Had a potential fire hazard occur a few days ago when a plug for an electri
c baseboard heater (the outlet is behind some furniture) apparently develop
ed a bad connection. Perhaps the plug was not seated properly or was dislod
ged by a tug on the cord? Anyway at some point things got hot. Two smoke al
arms failed to go off and in the morning I was told by a family member that
"the heater isn't working and apparently has fried itself."
There was a heavy smell that something had fried. Upon inspection the outle
t and heater plug was fried and there was a burnt spot on the back of the f
urniture. The white wire insulation inside the outlet was totally crunchy,
so there was considerable heat at the connection and the heat conducted up
the white wire.
I guess an AFCI breaker might have shut this down, and I see one company se
lls temperature sensing outlets. I wish the outlets were more commonly avai
lable. The breakers would never fit my old box.
You were well past a "potential" hazard at that point.
More likely there was a loose connection on the outlet...isn't the
"backstabbed" type by any chance't?
That sounds like a very risky installation/usage pattern, anyway, of
placing an electric heater behind furniture...
Yes indeed. Congratulations Davej. You dodged a bullet
I think only the plug and cord were behind the sofa.
When I had this problem, it wasn't behind furniture, but I was sleeping.
I also had a very old receptacle. Mine was covered by 20 layers of
paint, but in the case of Davej, if it was more than 10 years old, maybe
he should have replaced the receptacle before using it for this,
certainly if it was at all loose or more than 15. I presume you've
replaced it now. Make sure it grips it firmly and after the heater
has been running for 20 minutes, hold the plug in your hand to see if
it is warm. It shouldn't be at all warm.
I've also seen clamps to hold a plug into a recept. Held on by the
screw in the center. Or you can make one
On Sunday, February 16, 2014 3:58:50 PM UTC-6, Stormin Mormon wrote:
The outlet was old, perhaps 20+ years, but not coated with paint. The outle
t is behind furniture but the heater isn't. The cord is not in a place wher
e it would normally be touched or tugged. I remember unplugging the heater
last Spring but I believe a family member plugged it back in this Fall, and
that I suspect may have been the problem. Perhaps they did not plug it in
carefully and fully? I never use the "backstab" holes in outlets because I
don't feel they provide a reliable connection. I think I will be installing
a dedicated outlet for this heater in the near future. Thanks for the comm
coated with paint. The outlet is behind furniture
but the heater isn't. The cord is not in a place
where it would normally be touched or tugged. I
remember unplugging the heater last Spring but I
believe a family member plugged it back in this
Fall, and that I suspect may have been the problem.
Perhaps they did not plug it in carefully and fully?
I never use the "backstab" holes in outlets because
I don't feel they provide a reliable connection. I
think I will be installing a dedicated outlet for
this heater in the near future. Thanks for the
I'm pleased the outlet did not catch fire, and you
and your family are safe. Might be a good socket to
install a two dollar socket, not the bargain basket
59 center. Cheaper than rebuilding.
They are starting to sell AFCI receptacles and the code is evolving to
allow you to simply replace the first one in a string to protect that
receptacle and all down stream from there. It is a compromise since
the first small leg of the circuit is not protected but it is a
compromise in the right direction for people who are not going to do
any extensive rewiring.
The smell produced by overheating was probably not smoke. Not exactly
reassuring, but if there had been an actual fire the smoke alarms should
have gone off.
There can be a loose connection that produces a "glowing connection"
that is stable, and an AFCI won't trip on arc detection because there is
In 2001 UL did testing for Cutler-Hammer on glowing connections. Steel
connection screws (plated)were the easiest to create a glowing
connection. A few amps could produce a glowing connection. The heat can,
obviously, start fires. And the heat can carbonize the plastic, which
can create a leakage path to a ground terminal or the yoke. Since AFCIs
also have a ground fault detector (at 30 mA), that can cause an AFCI
trip. Of 16 receptacle trials, 9 tripped the AFCI. (In 6 the wire burned
open, and in 1 the test was ended after 35 hours.) A glowing connection
might also eventually become an arc.
The earliest AFCIs would only detect high current arcs - between wires.
They would not detect a loose connection, which is lower current than
the circuit breaker rating. More sensitive AFCIs were required in the
2002 NEC, and they can detect a series arc (?above 5 amps). The older
ones will still trip on a ground fault.
As gfretwell wrote, AFCI receptacles are now available. In addition to
installing them as in his post, you could put AFCI receptacles next to
the panel and use the wire-through feature to protect existing circuits.
