Please create an account to participate in the Slashdot moderation system


Forgot your password?
For the out-of-band Slashdot experience (mostly headlines), follow us on Twitter, or Facebook. ×
Technology News

New York City Street Lights To Go LED 303 303

eldavojohn writes "Wired has a short piece on NYC's new street light project. I don't think we need to belabor the many benefits that LEDs hold over traditional light bulbs, but the finishing touches are being addressed, and they will hopefully be put into place sometime next year. This design won a competition back in 2004, and OVI has been whittling down the prototypes. At $1.175 million, this sounds like a pretty cheap deal considering the DOE forked over $21 million to 13 R&D projects along the same lines."
This discussion has been archived. No new comments can be posted.

New York City Street Lights To Go LED

Comments Filter:
  • flicker crashes (Score:5, Interesting)

    by goombah99 (560566) on Saturday December 20, 2008 @12:20AM (#26181331)

    The thing that is awful about led lamps is that most of them are run straight off the AC voltage and have massive 100% brightness flickers. If you are moving it's like a strobe. You don't see it in car lights since they are run off DC. but most, perhaps not all, AC socket lamps I've seen have really bad flicker.

    I also how they have secondary lenses since LED's can be very directional the way they are typically resin cast.

  • by pwizard2 (920421) on Saturday December 20, 2008 @12:23AM (#26181345)
    This is something I've been wondering about for awhile. LEDs (especially the white ones) are really bright for being so small, and they don't have that yellow tint that incandescent bulbs do. Compact florescent bulbs are nice, but they aren't perfect for every situation. I'm not an expert on the subject, but I've always wondered why they don't make giant LEDs that can replace ordinary light bulbs. It seems like 220 AC would be more than enough to power them.
  • Re:flicker crashes (Score:5, Interesting)

    by digitalunity (19107) <`digitalunity' `at' `'> on Saturday December 20, 2008 @12:30AM (#26181401) Homepage

    Assuming the line voltage is run through a full wave bridge rectifier, there would be a 120 Hz flicker, imperceptible to most people. Toss a large capacitor across that DC output and you've got dramatically less ripple.

    Your directionality comment is apropos. It's also worth noting that some people don't like the light spectrum output on white LED's. Personally, I prefer the pink tint from high pressure sodium lamps.

  • by Anonymous Coward on Saturday December 20, 2008 @12:36AM (#26181443)

    That's not what he means. The replacement 120v led lamps are a collection of a bunch of little white led's. Why can't they make a single led the size of a lightbulb instead of 100 small led's.

    Is it possible to make a single, huge led?

  • by dgatwood (11270) on Saturday December 20, 2008 @01:06AM (#26181609) Homepage Journal

    Yeah. I've noticed that. What I don't get is why they choose to set the color temperature that way. Red LEDs are extremely cheap compared with producing light at the other end of the spectrum. Why in the world would they balance them towards the blue (expensive) end of the spectrum when that is both more expensive and visually unpleasant? About the only thing I can imagine about the current LED designs is that they were designed to be used in combination with standard incandescent bulbs. If you blend the two, you should get a fairly nice looking light spectrum, albeit probably a bit heavy in the yellows....

    I'd buy LED lights instantly if they actually used three emitters. Unfortunately, most don't. They use two---one yellow, one blue. Because the yellow LED has a relatively narrow light spectrum compared with an incandescent, you end up with basically no light output down near the bottom of the visual spectrum. The result is light that is downright unpleasant to deal with in every way. The bluish light makes it hard to see color accurately, makes colors not reproduce well in photography or video, and really isn't good for you mood-wise. Basically, the current crop of LED lights have all the problems of CFLs except the mercury (well, and the LEDs should last a lot longer, I believe).

    The question, then, becomes this: "When are we going to see properly designed white LED bulbs?"

    On the other hand, while they suck for homes, the existing LED lights are perfect for street lights. First, there was one experiment [] that suggests that suicides and crime may decrease when street lights are replaced with bluish lighting. Second, the color temperature of blue LEDs are virtually indistinguishable from the mercury vapor lights (~6000K) that are already used in a lot of places.

  • by cathector (972646) on Saturday December 20, 2008 @01:10AM (#26181639)

    > and they break when they get too hot.

    actually, high-power LEDs such as Philips's Luxeon series are quite robust in the face of surprising amounts of heat. I've run enough current through them so that they melted their soldering several times, and while its true their efficiency declines with heat, they suffered no permanent damage. When you put an amp and half through one of those suckers, they're literally stunningly bright.

  • by Anonymous Coward on Saturday December 20, 2008 @01:17AM (#26181683)

    Here in Portland, OR, we have already started to use LED street lights. And now that we're in a snow storm, these lights aren't working. LEDs don't produce heat (that's why they're efficient). By not producing heat, they don't melt the snow away from them. So all the LED streetlights in Portland are covered in snow and cannot be seen.

