Draw your circuit, including scope and generator. Now connect a big ol wire between where the scope probe ground is connected and where the shield side of the function genny is connected. That wire is the shield of the bnc cable connectin the FG to the scope. If scope is not floated, there is another parallel wire that is the third wire of the power cords and the building ground wiring.
You have to break all the connections between the chassis of the generator and the scope.
Regardless of whether or not any instrument is floating, keep in mind the alligator clips (for the scope probes) are connected together. So if you connect Channel 1’s alligator clip to point A in a circuit, and connect Channel 2’s alligator clip to point B in a circuit, you are creating a short between points A and B.
Seems that part of the problem is that I have a big leak to the ground in my house’s electrical system. And the RCD on the fusebox is bypassed because it tripped all the time.
Sometimes I would pick up (along with my signal from the generator) something resembling the mains frequency, although quite mangled up (too bad I didn’t take a screenshot).
Anyway, I managed to get the correct output by floating both devices. But I am picking up some spikes when the voltage drops to zero between the lobes. What could cause that?
Here’s the screenshot:
With everything floating, you are bound to pick up a lot of noise!
What does this mean?
I am guessing that he means that since the RCD (same thing as a GFCI) kept tripping there must be some leakage somewhere that is setting it off, although I don’t see how this is relevant to the problems he is experiencing, especially after floating everything; I have done pretty much the same thing without problems with noise (e.g. when designing a SMPS, the oscilloscope is floated so I can probe the primary side, which has a bridge rectifier to the AC line, effectively isolating it since current won’t flow through the diodes unless the line voltage is higher than the output).