Analog multimeter shows voltage even with 3-way switch off

Friends,

Since this my first post I would like to so I have read the posting rules so I would to start off with an introduction of my background and knowledge.

I currently work as an IT Systems Administration and Engineering, or in slightly more lamens terms; I am an Information Technology Consultant for a company that specializes in designing, implementing and building computer and server networks for small and medium businesses. I have been doing it for over 15 years and I have extensive knowledge in the low voltage industry. And I enjoy it and am a pretty quick learner.

Big change came a few years ago when I finally became a homeowner. And of course I jumped right into the Mr. Fixit role and am loving it. I do have to admit that I am not a licensed electrician and I am aware of the liability and limitations of what I can do. That being said while I will always call in a professional (and have on several occasions) once I have fully understood the problem so I can in 100% accuracy describe what I need the repair technician needs to do. That has resulted in a 100% satisfaction rate with the work done as well as good payment to a professional that deserves it. This means I wish for a full and complete explanation (if possible) of what my problem is and how it can be trouble-shot and corrected (if possible with renovation or w/e); not so that I can do it but so I can get the right person to do what I need.

Hopefully no one's' eyes are bleeding yet, so now I bring you to my problem.
I live in Colorado, USA in a home built in 1971, it has a single breaker box where it receives 2, 120V lines feeding the house a total of 150A to a single breaker box. The feed lines and almost all interior wiring is original aluminum from the early 70s. Initially it was built with a fully unfinished basement. Apparently a previous owner fancied himself as Mr. DIY, with the slight difference in that he had no idea what he was doing. Built out a good portion of the basement, about 40% to code, as well as added Air Conditioning to the central air. One of the biggest fouls being the A/C wiring especially where splicing copper and aluminum without taking the required precautions for the mixed wiring. Nor proper grounding or marking such as finding red coming into a box, splicing with a white, whole thing painted over white so no color visible, and finding 120V on it. Several times.

Ok, so the problem I have is I am replacing the chandelier in the kitchen because it keeps frying light bulbs and we (myself and my neighbor, who is not a licensed pro but has extensive experience in home renovation/handyman and shares my views on hiring professionals) made a discovery. We found that a dimmer on the patio-door wall actually also controlled the chandelier that we assumed was only connected to a normal switch in the center of the kitchen. When we discovered this took an analog multimeter and checked the voltage going to the light while toggling the two switches around(we had not yet removed the light, we were testing using the connections to the light). We found that even when both switches were off there was about 30-50V still going to the light socket. Having not encountered this before we immediately flipped the breakers to the floor lighting and the kitchen and checked voltage until everything read 0 to be sure. After this we used the multimeter to check continuity between the switches, the lines and the various possible outlets/lights. We discovered that, the switch by the door had the 'source' hot from the panel or perhaps the other light circuit on the floor, not really sure atm. We also found that the he was using a 3-way switch in the room and a dimmer to control one fixture by wiring the dimmer like a 3-way and splicing a bunch of things. Another important note is that the wiring going into, and out of this box is aluminum and none of it connects to the light fixture. The 3-way switch on the wall however, has 3-way copper wire coming into it and by continuity tests it is the feed from the dimmer for the 3-way and. Also the 3-way switch is the last switch before the light socket and that wire is aluminum to the light. We then removed both the switch and dimmer and capped off misc exposed wires.

We made note of the continuity and for the last test we flipped the breaker back on and checked the voltage going to and from each box and discovered that without the switches there is voltage going to the dimmer box on the black wire and also power going to the 3-way box, but there is no voltage going to the light socket, as expected. So with this verified, turned the power off, diagram noted, we triple checked how 3-way systems should work, purchased 2 new Levitron 15amp 3-way switches. We then wired them back up exactly as they should work based on a wiring diagram and verified again by multimeter. Then before hooking the light up I wanted to test again, so we turned the power back on and checked voltages. And the voltage on the line remained around 40-50V even with both switches off, with a proper change to ~120-123V when the switches were on.

Shocked to find that putting in the proper switches didn't change the voltage we checked the 3-way set on the normal stairs that were built with the house to find the same issue, ~45V on the 'Off' side of the switch.

Ok, I deeply apologize for the severe long-windedness but I said I take this stuff very seriously, in an OCD-sorta way lol. Now then, I can say I provided enough background I think.

Now I have been trying to do my diligent research like I should into this problem and I'm seeing mixed messages about what may be wrong. I have read things such as, an open neutral, open or no earth-ground (which, as near as I have found, there isn't one), or its caused by the high built-in capacitance of digital multi-meters causing them to show a "ghost charge", I have also heard it could be caused by induction into the line from neighboring ones and is a normal thing and the low voltage doesn't hurt anything but just remember that when you touch things, etc.

So then, given the severe difference in results I have some concerns and questions.

