An electrical question

[countryguy]

A couple of days ago the light fixture on my outside lamp post quit working. I assumed the bulb was burnt but I tried 3 new bulbs and it still doesn't work. I checked both the old and new bulbs in other fixtures and they all work. There are other electrical outlets on the same circuit and they all work so I know there is electricity getting to the post. The lamp post, fixture, motion sensor and underground wiring were all installed by a qualified electrician. Today I tested the light bulb socket with a circuit tester and there is power...I put the bulb back in and it doesn't work???? The bulb is a 150 watt incandescent. I then tried a 100 watt bulb and it worked! I don't understand why the 150 watt bulb just suddenly quit working. Any ideas?

Thanks,
Gary

 

[paulg]

I have two ideas

Indeed this is a really strange symptom. The cause will likely also be very strange. I'll give you two of my ideas; take them with a grain of salt.
Concept 1: The contacts of the socket are misshapen.
If the bottom contact of the socket is misshapen (probably bent too far down), some lamps may work and other not because the bulb's bottom soldering is sometimes larger, sometimes not. Larger solder pads on the bottom of the bulb may reach the contact and work, others not. Small solder pads on the bottom of the (presumed Edison-type) lamp may work for a while then arc itself out of the circuit.
Concept 2: You have a CURRENT problem.
Here is a case where the voltage is ok, but the current is low. The voltmeter may read the correct voltage but with low current, the "punch" is nil. Some lamps may light and others not.
You mention you have a light sensor in the circuit. Perhaps it has failed allowing voltage to pass but current to be very limited. In that case, some lamps may light and others not. CURRENT problems yield very strange symptoms.

 

[countryguy]

Concept 1: I can see how that would be a problem but it seems strange that the light just quit working. It was on when I went to bed but off when I got up in the morning. How could the contacts change on their own during that period? I did bend the lower contact up but that didn't help.

Concept 2: How do you test for a 'proper' current. I'm not an electrician so am just flying by the seat of my pants. I was thinking that perhaps the motion sensor had failed and I was going to replace it but when the circuit tester worked I just figured the sensor must be OK....perhaps I should still change it.

Gary

 

[cfz2882]

motion sensor probably faulty

sounds like the motion sensor may have bad contacts in it's relay,bad solder joint,
etc.

 

[whirlcool]

Gary:

Just use your meter to check the amperage in the socket. That'll tell you how much current is in there. Have you removed the light sensor and just plugged the bulb into that socket? That will check to see if it is light sensor or the light fixture socket itself.

 

[qualin]

OK, let me straighten this out for all of you.

If you see about 120 volts across the center conductor of the light bulb socket and the outside ring, then everything is OK. However, I'd like to mention that measuring voltage this way is dangerous. You should be using a socket adapter to measure voltage, such as this: http://amzn.to/xWSQ74

I've never heard of an electrical circuit not supplying enough current when there is enough voltage. Usually the circuit breaker kicks in well before you'll reach the current limit of that circuit.

You _CANNOT_ test current draw unless you have a way to interrupt the circuit.

Voltage is a differential of electromotive force between two conductors. Since it is a differential, you measure "Across" things. ie. The Positive and Negative terminals of a battery, or the Hot and Neutral Line of a wall plug.

Amperage is a measure of Electromotive flow. One of the most dangerous things you can do is measure "Across" conductors with a multimeter in Amperage mode. That will put a direct short on the circuit and usually result in a lot of heat and light, not to mention a tripped circuit breaker and a fried meter. If you are lucky, the meter will have a fuse in it and you'll blow the fuse before you blow the meter.

Instead, you have to measure "Inline" with the conductors. Doing that with a light bulb socket isn't possible. You literally have to find the wires leading up to the light bulb socket, cut one of them, put one end of the wire on one side of the meter and then put the other end of the broken wire on the other side so that the meter is in-line with the circuit, rather than across it. Of course, you won't see anything on the meter until you screw in a light bulb... and if the light bulb doesn't light, you most likely won't see anything on the meter anyway.

For education purposes, I've included a link to a video below.. it's a little dry though.

It's much more likely that you probably just have a socket which has become defective over time. Now, if the center conductor has built up a layer of carbon because the contacts weren't making good contact, that can create a resistance which can lower current flow.

If there really was truly additional resistance on that center conductor preventing optimal current flow, you would see a lower voltage than what should be there, because there would be a voltage drop across that resistance.

 

[gizmo]

The centre conductor can fail - at work we have ES (US Style) bulb holders and two have failed centre contacts. The contact is a small piece of bent copper, it has cracked and the tip has fallen out. There is still a bit showing when you look into the bulb holder, but it is just a stub and doesn't reach the centre contact of the bulb.

Also I have seen them where a previous bulb has been screwed in too hard, crushing the centre contact of the bulb holder. Later bulbs, it has been hard to get it to make a reliable contact. Particularly as modern bulbs seem to be very skimpy on the blob of solder that makes the centre contact of the bulb.

Personally I prefer BC bulbs, which are "normal" here in Australia - seem to be more reliable contacts.

 

[paulg]

Adjustments needed to comments

Qualin's comments appear to need some adjustment:

"If you see about 120 volts across the center conductor of the light bulb socket and the outside ring, then everything is OK."

