Interesting Pacifica feature

although I would nitpick a bit of it:

As for instrument panels - try putting 1 mW into one of the better InGaN
green or blue LEDs. That's 300 microamps. Light output of a good InGaN
green at 300 microamps is about that of many cheap old-fashioned GaP green
LEDs at 15-20 mA.

- Don Klipstein ()

 

My bad, I was very unclear.

My experience is that if driven at normal levels, any loss of output
that would cause a noticeable reduction in the effectiveness of the
signal will take YEARS. So long that many of the LEDS in the array will
have begun to fail due to wire de-bonding and other failure modes.

 

More Detroit gimmickry. A useless "feature" (what else is new?).

 

I'm not sure what you mean by "normal" levels. Normal for what? Signal
level LEDs or high brightness LEDs that are in competition to replace
incandescent and fluorescent light sources for general lighting
applications? You cannot compare your experience with signal level
LEDs with the demonstrated performance degradation of high brightness
LEDs, which you may consider "overdriven".

And, let me preempt another discussion here. Sooner or later someone
is going to say that light output from the LED die itself does not
depreciate; all the depreciation in output is due to the package
becoming less transparent or the phosphor used in white LEDs
degrading. Research at Sandia and other places tends to support this
argument, but I want to point out that the same is true for most
"conventional" light sources.

There is no decrease over life in the visible output of high pressure
mercury, metal halide or high pressure sodium discharges. And there is
very little to no decrease in visible light output from an
incandescent filament over life. In all these light sources the
measured lumen depreciation is due entirely or almost entirely to
absorption of light the light generated by the basic "source" by the
arc tube or bulb. These sources are then like high brightness LEDs
that do not use phosphor.

A somewhat similar situation occurs in fluorescent lamps. The UV and
visible output of the low pressure mercury rare gas discharge remains
constant over the life of the lamp. All the depreciation in output is
due to degradation of the phosphor and loss of transparency of the
phosphor and the lamp bulb. You can then compare the degradation
problems of fluorescent lamps to white LEDs that use phosphor. They
suffer degradation due to both darkening of the material used to
encapsulate the LED plus degradation of the phosphor.

 

Come to think of it, I think I've seen this feature on some city busses, too...

 

Interesting, Don. Must be some recent development. At what voltage do
those InGaN's run with 300µA?

Bill Putney
(to reply by e-mail, replace the last letter of the alphabet in my
address with "x")

 

A shade under 3v if the ones round here are any guide.

More interesting is that you can have always on faint emergency lighting
using a modern white LED drawing just 10uA and with 2.5v forward voltage
(or 100uA and 2.8v). Enough light to see by when properly dark adapted.

Regards,

 

These have been around a couple years. Voltage at 300 uA is about 3
volts.

- Don Klipstein ()

 

Blue and blue-green ones work even better for use that requires dark
adaptation.

- Don Klipstein ()

 

White LEDs, which you keep bringing up, are a red herring. They
aren't used in traffic signals.

Incorrect in the case of metal halide. One failure mode which I've
seen has the output decrease very significantly, apparently due to
uneven redeposition of tungsten on the filament, causing thick spots
which don't glow as brightly.

In any case, all those lamps (as opposed to the arc discharges) do
decrease in output, and discharge lamps aren't suitable for traffic
signals anyway.

 

Agreed - white LEDs are not used in traffic signals. However this
discussion has shifted toward issues regarding lumen depreciation of
LEDs.

It seems you are talking about tungsten halogen incandescent lamps
instead of metal halide lamps - which are discharge lamps and do not
use a filament.

 

I have yet to see an LED for which a range of drive levels vs. lifetimes
was not available from the manufacturer. There would be absolutely no
need to drive traffic signal LED arrays at the ragged edge by any
stretch of the imagination.

 

Here in Sweden we use white LEDs for bus traffic signals
(and for Trams).
There are symbols and letters instead of colors (S for Stop).
(I'm a bus driver since 3 years)

 

Seems to me they're used in a great many pedestrian signals (for the
"WALK" part, or its "icon of a walking humanoid" pictogram equivalent)

DS

 

We have some crosswalk signals that use them. I've also heard of their
use in illuminated lane demarcation 'buttons'.

 

OK, OK and OK - Thanks for the correction.

I was just trying to be polite. I really don't know or care if white
LEDs are used in traffic signals as it is not germane to the
discussion

 

Ahh - the march of technology! 8^) Amazing developments.

Bill Putney
(to reply by e-mail, replace the last letter of the alphabet in my
address with "x")

 

If I understand a previous poster it appears that the color is not seen to
be too important in LED traffic signals (which are designed with an
increased unreliability to ensure that they fail catastrophically before any
degradation would take them out of specification).
I guess some of the Ambers will turn to Orangey-Red but I have never seen a
'white' used for any of RYG. .

 

I can't speak from intimate knowledge, but I would assume that the color
degradation and life curves get approximately equally compressed when
the LED's are overdriven. IOW, if you intentionally designed with that
philosophy, it would be wishful thinking to believe that color
degradation would not take place before the catastrophic failure. Yeah
- the color degradation phase of the LED's life would be shorter, but
it's complete life would be proportionally shortened (I'm guessing - it
may not be exactly a linear relationship, but certainly as life is
sacrificed, color accuracy would be also).

Bill Putney
(to reply by e-mail, replace the last letter of the alphabet in my
address with "x")

 

My comments were an attempt at sarcasm - I certainly hope so!