ARC fault

[alfsauve]

In a worship center with a 45' high ceiling we had one 750W PAR (light) that flickered occasionally. The previous Lighting Director, had taken it out of the lighting controller configuration and didn't use it. The building was about 5 years old when I came on board. I notice the light that was out and tried everything I could think of to spot the problem. Swapped dimmers, hot wired the circuit directly, etc. It wouldn't always flicker. Finally it was flickering on a day the lift was available. I found that the original installer had not tightened down the neutral wire on the socket. Depending on the buildings vibrations it sometimes would make adequate contact and other times it would arc. After 5 years of this it had melted the socket. I only got pictures of the plug but the socket itself was an unrecognizable mess. Now you know why arc fault breakers are important.

 

[alfsauve]

Side story about the original electrical contractor. Seems he ran out of cover plates for twist lock receptacle. Evidently he had plenty of blank plates though and made his own by drilling a large hole. One trouble though, the mounting holes for a blank plate aren't the same spacing as a receptacle cover plate. On a regular plate you're securing the plate to the receptacle. On a blank plate you're securing the plate to the box. This means you can't really tighten the socket in place without cracking the plate. Safety wise it means when you take the cover plate off, the socket is now loose in the box.

 

[RCinFLA]

AFCI's have their place. They work by detecting high frequency noise on AC lines created by arcing wires.

Guess what else creates high frequency noise on AC lines. A high switching frequency sinewave power inverter.

Not all AFCI's are created equal. Their algorithms are continuously evolving to reduce false triggering. First versions would often trip whenever you used a brush commutator motor like an older vacuum cleaner or electric drill.

 

[x98myers7]

Now you know why <not cutting corners and doing a crappy installation> are important.

fixed that for you.

AC power has existed safely for more than a century without AFCI.

 

[alfsauve]

Was just pointing out the dangers of not securing connections. not trying to make a case for or against AFCI. Cutting corners, sloppy work, has consequencies. Same electrical company 10 years later, new building, forgot to wire in the emergency lights. The initial building inspector didn’t catch it. A year later we had a blackout during an event…. No lights.

 

[RCinFLA]

AFCI for catching a frayed electrical cord because a piece of furniture was shoved over slightly and accidentally crushed the electrical cord is justified.

AFCI to compensate for an incompetent electrician is not so justified.

 

[edcoopered]

Its interesting as AFDD (as we call them) are just about coming to market here in the UK. They are predicted to become more common place in years to come, we have a lot of really old buildings with extremely dubious fire safety so its a surprise we're so far behind with this technology. However, I think the AFDDs we have are quite different beasts, as they seem to feature quite advanced signal processing. I just hope the internal power supplies in these breakers are designed to last 20 years plus.

 

[mrzed001]

I am not sure AFDD could catch little arks like these.

Here they tested (with DiY lego tester) after how many Watts will the AFDD disconnect if there is an arc.

Also ... epic AFDD song ?

 

[RCinFLA]

AFCI's have to thread the line between being a false tripping annoyance and a useful safety device.

Arcs happen at the positive and negative peaks of the AC sinewave so there is high frequency spectrum arc spirts at twice the AC line frequency repetition rate. This is what is looked for by AFCI detector.

Listening to an old AM radio is a good way to judge the severity of an arc to cause an AFCI trip.

The sensitivity to triggering is usually set based on breaker amperage rating. In the video he did not mention the breaker amperage rating he was using.

AFCI's can be a real pain in the ass when your AC source is a sinewave inverter. The inverter's high switching frequency leakage on the AC lines can cause the AFCI to trip. To cure a problem you can try a low pass filter on output of inverter. A common mode EMI filter will not help. The LPF inductors cores must be large enough to carry the AC line current without saturating the filter inductors which will diminish the filtering effectiveness.

 

[OffGridForGood]

These have been required in Canada for nearly two decades (since 2003 CEC, 1999 in NEC IIRC), for sepecific parts of a home bedrooms notably initially but has expanded to most other areas since it's introduction. And yes, these breakers cost considerably more than a standard breaker of the same amperage. Although cheap compared to cost of fire damage I suppose.
Interesting aside, roof top mounted PV above 80VDC also requires AFDD but these need to be DC, Some Inverters have this function built in, some do not. Trying to find a stand alone AFDD DC-rated seems to be about as common as unicorns...

 

[RCinFLA]

PV high voltage DC arrays can create some really hot arcing. It will melt back the tips of the wires increasing the size of the arc. DC arcing is broad frequency spectrum.

 

[OffGridForGood]

Wow!
I saw Ian do something like this too with DC from some PV strings on one of his videos.
It lead me to double check every MC-4 connection on my roof array...just to be sure no loose connections up there. The racking is unistrut and the cables run in side the struct with plastic snap in covers. Metal building and roof so at least not like asphalt shingles, but next spring my next array will be on a car port, with asphalt shingle roof...I am thinking metal conduits attached to the racking and run all the cables in the conduits. Thoughts?

 

[x98myers7]

Wow!
I saw Ian do something like this too with DC from some PV strings on one of his videos.
It lead me to double check every MC-4 connection on my roof array...just to be sure no loose connections up there. The racking is unistrut and the cables run in side the struct with plastic snap in covers. Metal building and roof so at least not like asphalt shingles, but next spring my next array will be on a car port, with asphalt shingle roof...I am thinking metal conduits attached to the racking and run all the cables in the conduits. Thoughts?

a detached carport is not a residential building so you don't need AFCI. worst case, the strings arc and you burn up the panels. no big deal. arrays on residential homes is where AFCI is crucial. the only real AFCI solution is to get a SCC with AFCI built in.

I would argue a metal roof on a residence should be excluded from residential AFCI requirements, but i'm sure the NEC would disagree with me. just more bureaucratic BS and oversight to make things complicated and expensive.

 

[OffGridForGood]

the only real AFCI solution is to get a SCC with AFCI built in.

I agree, but how do we know which AIO have this built in, and if built in, how do we know if the AFCI used is listed?

My metal shop roof has been in place since 1981, without issue, survived pleanty of lightning storms etc, but as soon as PV was installed, suddenly it had to be be grounded. So now it is. I guess climbing up there on an alum extension ladder will be safer now.

 

[x98myers7]

It’ll be advertised in the specs. There are very few AiOs that have AFCI. Probably only Solark and Schneider

Separate component systems that are UL listed are much more common. AFCI support is pretty easy to find in SCC. Midnite solar classic series AFCI, GFP built in and UL listed.

 

[400bird]

It’ll be advertised in the specs. There are very few AiOs that have AFCI. Probably only Solark and Schneider

Schneider doesn't make an AIO and their ARC fault is built into a combiner/RSD thing.
A Schneider system uses individual components allowing you to purchase components that fit your needs. VS being stuck with what comes in the AIO box.