I wonder if magnet wire is a good idea? Generally magnet wire is
designed to get the turns as close as possible. I'm not sure that is
best for high frequency use since you want to minimize stray
capacitance.

Charly wrote:
Thanks to all for the replies...

With the hints received here and elsewhere, I finaly made my firt unun with the
piece of ferrite I found...

I used magnet wire retrieved from an old broken AC/DC adaptor :
- 2 turns for the radio side,
- 8 turns of wire antenna side, making a 4:1 unun.
Currently the unun is just beside the radio (no coax or whatever : just a few
centimeter of wire connected in the radio), grounded via the outer radio/antenna
connector.

And yes it works ! At least connected to the ATS909 : it increases the signal
level by two bars and the sound is clearer. The result is not that obvious with
the SW100, perhaps due to a different antenna connection impedance (anyone knows
the Sony SW100 or AN100 antenna connection impedance ?).

Next step : mounting an inverted L antenna between my house the one of the tree
near-by : connection in the attic, going through the attic length (to have a
longer wire), passing through a tiny hole between the wall and the roof,
reaching the tree at something like 5 meters above the ground, then going down
the tree.

At what heigth should I stop the vertical leg above ground ?

The horizontal leg will around 12 meters (say 8 meters in the attic, 4 meters
outside : I can't go longer in this city environment).

Feeding will be a 75 ohms coax grounded via some house's water pipe : it seems
that the 909 has a 100 ohm antenna connection impedance... so it will be a
better match than 50 ohms (unless I put another 2:1 unun just before the radio
of course).

In radio with my (poor) electronic knowledge, the more fun is building the
antenna stuff... :-)

Oh yes : I need to try the snake-in-the-attic antenna as well...

Charly