Since we have a few electricians on the site I thought you might appreciate this one. I have fixed the problem, but it is interesting and the actual cause is a puzzle:

Basement bathroom has a heater/fan/light. There are three switches, a duplex and a pilot light switch. One circuit feeds them all. The top switch on the duplex then goes to the light and the bottom goes to the fan and the heater switch (so that the heater cannot be run without the fan). The heater switch is the one with a pilot light.

Anyway, I bought the house (build in the 50s) a couple years ago and we noticed one or more of these switches would spark when you turn them off. So I investigate and find the circuit has a 20 amp breaker, but the switches were rated 10 amp! So I special order 20 amp switches and replace them. When removing the old switches I find a toasted wire that was the jumper between the fan and heater switch. But then the new 20 amp switches also sparked when you turned them off--unless you turned them off slowly, so I figured it was not necessairly a problem. However, the switch plate would get warm in the middle when the heater was run.

A year later (two days ago) my wife says the light won't shut off until the switch is all the way on off and just a slight jiggle will light it. I said we have a problem and tore into it. I found melted plastic around the terminal that the hot wire led into, which happened to be the terminal adjacent to the top duplex switch which was wired for the lights. When I installed a new switch (15 amp and a new 15 amp breaker in the circuit, because I later had learned the fan/heater was rated for a 15 amp circuit), I noticed some corrosion or something on the copper hot lead and shined it up with fine sandpaper. Figured there must have been resistance in the connection (though it was tightly connected) to cause the heat. New hookup tested 0 ohms. Ran the fan and heater for a while and the switch plate no longer gets warm; also no sparks on shutting off the switches.

Tore apart the partly melted switch to find that the plastic stub inside that works the switch on and off was melted and crushed down--no wonder it would not shut off. So, was this just a defective switch (Levitron, not an off brand) or was the wire corroded and causing resistance (all copper wire and #12 gauge). Also, if the switch only got hot with the heater on, which was the bottom switch, why was it the top switch (lights) that melted? The lesson I take from this so far is 1) check new switches with ohm meter before installation and 2) if switch plate feels warm, check connections with ohm meter for resistance.

Also, they wanted 60-70 bucks for a GFCI breaker (Bulldog Pushmatic) so I just went with a regular breaker and installed a GFCI duplex outlet between the breaker and the bathroom ($10 breaker, $14 GFCI, saved $40). I wanted GFCI since this circuit also has a light inside the shower! Anyway all is working well now, but am just curious what you electrician folks have to say.

 

I'm having a little trouble figuring out how you wired this. Could you draw a diagram?

 

Loose or corroded connections produce resistance. This transfers into heat. Even pushing the switch back into the box can sometimes make a wire loose. The loose connection was probably on the top switch. And even then, heat will conduct down the conductor to the next device. With the bottom switch on, you're flowing current through the top switch, screw terminal.
Wow, bulldog breakers. They are old style.
I'm glad you put a gfi in that circiut. There's a specific way to wire them. You can wire them wrong, and the devices protected down stream, will still work, but not trip the gfi.
Those electrons like doing funny things.
I think you've learned a valuable trouble shooting lesson.
I hope you didn't receive any electric shock therapy with your repair.
zanny

 

BUBBA SHRIMP: I can draw a diagram, but would have to figure out how to post it. Maybe I can describe it. There are two switches, a duplex (two switches ganged with the hot terminals connected and the neutrals separate) and a single with pilot light. Hot comes into the duplex on the connected side to feed power to both switches. The neutral side of the top switch then goes to the light. The neutral side of the bottom switch goes to both the fan and the pilot light switch. The pilot light switch neutral side then goes to the heater. So when the top switch is on the lights are on. When the bottom switch is on the fan is on, and when the bottom switch and pilot switch both are on the fan and heater are both on. You might be able to draw it from that. Also see last paragraph below--turns out it is not fixed.

==============================================

ZANNY: I have had the connection loosen by pusing the switch back into the box. I usually try to crip the wire around the screw--not easy with the limited work space and the screws are made to not come out all the way.

The Bulldogs are a trip (pun intended), but they work.

Yes I had a GFCI at my last house that let to the refridgerator, so I hooked the refridge to the line side so it would not go out on the trip. Power must to the the GFCI LINE terminals and the subsequent circuit (load, additional boxes) must be connected to the LOAD terminals.

Unfortunately, I did not fix the problem. My test was no good. Didn't run the heater long enough. When my wife took a shower, she ran the heater good and long! And the center of the switch plate was warm like before. So back to the beginning. Suspect maybe the heater is bad, because we ran just the fan for a long, long time and it did not heat up.

