CYCLINGFORUM.COM - Where Cyclists Talk Tech --- Return To Home

 

    Register FAQ'sSearchProfileLog In / Log Out

 

****

cyclingforum.com ****

HOMECLUBS | SPONSORS | FEATURESPHOTO GALLERYTTF DONORS | SHOP FOR GEAR

Return to CyclingForum Home Page CYCLING TECH TALK FORUM
          View posts since last visit

OT - loose connection or faulty sockets?
 Goto page 1, 2  Next

Author Thread Post new topic Reply to topic
April
Joined: 13 Dec 2003
Posts: 6593
Location: Westchester/NYC

12/28/14 3:28 PM

OT - loose connection or faulty sockets?

I have TWO light sockets that would intermittently stop working. One of them is in the living room which has a vaulted ceiling. In other words, I can't get to it easily. :-( It's so high that I want to fix it in the shortest time and least amount of ups and down. The last thing I want is to try this and try that to see if it solves the problem...

It's just a little odd that I have TWO of such. I wonder whether it's faulty hardware (socket), or loose connection on both? Anyone want to bet which is more likely?

Oh, before you ask. Both light bulbs are solidly screwed into the sockets.

 Reply to topic     Send e-mail

Jesus Saves
Joined: 16 Jun 2005
Posts: 1150
Location: South of Heaven

12/28/14 3:57 PM

It may be one or both...however, your time and an electrician's time is not cheap, while new sockets are. While you replace the sockets, or in the process of doing so, you will know if the connection is loose, too. The short answer is buy new sockets before you climb up the ladder. And keep the receipts.

 Reply to topic    

Steve B.
Joined: 19 Jan 2004
Posts: 769
Location: Long Island, NY

12/28/14 4:50 PM

More likely a bad socket. The standard medium screw base sockets and lamps can get a build up of corrosion on the "button" which is the hot connection between lamp and socket.

A short term fix is to take out the lamp and then take a flat screw driver blade - and while the circuit breaker is dead, scrape off some of the corrosion, also bending the button back downward to help the lamp make contact. Not worth doing this to the the lamp, just put in a new one.

Note that unless there's only one switch for the circuit, best practice is to turn off the breaker. If there's a 3 way switch then socket may still be hot and you won't know it by switch position and the lamp won't indicate that the circuit is off, obviously.

Long term for recessed high-hats is to pull off the trim ring, then see if the unit housing can be removed by pulling the fixture down. Then you can sometimes access the socket. If you can remove the socket, take it to an electrical supply house (or sometimes a big box will have them). Typically the sockets have 2 small mounting screws that make replacement easier.

This is also a good time to buy that 8ft A frame ladder.

Note that I'm a theater electrician so to me, this is all easy. Others not familiar or comfortable around electricity might balk at the process. Perfectly understandable.

SB

 Reply to topic    

April
Joined: 13 Dec 2003
Posts: 6593
Location: Westchester/NYC

12/28/14 5:14 PM


quote:
More likely a bad socket. The standard medium screw base sockets and lamps can get a build up of corrosion on the "button" which is the hot connection between lamp and socket.

Why hadn't I thought of that!!! (duh, because you're an electrician and I'm not)

I bet you're dead right. I'll start with the one in the spare bathroom. Except I won't know if it worked since that bathroom only got used once or twice a year!


quote:
This is also a good time to buy that 8ft A frame ladder.


I'm not afraid of height. Not irrationally anyway. But there's rational reason to be afraid in this case.

The "top hat" socket is at the foot of a stair some magical distance away from any walls. What that translate to, there's tight space to spread the legs of the ladder. But not close enough to any wall to use as a brace.

If something goes wrong in an open space, I can trust my cat-like reflex to land on my feet. At worst a sparing ankle. (I'll only be 6' off the ground) But with stair at proximity, things get complicated. The last time I climbed up that high, I was nervous, which increase my chance of falling off. Not to mention I really can't work for long up there being so nervous.

OK, I'm getting old. I used to do that without a thought. (my cousin was a electrician. I used to help him install ceiling lighting fixtures. But that was 20 years ago). But if the same treatment works in the bathroom socket, that'll boost my confidence to do the one on the vaulted ceiling.

Last edited by April on 12/28/14 7:40 PM; edited 1 time in total

 Reply to topic     Send e-mail

Sparky
Joined: 08 Dec 2003
Posts: 19083
Location: PDX

12/28/14 6:17 PM

Funny you should mention it. Just pulled this intermittent firing bulb.

