The following scan is for a 4:1 balun with 14 turns of #14 covered
wire. The band width is shown below and is narrower than wanted.
The picture below shows the coil wound on a 1 1/4" PVC pipe. The coils
were covered with electrical tape (not sure what the dialectic is) and
tacked together with solder. The "V" at the bottom is two resistors
(scanned) with a value of 101 Ω and 104 Ω. The bifilar, air
core, winding diagram is shown below with the 205 Ω resistor across the
balanced output to allow a balanced load on the input to the 50 coax.

Balun picture

In the above picture the coax is connected at the top. Two
resistors, a 101
Ω
and 104 Ω for a total of 205
Ω, that have a nearly flat response, are at the
bottom.
- - -
Coil data
| Description |
Dimension |
|
Form Diameter |
1.125" PVC |
|
Wire |
14 awg
THHN (covered) |
|
Wire diameter |
0.0641" |
|
Coil diameter(OD) |
1.875" |
|
Turns |
14 bifilar |
|
Coil Length |
3.0625" |
| |
|
- - -
Inductance - Calculated (based on 4:1 or 205/4=51.25) and Measured
Frequency
(MHz) |
Rs
(calculated) |
Rs
(measured) |
| 3.5 |
51.25
Ω |
43.559 Ω |
| 7.1 |
51.25
Ω |
48.433 Ω |
| 14.2 |
51.25
Ω |
30.189 Ω |
| 21.0 |
51.25
Ω |
12.440 Ω |
| 28.0 |
51.25
Ω |
12.610
Ω |
- - -
Scan - HF

- - -
Band Width
scan
(Right-click, save as) - 3 to 14 MHz

The coil is resonant at 20.0393 MHz. Above this frequency the coil
looks capacitive, most likely due to the capacitive coupling between
windings due to the thickness of the insulation.
- - -
Scan - Full

Obviously many issues with this balun. Need to try
winding with non-covered wire.
Links:
http://www.qsl.net/wb6zqz/3in1balun/construct.html
If you have any graphs that could be added to
this site please send them along. I'll post and credit your
information.