileds.com/pdfs/DS25.pdf

Remember, it's not just a matter of putting out light over
a given range, either, but rather maximizing the light output
within a very narrow range of wavelengths. Again, the individual
RGB emitters matched to the color filters in use (actually, the
matching generally goes the other way around) is the best way
to achieve this.

LumiLEDs also has an excellent paper on the whole solid-state
backlighting notion in general, which you can find at:

http://www.lumileds.com/pdfs/techpaperspres/HiBrtDirLEDTV.PDF

and second, the difference
between an LED and a laser is a little bit of trimming and some mirroring.
You don't need a laser to get monochromatic light, you need a laser to have
_coherent_ light.

As you'll see in the emission spectra shown in the documents
above, plain LEDs are NOT truly monochromatic sources. They
have a strong dominant wavelength, but there is also significant
emission to either side of this point. A laser does produce light
which is coherent (whereas the "plain LED" does not), but the same
phenomena which cause it to be coherent also mean that 's it a
much better monochromatic light source. The difference in
theory may be "a little bit of trimming and some mirroring," but
so far there's no good way to use laser sources in backlighting
applications in a practical sense.

Again, how? You have explained how it is possible to obtain broader gamut
if you are using the LEDs as the pixels, but we were discussing
backlighting. Or have you changed the topic without making it clear that
that was your intent?

Nope. Using the LEDs as pixels and using the LEDs as
backlighting gives very similar results - remember, the point
is saturated primaries, and putting the color filters over the
LED backlight (which admittedly appears to be making "white"
light) gets a lot closer to that than those same filters over a
more broadband source, such as a white-phosphor-based
emitter. Actually using the LEDs as the subpixel sources would
be even better, but clearly isn't practical for anything other
than very large displays. (You can also do better by going
to a field-sequential drive and illumination scheme, simply
because then the filters won't ever have to deal with the
"wrong" colors, but that's another can of worms.)

Well that's nice. Next time you change the subject would you make it clear
that you are doing so?

Sorry, no change of subject made. The point of a
backlight system for any color LC display, if the goal
is brightness and wide gamut, is NOT to make white,
but to provide the most light out through the color
filters. A true (broadband) white source will not be
the best way to do this.

Bob M.

 

[J. Clarke]

I'm not sure how I can make it any clearer without being
able to show you the spectral curves for the two types
of sources. You DO understand that the LEDs are much
"purer" sources of light of a given color than a white phosphor.
right? (White phosphors are actually generally a mix of
multiple different materials, and have emission peaks all over
the place. If they were truly broadband "white" emitters, as
an incandescent bulb is, the situation would be even worse.)

Take a look, for instance, at the emission spectra of various
"white" LEDs in:

http://www.lumileds.com/pdfs/techpaperspres/IEEEJSTQE-HighPowerPCLEDs.PDF

and then compare these with the spectra of individual RGB
emitters, as in:

http://www.lumileds.com/pdfs/DS25.pdf

Remember, it's not just a matter of putting out light over
a given range, either, but rather maximizing the light output
within a very narrow range of wavelengths. Again, the individual
RGB emitters matched to the color filters in use (actually, the
matching generally goes the other way around) is the best way
to achieve this.

LumiLEDs also has an excellent paper on the whole solid-state
backlighting notion in general, which you can find at:

http://www.lumileds.com/pdfs/techpaperspres/HiBrtDirLEDTV.PDF



As you'll see in the emission spectra shown in the documents
above, plain LEDs are NOT truly monochromatic sources. They
have a strong dominant wavelength, but there is also significant
emission to either side of this point. A laser does produce light
which is coherent (whereas the "plain LED" does not), but the same
phenomena which cause it to be coherent also mean that 's it a
much better monochromatic light source. The difference in
theory may be "a little bit of trimming and some mirroring," but
so far there's no good way to use laser sources in backlighting
applications in a practical sense.


Nope. Using the LEDs as pixels and using the LEDs as
backlighting gives very similar results - remember, the point
is saturated primaries, and putting the color filters over the
LED backlight (which admittedly appears to be making "white"
light) gets a lot closer to that than those same filters over a
more broadband source, such as a white-phosphor-based
emitter. Actually using the LEDs as the subpixel sources would
be even better, but clearly isn't practical for anything other
than very large displays. (You can also do better by going
to a field-sequential drive and illumination scheme, simply
because then the filters won't ever have to deal with the
"wrong" colors, but that's another can of worms.)


