LED lighting

[Don Klipstein]

I thought Daniel Stern told me a few years ago that ITE calls them yellow and not amber...

Meanwhile, all the red and yellow ones surely look in spec to me for color, whether LED or incandescent. Maybe a few yellow LED ones are barely out of spec in the direction of orange in hotter conditions mainly in unusual conditions such as "on flash" or stuck on yellow... Easy enough now to avoid now that at least some LED manufacturers are putting yellow LEDs into color ranks and one can order a color rank that stays in spec if it goes a bit orangish from baking in Arizona sunshine in July. Then again, yellow LED traffic signals are not used much yet because it is harder (more expensive) to meet the specification for brightness and the duty cycle is so low that energy savings from using LEDs is not much.

Some green ones appear to me off spec. Some LED green ones appear to me too pure green, and I see a fair number of incandescent green ones that appear too lime green. But it is now easy enough getting "blue-green" or "traffic signal green" LEDs whose color is in spec.

Now I see a draft proposed standard by ITE for LED traffic signals:

Proposed color requirements there are different from the existing ones. Fairly to really pure reds with dominant wavelength longer than about 632 nm would not meet this one despite (at least as far as I thought I knew) meeting the current spec. The limit towards orange is roughly 620 nm. And I have seen enough red traffic signals, both incandescent and LED, that are more pure red than "HeNe laser red" is. The new proposed one for LED traffic signal green appears to bar shades of green with dominant wavelength over about 505 nm, and the old one now looks to me to allow whiter but not much longer dominant wavelength unless whiter than traffic-signal-green LEDs. I see a few green LED traffic signals that appear to me to have dominant wavelength around 510 nm, and I see quite a few incandescent green traffic signals that appear to me to have dominant wavelength around 520 nm and a few incandescent ones at least 525 and appearing a bit too "lime green" (as opposed to whitish) to meet the spec...

has the CIE chromaticity diagram with the existing ITE specification color regions shown, but lacking wavelength as well as X/Y notations.

---------------------

I have tried making the color of yellow and traffic-signal-bluish green LEDs vary with temperature, and they don't do so much unless the temperature gets truly unreasonably high.

Eventually... And some drivers are waiting for some of those green ones to fade!

I have yet to see any LED ones fade from adequately bright to inadequately bright without being replaced at least as hastily as burnt out incandescents are replaced. Although I have seen a small number of LED red ones that had partial failure to the point of being non-uniform to an extent that is out of spec while being definitely adequately visible as "a red light", and seen them take longer to be replaced than burnt out incandescents, and then only in one municipality that has been mentioned a bit as "the city that doesn't work"... Furthermore, most of those were ones that I suspect to be pilot program test units - the odd partial failures were on a specific street or two where I noticed a pilot program of some sort apparently for testing LED red traffic signals. And given how much electricity cost they saved the city (along with probably one relamping halfway through that incandescents would probably require, with labor cost at big USA-rustbelt city union rates) over the roughly 5 years they lasted, I would think even these flunkers were better than incandescents!

- Don Klipstein ( [e-mail])

[Daniel J. Stern]

H'mm. That's certainly a start...

OK, that's definitely going to help...IF manufacturers implement it/municipalities specify it.

It will be interesting to see what kind of influence this exerts on the design of the lamp heads. It may tend to reinforce the current North American practice of using a matrix of many emitters, it may clear the way for the highly optical Swedish(?) units with only *TWO* emitters per lamp head, or it may exert no influence at all.

H'mmm. I can't see it happening too quickly, I'm afraid. Not with the large number of recently-installed non-spec units.

DS

[Daniel J. Stern]

Don't think it was I. I am not very familiar with ITE's lexicon. SAE calls turn signals "yellow", while ECE member countries squabble over whether it should be "Jaune auto" ("automotive yellow") or "Ambre", or the equivalent terms in English.

I keep hearing that yellow traffic signals don't "yet" meet standards, but haven't yet heard a satisfactory explication of the noncompliance, and I do see yellow wafers going up. Slower than reds and greens, 'cause the payback time is longer, but I *do* see them.