I wired in three new 20A circuits to service two space heaters and a
convection combo oven. Since I knew where the first outlets were on the
branch circuits I simply replaced those outlets with Leviton AFTR2-W
SmartlockPro Outlet Branch Circuit Arc-Fault Circuit Interrupter
Receptacle, 20-Amp units ($25 each at Amazon) and connected the downstream
outlets through the load connections.
The only real downside of doing that is that when an AFCI outlet trips, you
have to roam around the house looking for it. I've made an addendum to the
ever-growing list of notations inside the panel door of where they are
(along with GFCI's wired the same way) but it still complicates things.
There are circuits I'd like to protect where I don't necessarily know where
the first outlet in the chain happens to be (is there a way to determine
that?) so they would have to be protected by AFCI breakers instead of AFCI
The circuit breaker AFCI's are almost twice as expensive and require a
neutral pigtail connection in the circuit breaker panel. Since my neutral
bar is already at capacity, the outlets seemed to make more sense cost-wise
I have a question for you. Is it "code" to use two neutral buss bars in a
circuit breaker panel? The one I just bought is rated for putting two 12
gauge wires under one screw but I don't like to do that if I can avoid it.
I could remove the old one and replace it with the one rated for two wires
per terminal and double up all the neutrals but I would rather install a
second one on the opposite side of the panel and still connect one wire per
terminal. Just not sure that two separate neutral buss bars is permissible.
I have to admit since reading about what you wrote about the UL tests I am
not sure arc fault protection devices are worth all the effort. I've had
two meltdowns already from plugs that were either not fully plugged in or
had come loose and one because an six-outlet adapter's crimped internal buss
bar worked itself loose. It would be nice to know if the AFCI could have
detected any of those events. I am beginning to think they might not have.
New in the 2011 NEC - if you extend an existing circuit in an area where
AFCI protection is required the extension (including wiring) must be
AFCI protected. Wiring through an AFCI receptacle (as you did) is one
way to provide that protection. Almost every receptacle in a dwelling
must now be AFCI or GFCI protected (when originally wired or replaced).
In the 2014 NEC some must be both.
Also new in 2011
- replacement receptacles in areas where AFCI protection is now required
have to be AFCI protected (several methods)
- replacement receptacles in areas that now require tamper-resistant
receptacles must be tamper-resistant (that is most of the general
purpose receptacles in a dwelling)(keeps kids from putting paper-clips
in the receptacle)
- replacement receptacles in areas that now require weather-resistant
receptacles must be weather-resistant (damp locations, like outdoors)
And replacement receptacles where GFCI protection is now required must
be GFCI protected
AFCI or GFCI receptacles may have to be tamper-resistant or
Or install AFCI receptacles at the panel as in my last post.
You can double-up wires only when the manufacturer says it is allowed.
Look at the label for the panel. Neutrals are never allowed to be
doubled-up - it is a code violation (408.41).
It is relatively easy to add ground bars. The label for the panel should
say what accessory bars can be used. If you have ground wires on the
existing neutral bar they can be moved to the ground bar (and maybe
doubled-up). Neutrals can not be landed on a ground bar (the enclosure
can not be used as the neutral-path connection from ground bar to
neutral bar in a service panel).
I don't know if accessory neutral bars are available. They are not as
easy to install. Other than in a service panel, they have to be
insulated from the enclosure, and you have to figure out the size of the
conductor to connect it to the existing neutral bar.
Was it a high resistance or arc? Some loose connections are an arc. In
the research done for Cutler-Hammer about 1/2 of the high resistance
connections resulted in a ground fault trip when leakage to ground
resulted from the heat. The 30 mA ground fault trip in an AFCI is there
for that reason. And a "glowing connection" may turn into an arc.
The NEC keeps increasing the locations where AFCI protection is
required. Would be nice if there was data that showed they have a major
effect. But they will likely be most useful as the wiring ages.
<good stuff about new 2011 requirements snipped to be moved to a new thread
later> > > The only real downside of doing that is that when an AFCI outlet trips,
There's really not room for that in my panel area. I suppose I could
eventually figure out where the first outlet in a branch circuit is with a
fox and hound. I could ID all the outlets in circuit that way (or by
breaker switching) and then disconnecting the outlet that's the most likely
candidate and testing for power at the downstream outlets but that's a lot
of work and a lot of torquing around of old, cloth covered wiring that is
old and doesn't like to be bothered much anymore. Just like me. (-:
Which clearly makes what I have ground bars.