Sorry, no change of subject made. The point of a
backlight system for any color LC display, if the goal
is brightness and wide gamut, is NOT to make white,
but to provide the most light out through the color
filters. A true (broadband) white source will not be
the best way to do this.

Could you describe the geometry of whatever it is that you are proposing?

 

[Bob Myers]

Could you describe the geometry of whatever it is that you are proposing?

Well, it's not exactly what I am proposing, personally, but
rather the way solid-state backlighting is currently being
planned for TV and high-end monitor panels. Again, the
most concise resource I can point you to with a good
coverage of the basics of this is the LumiLEDs paper
referenced in my last posting. However, the basic
geometry of this, for large-screen LCD backlighting, is
generally an array of LEDs (either separate R, G, and
B packages, or possibly tricolor single-package emitters
(if the thermal issues can be dealt with), either located
behind or embedded within a diffuser. (LumiLEDS has
a rather elegant "side-emitting" package designed for
just this application - you embed a bunch of those into
the diffuser, and the light goes almost entirely out the sides
into the diffuser before eventually being scattered "up"
through the LCD. This better ensures that you get a
uniform backlight, rather than annoyingly intense point
sources!)

Smaller panels, such as cell phones, PDAs, etc. (where
solid state backlighting is already becoming popular) and
notebook PCs are more likely to use a side-light sort of
configuration, with the LEDs arranged as a strip along one
side (or maybe two sides) of the module, sending light into
a diffuser/reflector on the back of the LCD panel (the diffuser
then mixes the light for uniformity and sends it "up" through
the LCD).

Note that the end result is a backlight which, if seen without
the LCD over it, would still look to the eye like a bright,
uniform, white plane - the difference between it, though, and
a normal "white" backlight is that the light coming out actually
has the energy primarily concentrated in three narrow peaks,
as opposed to being spread over the visible spectrum.

I hope that clarified things!

Bob M.

 

[J. Clarke]

Well, it's not exactly what I am proposing, personally, but
rather the way solid-state backlighting is currently being
planned for TV and high-end monitor panels. Again, the
most concise resource I can point you to with a good
coverage of the basics of this is the LumiLEDs paper
referenced in my last posting. However, the basic
geometry of this, for large-screen LCD backlighting, is
generally an array of LEDs (either separate R, G, and
B packages, or possibly tricolor single-package emitters
(if the thermal issues can be dealt with), either located
behind or embedded within a diffuser. (LumiLEDS has
a rather elegant "side-emitting" package designed for
just this application - you embed a bunch of those into
the diffuser, and the light goes almost entirely out the sides
into the diffuser before eventually being scattered "up"
through the LCD. This better ensures that you get a
uniform backlight, rather than annoyingly intense point
sources!)

Smaller panels, such as cell phones, PDAs, etc. (where
solid state backlighting is already becoming popular) and
notebook PCs are more likely to use a side-light sort of
configuration, with the LEDs arranged as a strip along one
side (or maybe two sides) of the module, sending light into
a diffuser/reflector on the back of the LCD panel (the diffuser
then mixes the light for uniformity and sends it "up" through
the LCD).

Note that the end result is a backlight which, if seen without
the LCD over it, would still look to the eye like a bright,
uniform, white plane - the difference between it, though, and
a normal "white" backlight is that the light coming out actually
has the energy primarily concentrated in three narrow peaks,
as opposed to being spread over the visible spectrum.

I hope that clarified things!

Now I understand a number of things. The big one is that you don't seem to
understand how Luxeon's LEDs work. They are high-brightness and they
obtain that brightness by pumping a phosphor with a blue or UV LED. Their
red, green, and blue LEDs just use different phosphors, which is also why
they aren't monochromatic. If you don't believe me give them a call and
ask them and they'll tell you exactly what I'm telling you.

Once that piece is in place it starts to make sense. It is reasonable that
a single color phosphor might give higher brightness than a white phosphor,
and by tweaking the spectrum to match the response of the LCD even more
efficiency could be obtained.

 

[Bob Myers]

Now I understand a number of things. The big one is that you don't seem to
understand how Luxeon's LEDs work.

Sigh. No, apparently all we've discovered is that you are unable
to follow links to relevant material. Should you actually read the
LumiLEDs paper I pointed you to, you'd learn very quickly that the
devices being proposed for TV (and eventually, monitor) LCD
backlighting application are not UV/phosphor devices, but rather
separate high-efficiency R, G, and B LEDs - new side-emitting
devices developed specifically for this application, which LumiLEDs
refers to as the "Luxeon Direct" backlight system.

