GFI outlet Thermal mystery

[CrystalVoid]

good day everyone!

so the other day I picked up this cool new toy.. the "FLIR one" (loads of fun) and I was looking around my apartment and noticed this..

the GFI outlets in my apartment .. with nothing plugged into them are 10 to 20 degrees Fahrenheit  (+15c ish) warmer then the walls.
I have attached a sample image as well as uploaded it here https://dl.dropboxusercontent.com/u/4185122/grumble/GFIflir.png
is this normal? should I be concerned?
would anyone with a thermal camera mind confirming my findings?

thank you all

[helius]

Do you leave the outlets in the "reset" position? Neon indicator lamps become warm when lit.

[IanB]

I think a GFCI is an active device with an internal power supply and some electronics. If so I would expect it to consume some power while energized.

[CrystalVoid]

I do not think this model has a lamp (none visible) .. however I will trip one and see if it cools down .. maybe there is one inside that's not visible.
thanks

[Chris C]

Interesting.  A quick Google search list examples of GFCI outlets tested to consume power between 2-7W.  I had no idea they consume that much.  Seems your thermal imaging is credible.

Now if you see that kind of temperature rise on an extension cord, or where the plug enters the socket, that's an issue.

[CrystalVoid]

hmm interesting this is after ~2hours of it in the tripped off position   https://goo.gl/photos/TgfFzWq61x1AC6XR8

[Monkeh]

hmm interesting this is after ~2hours of it in the tripped off position   https://goo.gl/photos/TgfFzWq61x1AC6XR8

Output off != circuitry off.

[nbritton]

GFCIs are fairly complex circuits, I believe they even have microcontrollers in them. It would not surprise me if they ran hotter then regular receptacles.

[Monkeh]

GFCIs are fairly complex circuits, I believe they even have microcontrollers in them. It would not surprise me if they ran hotter then regular receptacles.

Uhm, a regular receptacle is nothing but some brass stampings secured in a plastic box, so they don't generate heat at all..

[Zero999]

Interesting.  A quick Google search list examples of GFCI outlets tested to consume power between 2-7W.  I had no idea they consume that much.

That's terrible. It probably wouldn't be allowed in the EU!

Why not just have one breaker where the power enters the property? Having a separate GFCI in each socket seems silly.

[IanB]

Why not just have one breaker where the power enters the property? Having a separate GFCI in each socket seems silly.

Because if that one breaker tripped your whole house would be plunged into darkness. Not to mention your fridge and freezer would stop as well. That would be a terrible idea!

Robust designs should follow the principle of least action. If one circuit develops a fault, isolate the one circuit. Don't cause collateral damage...

[cs.dk]

Why not just have one breaker where the power enters the property? Having a separate GFCI in each socket seems silly.

It done that way in Denmark. I think most of Europe is doing it that way to?

[HackedFridgeMagnet]

In Australia, in new installations we normally have one RCD (GFCI) combined with an overload circuit breaker per sub circuit. So maybe between 5-20 per household.
Dont know what power they consume.

[fubar.gr]

GFCIs are fairly complex circuits, I believe they even have microcontrollers in them. It would not surprise me if they ran hotter then regular receptacles.

This seems to be the case with US spec GFIs.

European ones are entirely electromechanical, no electronic parts inside. Also, they do not consume any power and in theory they can work at very low voltages and even DC.

I just tested an ABB made one, I can trip it using a AA battery.

[Monkeh]

GFCIs are fairly complex circuits, I believe they even have microcontrollers in them. It would not surprise me if they ran hotter then regular receptacles.

This seems to be the case with US spec GFIs.

European ones are entirely electromechanical, no electronic parts inside. Also, they do not consume any power and in theory they can work at very low voltages and even DC.

Errr... no, certainly not all of them..