I was under the impression, probably wrong, that if you connected a
secondary buss bar with a large gauge wire (#8) to the neutral buss bar that
it was acceptable to land neutral wires on the secondary buss bar but now I
am not sure where I read that or if it's true. I realize now that the OEM
neutral bar on my Square D panel sits above all the breakers and looks to be
isolated from the enclosure.
I just looked and it seems that the small buss bar I have can be
piggy-backed to the existing neutral bar, but I am going to have to look up
the manual for the panel to be sure. Now a number of things have become
clearer to me, which is probably good safety wise but is going to mean a lot
of work and/or expense to bring it up to code.
Electrically the panel is connected to the neutral conductor just because
ground wires are connected to the neutral bus, correct? I think you're
saying that the issue is that a secondary neutral buss can't depend on that
connection. The Square D panel has mounts for ground buss bars on both
sides of the panel. Even moving all the ground wires from the original
neutral buss to a auxiliary ground bar won't help because there aren't many
grounds (this is a two-pin outlet house, at least on all the older
I installed a number of dual skinnies just to be able to take high current
outlet loads off the old circuits which now primarily serve low wattage CFL
lights. I also installed one AFCI breaker with a pigtail before moving on to
the cheaper AFCI outlets, in part because of the lack of neutral buss bar
I am not sure what you mean by "other than in a service panel, they have to
be insulated from the enclosure." I thought I did this correctly because I
recall going out and having to buy one foot of 8 gauge wire to connect the
secondary bar going alongside the rows of breakers inside the circuit
breaker panel. I'll have to review my notes about that. I did notice
Square D make two identical forms of the buss bar except one was tinted
green, the other was all silver which made me also believe that they could
be used to ":land neutrals" (I like that phrase) as well as grounds.
One was an arc because you could hear it. The others were most likely high
Someone mentioned thermally protected outlets that cut out when they get too
hot. I've never heard of them. Are they something standard in the
industry? Would they offer protection against high resistance connections
that the AFCI's don't seem to provide?
I suppose that the data will fill in over time. Based on the number of
breaker trips I've experienced in my life, I suspect that breakers have
saved my life a lot more often than most other protection devices including
my firearms. (-:
Thanks for your input, Bud.
Making a map with device locations and circuit number is real useful,
but a PITA. Would give you an idea where the first device is.
If you AFCI the first receptacle and wire through the rest, the device
protects the wiring downstream. All the load is at or downstream. The
AFCI receptacle will protect the wiring back to the panel from series
arcs (loose connections) by disconnecting the load that maintains the arc.
I assume that if there is a wire-connection that would work. You have to
figure out what size wire is appropriate. Slight chance an inspector
might object to using something the manufacturer calls a "ground bar" as
a neutral bar.
A subpanel near the service panel is another possibility. (But then you
will never be able to sell your house - according to one opinion. And
you won't be able to get insurance. And a house inspector may say the
house should be torn down. And it should be torn down. And that does not
include the shame of having a subpanel.)
In a service panel the neutral and enclosure are connected together.
Could be a wire. Often is a screw and looks like a mounting screw but
may be green. If the neutral and enclosure are bonded, an added neutral
bar would not have to be insulated from the enclosure (but may not rely
on the enclosure as the electrical connection back to original neutral
In a subpanel, an added neutral bar must be insulated.
I have never heard of thermally protected outlets. Is that feature UL
And AFCIs may trip on a ground fault caused by a "glowing connection",
or if a glowing connection turns into an arc.
After thinking about the for a while, I realized I could use my X-10 signal
strength meters to discover the first outlet on a branch circuit. If I
place a test transmitter on another branch, the X-10 meter will show
noticeable signal attentuation as measurements are taken at each outlet on
the branch circuit in question. I often use that technique to "zero in" on
noise generators. As you move closer to the noisemaker the noise reading
function of my XTBM meter shows an increase in the noise signal level (in
millivolts). Never thought the technique could ID the first outlet in a
circuit, but I think it will.
I had the panel inspected after I installed the first set of dual skinnies.