Picture self explanatory?


<img src="http://coupekiss.host-ed.me/images/ttf/WIN_20141228_161052.jpg" width=320>

 Reply to topic     Send e-mail

Andy M-S
Joined: 11 Jan 2004
Posts: 3377
Location: Hamden (greater New Haven) CT

12/29/14 6:30 AM

One other thing.

Bulb sockets seldom go bad because there are few mechanical stresses on them and the electrical stresses to which they're subject are very low (it's rare to have as much as 1A [which is already >100W] going through a light socket, and I suspect they're designed to handle up to 15A--most look like it in my experience). If it's not corrosion (and depending on where you live, etc., it could be) then I would suspect the switch or switches controlling the sockets. These get cycled regularly, and if they're not "silent" switches (which at least used to just use mercury capsules) I would suspect failure either in the switch mechanism itself OR in the wire-to-switch connection (which gets stressed through vibration with every switch operation).

Switches are cheap, you don't have to work upside down, and you don't need a ladder. I'd consider starting there...

 Reply to topic     Send e-mail

April
Joined: 13 Dec 2003
Posts: 6593
Location: Westchester/NYC

12/29/14 7:30 AM

It's not the switches.

the same switch controls 5 lamps. The other 4 comes on and off consistently just fine.

 Reply to topic     Send e-mail

Andy M-S
Joined: 11 Jan 2004
Posts: 3377
Location: Hamden (greater New Haven) CT

12/29/14 7:58 AM

Then it's PROBABLY not the switches

Then you're right, it's probably not the switches, unless the original electrician did something dumb, like try to wire the switched side of all five sockets directly to the switch. In that case, you could be having mechanical issues.

 Reply to topic     Send e-mail

April
Joined: 13 Dec 2003
Posts: 6593
Location: Westchester/NYC

12/29/14 8:57 AM

I already looked. They are not.

 Reply to topic     Send e-mail

Steve B.
Joined: 19 Jan 2004
Posts: 769
Location: Long Island, NY

12/29/14 10:18 AM

"Bulb sockets seldom go bad because there are few mechanical stresses on them and the electrical stresses to which they're subject are very low"

Sockets are the weak link in the circuit. In a medium screw base socket ( and other sizes of similar design), the button, being a bent spring loaded contact, flexes due to heat, and even with a socket rated for 100 watts - .83 amps, there's still all that heat an incandescent lamp generates. Fluorescent and LED are not much better as they induce transformer heat in the lamp base area.

Thus the button moves and contracts due to heat. It's that movement that causes corrosion as a mini arc can form were electricity is trying to flow against a floating contact. This is where and why the corrosion occurs.

I maintain a facility with over a thousand lamps using medium screw base sockets and this almost always the first thing we look at.

Switches and wiring can and do fail, but we rarely see failure in wiring and wiring junctions as the loads are by code, kept to 80 % of circuit design. Switches fail more frequently just due to age and life cycle usage. We occasionally have to replace switches, but nowhere as often as failed lamp sockets.

 Reply to topic    

Andy M-S
Joined: 11 Jan 2004
Posts: 3377
Location: Hamden (greater New Haven) CT

12/29/14 1:04 PM

Interesting.

I hadn't thought of the heat load, since I've been running LED bulbs for the past few years (which in my experience are *much* cooler), but I can see how it would be a mechanical stress source if you were running incandescent or CF bulbs. I also prefer ceramic (as opposed to plastic) sockets, which seem to me to be a little less likely to fail.

My thought on mechanical failures near the switch(es) is that I've seen so many craptastic "home wiring" jobs over the years--one place had wires that were "just sort of touching" on the side of the switch...they'd been cut so short that they couldn't be properly looped for screws, and the switch in question had (IIRC) only side-wiring connections. Could have been worse--it could have been installed that way in a metal box.

 Reply to topic     Send e-mail

Sparky
Joined: 08 Dec 2003
Posts: 19083
Location: PDX

12/29/14 1:08 PM

"I also prefer ceramic (as opposed to plastic) sockets, which seem to me to be a little less likely to fail. "


As long as some ham fisted screwer is not cracking them. ;)

 Reply to topic     Send e-mail

Andy M-S
Joined: 11 Jan 2004
Posts: 3377
Location: Hamden (greater New Haven) CT

12/29/14 1:48 PM

Hmm

Dost thou speak from experience, forsooth?