 

Okay, got the diagram drawn. Probably could have drawn it from your first post if I would have paid more attention.

You said you cleaned off the copper wires. Are they copper, or could they be aluminumn? Old houses used to use aluminumn wire. Copper and Aluminumn don't mix uless you use a special paste on the connections.

Zanny already pointed out that the current for both the fan and heater are going through the first switch. If you got a 20 amp switch and a 15 amp breaker then I would assume the breaker would go before the switch would break down, however, all the current is going through the connection to the top switch. Zanny also pointed this out, and like Zanny said you really need to have a good connection.

If you can get enough wire out of the wall, you should unscrew the top screw as far as it can go, then using needle nose pliers bend the wire into a loop and put it over the screw. Then, using the needle nose pliers squeeze the loop shut. Be sure you have open end of the loop in the right direction so that when you tighten the screw it is tightening the connection, rather than opening it back up.

If you can't make a real good connection, things are going to get hot.

If you've done all that and it still gets hot, then I'm at a loss as to what to tell you. I think if the heater was bad, it would just not heat. Any loose connections in the heater would cause those connections (in the heater) to heat up and probably result in less current going through the switch, not more. Zanny has more experiance with these kinds of things so maybe there is something I am missing.

You might try wiring the heater switch direct instead of going through the fan switch, just to see what happens.

 

Thanks, Wire definitely is copper (by visual inspection, but also house built 1950) and 12 gauge (which makes is a pain to bend and crimp around the screw). Possibly I mixed a #14 wire in, but with a 15 amp circuit that should be OK. After the hookup and pushing the switches into the wall I checked with the ohm meter and got zero resistance between the copper wire and the terminal on each connection for the duplex switch. Can't get wire out of wall any more. Am able to wrap the wire in a U around the screw and always clockwise so tightening dosen't loosen. But logically it has to be in the connection and a 1.5 volt battery in my ohm meter may not meet resistance, but when 10 to 15 amps are flowing, that may meet resistance, so the ohm meter may be useless for this. Your idea of wiring the heater switch separate from the fan is interesting. I can either run a jumper from the other terminal where the hot comes into the duplex, or bundle two leads with the hot wire using a screw on connecter, which would then isloate the heater even moreso from the fan switch. However, I am not sure I want to leave it this way since I don't think the fan is supposed to run without the heater leat it overheat the heater surroundings. Hmmmm, Could have the fan switch as is and the heater switch operate both the fan and heater independently of the fan swtich. But all this re-wiring does not seem like it would help. I had better cut the power and make sure the wires are mechanically firm to the terminals and double check that the wiring diagram I gave above is correct.

BTW: I just posted this problem to an electrical forum at http://forum.doityourself.com/ to broaden the possible responses.

Her own mother lived the latter years of her life in the horrible suspicion that electricity was dripping invisibly all over the house. - James Thurber

 

Your current draw may be too much for the switches you are using. Ever hold a 18/20AWG(the household type with 3 or 4 plugs at the end) extension cord and use a circular saw for a few minutes? The cord will get hot.

Even with 20amp breakers and switch, you may just have too much current draw. Obviously, the 14ga wire will not conduct as much as the 12ga and will get warm before the 12ga. I would see if you can find out what the current draw is on the fan, heater, and light. It sounds like your power source is one 20amp line. I would think that the fan is 3-4amp, the light is probably about the same, but that the heater is 15amp. That puts you over the 20amp fuse and you are drawing more than 20amps through the duplex switch.

Just a few thoughts. I am not electrician by trade, but have done some wiring of garages, re-wiring aluminum houses, and trouble shooting mobile home wiring.

Check for reverse polarity. You can buy a tester for a few $$ at the local home improvement store. Someone might have switch wires on the heater and that might be the whole cause.

 

Wouldn't the breaker kick if the draw exceeded 20 amps (which it was until I put the 15 amp breaker in)? I put the 15 amp breaker in because when I pulled the heater/fan/light assembly out to clean it last year the sticker on the back said it should be installed with a 15 amp fuze/breaker (that's the heater/fan/light combined). All the wiring is 12 gauge, but the one switch is 15 ga, which should be fine with 15 amp breaker. Reverse polarity check is a good idea, but I already checked all wires in this box and there is only one hot wire, which is a black wire. Still, re-evaluating the way it is wired and re-checking all this is in order. I have both the neon light tester and the multimeter tester. As for how this thing was wired when I bought the house, they had a 20 amp breaker with device that stated to use a 15 amp breaker and 10 amp rated switches.

