Interesting Pacifica feature

, but slow-moving

Which can be done just as well with a rotating beacon using LEDs as can be done with an omnidirectional beacon using a xenon strobe....

Actually, its quite well understood, and the products have already penetrated the market heavily. LEDs have probably grabbed more than 50% of the commercial vehicle signal lighting market in recent years (Dan, got a number?), and- at least in my metro area of about 1 million- 100% of the traffic signal market.

Efficacy- how well does the lamp convey the intended information? It doesn't matter if it takes the whole Comanche Peak nuclear generating station to run the lamp, it can be very efficacious so long as it conveys the signal well.

Efficiency- how much input power does it take to achieve efficacy?

LEDs are much more efficient and COMPARABLY efficacious to incandescent lamps in most signalling applications. For some applications (eg. headlamps, stadium lighting) LEDs aren't efficacious enough yet. For others (traffic signals, turn signals, brake lamps, small hand-held flashlights) they are.

Efficacy is a measure of how well a source produces visible light. It is measured in lumens out per watt in. I don't understand how you measure "conveys the signal well."

Efficiency for any energy conversion device, including light sources, is simply power out divided by power in.

LEDs are much more efficacious than incandescent lamps ONLY when compared to filtered incandescent lamps. That is, when most of the light generated by the incandescent lamp is absorbed by a filter that transmits only a relatively narrow band of light, such as in almost all signaling applications. Since LEDs are naturally narrow band sources, they do not use absorptive filters and do not suffer from the filter loss.

When competing with white light sources, current generation LEDs have about the same efficacy as good incandescent lamps, since in this case the incandescent lamps are not hobbled by an absorptive filter. I agree that for very low lumen applications LEDs may have higher efficacy than some very low power incandescent lamps used in devices such as hand-held flashlights, but even this is not certain. I have seen efficacy data on flashlight bulbs that is equivalent to typical LEDs. However, there is no question that LEDs have longer life and are more rugged than these low power flashlight bulbs.

Have the (local) specifications for the traffic signals been altered to accommodate LEDs or are they, as I understand is not uncommon, just 'under test' ?

Within a small community, it may indeed have such a rigorous definition. So might the word "snorklefroodle," but who cares? I've got two degrees in electrical engineering, but have never been subjected to such a tight definition of "efficacy," and I suspect that no one in any of the newsgroups other than s.e.l. has either. In the general sense, efficacy is simply a measure of the capability to produce a desired effect. A signalling device that uses the entire power output of a Westinghouse pressurized water reactor to convey the signal "stop at this intersection" (or a bazooka to swat a mosquito) may be 100% effective, but horribly inefficient.

Efficiency, however, has a rigorous definition of power out divided by power in, and that applies to all audiences. Effectiveness is by no means strictly tied to efficiency in any sort of general sense.

And since we're talking specifically about signalling (subject is 'emergency lights'), name me an incanedescent lamp that can be used to produce red, blue, green, or amber light without filtering.

Yes, what's your point? If you had read the earlier post, you would have read that I said there aren't any viable LED beacons in the market yet. But that doesn't mean there won't be in the future. I just found this:

I've also seen several other prototypes from many other manufacturers.

You are absolutely correct. I should have said that the "efficacy of a LIGHT SOURCE is defined as .... " Since I am answering this question in sci.engr.lighting, I assumed, incorrectly, that this qualifier was not necessary. :-)

None, and that is the point of the discussion. The only way to get narrow band emission from an incandescent source is to use absorptive filters, which dramatically reduce the efficacy of the incandescent lamp. Under these conditions current generation LEDs have higher efficacy than filtered incandescent lamps, but, as I stated before, this advantage disappears in almost all cases when comparing white light sources because the filter is no longer used with the incandescent lamp.

The point is that if your in a smaller car following a Pukon, the light is so high that if your tailgating it's out of your field of vision.

Of course I'm not endorsing tailgating here, but since the CHMSL is supposed to prevent rear end collisions by morons who tailgate, putting it that high makes it pointless to have.

And if your not tailgating but your the second car behind, well I don't know about where you live but here there's a lot more SUV's and trucks and minivans than Caprices that would entirely block the CHMSL no matter where it was placed. You claimed one of the reasons for the thing was so that vehicles behind could see through the windshields of the ones in front of them.

Almost certainly any study that claims that the CHMSL is normally visible through the windshield and backglass of interceding cars must have been done in Europe somewhere. In the US at least half to 3/4 of vehicles on the road you can't see squat through.

It may not be intended to cover that position. But you can see other vehicles CHMSL's at that position. So once again why were the stylists allowed to move it to a position that makes it less safe than the competition? (I know, they wanted to produce an even uglier SUV than the competition) OK, so this example may be more one of deliberately not taking advantage of a "safety" feature that your mandated to have, (ie: stupidity) than any argument for or against the CHMSL. But it does illustrate the silliness of the regulations that they are so lax as to allow the Pukon to move the lamp to such a dumb place - it's definitely not in the axis of the drivers field of vision.

cost protection on a ten cent lamp in the back of the car? (in quantities of ten million or whatever that the automakers buy them at) Automakers spend more money on fancy logos that they plaster over the ass-end of the car. OK, as for the other reasons, I guess I can't argue that people that would believe those kinds of stupidity don't exist.

