On Sun, Jul 14, 2013 at 06:37:37PM +0200, Stefan Schäfer wrote:

> I just came back from a field experiment, using the short active
> dipole which is resonated to 475 kHz (4.7 mH and 10 mH fixed
> inductors in parallel plus a few pF (<10) for fine adjustment), with
> fiber optic cable.

> As i said, i have to repeat the test because a few of the
> measurements were to close to the soundcards ADC limit and caused
> nonlinearity...

I think these results already make a lot of sense.

Without ground, your antenna is a short dipole, measuring the potential
difference between two points about a meter besides (horizontal) or
above (vertical) each other.
Clearly, your signal is vertically polarized, and badly received
when the antenna is horizontal.

When you ground the lower end of the _vertical_ dipole, the signal
increases a bit (2 dB, apparently), because you then measure the
potential difference between the top half of your dipole and the
ground; at the same field strength (V/m), the larger vertical distance
gives a larger potential difference.

When you ground one end of the _horizontal_ dipole, your antenna
becomes _vertically_ polarized: it measures the potential difference
between a piece of metal (namely the not-grounded half of your dipole)
a few meters above the ground, and the ground itself. (The other half of
the dipole is now grounded and will of course distort the field a bit.)
Since the antenna is now vertically polarised, you get good reception
again.

73, Pieter-Tjerk, PA3FWM