    The old lights produce enough heat to melt all the snow. Snow in Portland is rare, so it's not that big of a deal. In NY, it's quite the opposite.

  • by hardburn (141468) <<ten.evac-supmuw> <ta> <nrubdrah>> on Saturday December 20, 2008 @01:17AM (#26181685)

    NYC is a lost cause as far as astronomy is concerned, but I have hope that smaller cities and towns will see this and adopt it. LEDs are inheirently directional, whereas most fixtures tend to waste a lot of their light going out and up. So LEDs should be a win for astronomy.

  • by adamjaskie (310474) on Saturday December 20, 2008 @01:27AM (#26181737) Homepage
    LPS/SOX is better, really; the spectrum of LEDs is pretty intrusive to observations. LPS/SOX is also more efficient IIRC, but the bulbs don't last anywhere near as long.
  • Sodium Vapor vs LED (Score:3, Interesting)

    by lobiusmoop (305328) on Saturday December 20, 2008 @02:21AM (#26182015) Homepage

    Personally I'd miss sodium vapor street lights if LED replacements became fashionable. Perhaps it is a romantic notion, but it seems to be that one of the reasons sodium lamps have become so popular is that the orange light they emit is reminiscent of fire, and in colder northern climates their warm glow is comforting to people at some deep instinctual level.

  • by B4RSK (626870) on Saturday December 20, 2008 @02:32AM (#26182051)

    We've had LED signals here in Osaka for 5+ years how and they work very well. Here are some links (in Japanese) with photos showing what they look like:

    Red Light, Green Arrow []

    Pedestrian Crossing []

    Green, Amber, Red [] (the amber is actually brighter than it seems in this photo)

    I haven't experienced any problems with them and I drive daily here. There is no noticeable flicker and they are a lot brighter than the traditional signals they replaced.

  • by Sycraft-fu (314770) on Saturday December 20, 2008 @04:35AM (#26182507)

    Then it is psychosomatic, not real. Cars are DC systems, not AC. It would be rather stupid to go through the trouble to take that DC and convert it in to AC just for the tail lights.

    This doesn't surprise me, as I've found a number of the "I can see flicker," people have it mostly in their head. A former coworker had a wife like that. I've no doubt she was more sensitive than the average person, but most of her problems were in her head. She complained she couldn't stay in our office long because of the flicker of the lights... Except I checked, our overheads were powered by electronic ballasts that operated in the 30kHz range. So she wasn't seeing flicker, she was seeing florescents and assuming they were flickering.

    At any rate automobiles are DC powered. Check one with a multimetre if you don't believe me. Thus they are not going to be pulsing their lights.

  • by serviscope_minor (664417) on Saturday December 20, 2008 @04:44AM (#26182517) Journal

    "We don't need to belabour the advantages of LEDs over traditional lightbulbs"?

    Actually, we do, since we've had lightbulbs other than incandecent for over a decade, and incandecents are never used to light streets. LEDs manage about 100 lumens per watt, similar to high pressure sodium lamps. The old orange low-pressure sodium lamps are still king of the hill at 200 lumens per watt.

    So what were those advantages again? Compared to high-pressure sodium lsmps, they're the same efficiency and lifetime, but a lot more expensive. The only advantage to low pressure lamps is colour, but they loose a factor of 2 on efficiency.

  • by Threni (635302) on Saturday December 20, 2008 @06:01AM (#26182671)

    > Monochromatic lighting is somewhat dangerous since details don't stand out a

    You don't need details - you're just trying to avoid bumping into things when you're moving fast. It doesn't matter what the colour of the thing you're trying to avoid bumping into is, and in the event of crime/accidents, people don't remember what colour things are anyway.

  • Re:flicker crashes (Score:3, Interesting)

    by Gordonjcp (186804) on Saturday December 20, 2008 @07:26AM (#26182943) Homepage

    What you say is of course obvious to any EE, and yet i've never actually seen a single 120v LED lamp made that way. One wonders why.

    I actually have a broken LED GU10 lamp on my bench at the moment. Inside is a small bridge rectifier, a smoothing capacitor and a tiny chopper PSU. The LED string is run at around 50D. The bridge rectifier and capacitor provide around 340V (240V mains) with about 10% ripple, but the chopper compensates for this and provides a 60kHz pulse-width modulated supply which is then smoothed and fed to the LEDs. With the smoothing cap off the output of the chopper you can see the PWM output modulated at 50Hz, to compensate for the ripple on the input. With the cap in place there is no measurable ripple on the 50V supply.

    This is just an el-cheapo one (five quid from Screwfix, kind of thing). I would have thought that more expensive LED lamps would have something similar if not better.

