Help explain free energy in my house light circuit?

[BooMod]

Pulling hair out measuring voltages on a circuit I was mapping in my new house, until finally taking out the compact fluorescent light bulb in my fixture and replacing it with an incandescent to observe finally no voltage coming from the dead-side of a circuit I broke the live-feed wire from.

I first thought that one of my three-way switches was toasted, leaking electricity as I was measuring 120V and 25V instead of 0V in some combinations, but I replaced each in the pair with newer ones and used ones that I had replaced with rocker-panel threeways in other spots in the house earlier. The switches aren't the problem. I seem to be recording 20-28V at the common terminal on the "outgoing power" going TO the light bulb, as well as on the travellers in some combinations. I have even disconnected the live power to the common terminal powering the light switches, and still am recording a 25V average at the terminals. Then, I noticed if I disconnected the "to light" wire from the common terminal, the wire itself was only 1.5-2.0V, but if I push it with the common terminal, it jumps to 25V.

After looking for other crossed wires and similar effects "downstream" of the switch / light here, then I did the same tests with no bulb and an incandescent bulb and it looks normal (0.0-0.2V on my multimeter, free or attached to common terminal) which is exactly what I would expect to see in my circuit!! I can almost plan on putting everyting back in the wall, scrwing the plates back in, and saying "all is okay" and pretending nothing happened, but I really want to know why a compact fluorescent bulb first of all does something different with this circuit from what is expected, and second, assuming I bought a special bulb that is really a generator generating 1.5V on my wire, why are my simple 3 way switches acting like transformers and turning it into 25V.

In theory, if I hook enough of these bulbs and switches together correctly, I can probably power another bulb! I seriously doubt I have stumbled upon Free Energy, so an answer would make me sleep a lot better.

You can see a schematic of my circuit:


and a sketch I drew up mapping voltage measurements with the different three-way combinations for the light switch, but this was before I realize the difference between voltages when the 'out' wire was attached to the common terminal or not.

[Chris Smith]

A simple test for bad grounds and returns is to find a great absolute ground. [or pound one in]

A gas main or water pipe can be used for this TEST. [use phone wire or single strand 18 -22 gage long wire]

Hose clamp the long wire onto this good ground, sand if necessary and string the long wire back to your “lights in questionâ€

[BooMod]

Am I losing 20V?
I don't think so - I'd be measuring 100V on my live powers by that measure. Losing 119V (or 100V when attached to switch terminal) maybe, if the power feed of 120 were coming from that direction, which is not supposed to.

AS much as I can tell, it is JUST at this one point in the entire circuit things are measuring strangely. Its 119-122V everywhere else.

The house is 10 years old everything grounded with ground wire pigtails and screwed to boxes, although this light fixture is one of those simple, black and white wires only sockets, there is no ground wire in the device.

[dyarker]

"I seem to be recording 20-28V at the common terminal..."

In reference to what? One meter lead on common, where was the other meter lead? I agree with Chris, a good reference is needed.

[BooMod]

one lead lodged between a groundwire and the screw clamping it to the box, the same spot wedged for recording all the line voltages of the live incoming, travellers back in, etc.
I had even removed the switches earlier, and put some nuts on the live power + one of the travellers only, initially thinking that it was a problem in the 3-wire cable between switchboxes, but would read 120V and 0V when powering one of the travellers at a time.

Would the ground be satisfactory for measuring all these other point voltages within this box except for the 'output' point?

[Edd]

.
.

Pulling hair out measuring voltages on a circuit I was mapping in my new house, until finally taking out the compact fluorescent light bulb in my fixture and replacing it with an incandescent to observe finally no voltage coming from the dead-side of a circuit I broke the live-feed wire from.

Working with that key initial statement, wherein a common ground metering reference is utilized across the board:

Well…HELLOOOO!…..looks like you have just been introduced to the eccentricities of NOT having an almost pure resistance (.05% L) in your load line as was experienced while utilizing an incandescent lamp, as compared to your previous utilization of a CFL as being the load in the same circuit line.
In all probability, you are using a modern day digital DVM as compared to an older analog unit with its lower sensitivity, which, in itself, would tend to load down the circuit a skosh more whilst taking a reading. Also, what you are getting is a “staticâ€

[BooMod]

Thanks for the detail - however, its right up there at 10 and I need it dialed back to about 6... some dummying-up would be appreciated.

Are you indicating that it is the mere act of hooking my multimeter up that is going to cause some trivially non-zero to report as a much greater value? I can fully understand how when a fluorescent bulb is shut down it no longer is operating at above threshold values, but some of the energy in the system bleeds below that threshold and then remains non-zero, turning the bulb into either a microbattery or the circuittry in it is working as an amplfier...

If it all comes down to "no matter what house, no matter what circuit layout, no matter what combination of feeds, a CF bulb will always read a small voltage on the unpowered live-in side, okay. And I'm getting the feeling that this will indeed be the case.

But still not understanding how that 1.5-2V wire when touched to the screw of the unpowered switches (or, powered, see diagram of combinations) the 2 becomes 25? This, this part is the part that I am really interested in figuring out why. Mostly because 25V in a 'dead' line seems a bit unhealthy to me.

If that was explained above, lets give it a try at "6" and hopefully won't have to go lower than that

[positronicle]

--Edited by Positronicle--

[BooMod]

I already did readings with an incandescent bulb and no bulb at all - 0V as I had expected.

>>After looking for other crossed wires and similar effects "downstream" of the switch / light here, then I did the same tests with no bulb and an incandescent bulb and it looks normal (0.0-0.2V on my multimeter, free or attached to common terminal) which is exactly what I would expect to see in my circuit!!

[Robert Reed]

Boomod
What happened to your schematic? Will not download. I wanted to take a look at that switch in the lower left and see if that could probably cause a problem. It breaks the neutral wire rather than the"hot" wire. Just going from memory here from a quick previous look, when the prints were posted originally.

[BooMod]

it has been magically fixed - the images, not the problem

[Chris Smith]

NOT TRUE

My house delivered 120V, and lost 20V on the returns.

Now I always place a dedicated ground on all my work shops.

Short, and to earth.

[Robert Reed]

Are you measuring this 25 volts common to ground , or common to hot? you never mentioned your reference point. If in fact you were dropping 25 volts somewhere, even with a 60 watt bulb, you would be dissipating 10 watts at some other point. this is a lot of heat and would soon show itself. If you are measuring this voltage on open circuits then it is quite possible that the DVM's input impedance (10 megs and higher) is acting as a voltage divider. Consider this- total open circuit impedance is 40 megohms (a realistic value). DVM bridging open circuit would constitute 20% of that giving you 25 "phantom" volts across the DVM's internal resistance. Quite possibly with an analog meter 's low impedance you never would have seen this. One way to prove this is if you can bridge the meter probes with a 100K resistor to see if all readings return to normal.I didnt see any obvious problems with your low side switching-unusual, but apparently functional.

[BooMod]

Reference?

>>one lead lodged between a groundwire and the screw clamping it to the box, the same spot wedged for recording all the line voltages of the live incoming, travellers back in, etc.

Is that not meaning reference is ground? Its the same reference for EVERYTHING, one probe jammed into the same spot, wedged between the groundwire and the screw screwed to the metal box.

[Chris Smith]

Find a real or absolute ground [A known ground, Not a perceived one] , and re check with your meter.

If you have still have a 20 volts leak, fix your ground return like I mentioned.

[circa 1980, de javu]