Hi Earl,
I had to "test" a new analyzer prototype using a 16 bit A-D converter
today, and one requirement is I like to use is some real world devices. As
such, I wound some transformers on the 73 material binocular core Ameritron
uses in HF PA's. This is a common core form baluns and broadband
transformers, it is commonly stocked.
Since applications like this are not critical to exact impedance, I used
your 800 ohm secondary suggestion. Additional VSWR by this would only be
1.125 to 1 in an ideal transformer.
The core is about 1" long and 1/2" thick. The winding was a two turn
primary with a 8 turn secondary. Wire size was #26. No special winding
techniques were used. The core was sprayed with Krylon, primary was wound
first and sprayed with Krylon, and the secondary wound over the primary.
NOTE: This winding method has slightly more loss, but has MUCH better
feedline to antenna isolation. Isolation is much more critical with
receiving antennas than loss, so I avoid trifilar and other windings that
provide direct common mode paths. This makes the transformer more "core
critical", so substitute at your own risk.
Using an HP RF Impedance test set (1990 vintage) and Wiltron Network
Analyzer, I measured the following with 800 ohm termination. Frequency is
first, R second, J third, and loss fourth. Loss is corrected to factor out
mismatch loss, so we know true transformer loss (amount of correction is in
brackets).
2MHz, 47+j2, -0.38dB, (.035dB)
4MHz, 47+j0, -0.41dB, (.035dB)
8MHz, 47+j0, -0.44 dB, (.035dB)
32 MHz, 44-j3, -0.65dB (.08dB)
> Can anyone provide info for winding a balun transformer for 1.8 to 7.5
> mHz with a 1:16 impedance transformation ratio, i.e., to match 50-ohm
> coax to an 800-ohm load?
I'd like to offer a few more suggestions.
Conventional transformers can use a "Faraday Shield", bifilar and other
transmission line transformers can not. BUT...
A Faraday shield only works if it is separate short shield that is slotted
rather than a shield wound with the same turns as the winding it shields.
If it is wound with the same number of turns, it acts like it isn't even
there. Same as with a loop antenna, the shield simply becomes the actual
antenna. In a transformer a solid shield over a wire actually becomes the
winding and coveys energy to the other winding. No "shielding" takes place.
The advantages of a real Faraday shield are minimal. It aids in balancing
the windings (same in a loop). If you do use one, it needs to connect to
it's own ground or a balancing circuit (like a pot across the primary,
similar to what we did in old tube filament systems in hi-fi amps).
I believe the Pennent would be sensitive to degradation from common mode
coupling, and there is no easy cure for that. Using an isolated primary
conventional transformer would help, but I think additional decoupling
would be advisable.
You can model the effect I am concerned about by connecting a wire to one
terminal of the antenna in the model, and varying the length somewhere
around 1/4 wl long. This assumes you have a typical floating Eznec source.
If the pattern changes, the feedline must be isolated from the antenna at
the antenna in the real world so you have that "perfect" Eznec source. You
can determine the amount the amount of decoupling required by placing a
"load" between that wire and the feedpoint terminal.
Second step, move the wire to the other terminal.
Third step (and in this one ignore impedance changes), use two parallel
wires to each terminal. Leave the far end of the wires shorted. If you see
a pattern change, the antenna is sensitive to common mode impedances by the
feedline and you should determine how much feedline isolation you need by
adding two loads. R is ok instead of X for all isolation loads.
If we have some idea of effects and series impedances required to isolate
the feedline, we know what type of additional choke balun the system needs
(if any).
Roy Lewallen suggests this test for some applications, and I agree with
him. If you fail to make this check, the antenna might not work as modeled
when a feedline is attached. Remember, we do not have perfect ground
independent current sources in the real world.
73 Tom
--
FAQ on WWW: http://www.contesting.com/topband.html
Submissions: topband@contesting.com
Administrative requests: topband-REQUEST@contesting.com
Problems: owner-topband@contesting.com
|