51.4 volts on 3-way receptacle

Status
Not open for further replies.

stephena

Member
Location
oregon
Electrical engineer here - question on my own home for a change. Newer home, with a 3-way switch-controlled receptacle in one room. Over Christmas my wife unplugged the lamp that was being used on the recept, and plugged in some LED lights. We noticed that when the switch was off, they still lit (dimly). Finally got around to metering the receptacle today: 123 volts when the switch is on (OK). However, still 51.4 volts when the switch is off. Standard outlet tester indicates (red light off / 2 yellow lights on) that the outlet is ok from that standpoint. Thoughts on what to do / where to start 'troubleshooting'? I'm hoping this isn't a wiring issue. I'm a handyman also, but typically prefer projects by choice! Thanks in advance.


If the romex was stapled too hard or pinched the voltage could be coming fromhot feeding the switch if it's in the same cable. If it's a switch loop that's most likely the cause
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
If the source is capacitive coupling in the traveler section of wire the observed results should not depend on which switch combination is used. If there is leakage across one contact pair in one switch then reversing both switches should cause a different result.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
160127-1431 EST

jonathankpa:

Your meter is a high input impedance meter with an apparent input resistance of 10 mrgohm.

Making current measurements in the amp or mA range can be a problem if the current is much greater than expected. A meter current protective fuse is typically in the range of $15. So use a resistor and the meter in volt mode. I suspect your current is less than 5 mA. Use a 1k 1/4 or 1/2 W resistor as a current shunt. 1 mA = 1 V.

Measure the short circuit from your hot socket terminal when that terminal is de-energized. Measure the current for both switch positions that make the socket terminal off. Do the open circuit voltage measurement for both combinations.

I will guess at between 0.02 to 2 mA for shoprt circuit current. Next determine the phase angle of the current relative to the voltage waveform of your 120 V source.

.
 
If you'll be doing routine work with power circuits

If you'll be doing routine work with power circuits

A tool to get is a low-impedance solenoid meter, colloquially a "wiggy". If you won't be doing routine work with power circuits, hey, it's still an excuse to buy a new tool :)

Builder-grade components are the bottom of the quality ladder. Spec-grade replacements for the switches are much cheaper than your time is. Swap out the switches and you solve the noise problem at the same time.
 
Last edited:

Dark Sparky

Member
Location
USA
Occupation
Electrical Engineer
If the source is capacitive coupling in the traveler section of wire the observed results should not depend on which switch combination is used. If there is leakage across one contact pair in one switch then reversing both switches should cause a different result.

Understanding you mean switching (toggling) both switches? We tried all 4 combinations of both switches. Both combinations yielding "off" were 51.4 volts. Both combinations yielding "on" were 120V.
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
Understanding you mean switching (toggling) both switches? We tried all 4 combinations of both switches. Both combinations yielding "off" were 51.4 volts. Both combinations yielding "on" were 120V.
That makes it less likely that the problem is a switch with leakage unless it is leakage between the two traveler terminals rather than between the common terminal and one traveler terminal.
 

ritelec

Senior Member
Location
Jersey
That makes it less likely that the problem is a switch with leakage unless it is leakage between the two traveler terminals rather than between the common terminal and one traveler terminal.

Commons. Back stabbed. Maybe.
Either way the op has to do something and not just keep taking meaurements.

Rule out the switches first.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
160127-2014 EST

jonathankpa:

I will make a wild guess that the wiring distance between your two switches is about 15 to 30 ft (I could guess this because the wiring is in a house, but that is not how I came to that conclusion.), and your short circuit current is about 5 to 25 microamperes as measured by your meter. I calculate 18 ft of wire, but that has potential for a lot of error.

A problem at the levels you are measuring is that the actual meter input impedance is not just 10 megohms of resistance.

.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
160127-2053 EST

ritelec:

I don't believe there is anything wrong with his switches or wiring.

A "wiggy" or the voltage across a 25 W incandescent will read essentially 0 V.

.
 

ritelec

Senior Member
Location
Jersey
With the switch off reading 54 volts. He should be able to take a piece of wire and go hot to neutral or ground with no arcs and sparks ??

:- )
 

kwired

Electron manager
Location
NE Nebraska
With the switch off reading 54 volts. He should be able to take a piece of wire and go hot to neutral or ground with no arcs and sparks ??

:- )
If it is capacitive coupled voltage - most likely yes, there shouldn't be enough energy available to make much of an arc when it gets shorted.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
160127-2137 EST

ritelec:

He has already run the test with a low impedance load, namely his original incandescent bulb, it dosen't light. The problem was with an LED light load that when the switches were in the off positions the LEDs lit dimly.

.
 

Dark Sparky

Member
Location
USA
Occupation
Electrical Engineer
Thanks for everyone's suggestions. Life rolls along (hence replying at almost midnight) - but I'll endeavor to do some testing with your ideas shortly. Will post back when I do...
 
Status
Not open for further replies.
Top