[Lowfer] antennas

[Lyle Koehler]

Dave, what software are you using for coil modeling? Some of the software
I've seen has the opposite problem, in that it predicts unrealistically high
values for coil Q.

Here is an example of the results I get with Reg Edwards' SOLENOID.EXE and
Brian Beezley's COIL.EXE software. The coil is made of 72 turns of #14
copper wire with 1/8 inch turns spacing (winding length 9 inches) on an 18
inch diameter form.

SOLENOID.EXE: N=72, L=228.6 mm, D=457.2 mm, Ratio=0.51, F=185 kHz

Predicted L = 2.463 mH; Q = 1013

COIL.EXE: Wire size = #14, N=72, L=9", D=18", F=0.19 MHz

Air core       L = 2.678 mH, Q = 746

Polystyrene core      L = 2.874 mH, Q = 692

PVC core      L = 2.918 mH, Q = 340

Reg Edwards' SOLENOID.EXE program predicts the "true L" and "True Q", which
is what you would measure in a parallel tuned circuit with corrections for
the effects of distributed capacitance. The predicted L and Q values from
COIL.EXE are the "apparent L" and "apparent Q" values, which you would
measure in a series-tuned circuit such as that used in the Boonton Q meter.
The two programs show much better agreement at very low frequencies, where
distributed capacitance is not a significant factor. Since a LowFER loading
coil is in a series circuit, the apparent L and Q are the parameters we
really need to know. A coil made of #14 insulated building wire should come
out fairly close to the L and Q values given by COIL.EXE for a PVC form.
Basket-weaving can improve the results, and with large Litz wire, it is
possible to get a Q of 1000 or so. By the way, COIL.EXE was included in the
software that came with the ARRL Handbook a few years back; I don't know if
it is available on-line.

It is hard to get a Q of more than 300 with a ferrite-loaded coil on the
LowFER band. Another problem with ferrite cores is that they will saturate
at very low power levels, and the Q may degrade considerably even in a
1-watt LowFER application.

Lyle, K0LR