Led Traffic Lights Dying Already.

[Stingray]

[ QUOTE ]
Saaby said:
Saab uses a system on their engines called Direct Ignition. Basically a mini-coil over each plug. The advantage is less moving parts, less parts to wear out period, better control over the firing process, and thus better emissions.

1 big cartridge, 4 plugs, all epoxied in. Not a terrible product, but more failure prone than it should be. The suspected cause? Heat.

Striking similarity to these LED modules, supposed to be simpler and cheaper but it's not as perfect as one would think. When they figure out how to keep the LEDs happy somebody pass the knowledge onto Saab would ya?

[/ QUOTE ]

Those types of systems have been on various cars for the last decade or two. My brother had a Nissan, I think it was a 300Z, with individual coils quite a while back, he didn't have any problems as far as I can remember.

 

[PhotonWrangler]

[ QUOTE ]
PlayboyJoeShmoe said:
Lots of LED traffic signals around Houston have a few to several individual red or green LEDs out. Either I don't see a lot of yellow LEDs or none have failed.

[/ QUOTE ]

Interesting. I wonder if this has something to do with the relatively short on-time for the yellow light?

 

[3rd_shift]

In Dallas Tx. the amber parts of the lights are still incandescant (almost never on, why bother converting those?), just the red and green are leds.

 

[jtr1962]

Same thing in NYC. They left the amber lights incandescent.

I second what everyone else said about overdesigning traffic signals. It's a mission critical application, and it costs big bucks every time you need to call a cherry picker to change the lights. No reason these should be failing so soon.

 

[Sigman]

Our crossing signal lights seem to be holding up (except when someone shoots one...and even then we just lose a few LEDs and 90% or more of the LEDs still work. It's the hole in the housing that concerns us, allowing moisture & dirt in...holes can be plugged though!). Additionally, though the LEDs are red, we still have a plastic lens panel (to make the LEDs more visible over a wider angle) and another red plastic cover (for additional weather proofing & red awareness) that tends to shatter when shot. It's more cost effective to replace the whole assembly in the long run. I think they could be designed differently.

Seems the failures we get on installed LED lamps are the ones on our crossing gates. They aren't failing on their own...it's when a vehicle runs into one. The "weak link" seems to be the sockets, conventional in design - the internal soldered wire seems to be breaking when impacted.

Summing it up...we don't seem to be having the problems that the traffic signals seem to be as described here. I've got a friend that works for the city's traffic signal department as well. I'll give him a call and see what he says. Of course, here in Alaska...we aren't having heat much over the mid 80's (and believe me - THAT'S TOO HOT FOR ME!!)

 

[Saaby]

[ QUOTE ]
Stingray said:
[ QUOTE ]
Saaby said:
Saab uses a system on their engines called Direct Ignition...

[/ QUOTE ]

Those types of systems have been on various cars for the last decade or two...

[/ QUOTE ]


Ok so I stand corrected. I've got to hire a better fact checker. Last time today I act like I know anything, I swear!

Anyway I still stand by my original conjecture that heat is a major factor here, either for the LEDs themselves, or the electronics that drive the LEDs.

 

[iddibhai]

that's what CPF needs, a Fact Checker. I like the sound of that..

back to regularly scheduled programming. none of the LED traffic lights in my city show signs of stress, been running couple years.

 

[James S]

They were just putting in these in the signals in Richmond when I was living there, so I didn't get to see them over a long period of time. They were really blinding at night though, considerably brighter than a regular bulb one.

Here they had them (even for amber and arrow signals) almost everywhere when we moved in a year ago, so they are over a year old now at least, possibly 2 or 3. I have not seen a single dead LED or any other obvious problems with any that I've driven under. And I've been looking closely the last few days since noticing this thread /ubbthreads/images/graemlins/smile.gif

It's possible that a manufacturing defect is messing with a whole batch of them somewhere, but more likely it's like everything else, you get what you pay for. There are cheap ones, and there are long lived ones and any information on the overlapping of the 2 categories came from the marketing department /ubbthreads/images/graemlins/grin.gif

 

[ledaholic]

They are over-engineered. If I can find the control board out of one, I'll post some pics. There is a governing body for traffic signals, local governments can't just throw in some kind of lamp, they have to meet pretty tight requirements. LED lamp assembly's are still fairly new to traffic lights. In time they will get better or will be replaced with something else.

 

[gadget_lover]

Several posts stated that heat would only cause LEDs to dim. Unfortunately, this is not true.

There are three things that bother semiconductors; heat, incorrect voltage and vibration. The advent of potted electronics (ICs, LEDs, etc) has made vibration much less of a problem, but the other two are still killers.

When a semiconductor heats up it's electrical properties change. That property is actually one of the definitions of a semi-conductor according to a 1970's textbook. If the parts' characteristics change too much the current will excced the design's capacity and the part will melt internally.

Overvoltage will usually cause death by pushing too much current. See above Way over voltage will cause internal arcing, again death by melting.

Heat builds up quickly in an enclosed container. If you think that there's not much heat from a "few watts" used in the traffic light you can do your own experiment. Drop an arc AA or other single LED light in a thermos bottle, then seal it and watch TV for 1/2 hour. I'm pretty sure it will be uncomfortably hot by then.