 Reply to topic     Send e-mail

Steve B.
Joined: 19 Jan 2004
Posts: 769
Location: Long Island, NY

12/29/14 1:48 PM

LED's have possibly more heat issues then CFL's and incandescents.

The design crams hundreds of lo-voltage heads in a cluster, which frequently requires a lot of engineering to deal with the heat put out by all those tiny lamps, as well as the transformers. Heat is the number one issue for LED lamp designers to deal with, that and getting acceptable dimming, which requires better and more controlling electronics, all which adds heat. Etc.... We used to see rated hours on LED's once specified as having 50,000 hours. That's now down to 30,000 and some are now at 20,000, all as the manufacturers gain better experience with the lamps.

There's little noticible heat coming out the front as there's no infra-red being produced. All the heat is at the socket end.

 Reply to topic    

Sparky
Joined: 08 Dec 2003
Posts: 19083
Location: PDX

12/29/14 2:06 PM

"Dost thou speak from experience, forsooth?"

I am actually not too heavy for light work.

I suspect that bulb got the contact arced down/off due to the photo sensitive on/off adapter not allowing the bulb to seat down enough due to that shoulder on the housing that the circuit is encased in. Run on sentence, I know...

 Reply to topic     Send e-mail

Andy M-S
Joined: 11 Jan 2004
Posts: 3377
Location: Hamden (greater New Haven) CT

12/29/14 2:10 PM

Yabbut

Oh, they are by no means zero-heat, 100% efficient devices. But given the comparative efficiency of LED lights to hot-wire bulbs (i.e., the proportion of power radiated as visible light rather than infrared) and the much lower input power (which is largely a function of the above) there is much less heat sitting in the can from LEDs than from incandescents. Heat rises, so you put a 60W indy bulb in an inverted can fixture and you get almost everything that isn't light (which amounts to about 50W, IIRC) rising back into the can as very hot air.

In comparison, take an LED light and if it's a 10W light (60W equivalent), even if it's grossly inefficient (probably made moreso by AC to DC electronics & c), there aren't nearly as many watts to heat up the can. Granted, they *radiate* near the back of the can, but I suspect that the heat rising from the indy is a greater issue.

I have LEDs all over my house, some in inverted fixtures, and the heat level at the back of lamp is significantly lower than it is with an incandescent of similar output.

OTOH, I am by no means a pro, and you are. :-)

Someday it would be nice to see lighting fixtures and wiring designed for LEDs--low voltage DC systems fed from a common xformer, perhaps with connectors closer to the center of the bulb and heat sinks in back or built into the fixtures. If LEDs have nearly the life expectancy that is being touted for them (and IME they do pretty well) then it wouldn't even be too weird for people to apply some conductive paste when installing the bulbs in their sockets/heatsinks.

 Reply to topic     Send e-mail

Steve B.
Joined: 19 Jan 2004
Posts: 769
Location: Long Island, NY

12/29/14 3:24 PM

What we find with LED's and fluorescent's is that the base of the lamps generates much more heat then you would expect, due to all the electronics and transformers being located inside the socket with no place to radiate heat, except inside the socket. This heat is a huge design issue for LED manufacturer's as stated prior.

Incandescent's get hot all over, with some of the heat being convection carried and some being by transferred in the glass and metal socket into the base. The majority of the heat in an incandescent is just loss in heating the filament with a lot of additional heat being infra-red and is being radiated outwards from the filament.

Thus in practical usage and experience with lamps being changed while hot, the bases of an LED and Fluorescent can be every bit as hot and hard to handle as incandescents.

Note that it is not typical of a home-owner to be hot-swapping a lamp, usually you are replacing a dead lamp and thus don't notice the heat.

 Reply to topic    

April
Joined: 13 Dec 2003
Posts: 6593
Location: Westchester/NYC

12/29/14 4:00 PM

Steve, do you see LED bulbs more sensitive to voltage fluctuations?

 Reply to topic     Send e-mail

Sparky
Joined: 08 Dec 2003
Posts: 19083
Location: PDX

12/29/14 4:29 PM

They should be less sensitive due to a transformer in circuit I would imagine..

 Reply to topic     Send e-mail

Andy M-S
Joined: 11 Jan 2004
Posts: 3377
Location: Hamden (greater New Haven) CT

12/29/14 9:29 PM

Swapping


quote:
Note that it is not typical of a home-owner to be hot-swapping a lamp, usually you are replacing a dead lamp and thus don't notice the heat.


