The original question (puzzle) is still not answered (explained) satisfactorily (to me at least). That is ---
[Open question 1]Why turning the leads of the probe around will hide the signal (about 40V peak, when the load is not connected)?
Note:
1) when there is load (a dummy pure resistive load), this phenomenon does not exist, a signal can be detected both ways (though without fluctuation but rather stable +/- about 40V).
2) the circuit under test is fully isolated from the mains and mains PE.
Another question that has come out of the discussion and has also not been satisfactorily answered ---
[Open question 2 --- edit: this has been partly answered]When an isolation transformer is needed in a repair job, which I understand is usually connected to the equipment under test, can the isolation transformer be used between mains and the scope to achieve the same effect of isolation?
[Answer]
According to the contributors to this discussion ---
This was probably once used by experienced people decades ago. But is definitely dangerous so must be avoided, particularly because of the availability nowadays of isolated scopes and differential probes which make this dangerous trick unnecessary.
The major risk is that, by doing this, when you clip the ground lead of the probe to a point in the circuit of the device under test (DUT) that has a voltage above the earth (say 110V, DC or AC), the metal case of the scope, as well as the ground lead of other probes of the same scope with become live as well (at the same potential). This is not only a threat to your own safety but also to people around you, people who o rescue you (in case the worst case eventuates), none of who would likely be aware of this hazard.
The key is that this does not require an earth fault inside the scope. This risk exists with a scope in fully working order without any fault because of the way scope (that is not an isolated type) works. This is different from using the isolation transformer with the DUT, in which case the metal case of it will not become live unless there is an earth fault inside it (that is, contact of the live wire with the metal case).Yet another question related to earthing system is ---
[Open question 3 --- edit: this has now been answered satisfactorily]In the TT system is there a connection at the household end between N and PE as indicated with a green dash line in the figure below?
[Answer]
A big, fat, firm NO. In TT system, there is no such direct connection in the household switch box.
Furthermore, in the five earthing systems described in the IEC 60364 standard, which are TN-C, TN-S, TN-C-S, TT, and IT, only TN-C or TN-C-S systems have a direct connection between N and PE at the consumer end (usually in the household switch box).