Scratching my head after trying to troubleshoot some failed outdoor lights. It is a long story.

The issue was with the outdoor lights outside my garage door. There are 3 sconce lights, and one overhead light that is controlled by a motion detector. All 4 lights are controlled by a single switch. There are other lights inside the garage (including garage door opener) that shares the same circuit breaker, and they are working fine. This is a relatively new house, originally built is 1996, but significantly upgraded in 2006.

Wife complained that the overhead light always stays on (i.e. not controlled by motion detector), and the sconce lights never turned on.

I first tackled the overhead light. The motion detector was bad; so I changed the motion detector, and it started working correctly. So far so good!

Then I started checking if any power is coming to the bulb holders of the sconce lights. Verified that all 3 bulbs are good. Also put a "Lamp Holder to Outlet Adapter" in the bulb holder, checked the voltage using a multimeter, and got ~120V between hot and neutral.

However, when I connected a LED light to that adapter, I got a very dim output. Also tested it using a "Klein Tools Digital Circuit Breaker Finder/Receptacle Outlet Tester", making sure to connect the ground wire to it as well. It showed correct wiring (2nd and 3rd yellow lights were on), but these lights were dim and flickering. I saw the same issue at all three sconce lights.

In addition, if I connect a light bulb to the bulb holder, then the outlet voltage drops down to almost zero, even though the light bulb does not turn on. If I use a CFL bulb instead of filament bulb, I see a faint flickering.

At this time, I found that the overhead light has stopped working altogether. I removed the overhead light, and saw the same issue there as sconce lights; i.e. 120v but outlet tester was showing dim/flickering lights.

To isolate any issue with the devices, I completely removed all the 4 lights, leaving only the wires coming out of the outside wall. Still seeing the same issue.

As the next common point was the switch, I removed the switch completely, and connected the black wires together. Still no change. Also, if I use the outlet tester at the switch, then the two yellow lights are bright and steady. I also checked that the neutral wire at switch box is securely attached with the other neutral wires (there is another switch in the same box).

At this point I am at a loss. I hate to think that this is a wire issue, because the wires go above the wall of the finished garage, and not easily accessible.

Is there any other experiment that I can do?


PS: Adding more datapoints based on the suggestions below.

Tried these experiments:

(a) Connected an Edison bulb to one of the sconces, and on 2nd scone bulb-socket, attached an extension cord to bring the point near the switch. With the light-bulb connected, the voltage between the hot and neutral of extension cord drops to 1 volt. Now, I checked the voltage between neutral of extension cord and that of the switch. It was 0 volt. Next, checked the voltage between hot line of extension cord and that of this switch. It was 120volt.

(b) With bulb removed and power turned off, I measured resistance. Between neutral of switch and extension cord, it was 0 ohm. Between hot of extension cord and that of switch, it was about 6 M-ohm. This should also ideally be close to 0, right?

  • 2
    Plug a “multi-socket” extension cord into that Edison adapter. Into one socket, plug any load (a nightlight will do). Into another socket, stick your voltmeter probes. What do you get now? Also, find a third location with a reliable safety ground, and run a wire from there, and measure voltage hot-ground and neutral-ground. – Harper - Reinstate Monica Jul 9 '20 at 20:30
  • I had done this test as well, using 2-outlet socket adapter with pull-chain. When there is no bulb, or pull chain was not pulled, I get 120v at the outlets using multimeter. When the bulb is there and chain is pulled, the voltage drops to ~3v. – JBH Jul 10 '20 at 0:25

Everything you have observed points at a bad connection with high resistance. When the load is small (voltmeter or neon test lamp), the measured voltage is high, and when there is a larger load, the voltage drops. LED and CFL lights may flicker because they can start with very little current and they have capacitors in their power supplies enabling momentary running on the tiny available power, but an incandescent bulb will stay dark because it never gets hot enough to be visibly lit.

You will need to find the bad junction. One way to do this is, while there is a load (the sconce lights), to put a voltmeter across two different points on the same wire (both hot or both neutral). Good wiring should see no more than a few volts, and usually much less; a bad connection will show somewhere up to the full 120 volts. If you see a large voltage, the bad junction is somewhere between those points, and you can then test further within that span, or inspect all the joints. You may need extensions for your meter's leads to be able to reach.

  • If this is a high-resistance issue, shall I check the resistance of the wire between outlet and switch using multimeter? What is considered a high-resistance value? – JBH Jul 9 '20 at 20:16
  • I was thinking similarly. The motion light worked then failed so trying to identify the first point of failure is the trick here and getting things re connected. – Ed Beal Jul 9 '20 at 20:20
  • JBH it may be faster to pop the fixtures and look for a loose wire, in some cases you will have a good value when not under load 10 ohms or less but, when under load the bad connection heats up and starts arcing or totally looses contact, it could be on the hot or neutral wire. – Ed Beal Jul 9 '20 at 20:23
  • 1
    @JBH I'm afraid I don't have a good range of numbers to give you, but note that the method I suggested is to test the voltage on the wire, with the circuit on, not the resistance with it off. Given the non-functionality of the lights, you should see close to 120 volts across the bad junction. Also, a resistance measurement might be unrealistic as a bad connection can change its resistance under load (as it heats up). – Kevin Reid Jul 9 '20 at 20:24
  • @KevinReid, I added the results of the experiment to the question. – JBH Jul 10 '20 at 1:17

After some experiments, I believe I found what was happening. The 3 sconce lights, while in parallel with each other, are in series with the motion-controlled overhead light. So, when I removed all 3 Edison bulbs, the overhead light also stopped working. When I add one or more Edison bulb, the circuit is formed through the bulb(s); and the overhead lights, which have LED bulbs, started working again. But the current through the Edison bulbs were not high enough to light themselves up. I believe the previous owner of the house connected them in series in an attempt to control all the lights through a single motion-detector, but that did not quite work out.

I did multiple resistance calculations with the power turned off, and adding/removing one or more bulbs to the circuit, to come to this conclusion. I could not check the actual inter-connections, because the wires go through an inaccessible area of the attic on top of the garage.

As a work-around, I have replaced all the edison bulbs in sconce lights with dimmable LED bulbs. When no motion is detected, the LED bulbs glow very faintly. When motion is detected, everything powers on brightly.

Many thanks to the forum members who responded. Talking with you helped me a lot.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.