Author Topic: Arc fault trickle lights LEDs (Read 11679 times)

Chuck · « **on:** March 04, 2009, 04:51:37 PM »

I'm just finishing up my house, getting ready for the final inspection. When I did the electrical I was required by code to install and arc fault breaker for the bedroom circuits. Newer houses are soon to require arc-fault circuits everywhere, at least in my county. Not a big deal other than a bit more expensive. So far universal motors haven't popped the breaker.

Then the other day I picked up some LED lights and was eager to make use of 1.3 watt luminescence. I put them in light sockets on the bedroom circuit. They work fine, although they aren't as bright as an equivalently rated (40 watt equivalent) CFL.

I discovered that I now have night lights. Even when turned off they glow softly. I'm not sure how exactly, since the circuit is not complete. If I pop the breaker, the light goes out completely. If I put the LED bulbs on a different, non arc-fault circuit, there is no light produced. As soon as I put them back on the arc-fault circuit, they glow. When turned on they put out the normal amount of light.

I'm going to assume this has something to do with how arc-fault breakers operate, but haven't found a good explanation anywhere. Does anyone know the details of how arc-fault breakers work ?

ghurd · « **Reply #1 on:** March 04, 2009, 03:32:48 PM »

LEDs will glow fairly well with 1/100th of a ma.

US and UK are not the same.

Need to know where you are to begin guessing at the solution.

I am guessing it is UK, because I do not know the term "arc-fault breakers"

paulb662000 · « **Reply #2 on:** March 04, 2009, 03:57:22 PM »

US electrical code now calls for these. A note of the code item and a short description is here. http://arcfaultcircuitbreaker.com/

Looking at the section "HOW THE AFCI WORKS" partway down this page http://arcfaultcircuitbreaker.com/afci_fact_sheet.htm I wonder if the sensing circuit could be feeding the tiny amount of power needed.

Paul

Chuck · « **Reply #3 on:** March 04, 2009, 04:05:12 PM »

Nope, USA. I hadn't heard about them either until a friend who's an electrician clued me into the code change.

ghurd · « **Reply #4 on:** March 04, 2009, 04:11:07 PM »

"GFCI"?

Yes, they leak.

G-

Chuck · « **Reply #5 on:** March 04, 2009, 04:25:52 PM »

Hi Paul,

Yes, right on all counts. Unfortunately the how it works section is very short on specifics.

I surmised that there must be a trickle of current being utilized by the sensor, but I still don't know how a current, no matter how small, can be flowing in a circuit that has the switch turned off. Both hot and neutral connect to the breaker, but the other end (switch) is open, no contact, thus no current flow. The sensor circuit can only flow within the breaker. Or so my logic leads and dumps me. So I'm still confused.

Chuck · « **Reply #6 on:** March 04, 2009, 04:32:09 PM »

Ghurd,

Arc-Fault are not GFCIs. They are about twice as expensive. AFCIs deal with normal overloads and also arcing or sparking. Somehow they are supposed to allow brushed motors to function, but pop when sparks happen in the house wiring, like a loose connection might cause. Both types can be touchy. GFCIs certainly can be when running several on the same circuit.

Madscientist267 · « **Reply #7 on:** March 04, 2009, 06:56:00 PM »

Your LED nightlights are probably not truly 'off' but being turned off by a transistor/SCR/triac.

If there is sufficient noise produced by the sensing circuit in the AFCI, there may be transients that are sufficient to pass through. One of those cable sniffers may be able to provide some insight, if straight out ripping one apart doesn't reveal the answer.

If the LED lights have a CdS cell in them to detect light, it is highly possible that this is how they are set up, and so will not light up when plugged into a regular outlet. Small power supplies inside such things are notorious for using capacitive-transformerless designs, which are rather 'sensitive' to high frequency switching like what may be happening inside the AFCI.

Not 100% on this, but it's likely.

Steve

dnix71 · « **Reply #8 on:** March 04, 2009, 08:26:35 PM »

Arrgh. Will these things trip if you are a sloppy when arc welding? GFCI's can be touchy enough.

cardamon · « **Reply #9 on:** March 04, 2009, 08:58:26 PM »

I hate arc fault breakers. As an electrician, I am aware of the extensive debates on these things. Personally I feel they were rushed into the code before they were proven to function or provide sufficient extra protection to be worth the cost. I think pre-twisting connections before wire-nutting and not using the back stabs on switches and receptacles would provide far more safety that an AFCI. Sorry this doesn't address the OP, rant over.