1. My primary concern obviously is; do I have a serious problem here that warrants action before "the day comes" so-to-speak? I mean I have been here for a couple years now and while I have repeatedly come across evidence and stories of how this guy was like the king of corner-cutting, nothing has actually been damaged that I recall, just completely out of code. So I wouldn't be surprised if his horrible basement renovation created this problem, but again nothing 'bad' has happened yet other than killing 15 light bulbs in a year lol.

2. A majority of the basement wiring is in the unfinished area including the kitchen so tracing is relatively possible. If this is an issue simply caused by him splicing stuff to cut some code corners in the basement wouldn't it realistically be easy to solve by turning off the power and undoing the mess(and hiring an electrician to put it back right! )? This would be assuming he somehow cross-wired something or left a line disconnected somewhere in the mess he made.

3. If its severe, what kinda 'damage' am I looking at? I can give further details just for even a guess ballpark, just really need to know if this is bad and if so, how bad. Thanks.

 

There are many different ways to wire a 3-way circuit so the diagram you used may be wrong for your situation. To boil it down to simplest you need to identify the power in and connect the black of that cable to the common terminal of one 3-way switch and you need to identify the hot lead going to the light and connect that to the common terminal of the other 3-way switch. Then you connect the travelers of each switch together. Given what you have described and being that part of the wiring is aluminum I wouldn't even try to figure it out, just abandon in place and run new.

Tech note: Your house is not supplied with two 120 volt lines it is supplied with one 240 volt feed. There is a center tap from the secondary of the supplying transformer that is used with one of the 240 volt lines to derive 120 volts.

 

I gleaned a small bit of pertinent information from your post that needs to be addressed if you are going to be doing your own repairs; that is that your original wiring is aluminum. There are specific ways to handle connections when working with aluminum wiring that you are probably not aware of as I saw no chapters in your book that addresses them. The best thing to remember about working with aluminum wiring is that aluminum wire and copper wire should NEVER come in direct contact. The next best thing to remember is that wire nuts are only approved for copper to copper connections.

There is a CPSC publication you should take the time to read before you go too far.

http://www.cpsc.gov/cpscpub/pubs/516.pdf

 

To add to what Ray said, you have single-phase 240V service. The single-phase 240 is split into two 120V legs. Connecting to both of the split-phase legs gives you the 240V power. Connecting to one of those legs and a grounded conductor (a neutral) gives you 120V.

What are you doing to correct that, and to avoid having the CU and AL come into direct contact?

We can help you troubleshoot your problems. Butdoes not sound like the right approach, if you want to save time and money. I would only undo those things I was prepared to resolve. If I was going to hire an electrician, I would leave everything in place to aid that professional in his or her diagnosis.

Would you rather someone brought you a complete, closed PC to troubleshoot, or all of the parts that (they think) used to be part of that PC, in a box?

 

@ray2047, That is exactly how we wired the 3 way switches back together, and from a test standpoint they function like they are supposed to.

Also thank you for the info on the feed line, I certainly only look and not touch so I just counted the 2 lines + ground going to the box and figured thats what it meant but your description makes more sense.

@CasualJoe, We use the "Alumiconn Method of Repair" mentioned in that publication. We have used this method on any repairs we have done involving aluminum and did the same again with these 3-way switches.

@Nashkat1, as I mentioned to CasualJoe I am using the "Alumiconn Method of Repair" from the CPSC page he posted. And your point about removing the wiring is mostly correct, let me expand on what I meant though. The unfinished part of the basement contains fully exposed wiring that the DIY maniac did. My thoughts were simply, if his basement remodel caused this wqhole house problem then I could easily remove the offending stuff so the rest of the house is fixed. And then, since it wasnt to code or in a good location at all, have a real electrician REDO the basement to my specs. So yes removing it would make his troubleshooting job harder but I only intended to do that if it was most likely the problem.

I also intend to eventually have an electrician replace and update wiring as I fully renovate rooms though.

 

Apparently you cant edit posts after a given time, heh. This can be marked as resolved, the voltage found on the line was phantom voltage and was normal based on the way we were measuring and checking.

 

When you made a point of saying analog meter in the title I didn't suggest phantom voltage. Had you said digital it would have been my first comment. Thanks for letting us know what was happening.

 

Ya I was getting mixed messages between searches involing digital multi's but never saw much on analog. However when connecting everything and testing everything there was no phantom voltage when measuring from the correct side of the switch while there was a load on the circuit (having the fixture wired) So all is good on that side. But still need to get someone out to put an earth ground in and do some other fixing on the panel side. Either way its all good now and I dont have any major problems in the end, aside from earth ground that is.

 

Glad you got it, and thank you for the feedback.

Just for future reference, measurements need to be made across open circuits; that is, with the devices disconnected.