 

I was thinking that maybe the wires got crossed at the heater/fan/light combo in the ceiling.

Just an idea.

I hate electrical. Those pesky electrons are hard to see.

 

You are on to something. Worth a check. Not that I think it would be the cause of my problem, but I think I will pull the unit out of the celing and check all the wiring. Never know what screwball mix up could be up there. The installer obviously has a poor track record from what I already checked out.

This house has multiple wiring mixups, including reverse polarity on some outlets, a lot of thee prong outlets with no ground. When I asked my home inspector why he did not tell me about the reverse polarity, he said it dosen't matter because it is alternating current--hello dude? Where did he get his license (if required), in a Cracker Jack box? May be AC, but there still is a hot and a neutral and the two should never be reversed. Do I want the screw part of a lamp socket hot?

 

We had a mobile home that had lots or problems, lots of reverse polarity plugs and switches. If you touched the dishwasher and the sink with the water running, you would get shocked. It took a professional electrician 2 days to find it. Got the switches and all fixed and the shock went away.

Of course, my electronic tech brother thinks that if something is wired reverse polarity that it will run backwards. I had to show him different with a can opener and a hair dryer. He believed me after that.

 

Would a poor or bad ground also cause overheating? I know a lot of older homes didnt have ground wires run when they were built.

 

See, all those little electrons do funny things. Let's start at the f,l,h combo.
1. switch is for light. 2. switch is for fan 3. switch is for heater.
What kind of fan is in it? Exhaust or blower? It sounds like you are pulling the load for the heater and the fan through one switch. You should have three hots a neutral and a ground going to the f,l,h combo. Put those three hot wires on the top of each switch. One wire for each switch. Then on the bottom of each switch put a wire about 8 inches long on each terminal. Now take those three wires on the bottom of the switch and wire nut them to the hot coming from your panel.
zanny

 

zanny,

Since he states that the heater won't come on without the fan blowing that sounds like it might be a blower fan. I don't think an exhaust fan would/should be tied to a heating unit. The purpose of the fan I would think, in this application, is to blow the warm heated air into the bathroom. Without seeing the fan, I'm speculating on this, please cut me a liitle slack here. Since the heater shouldn't come on without the fan, couldn't he eliminate one swith here and put the heater and fan on one good switch. Since it doesn't appear to be an exhaust fan, I can't think of a time when you would run just the fan by itself. Does this make any sense?

 

Things will run backward if DC and the battery is installed backward. Found this out when we put a battery backward into my son's "Stomper" dump truck we got for a dime at a garage sale (later the Stomper was knotted into my daughters hair--pretty sticky tires on that thing ).

Ground is good because I tested it last year when I accidently hit the hot terminal to the box and there was a POW/flash and the lights went out. Kicked the breaker (20 amp at the time) and left a blackened area on the box edge. Hmmmmmm, that was the pilot light switch, which controls the heater. But I don't suppose that little grounding test could have damaged the inside of that switch to cause my hot terminal problem on the other switch? Unlikely, especially since it had this problem before that.

Wiring is one hot to the two switches (light and fan) and the fan neutral becomes the hot to the heater. All three go to their respective devices, from which come three neutrals to a single terminal in the fan, whence they return to the switch box and then on to the service panel. The diagram inside the panel housing says to wire it exactly this way (perhaps appropriate for 1950s). Shows 12 ga wire in the heater circuit, but 14 ga on fan and lights. Fan is strange. It blows the heated air back into the bathroom, but also has a vent out the back wall of the house. There is a damper that automatically adjusts, presumable to maintain a set temp in the bathroom (though no visible thermostat that you could set). Also has a link that will burn out if it gets too hot in the fan housing. I am reluctant to put the fan and heater on a single switch since we normally run the fan after shutting down the heater until it cools and sometimes run just the fan.

Then it lists the heater as 1478 watts, the motor as 50 watts, and the light as 100 watts max, all at 115 volts. There happens to be another light connected in with the heater assembly light. The two together were 175 watts (perhaps why it was installed with 20 amp breaker), but I changed them to total 120 last night. So the total is 1648 / 115 = 14.33 amps. With the other bulbs it was right at 15 amps. Still, this should not matter unless it were kicking the breaker.

Here is the lead I got off the DoItYourself site I posted at: "I think you may mean the opposite of a poor connection (in a way), I think somehing has shorted, and has less resistance that it should(higher current as a result)." Only thing is all connections tested out to 120 volts (hot to ground, hot to neutral, etc.). So that does not seem likely, but I guess I should test the resistance though that pilot switch that took the "ground test."

Anyway, I know an electrician that I can call if I can't figure it out.