But I think at that time I had moved all the ground wires from the panel's
legit neutral bar to the ground bar. I was mostly concerned at the time
that the inspector would disallow the dual skinnies because of the potential
to overload the panel. It was obvious from what I had done, though, that I
was relieving the older cloth-wire circuits of heavy loads (like toasters)
and using the new 12/2 romex instead. The overall load on the panel wasn't
changing, just the distribution of that load. Which brings up another
subject. Today with my tong meter I realized that one side of the panel was
pulling way more amps than the other. I guess that's for another thread,
too. What happens when a circuit panel is unbalanced and has substantially
more of a load on one leg than the other?
Since the last inspection I have added more circuits but the overall load is
still way less than the total the panel is rated for. Unfortunately now the
neutrals are connected to the auxiliary buss bar and not just the ground
wires and I recall that was new work, post inspection.
No subpanels for me! A long time ago I decided that this house would
probably have to be "gutted" when sold so it's not a big issue. Part of the
problem is that it was built in 1941 when there was a shortage of everything
and those shortages were reflected in the materials used in the house. No
building paper between the floors, badly cured wood on the roof, etc.
Fortunately where I live the land's what's worth the most $ and houses like
mine are often knocked down so that larger homes can be built to replace
Gotcha. I may be safe after all. The question now is whether to double up
all the ground wires on the buss bar I installed (with the number 8 wire
connected to the panel's original neutral bar) or to install another ground
bar on the left side of the panel. I am leaning toward the latter because I
don't like the idea of disturbing the old wires yet again and because
there's a lot more free space on the left side.
Another reason not to endure the shame of a subpanel.
Dunno. I'll first have to look up where someone mentioned them here and
then go on to Google to find out more. They have to be pretty simple to
build - just needs a klixon internally to interrupt the circuit in case of
an overheat. Look for that in a separate thread as the initial search
turned up a number of likely candidates. There's probably a third thread to
evolve from this. Is UL listing necessary for approval of the installation
of auxiliary equipment like an X-10 coupler/repeater? Now I remember why I
haven't invited the inspector back . . .
A temperature "aware" outlet seems to be the solution to glowing
connections. More in a new thread.
Thanks for your input, Bud.
On Wednesday, February 26, 2014 5:59:43 AM UTC-5, Robert Green wrote:
How much is way much more?
I guess that's for another thread,
Nothing unless the neutral service conductor was undersized
when it was installed and you put such an unbalanced load on
it that it exceeded it's rating, eg managing to put close to
the full service current rating on just one leg, which is highly
unlikely with typical 120V loads.
I don't know what's shameful about using a subpanel. They
are installed frequently and properly done, nothing wrong with
them. It sure beats tearing out a perfectly good main panel
just to add two more circuits. And having the neutral not grounded
in the subpanel is trivial, they are made to do that. On the
other hand, if you have an old main panel and there are other
reasons to upgrade the whole thing, then I can see going to a new
I think it's required to be listed and while it's typically listed
by UL it doesn't have to specifically be UL, just a recognized testing
lab. Given the crap that most X-10 products are, and since you're
worried about getting thermally protected outlets, AFCI's etc, I'd
think you'd want the X10 repeater listed. Is this an active one or
just a passive one? The passive ones, at least some, are just caps
OK. I remember a long, long time ago I had problems with a Sony monitor at
work that an electrician attributed to an unbalanced panel, but balancing
the loads didn't solve the wavering on the monitor so while rebalancing at
that site didn't help, it probably didn't hurt, either.
I was just "playing along" with Bud who I assumed was recapping odd comments
he's seen people make about subpanels. If I misunderstood you, Bud, I
apologize in advance.
A total rewire is probably what's required here but as I said, when we leave
this house, it will probably be gutted and renovated into a much larger
house with all new wiring. Putting in a new service panel at this late date
wouldn't be cost effective. I would probably have been forced to upgrade a
long time ago except that the advent of CFLs and LEDs really reduced the
overall load. Same with a new fridge and AC. Much lower total current draw
than the old equipment.
Different concerns. I'd already melted more than one outlet with a space
heater whose plug had worked slightly loose. That's a serious fire hazard.