Please also note that "Luxeon" is not the name of the company;
their name is "LumiLEDs" (which I've seen either with the ED
capitalized or not; I prefer to capitalize it, since these ARE
LEDs we're talking about), a joint venture of Philips and
Agilent Technologies. "Luxeon" is a brand name applied to a
wide variety of LED products they've developed for illumination
applications - some of which are UV/phosphor types, and others
are direct emitters.

You might also, were you inclined to actually look at some of
this information rather than simply argue about it, want to review
the papers presented by both LumiLEDs and NEC-Mitsubishi
at the Society for Information Display symposium this year.
The LumiLEDs paper (41.3) goes over their plans for LCD
backlighting in general, including the THREE different forms of
such backlighting they've proposed (basically, these are for
small-size, medium, and large-area LCDs). The NEC-Mitusbishi
paper (41.4) concerns their prototype 23" LCD monitor which
was constructed using the LumiLEDs backlight system. Both
papers go into considerable detail on the color gamut advantages
such backlighting provides (the NEC-Mits prototype, for instance,
provides a gamut of over 101% of the 1953 NTSC spec, which
is well beyond any current LCD or CRT television). Both papers
may be found in this year's SID Symposium Digest; I don't know
if they're on-line anywhere or not.

Since you still seem to belive that the typical LED IS truly a
monochromatic light source, I will also refer you to the following;
look for the words "spectral width" and "chromatic dispersion"
here to find the relevant information:

http://www.fiber-optics.info/articles/LEDs.htm#Table_2

http://www.mtmi.vu.lt/pfk/funkc_dariniai/diod/led.htm

http://zone.ni.com/devzone/conceptd.nsf/webmain/90B8D6058CB7C18386256C6A00598C96

In theory, the LED WOULD be a monochromatic light source if
all you had to worry about was the emission resulting from a
transition of an electron over a single, stable band-gap energy.
In practice, of course, you get something a bit different, and so the
spectral width (as measured at the half-power point) of even the
non-phosphor, direct emitters is typically on the order of 50-100 nm
or more (i.e., a fair fraction of the visible spectrum), vs. a width of
1 nm or less for a solid-state laser source. Again, if you want to
get more into the whys of this, the above are a good place to start.

Since, though, you are apparently more interested in simply arguing
rather the reading the material, I think this is a good time to close
this discussion. There's certainly been more than enough information
and pointers to information provided for anyone wishing to learn
more about the subject.

Bob M.

 

[J. Clarke]

Sigh. No, apparently all we've discovered is that you are unable
to follow links to relevant material. Should you actually read the
LumiLEDs paper I pointed you to, you'd learn very quickly that the
devices being proposed for TV (and eventually, monitor) LCD
backlighting application are not UV/phosphor devices, but rather
separate high-efficiency R, G, and B LEDs - new side-emitting
devices developed specifically for this application, which LumiLEDs
refers to as the "Luxeon Direct" backlight system.

No paper to which you pointed me describes such a device. If there is
another paper please post the link.

Please also note that "Luxeon" is not the name of the company;

their name is "LumiLEDs" (which I've seen either with the ED
capitalized or not; I prefer to capitalize it, since these ARE
LEDs we're talking about), a joint venture of Philips and
Agilent Technologies. "Luxeon" is a brand name applied to a
wide variety of LED products they've developed for illumination
applications - some of which are UV/phosphor types, and others
are direct emitters.

Please provide a link to a data sheet on a direct emitter sold by this
company for illumination purposes. And if you really care how someone else
capitalizes the name someone needs to buy you a life.

You might also, were you inclined to actually look at some of
this information rather than simply argue about it, want to review
the papers presented by both LumiLEDs and NEC-Mitsubishi
at the Society for Information Display symposium this year.
The LumiLEDs paper (41.3) goes over their plans for LCD
backlighting in general, including the THREE different forms of
such backlighting they've proposed (basically, these are for
small-size, medium, and large-area LCDs). The NEC-Mitusbishi
paper (41.4)

I'm sorry, but your numbers there have no meaning for me.

concerns their prototype 23" LCD monitor which
was constructed using the LumiLEDs backlight system. Both
papers go into considerable detail on the color gamut advantages
such backlighting provides (the NEC-Mits prototype, for instance,
provides a gamut of over 101% of the 1953 NTSC spec, which
is well beyond any current LCD or CRT television). Both papers
may be found in this year's SID Symposium Digest; I don't know
if they're on-line anywhere or not.