* This is incorrect on a number of levels. Any electrician will tell you that voltage can remain normal but current can be severely limited by resistance in series with the load. Also, testing the lamp socket with probes does not exactly emulate the socket-to-lamp physical contact. The voltmeter test to the lamp socket contacts is useful but likely misleading.

It is odd that you've never heard of a circuit supplying correct voltage but insufficient current. A resistance in series may leave the voltage stable but decrease the current. Putting a voltmeter in place of the lamp will not reveal this symptom when testing at the lamp socket. With the lamp missing, the voltage will rise to its limit but the current limitations will not be detected.

This case seems to dictate the need for an electrician.

 

[countryguy]

Thanks for all the information guys and different suggestions. As I said yesterday I put the 100 w bulb in the socket (which worked) and left everything else as previously installed i.e. fixture installed to the dusk to dawn motion detector and waited for darkness to arrive. Sure enough, the light bulb came on and worked all night. I still don't understand why the 100 watt bulb works but the 150 watt bulb does not any more after having worked fine for 4 years. I changed the bulb about once a year when it would burn out.

Gary

 

[sudsmaster]

I suspect the motion sensor might be the cause of the problem. It might just be switching itself off when the current required for more than 100 watts is reached.

Down this way, electricity is expensive enough that I would not contemplate fitting an outside fixture with a 150 or even an 100 watt light bulb that would be on about 12 hours a day/night. I'd probably use a CFL instead (with a motion sensor that is CFL friendly). Recently I replaced the CFL on my porch light with an LED bulb. The bulb wasn't cheap (about $6), and it's not overly bright (only about 130 lumens) but it's enough to light the porch well enough to find the keyhole or door bell ;-). And at 2.5 watts it saves a fair amount over the previous 13 watt CFL. It's on a light sensor, not a motion sensor, so it's on from dusk to dawn, and then some, usually.

Another possibility is that there is a slow leak to ground through the underground wiring. An electrician should be able to check this out fairly quickly with an ohm meter (provided the post light circuit is on a switch).

Good luck!

 

[paulg]

Something changed

Your question is a good one however nobody will likely know the true answer until the issue is fixed.
I suppose we know a few things:
Clearly something in the circuit has changed.. over time. (It worked before but not now)
And the symptom varies with the load. (100W lamp works, 150W lamp does not).
Tough nut.
So what can stop the lamp from lighting when it's loading effect upon the circuit increases?
I'd suspect something in series with the lamp is faulty in a borderline way. Something is becoming electrically open or of increased resistance as the stress of the 150W lamp is applied.
That leaves the triac circuitry or relay within the motion sensor as potential culprits.
Yeah, totally speculative. Components that fail under load and especially those that impede current are tough issues to solve. And the components can test good!
In the electronics world, we've all run into these situations of circuits that fail under varying loads. They'll turn your hair gray and teach you how to swear.

 

[countryguy]

Sudsmaster,

I'd like to use a CFL bulb but unfortunately it can go down to -30 F here and colder so a CFL will not work, at least not any of the ones I have found.

Paulg,

I'm leaning towards the motion detector unit as being the culprit. If the problem happens again, I'll just replace the motion detector. Since I don't like ladders, I took the lamp post down (it sits on a wood retaining wall, with the base screwed into the wood) to work on the problem. Now that the post is back up (not easy to do with just 2 hands) and the light is functioning, I won't worry about it until the problem re-occurs.

Thanks to all for the info.

Gary

 

[qualin]

To PaulG

> Any electrician will tell you that voltage can remain normal but current can be
> severely limited by resistance in series with the load.

That's true. However, there will always be a voltage drop across that resistance. This is called a "Voltage Divider". http://en.wikipedia.org/wiki/Voltage_divider

>Also, testing the lamp socket with probes does not exactly emulate the socket-
>to-lamp physical contact.

That is certainly true and I agree with that. Read my further explanation below.

> A resistance in series may leave the voltage stable but decrease the current.

Yes, but I don't think you understand that if you have a few different spots of resistance, the voltage drops across them must be the sum of the total voltage.

> Putting a voltmeter in place of the lamp will not reveal this symptom when
> testing at the lamp socket.

The problem is, if the volt meter reads 115 volts, a person may not think there was a problem, even though there really is a 2 volt drop across the carbon built up on the contacts... of course you can't measure that.

Let's just use some simple math though..

Let's assume that the bulb isn't lighting because a maximum of only about 0.84 amps of current are being supplied to the socket with the light bulb installed. That's enough for a 100 watt bulb but not for a 150 watt. In other words, the total amount of resistance on the circuit can't exceed 142 ohms.

So, let's say that you plug in a 100 watt light bulb. It has a resistance of about 114 ohms when it's hot. That means that the wiring is supplying the remaining 28 ohms of resistance, which is really high.

Assuming that there really is 28 ohms of resistance in the wiring and not the light bulb socket itself, the maximum supplied current draw for that wiring would actually be 4.2 amps. Which means that this wiring would be acting like a 504 watt space heater!

If you had the light bulb in circuit and you could measure across the terminals of the light bulb when lit, you would see 95.76 volts across the light bulb and you would have 23.52 volts across the wiring.. Or a total of about 120 volts.

So, ideally, the best place to measure would be on the screw terminals of the light bulb socket with the light bulb installed.

I felt I had to kind of clarify this. Of course, if your wiring is acting like a 502 watt space heater, I think you would have noticed by now.