Opera House · « **Reply #10 on:** March 05, 2009, 02:47:53 AM »

It possibly has nothing to do with the breaker but with the wiring. If you temporarily replaced that breaker with a standard one it will likely still glow. A long parallel line going to the switch from capacitance or wiring leakage. I remember suggesting using common house cable for a resistive sense application and it was pointed to a wire to wire resistance. Older mercury disk switches also can leak with age as the glass begins to plate with mercury.

halfcrazy · « **Reply #11 on:** March 05, 2009, 04:16:22 AM »

Cardamon I have to agree 100% they wherent ready for prime time

bj · « **Reply #12 on:** March 05, 2009, 02:12:07 PM »

Ghurd just for info--they became code in Alberta in 99 when I was building this house. Don't know much about them except at that time

they were about triple the cost. Less now I think. Just tripped mine,

and could not read anything with meter on MV, except usual noise on this

setting.

bj

paulb662000 · « **Reply #13 on:** March 05, 2009, 03:25:04 PM »

The code calls for the arc faults to be on indoor living space circuits. Garage, outdoor, and circuits near water (kitchen, bath, laundry) are GFCI. So, as long as you are not welding inside you should be OK. I have talked to a number of electricians who have said their number 1 call back on new construction is arc fault breakers tripped by brushed motors in vaccum cleaners.

Paul

Madscientist267 · « **Reply #14 on:** March 05, 2009, 07:25:34 PM »

To add to this;

I think if they got rid of 'back-stab' connections altogether, there's 99% (ish?) of the problem to start with.

I moved into new construction, and can't tell you the number of outlets and switches I have had to go behind 'the professional' on, and fix the crap that he built my house with. In fact, if the wiring problems keep up, I'm going to look into going about reparations. Seeing what I am seeing, it reminds me of Quest plumbing. Why fix something that is inherently broken with some more broken crap?

I personally would be willing to pay an electrician for all the extra time it would take to turn screws like the old days and do the job right the first time, than to deal with the BS that comes from some short cut that some idiot developed somewhere. GFCI I agree with. Protect humans from humans. AFCI? Thats crap. If something isn't safe, it shouldn't be in the circuit to start with. PERIOD.

</end rant>

Steve

Madscientist267 · « **Reply #15 on:** March 05, 2009, 08:26:03 PM »

Oh yeah, and did I mention?

Not ONCE did an AFCI TRIP due to anything. Not from lack of valid circumstance, I promise.

Flux · « **Reply #16 on:** March 06, 2009, 06:53:27 AM »

Leds need so little energy to make them glow that many devices with any capacitive paths will do it. In this age with filters and capacitors everywhere I am not surprised that things glow a bit.

With all the switching power supplies that seem to be coming on line to replace transformers I think we shall have a lot of strange phenomena to explain.

Flux

Chuck · « **Reply #17 on:** March 06, 2009, 09:13:45 AM »

Thanks everyone for the comments. At this point I'm fairly happy with a mix of LED and CFLs. With arc fault on just two circuits I only have a few places where the LEDs glow softly while turned off. If they are truly glowing only on the noise of the arc fault breakers, which have to be there anyway, I'll accept the night lights. It does seem odd that the breakers would so noisy and/or amazing that LEDs are so sensitive. It reminds me of the old photos of people holding florescent tubes under power lines and watching them glow.

Gordy · « **Reply #18 on:** March 07, 2009, 12:38:47 AM »

Chuck,

I noticed the same thing with several strings of LED Christmas lights. Though they are on a regular circuit breaker. I assumed something was wrong with the switched cord I used. It looks like an extension cord but has a switch on one end and combination male/female plug on the other end.

It's interesting that the blue and green are the brightest, red barely glows, and the yellow and orange no glow at all. Although they don't glow bright enough to see where your going.

Grud,

The arc fault breakers that I have (in MN, USA). Look like a regular CB except they have a yellow test button and if you pull the panel cover they have a white wire coming from inside that is hooked to the ground / neutral bus.

Gordy

News:

Author Topic: Arc fault trickle lights LEDs (Read 11679 times)