With the XTB-II repeater, I am not worried that it will start a fire.
some, are just caps
This is an active one designed by a guy who built custom electronics for
NASA for several decades. He looked into UL listing but it would have made
the device impracticably expensive. He's provided detailed circuit diagrams
and component lists and enough information to make me comfortable using the
device. We had extensive conversations about worst-case scenarios and while
I'd like it to be UL listed, I don't see it as being a dealbreaker,
especially considering how useful the device has been. You can see for
yourself that this is not typical Chinese made X-10 consumer stuff:
Jeff's design and soldering work is impeccable and the XTB-II pretty much
eliminates all of the typical problems X-10 faces in the world of modern
switching-power supplies. Unlike other repeaters, this unit boosts the X-10
signal to over 25 volts which is usually enough to power through any
Investing in the XTB-II saved me from having to switch to another home
automation protocol and protected my substantial investment in X-10 gear. I
am sure you would change your opinion about X-10 if you saw one of these
units in action and the difference it makes to an "iffy" X-10 setup. The 5
to 10 volt signal strength of standard X-10 gear is just not enough to reach
across phases or fight signal attenuation. That's partly because modern
electronics are pretty noisy and are often designed to choke RFI which,
ironically, is what the X-10 signal looks like to many devices. A 120Khz
noise spike at the zero crossing.
Doesn't sound like much unless it is a 60A panel. It will vary depending
on what is on anyway
There is a certain party here who has that opinion of subpanels, telling
RBM he did schlocky work because he has a subpanel in his house. The
same person has strong opinions about k&t, insurance, and some other
On Wednesday, February 26, 2014 2:23:18 PM UTC-5, Robert Green wrote:
I hear you, but I don't think it would change my overall opinion of X-10.
I'm sure it helps increase the reliability and can make it work where
it otherwise would not. Part of the problem is that they are not upfront
about this problem. I'm sure a lot of people wind up not being able
to get it to work because it really should have a bridge/repeater across
the phases. Then if you want such a device, they are expensive and
as you've found out, have problems of their own, ie not UL listed and
there is no guarantee that they will make it all work either. Or maybe
you get it to work for what you're currently doing, then want to add
something else and that won't work.
The other big problem is that AFAIK, no one is doing anything new
with X-10. What you have now, you've pretty much had the same stuff
for more than a decade, ie no new products, no improvements, etc.
The outdoor sensors don't last, that's another problem. I'd still
use it for a simple application, where I put $20 in it and if it works
it works, but I would not choose it as a means to more extensive home
Panels will have a rating something like 10/16. There are 10 full size
breaker positions and you can put 16 "poles" in it. A tandem breaker is
2 poles, so you could install up to 6 tandem breakers. The label should
tell you the rating.
The panel is tested with the maximum number of poles - 16 above - and UL
does not allow installing more poles than that. That is done by only
allowing tandem breakers in certain positions (6 in the example above).
The label will indicate which positions tandem breakers can be installed
in. In a 10/20 panel it is all positions. It may also be no positions. A
SquareD tandem breaker has a bar on the bottom that has to fit into a
slot through the gutter rail, which exists only in the positions where
tandem breakers are allowed. These are class CTL panels and breakers
(circuit limiting) and they have been around a long time.
The inspector wants to know that tandem breakers were installed in
positions where they are allowed.
There are also non-class CTL tandem breakers that can be installed in
any position - for older panels that are before CTL came out. I know no
one here would ever install one of them in a class CTL panel.
Panel is a QO CAT QO BW - 20M - 100 - 5
I believe that the 20 refers to the max # of breakers and the 100 is the max
current. Otherwise the catalog number don't seem to agree with the listing
in the PDF. No idea what BW means. It's hard to read the label because
it's a) faded, b) upside down and c) obscured by neutral and ground wires.
This has to be a pre 1985 panel and I haven't found it yet although I found
a few like it in this PDF:
From what I can see, that's not an issue with my panel. As I recall, the
hot bar (not sure of the technical name of the alternating metal "fingers"
running down the center) was the same from the top to the bottom - no
specialized slots for tandems (why would there be?). The replacement
breakers have a very different connection mechanism that the original ones.
The ones that came with the panel have two pronged clips that make the
connection to the rail but the replacements have a single claw-like
"grabber" - so different looking that I was sure I bought the wrong
I am pretty sure that at least on of the labels on the panel has the letters
CTL but if it came out after 1985, then I definitely don't have a CTL panel.
From what I can see of the PDF listed above, the breaker panels that support
special slots for tandems look quite different from the box I have which is
a series of alternating metal fingers running down the heavy black plastic
"spine" of the panel.
Talks about CTL being in place since 1969 and replacement breakers being
non-CTL. Even if the panel ends up "dangerously overloaded" isn't the main
breaker supposed to trip before anything bad happens?
HomeOwnersHub.com is a website for homeowners and building and maintenance pros. It is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.