Since you still seem to belive that the typical LED IS truly a
monochromatic light source, I will also refer you to the following;
look for the words "spectral width" and "chromatic dispersion"
here to find the relevant information:

http://www.fiber-optics.info/articles/LEDs.htm#Table_2

http://www.mtmi.vu.lt/pfk/funkc_dariniai/diod/led.htm

Which point out that a laser is not a "truly monochromatic light source"
either. All that differs is the degree to which they differ from the idea.
Both are monochromatic for many purposes but not for others.

http://zone.ni.com/devzone/conceptd.nsf/webmain/90B8D6058CB7C18386256C6A00598C96

In theory, the LED WOULD be a monochromatic light source if
all you had to worry about was the emission resulting from a
transition of an electron over a single, stable band-gap energy.
In practice, of course, you get something a bit different, and so the
spectral width (as measured at the half-power point) of even the
non-phosphor, direct emitters is typically on the order of 50-100 nm
or more (i.e., a fair fraction of the visible spectrum), vs. a width of
1 nm or less for a solid-state laser source. Again, if you want to
get more into the whys of this, the above are a good place to start.

I'm sorry, but I got all of this I wanted in my solid state physics courses.

Since, though, you are apparently more interested in simply arguing
rather the reading the material, I think this is a good time to close
this discussion. There's certainly been more than enough information
and pointers to information provided for anyone wishing to learn
more about the subject.

Yeah, right, whatever. If you want to take your ball and go home be my
guest. Personally I find that you talk a good game until you're pressed to
support your arguments, then you come up with papers that are either
inaccessible or say something other than what you claim. Perhaps you
should follow your own advice and actually read the papers whose titles you
toss about.

 

[Not Gimpy Anymore]

instead

Nor did I say they were; but toxic compounds are
not as much of a concern IF they are, and reliably will
be, kept away from possible leaching into groundwater,
etc., whether it's through physical containment (i.e.,
the material in question is reliably sealed away) or
through being chemically bound so tightly that it's
extremely unlikely that any hazard will exist. That's
why the lead in CRTs, while it hasn't been an
ignorable hazard, has not been the concern that one
might initially think. It's contained within the glass, and
is about as likely to be a health hazard under typical
conditions as drinking from your grandmother's prized
lead crystal goblets would be.

Similarly, the arsenic that was mentioned as being a
concern for LEDs is both chemically bound up, and
very unlikely to leach out just for that reason, plus
it's embedded in a metal and epoxy package. The
mercury in CCFLs (or other fluorescents, for that
matter), is a LOT more likely to get out into the
environment and cause problems. The mercury content
in backlight tubes, specifically, has been rather strictly
limited by various regulatory agencies (the limits imposed
by the European Union, under directive 2002/95/EC,
are the most commonly referenced). However, no
requirement for mercury-free backlighting has been
imposed yet, because at this time the various national
and international agencies are more interested in
encouraging the move the LCD monitors, for energy-
efficiency reasons. And it's understood that the
mercury-free alternatives simply aren't economically
feasible at this time. Once solid-state backlights are
more cost-competitive, that situation may change.

Bob M.

I stand with you on this Bob, but unfortunately all
evidence to date shows that EU and EPA have NO
comprehension of "tightly bound".....so we consumers
will end up PAYING money due to their chosen ignorance.
Wonder how many Europeans are replacing their precious
leaded crystal with "unleaded crystal".....

Now that I got that off my chest (not the first time mentioned,
but first time in this group)...... back to work....

NGA

 

[Bob Myers]

:[email protected]...

I stand with you on this Bob, but unfortunately all
evidence to date shows that EU and EPA have NO
comprehension of "tightly bound".....so we consumers
will end up PAYING money due to their chosen ignorance.
Wonder how many Europeans are replacing their precious
leaded crystal with "unleaded crystal".....

True. Gawd knows there is little to no overlap between
"the way things really are" and "the way regulations make it
look like they are"....

Bob M.

 

[J. Clarke]

A high-end CRT gets fuzzy and dim anfter a couple of years.

How does a good LCD change with age and what is the limiting factor
for life ?

Does turning an LCD on and off frequently shorten the life. I have in
mind a laptop that has power saving set to trun the screen off after
15 minutes of activity.

For an LCD the limiting factor is generally the backlight.