And, as I've mentioned here before, Ontario uses 2-color LED heads for the schlock green arrow that then changes to a yellow arrow (same lamphead, same arrow, just a color change).

DS

[Owain]

"Clive Mitchell" wrote | >Using LEDs you could have a little animated car chugging | >on the green. | >Would be great for pedestrian crossings -- a stationary | >red man could change to an animated walking green man. | Likewise you could replace the red "don't walk" man with | the green "walk" man being hit by a yellow car with | animated LED blood spray.

You've taken that too far you have.

Owain

[Owain]

"Clive Mitchell" wrote | Yes here in Glasgow they use a generic LED illumination | panel and just mask off the unwanted LEDs.

So they wouldn't notice if the masked-off LEDs went missing :-)

| Seems a bit wasteful to me,

Glesga Cooncil wasteful, neffer!

Owain

[Mike]

Wasn't that a "Not the nine o'clock News" sketch ? (no LEDs of course)

[Simon Waldman]

Do you know whether the UK's portable roadworks LED traffic lights use the same dimming technology? I have a couple of sets near to me at present, and I find all three colours unpleasently bright when driving at night.

THen again, I tend to find the LED brake lights on modern cars rather dazzleing as well... maybe it's just me...

[Simon Waldman]

boh!

[Steve Firth]

[LED]

Just like those crappy EL fluorescent lights you mean?

[Dave Plowman (News)]

Yup. I have none indoors, and am unlikely to ever have.

[Clive Mitchell]

In message , Steve Firth writes

EL was never suited to internal illumination. Remember how we were all told that our houses would use EL sheets as wallpaper and the whole house could be lit up for virtually nothing?

They didn't mention the terrible intensity fall-off when the EL material is first used, and the tendency for any moisture penetrating the outer membrane to cause dark spots. (And moisture DOES penetrate through plastic.)

EL certainly has it's uses though. Like....... Ummmmmmmm.....

Flashy disco badges. Short life LCD backlights. Groovily expensive illuminated bill boards. Shitty EL wire for boy racer car lighting. Etc.

I wonder how many of the reliability issues have been fixed for the forthcoming printed matrix "OLED" (Organic Light Emitting Diode) displays. If they haven't fixed the fall-off issue then it's just as well the displays can be rolled up. It'll make them much easier to throw in the bin.

[Clive Mitchell]

In message , Simon Waldman writes

Have you tried squinting?

[Daniel J. Stern]

It's an excellent illustration of the variability in discomfort glare thresholds. There is much, much less glare in the UK (European) nighttime roadway environment than there is in the North American. UK headlamps typically produce less than half the dazzle to oncoming traffic. UK traffic lights are less intense. UK brake lights are permitted to be only half as intense as US brake lights.

And yet there are still complaints of discomfort glare.

[Tim Mitchell]

The temporary ones I have seen do not dim, they just blast you away at night.

Another interesting thing, the 2 ends appear to be linked by radio. I would like to know how they have made this work reliably and safely.

[Clive Mitchell]

In message , Tim Mitchell writes

The more important question is whether you can make a device that turns them both green at the touch of a button. :)

[Zak]

Well you can make it fail-safe... turn red or flashing yellow when the other side is dead.

Or probably even better switch to a timed mode with ever longer guard intervals. Given a 1 second per day maximum deviation each day the 'red' could be extended by a second and the 'green' shortened by a second... obviusly keep the cycle time constant, otherwise funny things happen if one side misses the other but the otehr doesn't miss the first for some reason.

Thomas

[Mike]

Who said they were reliable ? :-)

Or safe ? :-(

[Andrew Chesters]

Or even better, to make "yours" go green as you approach...

(and I'll bet Clive gives it a go!)

[Mike]

You could with the original Irish ones - two cars with 'Ambulance' overrides (common fitment in Eire) arriving at EXACTLY the same time did just that.