OT - Streetlight

A small distance from my house is a streetlight. The streetlight has been there as long as my house, about 8 years. In the last 8 years, the streetlight has behaved as proper streetlights do, coming on at dusk and going off at dawn.

All this changed about a month ago. The street- light started coming on at dawn, burning all day, then turning off at dusk and remaining off all night.

This event has shaken my faith in the embedded programming world. Is it possible that somewhere, a programmer wrote code for my streetlight that maintains a "nighttime" variable and then compliments its state on the dawn/dusk edge that the photo- detector sees? And never actually checks the reality of "day" or "night"? I can think of no other possible failure mode.

I'm considering an experiment consisting of shining a bright spotlight on the streetlight at night to see if that will trick it into proper operation.

Any thoughts on this bizarre behavior?

Reply to
Jim Stewart
Loading thread data ...

What makes you think there's programming involved?

I doubt it.

I doubt there's any software involved at all.

Let us know what happens.

Probably a component failure of some sort?

--
Grant Edwards                   grante             Yow!  If elected, Zippy
                                  at               pledges to each and every
 Click to see the full signature
Reply to
Grant Edwards

What makes you assume there's any programming involved here, in the first place? Why would a street lamp have a CPU, where some analog circuitry driving a relay would do?

Possible: yes. But also rather less than likely.

Wherever I've actually checked so far, street lights went on by the clock, not by the sun. I.e. the were controlled by a central switch somewhere. Now imagine somebody won the bid for installing those lamps by saving the money for a bistable relay with bidirectional active switching (possibly needing one more control wire from the central switch box), and just used a pulsed toggle relay instead.

--
Hans-Bernhard Broeker (broeker@physik.rwth-aachen.de)
Even if all the snow were burnt, ashes would remain.
Reply to
Hans-Bernhard Broeker

Maybe a variable intialized at power-up and stupidly never refreshed.

Was there an electrical storm a month ago?

Eg. something like:

int light_off = 0, threshold = 0x80, light_level;

while(1) { read_adc0(&light_level); if (light_level > threshold) output = light_off; else output = ~light_off; clear_watchdog(); // which does absolutely no good in this case }

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
 Click to see the full signature
Reply to
Spehro Pefhany

I did once see a port pin failure in a uC, that flipped the pin polarity. I did not believe the user, so he brought it round to prove it, and he was right : working, not stuck, and flipped.... All I could think of, was some punch-thu of an inverter ?

Street lights are rarely controlled in 'ones' - do others nearby do the same ?

Reply to
Jim Granville

It *is* possible to get phase reversal from a component failure, but MUCH more usually it would stick on or off.

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
 Click to see the full signature
Reply to
Spehro Pefhany

Isn't that pretty obvious? That it's broken.

Imagine that.

--=20 Fredrik =D6stman

Reply to
Fredrik :Ostman

snip

snip

Most of the city streetlights I have seen in the US use a photo-sensor to switch the light on and off. If you look at the lamp from the side, there is usually a small cylinder, about 3" in diameter and a 2" tall on the top side of the lamp. The last one I saw up close used an ancient cadmium sulphide phot-resistive sensor. It should be readily visible with naked eyes and easily studied with a pair of binoculars.

Commercial or parking lot type lights are probably centrally switched instead, but that is a guess.

As for why yours has lost it's little mind, I would guess a component failure.

Bob

Reply to
MetalHead

  1. Write down the pole identifier number.
  2. Call the electric utility and tell them their streetlight has a problem.
  3. Sit back and wait for electric utility to dispatch a repairman.
--
James T. White
Reply to
James T. White

  1. Go out and chat with the repair folks and find out what was wrong. A cold soft drink or bottled water might be appreciated.

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
 Click to see the full signature
Reply to
Spehro Pefhany

It's fallen into bad company, like my son who stays up all night and sleeps all day. The solution is to get an older streetlight to have a serious talk with it.

Paul Burke

Reply to
Paul Burke

Not necessarily, it could be a computing science student, or even a computing science graduate! Look for discarded 'energy' drink cans nearby...

--
Mark McDougall, Engineer
Virtual Logic Pty Ltd, 
 Click to see the full signature
Reply to
Mark McDougall

The street lights where I grew up in the UK were controlled by a light sensor in a substation. I knew because us mischievious young scamps used to put a can over it and giggle ourselves silly when the lights came on. It never occurred to us to shine a torch on it at night to turn them off. But then it never occurred to us how close we often came to dying horribly sparky smokey deaths in the sub station :-o

Perhaps if they were clock controlled as you say but if it was a photosensor then a regular SPST relay would probably be used. I'm sure they must have some kind of failure mode which would reverse the polarity. In my experience whole streets and block of streetlights come on together rather than individuals though.

Reply to
Tom Lucas

These photo detectors have a three prong twist-lock plug, wired to a NEMA standard (don't remember which one).

Many years ago, I was a field engineer on a large copper smelter project. We had hundreds of timers, all over the facility controlling the outdoor lighting. Everytime we lost power, all the timers would have to be reset. It would take maintenance several days to get around to resetting them, and we would lose power every couple of days. In other words, the lights were never on at the right time.

Maybe you have a timer that's off by half a day.

-- Hershel

Reply to
Hershel Roberson
14 messages on a broken street light controller so far. A dozen of them either explaining the problem or reverse engineering the street light controller with incomplete information to diagnose the failure.

It is an engineering problem all right, not enough time not enough information and not enough resources.

We have the resources of the internet and the local utility company - solved Time : we have all day essentially solved Information: The reverse engineering and fault diagnosis needs help. It seems to me we have missing information

1) What type of street light controller is it? 2) Who is the manufacturer? 3) Is this a known fault? 4) Do they have fault prevention retro fit?

Two of the posts are developing a work flow to get the street light fixed follow..

:)

w..

Spehro Pefhany wrote:

Reply to
Walter Banks

We'd better have a daily meeting to check progress. That'll fix it

Reply to
Tom Lucas

Not necessarily. Like your son and myself, the streetlight may have that cycle programmed into its DNA. For us, working "normal" hours is the abnormal thing to do. That poor streetlight was living a charade until it decided it could not take it any more...

Reply to
Roberto Waltman

Well, last night I took my big hulking hand spotlight and shined it on the sensor for a good 5 minutes and the light didn't come on. I didn't want to press my luck as I looked kind of silly, shining a spotlight on a dead streetlight and the local cops wouldn't grasp the significance of my experiment.

So, to everyone who suggested a timer/remote control, no, it has a sensor.

To everyone who suggested a hardware failure, I am

*very* hard pressed to visualize any sort of hardware failure that would invert the functionally of the lamp.
Reply to
Jim Stewart

This conclusion seems at odds with your experimental data.

Reply to
Dingo

It's not clear. I've heard that it may take as long as 10 minutes of darkness to get a normal lamp to come on.

Reply to
Jim Stewart

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.