  • Re:flicker crashes (Score:4, Interesting)

    by Tokerat (150341) on Saturday December 20, 2008 @08:10AM (#26183113) Journal
    1. Find a car with LED tail lights.
    2. Look at them as the brake is applied and they illuminate.
    3. "Shake" your eyes left and right rapidly

    You will observe, instead of a smooth trail of light (referred to on film as a "motion blur") like you normally would see, you can see individual "sets" of lights; very broken partial light trails. This effect is exaggerated enough to be realized in this case by the constant movement of your eyes.

    Every set of LED tail lights I've ever seen could produce this effect. How can this be? Apparently, I'm not alone in my experiences, either. I'm certainly not one of those nut jobs who says WiFi makes them nauseous, hell I can't tell teh difference between 320kbps MP3 and WAV.

    I can, however, hear a high pitched whine from an old CRT with no signal, 60Hz monitor refresh gives me a headache, and LED tail lights leave a strobe pattern instead of a smooth trail.

    I believe this is dangerous as it can make determining the point of origin for such types of lights difficult when split-second instinctive brain functions take over. Instead of a line leading to the current position of the tail light, in that slit second my mind has 3 "still frame" snapshots to piece the scene together with - not quite enough information.

  • Re:flicker crashes (Score:4, Interesting)

    by paul248 (536459) on Saturday December 20, 2008 @08:40AM (#26183193) Homepage

    I've seen a definite flicker in some cars' tail lights as well, so you're not crazy.

    I think the cars pulse the lights at ~50% most of the time, then switch to 100% when the brakes are applied.

  • Re:flicker crashes (Score:4, Interesting)

    by gfilion (80497) on Saturday December 20, 2008 @09:07AM (#26183265) Homepage

    In SoCal, we have LED traffic lights everywhere.

    In Québec we have them pretty much everywhere too. Sometimes when the snow is just the right consistency and falls in the right direction, it sticks to the traffic lights lenses, obscuring them. In the past, the heat generated by the incandescent bulbs would melt the snow, but the new LED lights don't produce enough heat. A city worker has to remove the snow with a kind of small broom attached to a long pole.

    Two steps forward, one step back...

  • Re:flicker crashes (Score:5, Interesting)

    by zenyu (248067) on Saturday December 20, 2008 @11:00AM (#26183705)

    *sigh* You remind me of the EE graduate student who was showing me and my classmates around his lab one day. To prove the point that humans couldn't hear a 16kHz tone very well he quickly turned the power up from 1 to 11, and through pain practically paralyzed the half of the class that still had their hearing intact.

    Many cars have the flickering tail LED lights, and it has nothing to do with DC-DC converters or other sources of ripple in the supply current. It's simply a matter of the duty cycle timer, the tail lights are "dimmed" not by limiting the current but by turning them completely on and off at a low frequency. The ones I've seen are in the 40-80Hz range, just stick an oscilloscope on there if you don't believe me. The flicker stops when the lights go to full illumination (i.e. the break petal is depressed).

    Now go outside and look at some LED tail lights! Even if you have very poor vision you should be able to see the ones flickering at 40-50 Hz.

    PS If you are legally blind, just don't comment on lighting. You are bound to suffer from foot in mouth at times; go take on those annoying audiophiles buying 1000 euro power chords.

  • by emmavl (202243) on Saturday December 20, 2008 @02:34PM (#26185119)

    A dutch company [] seems to have solved that problem (color recognition) by using a mix of mostly green and some red leds.
    They use green leds because the human eye is most sensitive to green light in the dark, so the green light gives the best visibility at night. But to enhance the color recognition (which is basically zero with the almost monochromatic green light from the leds) some red leds are added.
    Here [] you can find a nice presentation (with explanation) of the product.

  • by smellsofbikes (890263) on Saturday December 20, 2008 @09:22PM (#26187747) Journal

    White LED phosphors are blue-heavy with a yellow peak.
    People build them that way because they're cheap.
    Now that everyone's getting pissed because they look cheap, any LED module designer worth twenty cents is designing systems that have roughly 3 white to 2 red LED's to bring the spectrum down.
    However, every lighting designer I've talked to, when we suggest making multiple color LED fixtures, especially ones with adjustable color spectra, say "the customers *say* they want that but they won't pay for it." They say the public has almost no interest in LED lighting in general, and particularly not in premium color solutions: price drives lighting.

  • Re:flicker crashes (Score:2, Interesting)

    by Cougar_ (92354) on Saturday December 20, 2008 @11:11PM (#26188451) Homepage

    We've been using LEDs in traffic lights for many years in Australia. The whole system runs from 12v DC (it did so for years even with incandescent globes too), so flicker isn't a problem. The issue of differing brightness of each colour of LED is addressed by having differing amounts of LEDs in each colour light, depending on relative brightness. This works quite well, they all appear to be around the same brightness to the eye.

After an instrument has been assembled, extra components will be found on the bench.