Farmerik,

I have read the posts about your experiments, also with interest. I'm not saying that an unbalanced air-loop won't bring in strong signals. More precisely, I think it will be inconsistent, difficult to null, and prone to noise pickup.

The "whip effect" is also not limited to the lead-in wires. The loop itself, when unbalanced, acts like a fat whip. So in your experiments, it does help to eliminate the the lead-in wires, but since air loops are big, there is still considerable effect. In my experiments, even a few inches of wire on the "hot" side of the unbalanced, tuned circuit picks up a lot of RF.

For air loops with high inductance, directly matching the Si4734's tuning range, it is difficult to make a 1:1 balun transformer. It takes a lot of windings on both the primary and secondary, leading to more parasitic capacitance and series resistance. Also, a larger core must be used. The high-impedance loop will also be more susceptible to stray capacitance, so limiting your lead-in wire length. So my answer to your question is that a low-inductance loop (10-20 uH) with matching transformer is a better setup.

Here's the recipe for the balun transformer I'm using, which should work for air and ferrite loops in the range of 10-20 uH:

1) FT50-61 ferrite core
2) Secondary winding is 80 turns of 32ga wire. If you can measure it, the open-circuit inductance of this winding should be 440 uH. I needed 84 turns to get there, but it's not vital. Arrange the windings neatly with several millimeters gap between the ends of the coil. Secure the ends lightly with super-glue.
3) Primary winding is 28ga wire (easier to work with), two strands twisted loosely together. Wind 8 turns of this bifilar wire over the primary windings. Secure with super-glue.
4) Primary connections: Separate the twisted wires and identify the two wires A and B and two ends 1 and 2. Twist wires A1 and B2 together and connect to the radio's RF ground. Twist wires B1 and A2; each wire connects to one side of your loop lead-in.
5) Secondary wires. One attaches to the radio's RF ground, the other to the AM input. It is vital to keep these leads short. It helps further to keep the "hot" wire (AM input) and toroid shielded by RF ground.

This setup is designed so that a 16 uH loop will be transformed to an effective 220 uH as seen by the radio chip. I'm using it with a 10 uH loop and the chip sees 180 uH, just within the range tunable by the chip over the AM broadcast band. With a 60x133 mm, 7-turn, 10 uH rectangular loop, I get RSSI readings on local stations roughly 6 to 10 dB higher than with the stock antenna on my PL-380.

-Scott-