Good point, I wish the US was like this for a lot of things on vehicles.

Ted

Victor Roberts wrote: ...

I have not followed this entire discussion, but I won't let that stop me from commenting that The avowed point of LED stop lights is that it is FASTER AND therefore a more effective way of letting the person in the rear know that they should take action. - RM

The city has decided that they will no longer use ANY incandescent signal traffic signal lamps. All replacements are to be with LED arrays. That decision is probably 9 months old, and its already nearly impossible to find an incandescent traffic signal anymore. They were "under test" at selected intersections 5 years ago, and that is the basis for the wholesale conversion.

After 5 years the degradation will be getting significant. How do they measure/monitor them for specification or is it a time-planned automatic replacement.? Are they color controlled in any way?

Thats a lot more detail than I know about the program. My guess is that the city enineers buy the modules from some vendor, and they know that they don't have to replace them nearly as often as they do incandescents. Replacement probably is triggered more by percentage of elements in the array not working than by any degradation of each element's output or color.

[dire warnings of LED inadequacy, degradation, inefficiency, inapplicability and assorted other such piffle]

Look, Mr. Lewis, I donno why you're so rabidly against LED traffic and vehicle lights. They exist, they work well, and they save a lot of money in operating and maintenance costs compared to the incandescent and strobe devices they replace. While some of the problems you mention are indeed real technical issues with LEDs, they are addressed adequately in the units that are actually in service.

Virtually every locality that "tested" LED traffic signals went ahead and installed them right across the board, not even waiting for the incandescent ones to burn out. Yes, they really do cost *that* much less to run and maintain.

The only real remaining issue with LED traffic signals is that they can be excessively intense, causing glare after dark -- especially the green ones. This is handled in European regulations that call for traffic signals with a higher daytime and a lower nighttime intensity; I'm not aware of any such dual-level installations in North America. As the technology and the regulations evolve, manufacturers will probably feel less urge to use the most intense LEDs they can (which has in the past been done both for the "Look what we can do!" effect and probably to ensure the devices are still compliant after several years' service). Consider that the original red traffic signals contained 800 emitters and the latest ones contain 2 emitters(!). The technology is a great deal more advanced than you seem to be willing to grasp, for reasons unknown but to you.

Naw. They just use any ol' colour. They turn 'em all green for St. Patrick's day, and they're all pink in Castro and Dupont Circle. And in the Haight-Ashbury and Eugene they've got colour-changing ones what swirl around like a lava lamp.

DS (Never thought I'd say this, but where's whatsisface Lowrie with his Amazing Patent Pending LED Turnip Twaddler when you need him???)

It is unlikely that replacement is triggered by the percentage of LEDs in the array that fail to work. Or, if this is the method used, then it will lead to serious problems.

Unlike incandescent and discharge light sources, most LEDs will operate virtually forever, but with a continuously decreasing amount of light output. The old definition of "lamp life" as being the time required for 50% of a large sample of lamps to fail is not applicable to LEDs because of the different way in which they "fail".

The LED industry is moving toward a new definition of "life" for LEDs that is based on the time required to decrease to 50% of their initial light output. I believe the 50% cutoff gives an overly optimistic result for LED life, since incandescent sources and most discharge sources will have lost much less than 50% of their initial output by the time they fail.

Since good lighting systems are designed on the basis of mean lumens instead of initial lumens, using 50% loss of light as the definition of end of life for LEDs will further reduce the mean efficacy of LEDs when compared to existing white light sources.

ROFL. I want one of those swirling lava lamp traffic lights for my rec room.

I agree.

If you look at the LED manufacturer's graph for intensity over life, for a given LED (meaning individual LED die or commonly packaged array of dies), it is a series of curves with each curve being for a given drive current. Bottom line is that the intensity curve for a given end product that contains the LED's will be very much dependent on how hard the LED's are driven in the particular design (as percentages of their max. allowable ratings). If there was difficulty meeting some initial brightness spec. in the design phase (i.e., the technology was being pushed to the edge for the particular application), there is more likelihood that the engineer pushed it pretty hard, and degradation will be more rapid. As the technology matures (and it has already to a great extent), the designs can be less close to the edge, and degradation can be easily insignificant over many years.

Similarly, one manufacturer may gain an initial competitive advantage by driving them harder for higher initial brightness, but find that it's competition wins out in the long term if they (the competition) use more conservative designs (by using, say, more LED's and driving them less to achieve the same intensity - costs more, more parts - but lasts longer for lower overall long-term costs).

BTW (some LED trivia) - it was only within the last fifteen years that a blue LED was commercially available - that was a holy grail of the industry, and was quite a breakthru when it was achieved.

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

Much less. Nichia developed the first high brightness (1cd) blue LED in 1993. I don't have a date when they were commercially available but it must have been at least a couple of years after that.

Which as it turns out are Whelen LFL all-LED bars, if anyone wants to know. They do seem to have red and blue outer covers though, I'm not sure what that does to the color or brightness of the LED modules in question. Though from what I can estimate, based on the photos I've seen, probably not too much. (A blue filter tends not to block blue light after all, or red => red, etc.)

--Aardwolf.