[SeanB]

Definitely not the newer ones. The older ones used the imbalance current to drive the trip coil, but the new ones use a small capacitive dropper inside to power the single IC that does the work. The capacitor does draw a small amount of reactive current and the electronics run off a low voltage ( capacitor is a 47uF 10V tantalum) rail provided by the 330nF 300VAC capacitor inside.  Teeny little 8 pin IC inside, HML235, with no data available of the part number, but it does detect the imbalance current and provide the drive for the trip coil. The terst button operates a sense input on the IC, not a dedicated sense winding and 47k resistor like the older ones.

[Zero999]

Why not just have one breaker where the power enters the property? Having a separate GFCI in each socket seems silly.

Because if that one breaker tripped your whole house would be plunged into darkness. Not to mention your fridge and freezer would stop as well. That would be a terrible idea!

Robust designs should follow the principle of least action. If one circuit develops a fault, isolate the one circuit. Don't cause collateral damage...

I should have been more clear. I didn't mean one breaker for the whole property but one for each circuit, like how it's done in the UK. The kitchen is on one RCD, the lighting on another and rest of the house on another RCD. It may be more of an inconvenience because lots of appliances are turned off when there's a fault but at least it doesn't result in 2-7W wasted per socket.

[helius]

I should have been more clear. I didn't mean one breaker for the whole property but one for each circuit, like how it's done in the UK. The kitchen is on one RCD, the lighting on another and rest of the house on another RCD. It may be more of an inconvenience because lots of appliances are turned off when there's a fault but at least it doesn't result in 2-7W wasted per socket.

In the US (and I guess in most countries that don't use ring mains) there is one main fuse/breaker for the premises and 20-30 circuit breakers in a panel, each rated for the appliance it feeds. The outlet GFCIs are common, but they don't need to be in every outlet; a single GFCI outlet can protect the entire circuit if it is wired first in line. So I don't see moving the RCDs from outlets to the panel would save much energy.

[dom0]

Well that's interesting now, isn't it?

I never knew until now that there are electronic RCDs for domestic use, I only knew the usual electromechanic RCDs (power usage = 0 Watts). Honestly I'm surprised that people spend that much money on RCDs, the electronic ones I know are all very expensive industrial stuff...

[sunnyhighway]

In the US (and I guess in most countries that don't use ring mains) there is one main fuse/breaker for the premises and 20-30 circuit breakers in a panel, each rated for the appliance it feeds. The outlet GFCIs are common, but they don't need to be in every outlet; a single GFCI outlet can protect the entire circuit if it is wired first in line. So I don't see moving the RCDs from outlets to the panel would save much energy.

But wouldn't that make the line between the circuit breaker panel and the first GFCI outlet unprotected from ground faults?
This looks like playing Russian roulette to me when drilling a hole in the wall.

[Monkeh]

In the US (and I guess in most countries that don't use ring mains) there is one main fuse/breaker for the premises and 20-30 circuit breakers in a panel, each rated for the appliance it feeds. The outlet GFCIs are common, but they don't need to be in every outlet; a single GFCI outlet can protect the entire circuit if it is wired first in line. So I don't see moving the RCDs from outlets to the panel would save much energy.

But wouldn't that make the line between the circuit breaker panel and the first GFCI outlet unprotected from ground faults?
This looks like playing Russian roulette to me when drilling a hole in the wall.

The Americans have little concept of circuit protection like that. They also have no concept of running wires in predictable places, they just string them wherever is convenient in the wall.

[IanB]

The Americans have little concept of circuit protection like that. They also have no concept of running wires in predictable places, they just string them wherever is convenient in the wall.

That depends on whether the lobbyists with vested interests have managed to get to the politicians. In places where they have the regulations require all cables to be installed in metal conduits, for example, and also require arc fault circuit interrupters to be installed at the panel (the stock price of AFCI manufacturers skyrocketed when they got that one on the books).

[Monkeh]

The Americans have little concept of circuit protection like that. They also have no concept of running wires in predictable places, they just string them wherever is convenient in the wall.

That depends on whether the lobbyists with vested interests have managed to get to the politicians. In places where they have the regulations require all cables to be installed in metal conduits, for example, and also require arc fault circuit interrupters to be installed at the panel (the stock price of AFCI manufacturers skyrocketed when they got that one on the books).

