A Puzzle for Engineers

[Time2]

I don't see in your description the logical thing and that would be, to put the flickering LED bulb into a known good light socket. I have had LED bulbs flicker, the power supply inside goes bad, usually a bad capacitor.

 

[sdtech]

plug the lamp into another outlet that doesn't have a timer. If it works ok, it's the timer

 

[mf15]

The timer goes on at dusk as scheduled,but it is not dusk due to daylight savings time,there is a lot of light.
Since you may have not adjusted the timer to the correct current time.
Ok thats fine,but the lamp has a light sensor,which turns the lamp on in low light,the sensor defective or partially operates in low light conditions that are not quite dark enough yet.

old industrial chemist

 

[GrayHare]

I've seen light sensors do that but that's not the problem source this time. Another hint: the electromechanical timer plugs directly into a wall socket (yes, that is pertinent in this case).

 

[mf15]

Ground fault wall socket,and or light sensing wall socket.
oldmike

 

[Qs Laptop]

I've seen light sensors do that but that's not the problem source this time. Another hint: the electromechanical timer plugs directly into a wall socket (yes, that is pertinent in this case).

The electromechanical timer does not have a grounding prong. The timer is not properly grounded.

 

[GrayHare]

The electromechanical timer does not have a grounding prong. The timer is not properly grounded.

It lacks a grounding prong but I think one would not have made a difference in this case. A hint: by the next day, things returned to proper function (no flicker), all with the same equipment. So, the culprit was transitory.

 

[ncbill]

It lacks a grounding prong but I think one would not have made a difference in this case. A hint: by the next day, things returned to proper function (no flicker), all with the same equipment. So, the culprit was transitory.

something else electrical inducing transients on the circuit...dog could hear it when it happened.

 

[GrayHare]

something else electrical inducing transients on the circuit...dog could hear it when it happened.

Yes, the dog could hear it and probably smell it, though it's not something electrical. What was it? I gather this is rare enough that people have not experienced it often.

 

[like2]

Mouse chewing the wires in the wall?

 

[street]

It not being electrical but I was going to say Harmonics Frequency/Distortions on electrical system.

 

[GrayHare]

I'll post the puzzle solution soon, so take final guesses. Final hint: the cause was more them than it.

 

[Tetto]

You lost one half of the voltage coming into the house?

 

[mf15]

Ok loose wire on the receptacle, mice moving the wire.
oldmike

 

[like2]

Squirrel eating into the service drop?

 

[GrayHare]

Thanks for all the replies. Lots of logical thoughts, but here's the strange answer. Spring is ant season. For an unknown reason, a small colony followed wires inside the wall to the outlet where the electromechanical timer was plugged in. They moved into that timer. When the timer flipped on, they must have been grouped in the wrong spot, perhaps gumming up the power contacts. IMO less likely is that they formed just enough of an alternate current path to short the supply but not enough to trip the breaker. In either case, the light bulb flickered.

The dog ran to the timer and kept listening to it while I pointlessly swapped light bulbs. Eventually, I pulled the timer from the outlet and saw ants. A little bug spray took care of things. I let the timer dry then plugged it back in and everything has worked properly since. I am curious about exactly what went on inside the timer so plan to disassemble it. I've never seen this happen before.

 

[landover]

Thanks for all the replies. Lots of logical thoughts, but here's the strange answer. Spring is ant season. For an unknown reason, a small colony followed wires inside the wall to the outlet where the electromechanical timer was plugged in. They moved into that timer. When the timer flipped on, they must have been grouped in the wrong spot, perhaps gumming up the power contacts. IMO less likely is that they formed just enough of an alternate current path to short the supply but not enough to trip the breaker. In either case, the light bulb flickered.

The dog ran to the timer and kept listening to it while I pointlessly swapped light bulbs. Eventually, I pulled the timer from the outlet and saw ants. A little bug spray took care of things. I let the timer dry then plugged it back in and everything has worked properly since. I am curious about exactly what went on inside the timer so plan to disassemble it. I've never seen this happen before.

you discovered new type of semiconductor,

 

[AudiDudi]

Alex, I'll take ants in the timer for $200 said no one ever Interesting issue/solution!

 

[HFWR]

you discovered new type of semiconductor,

Uses the dual antescene process…

 

[ERD50]

I didn't see this until after you posted the answer, but in general, a poor connection (another poster mentioned something like this) can result in just enough power getting through such that the LED will store up energy for a while (fraction of a second), and light up which drains the energy, so goes out again. This explains the flickering. The other bulbs draw too much current so the poor connection doesn't provide enough power to light them at all.

The ants got between the contacts, creating a poor connection.


Poor connections aren't that rare - a failing switch will sometimes act this way. I've even read where an LED that is switched OFF will light dimly or flicker, just from picking up voltage from a nearby circuit.

 

[ncbill]

I knew it a soon as I read "Final hint: the cause was more them than it."

Ants also cause major problems with the contractor on outside AC/heat pumps.

 

[GrayHare]

I knew it a soon as I read "Final hint: the cause was more them than it."

Initially, I'd typed "Them!" but then figured that would make things too easy. Fun that you spotted it anyhow.

ERD50, cool info about LED operation. Interesting.

 

[Texas Proud]

I am just going to throw out that the people who said the timer was the problem were right...

They did not explain why it was the problem, but it WAS the problem...

 

[JoeWras]

Although this was an interesting thread, it reminded me too much of w*rk and the sh^! I had to deal with day to day! My last problem of my career wasn't ants, it was metal layers that were too thin in the semiconductor and occasionally leaked between layers in about 1% of the chips. This was not found in testing, but in the field. My software team was blamed for it. I didn't solve it before I retired. I later found out from a friend who still worked there that a few weeks after I retired, the hardware group admitted they specified the fab layer too thin and it was their problem.

It might as well have been ants.

Glad I retired. You would have thought someone officially would have told me that the last 6 months of BS I suffered through was not my fault and they are sorry for putting me through hell in my last months. No, instead corporate gave me some static in a different way (that I cannot discuss) that was another load of BS.

 

[AudiDudi]

Although this was an interesting thread, it reminded me too much of w*rk and the sh^! I had to deal with day to day! My last problem of my career wasn't ants, it was metal layers that were too thin in the semiconductor and occasionally leaked between layers in about 1% of the chips. This was not found in testing, but in the field. My software team was blamed for it. I didn't solve it before I retired. I later found out from a friend who still worked there that a few weeks after I retired, the hardware group admitted they specified the fab layer too thin and it was their problem.

It might as well have been ants.

Glad I retired. You would have thought someone officially would have told me that the last 6 months of BS I suffered through was not my fault and they are sorry for putting me through hell in my last months. No, instead corporate gave me some static in a different way (that I cannot discuss) that was another load of BS.

Having been Software Test Engineer in the early/mid 90's, the different philosophy now is amazing. Then, it could cost several $mil to fix a bug, mailing out floppies or CDs, so the products was well tested. Now, it's a simple matter of issuing a monthly update which costs nothing. The consumer is now the Test Engineer.