True. But I have LEDs in (e.g.) bedside invert mounts, where I can get a pretty good sense of the tail-end heat. And I do occasionally do things like hot-swap just because I'm interested in the tech (I home-brewed my first LED, generator-powered bike headlight nine or ten years ago).

 Reply to topic     Send e-mail

Steve B.
Joined: 19 Jan 2004
Posts: 769
Location: Long Island, NY

12/29/14 10:09 PM

Our village has local power generation and I rarely experience flicker, so can't say as to whether LED lamps are more sensitive or not.

At work, the LEDs are in our aisle fixtures and as the power is from Con-Ed and is an underground feed, we rarely see flicker and in truth when I do I'm not looking at the aisle lighting, but at the stage lighting system.

In theory, LEDs are no different then a fluorescent as they both have a transformer. You might get lag when they re-fire. I see my assorted recessed LED lamps lag about a 1/4 second when they get powered up.

 Reply to topic    

Andy M-S
Joined: 11 Jan 2004
Posts: 3377
Location: Hamden (greater New Haven) CT

12/30/14 4:44 AM

LED


quote:
I see my assorted recessed LED lamps lag about a 1/4 second when they get powered up.


What sorts of power in "watt-equivalents" are we talking about?

I have only one that does that--a large 110 "watt equivalent" we use in a fixture in our dining room (I should pull it to see what the actual wattage is, but I'm upstairs now and lazy).

None of the lower-watt bulbs seems to have any noticeable delay, but they are comparatively low-power (here's a Phillips 10.5W/800 lumen (c. 60 watt equiv). I don't think a *transformer* should have much impact on start up time, but a capacitor can take significant time to charge, and I suspect the larger LED lamps have correspondingly larger capacitors.

 Reply to topic     Send e-mail

Steve B.
Joined: 19 Jan 2004
Posts: 769
Location: Long Island, NY

12/30/14 8:32 AM

Mine are either a 90W equivalent in an PAR38 (outdoor rated yet used indoor) design, or 2 as 75W R30 indoor lamps. All in ceiling recessed hi-hats. It's typically the 90/PAR38 that has the start up lag. Not sure of the manufacturer. the 90w equiv. I got at my electrical supplier, he subsequently stopped carrying these lamps. The 75's I got a Lowes and are nice lamps costing about $12 ?, maybe $16 can't remember. They have a color temperature (about 2900K) similar to an incandescent, but a tad less spread then the incan. flood.

I'm sure you are correct that charging the capacitor is what causes lag.

 Reply to topic    

ErikS
Joined: 19 May 2005
Posts: 8337
Location: Slowing boiling over in the steamy south, Global Warming is real

12/30/14 4:00 PM

The heat kills the sockets. I would replace the fixture. If you don't know it already fixtures are rarely made to be repaired anymore. Most parts are press fit and don't come apart or worse go back together.

Replace the entire fixture. It is better than burning down your home or the apartment building you live in. If you are not qualified and renting, have an electrician do the work and bill the landlord. You don't want the liability of the entire building on your non-certified or bonded handy lady work. I sure would not do a dang thing to a rental property unless I was the landlord.

 Reply to topic     Send e-mail

dddd
Joined: 11 Jan 2004
Posts: 3345
Location: NorCal

12/30/14 6:10 PM

Did anyone mention a possibly-loose screw connection on the bulb socket?

It's something that doesn't apply to many modern fixtures, but thermal cycles can induce loosening there.
Same thing for those wire nuts, I recently found one leg of a multiple-termination to be open-circuited, and beat an on-site electrician to the punch on that one!

First things first though, and some kind of additional support for solid ladder footing is the first step toward getting the fixture sorted out. Perhaps a ladder with adjustable legs can safely position the ladder steps closer to the fixture by resting one or two legs the stairs.

Has the bulb been removed and installed a few times to displace any possible corrosion as alluded to by Steve B.? The bulb itself could have a loose, failing socket as well.

 Reply to topic    


Return to CyclingForum Home Page CYCLING TECH TALK FORUM
           View New Threads Since My Last Visit VIEW THREADS SINCE MY LAST VISIT
           Start a New Thread

 Display posts from previous:   


Goto page 1, 2  Next  
Last Thread | Next Thread  >  

  
  

 


If you enjoy this site, please consider pledging your support

cyclingforum.com - where cyclists talk tech
Cycling TTF Rides Throughout The World

Cyclingforum is powered by SYNCRONICITY.NET in Denver, Colorado -

Powered by phpBB: Copyright 2006 phpBB Group | Custom phpCF Template by Syncronicity