Please don't bring politics into this, they have nothing to do with it.

[Troy]

The Americans have little concept of circuit protection like that. They also have no concept of running wires in predictable places, they just string them wherever is convenient in the wall.

"But muh electrical system superiority"
GFCI circuits in the States are often fed from the first receptacle in the circuit, as it is more for protecting the person using an item plugged into that outlet rather than someone bashing a hole in the wall and chewing on the cable.


What the source of heat is likely from is a poor wire splice in the back of that box, or a loose connection.

[Monkeh]

GFCI circuits in the States are often fed from the first receptacle in the circuit, as it is more for protecting the person using an item plugged into that outlet rather than someone bashing a hole in the wall and chewing on the cable.

Thank you for proving my point.

What the source of heat is likely from is a poor wire splice in the back of that box, or a loose connection.

Modern GFCIs, as you call them, actually do produce heat. Some actually hold the contacts closed with a solenoid.

[bookaboo]

If it looks hotter than other similar sockets I'd take it apart for further inspection (assuming you are qualified to do so). Also you want to do thermal imaging under as close to maximum load as possible. You won't pick up anything like bad connections on a socket unless it is under or has recently been under load.

If there's some circuity in there it could be faulty or of just plainly bad design, I've seen numerous PCB components sitting at >100 degrees C under normal operation.  Clearly design faults and obvious with the benefit of TIC. Furthermore some of them I noted in reputable brand industrial products.

[Zero999]

So I don't see moving the RCDs from outlets to the panel would save much energy.

Because less RCDs are used. Instead of having one per socket. One RCD can be used to protect 10 sockets. In which case the power dissipation would be reduced by a factor of 10.

[Teledog]

My 2 cents;
The GFI on our older (Federal Pioneer) breaker box also shows up as "warm" with the Flir - we also have a few devices connected to that GFI circuit's outlets.
Neighbor's breaker box is the same make/model, but his GFI breaker shows cold/same as the other breakers, but then he doesn't have anything connected on his GFI circuit.

[helius]

Because [fewer] RCDs are used. Instead of having one per socket. One RCD can be used to protect 10 sockets. In which case the power dissipation would be reduced by a factor of 10.

The point is what do you trip when a current imbalance is detected? In the US, a 15A circuit breaker for (dual) wall outlets usually feeds fewer than 5 outlets. Whether the RCD is inside the circuit breaker, or the first outlet in line, only those 5 outlets can be protected by a single device.

The other issue is what the sensitivity needs to be to prevent nuisance trips. The greater number of appliances protected by the RCD, the higher trigger level that is needed. The standard for GFCI outlets is 6mA. In the UK, regulations provide for 30mA RCDs! How they can tolerate such an unsafe situation is a mystery.

Another fact that people don't like to talk about is that at sensitive trigger levels, many types of electronics generate ghost voltages high enough to trip the circuit. When the chassis ground of one device is floated to half the mains by its line filter, and it is bonded by an I/O cable to another device on a different GFCI, each GFCI sees an imbalanced current flowing. Because of this problem, some countries use special plugs for electronics that mate with special, non-RCD, outlets.

[gaijin]

Tested a older unused US Leviton GFCI.
At 122v it was drawing 6.58mA about 802mW on or tripped.

[CatalinaWOW]

In the US (and I guess in most countries that don't use ring mains) there is one main fuse/breaker for the premises and 20-30 circuit breakers in a panel, each rated for the appliance it feeds. The outlet GFCIs are common, but they don't need to be in every outlet; a single GFCI outlet can protect the entire circuit if it is wired first in line. So I don't see moving the RCDs from outlets to the panel would save much energy.

But wouldn't that make the line between the circuit breaker panel and the first GFCI outlet unprotected from ground faults?
This looks like playing Russian roulette to me when drilling a hole in the wall.