Of course, I could be wrong /ubbthreads/images/graemlins/smile.gif

Daniel

 

[elgarak]

Well, I referred to the error mode I observe. I see the following: An array of 5(let's say) 0 LEDs. 25 on the left side do not give off any light (when observed during daylight), the right half is very bright. Ageing the LEDs with overheating would not result in this error. One should see varying brightness levels over the array. If one side of the array is hotter than the other side, one should see gradually changing brightness from one side to the other. Not the sharp edge of LEDs on and off I observe.

 

[IsaacHayes]

I belive it probably is poor circuit design. And there are lots of manufactors out there. I know I've shipped some led lights back for D.O.T. that I assume were broke. Also it seems they replaced or upgraded to a newer model LED lights as they don't have the same ones anymore in the lights. Some even have diffusers so it doesn't blind you at night. Some may also be luxeon behind there, as it seems to be even more diffused and not a pin point source, also a slightly different color than the other leds...

We also have Yellow LED ones now...


One thing though, in winter, snow doesn't melt off of the LED's, making it hard to see the lights! Anyone eles notice this?

 

[The_LED_Museum]

As I understand, some LED traffic light fixtures have heaters to melt the snow. But not all of them do; that's the problem.

 

[ledaholic]

I found the controller-powersupply board for a 12in red led traffic lamp assembly. I have a picture if anybody is interested. If someone will post it, I'll email it.
Bob

 

[3rd_shift]

Better yet, go here;
www.photobucket.com
Once your pic is uploaded there,
(it's easier than I'm making it look /ubbthreads/images/graemlins/tongue.gif ) ,
then highlight the whole url to it, right click copy, then right click paste it here on cpf . /ubbthreads/images/graemlins/smile.gif
We then can see it. /ubbthreads/images/graemlins/ooo.gif
Done and done. /ubbthreads/images/graemlins/grin.gif /ubbthreads/images/graemlins/clap.gif
Anyways, on with the show. /ubbthreads/images/graemlins/wink2.gif
Edit: Oh yeah, if you need to resize the pic down to 640x480, go here for a good free photo editor called irfanview. /ubbthreads/images/graemlins/wink.gif
edit again: too much hassle to learn in one day? /ubbthreads/images/graemlins/banghead.gif
send the photo to [email protected]
I'll fix it up real good for ya pardner. /ubbthreads/images/graemlins/grinser2.gif

 

[NewBie]

[ QUOTE ]
PhotonWrangler said:
LED traffic signals

[/ QUOTE ]


Yeah, LumiLEDs is also one of the major players in traffic light signals.

http://www.dialight.com/contents.tpl?page=contents&sub=contents&sku=19

http://www.dialight.com/subs.tpl?page=transportation&cat=Traffic

Here is a study on LED traffic signals:
http://cogprints.ecs.soton.ac.uk/archive/00002486/01/led.pdf

I spoke to one of the presenters at a LumiLEDs (their luxeon and their superflux) conference about failures in LEDs in the field, and he actually used traffic signals as an example. For failures they had seen, there were two big issues. One of the major factors that he indicated was corrosion in the electronics from moisture intrusion, water intrusion, exposure to humidity, and poor cleaning after assembly (folks had used no-clean flux (which is a whole issue in and of itself)), some had used nothing but bare boards and did nothing to protect the electroncs (such as potting or conformal coating).

He indicated the heat failures seemed to be due to a mis-understanding of the environment where they were used (imagine how hot the fixture that is painted flat black gets in the sun), and poor understanding of removing the heat in the first place.

Here is a report from various cities and their comments of their experiences with them:
http://www.dot.wisconsin.gov/library/research/docs/tsrs/tsrledsintrafficsignals.pdf

 

[ViReN]

I think, Instead of potting... If they use Silicone Rubber RTV Sealent to seal off the parts (including the LED's in to a single block... will solve most of the problems ... and as far as the heat sinking is conserned... The Heat Sink Itself can be Outside... and other parts sealed IN...

If there is a need to rework... removing RTV isn't difficult... moreover... RTV Sealent is almost transperent... (though i dont know effects of Sunlight tht might be falling on it... (does it get yellow ??)

-ViReN

 

[NewBie]

During one of the LumiLEDs hands-on thermal labs, they had a traffic light some vendor was making from luxeons, where the outer cone had a molded in sheet of metal to dissipate the heat.

Viren, humm, RTV is potting (it can also be conformal coating.)

 

[ledaholic]

3rd shift,
Photobucket said I had to wait till later to sign up so I put it here http://tinypic.com/2ujiu . Just in case it doesn't work, I'll send it to you also. The lamp assy I took this controller from was a 12 inch red light with 177 5mm bright red LED's. There was no heat sinking at all inside the assy. The LED's were mounted directly to the pcb the rest of the components (transistors to turn on the individual strings of LED's)were on. I don't have the pcb anymore, but if I come up with another one, I'll post a pic

Bob

 

[3rd_shift]

Here's the pic of that circuit board powersupply.

[image]http://tinypic.com/2ujiu[/image] Rats, it didn't work from that url. /ubbthreads/images/graemlins/icon3.gif