Today in the BH break room, r0k offers some electrifying entertainment. I was working on some old house wiring and it was a real joke. I was going behind some rank amateurs. For one, they had connected two normal (SPST) switches in series where there really should have been two 3 way (SPDT) switches. Where there were supposed to be two traveler wires, there was only one. For the other traveler wire they stuck a wire nut on one end (at the top of the stairs) and tied the other end (at the bottom of the stairs) to what should have been a ground terminal. In another room they proceeded to do some mysterious stuff I couldn't decipher at first. Look at the clues and think about the diagnosis before you scroll down to see my solution. In the US, we use a 110v system with Hot, Neutral and Ground. Typical homes are 220v with two phases of 110v and a Neutral between them. Neutral is connected to earth ground (by driving a 48 inch copper stake in the ground) at the panel coming in the home and "on the pole" just below the step down transformer that makes 220v from 4800v or whatever is on the highest wires. It turns out that common soil is an excellent conductor at 60hz which is why we call it "earth ground". We have wires coming from the power plant but if one of the hot lines touches the ground, it is like a huge dead short and should blow a breaker upstream. There are $5 testers available to find ground faults in house wiring. The testers have 3 lamps. The lamps are wired between the three terminals of a standard "type b" plug. For a correctly connected plug, the red lamp is off and the two yellow lamps are on. -Y- = open ground --Y = open neutral --- = open hot R-Y = hot ground reverse RY- = hot neutral reverse -YY = good Some testers have a GGCI "test button" which causes some current to flow from hot to ground and "should" trip a GFCI outlet. At this point, I should mention that I was seeing RYY = ??? All three lamps lit up If I plug in a 60 watt load it works fine and each receptacle seemed to be working except that the ground fault tester said "something was wrong"... If I push the GFCI test button, the two LED's would be lit showing the circuit was good. ... no cheating ... More data... I measured 52 volts from Hot to ground and I measured 70 volts from Neutral to ground. ... I mean it ... What was the matter with my wiring? Was it swapped wires? Was it an open? Was there a partial "resistive" short (hence the wierd voltages)? Having the red led lit began to convince me that I had a hot/ground or a hot/neutral swap and I was inclined to find the first plug in the string and swap the incoming lines. Before doing so, I decided to do a little more work. Give up? Click here... I checked the voltage between hot and a known good ground. It was 115 volts. What the heck? The red LED was on but hot and neutral were not reversed. Well, it turns out the designer of this tester neglected to list one case on his chart: RYY = -y- = open ground Since ground was floating, there was a path through two of the leds that could light them up dimly with the third LED on full brightness. Pushing the GFCI test button altered this path and caused the red LED to go back off. To test my theory, I went over to the hardware store and bought a 30 foot length of green insulated 14 guage solid wire and connected one end to a known good ground. The other end, I tapped on the suspect ground. I was a little worried that sparks would fly and current would flow. After all, I had measured the suspect ground at 70 volts from neutral, which is 70 volts from ground. Nothing happened. Now the tester shows -YY with the "jumper wire" connecting the "bad ground" to a "good ground". So now all I have to do is find the box there the doofus neglected to bring the ground up from the main breaker panel, or a ground wire broke, and repair the damage. Even if the problem is a broken ground wire "in the walls somewhere" all I have to do is run a separate ground conductor to one of the outlets. Problem solved! ... Drag here.