   

The Americans have little concept of circuit protection like that. They also have no concept of running wires in predictable places, they just string them wherever is convenient in the wall.

Actually there are some rules.  Wiring has to be placed far enough behind to wall surface to make contact with nails and screws unlikely, and even reduces the chances of hitting with a drill. 

Then if you think about what it takes to cause an electrical shock by drilling into a wall you realize that lottery ticket winnings look likely in comparison.  First you have to hit the wire.  Based on area this is less than one in a thousand, unless you look at outlet placement and other clues to wire location and aim for the wire.  You then have to be using an old style un-insulated drill.  And have either a wiring fault in that drill, or be standing on an conductive floor.  And hit the conductor which is hot relative to your wiring fault, or whatever.

The more real danger is damaging the insulation on both conductors and causing a short.  In the immediate case with the drill bit the breaker is likely to blow, but there is also the possibility of dragging some foil from building insulation or other source into the area and causing a resistive short that starts a fire over time.

It is Russian roulette, but in a revolver with thousands or millions of empty chambers.  I personally just pay attention to likely wiring routes, turn off the power, drill shallow holes that don't reach the code required depth of wiring, and once the wall is opened verify that nothing was there.

[Monkeh]

The other issue is what the sensitivity needs to be to prevent nuisance trips. The greater number of appliances protected by the RCD, the higher trigger level that is needed. The standard for GFCI outlets is 6mA. In the UK, regulations provide for 30mA RCDs! How they can tolerate such an unsafe situation is a mystery.

30mA is not unsafe. Please remember that the devices have never, will never, and cannot limit shock current. A tested, functional 30mA device will not allow a harmful shock to continue.

Another fact that people don't like to talk about is that at sensitive trigger levels, many types of electronics generate ghost voltages high enough to trip the circuit. When the chassis ground of one device is floated to half the mains by its line filter, and it is bonded by an I/O cable to another device on a different GFCI, each GFCI sees an imbalanced current flowing. Because of this problem, some countries use special plugs for electronics that mate with special, non-RCD, outlets.

This problem is of almost no concern with 30mA devices.. odd, that.

Actually there are some rules.  Wiring has to be placed far enough behind to wall surface to make contact with nails and screws unlikely, and even reduces the chances of hitting with a drill.

And still it more or less runs wherever it likes. Sure, it may be ~2" back from the wall surface, but.. where is it? On this stud, that stud, strung between them? Does it cross studs at the box, at the ceiling, at the floor, or at some utterly random point in between the electrician found convenient?

[CatalinaWOW]

And still it more or less runs wherever it likes. Sure, it may be ~2" back from the wall surface, but.. where is it? On this stud, that stud, strung between them? Does it cross studs at the box, at the ceiling, at the floor, or at some utterly random point in between the electrician found convenient?

Can't argue with that.  Just haven't found it to be a huge issue.  If you are in the rework trade there are convenient devices for sensing wire location.  Feeds a little known frequency in, and has a receiver tuned to that frequency.

There is also an interesting safety philosophy question.  The US practice will occasionally result in harm to some fool who doesn't take precautions for wires behind the walls.  The practice in your country will occasionally result in harm to the fools who don't know where the wires should be, and also to the person who "knows" where the wires are and didn't think that some fool or expat American put the wires behind that wall.  Which results in lower total damages?  I certainly don't know.  It does seem to me that the practices in both countries result in very low problem rates.

[Monkeh]

And still it more or less runs wherever it likes. Sure, it may be ~2" back from the wall surface, but.. where is it? On this stud, that stud, strung between them? Does it cross studs at the box, at the ceiling, at the floor, or at some utterly random point in between the electrician found convenient?

Can't argue with that.  Just haven't found it to be a huge issue.  If you are in the rework trade there are convenient devices for sensing wire location.  Feeds a little known frequency in, and has a receiver tuned to that frequency.

I know of toners. Used a few, and usually the receiver isn't all that frequency locked (hold 'em up to something electronic, hear the noise.. you can pick up a big TV from a few feet).