The verdict:
Magnetic Longwire Balun
is too expensive

by Alan Johnson

The most common antenna type used by SWL's is probably the end-fed random wire, also know as an inverted-"L", Marconi or long wire.

The N8KDV 10:1 transformer is the best buy, based on price and performance. Further reading: MLB -- a con or not? MLB -- not really a balun

The term "long wire' 'is incorrect in most circumstances, as the random wire is usually less than one wavelength long, which is the minimum length required for a true long wire antenna - the mistake is common and I'm guilty of it as well.
The textbook definition of the inverted-"L" is a horizontal length of wire suspended as high as possible in the air with a single-conductor wire connected at one end to serve as the lead-into the receiver.
The total length of the antenna wire and lead-in should be 1/4 wavelength for the lowest frequency of interest (length in feet = 234/frequency in megahertz).
This length turns out to be just about 70 feet for 3.3 MHz. Another part of the definition is a quality ground or counterpoise system at the receiver. This type of antenna system is so popular because it is sensitive to a wide range of frequencies, is omnidirectional in pickup and, most of all, is simple and inexpensive to construct and install.

Randon wire problems

There are several tradeoffs in performance that must be made for this simplicity. One is that the lead-in wire picks up signals just as well as the horizontal portion. Antennas as a general rule should be placed as far as possible from interference sources, but, by necessity, the random wire lead-in has to run right to your hones with its powerlines, TV, computer, etc.-- all great sources of interference. Also, the lead-in should be run well away from metal structures such as gutters and window frames.
So, why not use a shielded, i.e. coaxial cable, lead-in?
This brings up the subject of antenna impedance. Let us forego a definition, and just say that for maximum signal transfer, the impedance of the signal source (the antenna) should match that of its load (the receiver).
For frequencies at which the length of the random wire is a quarter wavelength, or an odd multiple of it, the antenna impedance is low, on the order of 60-80 ohms.
For even multiples of a quarter wavelength, the impedance can be quite high -- 600 to over 1000 ohms. Coaxial cable and the coaxial inputs of most receivers are designed to match a 50 to 75 ohm impedance.
Therefore, there will be several points across the SW spectrum where there will be a severe mismatch between the antenna and the coaxial lead-in and the receiver's input as well.
In transmitting applications, any mismatch over about 1.5 to 1 is undesirable.
In receiving applications the tolerance is greater, but a less than 2:1 or 3:1 mismatch is preferable for maximum signal transfer.
If a single wire lead-in is used, an antenna tuner (Transmatch) can be used to transform the high impedance of the antenna system to match the low input impedance of the receiver's input. Some receivers have a high impedance input terminal for random wire inputs- but there are still the problems of bringing an unshielded wire into the house.

Enter the balun

If one uses an antenna tuner at the receiver, but runs coax to the antenna wire, there's still a mismatch at the point of connection between the antenna wire and the coax, as well the inherent capacitance of the coaxial cable, which adds to the signal loss.
So why not do the impedance transformation right at the point where the lead-in connects to the antenna?
That's just what the MLB and N8KDV do, within limitations.
These two products are transformers to reduce the wide swings of the antenna's impedance to a range that is a practical match for coaxial cable. Balun is a contraction of "balanced-unbalanced", as this type of transformer was first to transform the balanced condition of a dipole to the unbalanced configuration of a coaxial feedline, without any impedance transformation.
However, these transformers can also be designed to yield an impedance transformation as well.
An additional benefit of such a unit is that it connects the antenna to ground for DC voltages (but not for radio frequency signals) so static charges are bled off the antenna.
This bleeding action does not occur with a single-wire lead-in) There are now two products available to the SW hobbyist community for the specific application of transforming the high impedance of a random wire antenna to that required to utilize a coaxial feedline. To be technically correct they are '`un-uns", since both the antenna wire and the Redline are unbalanced systems.

The Magnetic Longwire Balun (MLB) is a product of the Dutch firm OF Systems, Inc. It's a plastic cylinder 1 3/4 inches in diameter and I 1/2 inches long, made from two PVC pipe caps glued together with a binding post at one end and an S0-239 coax connector at the other.
It's usable over the frequency range of 100 kHz to 40 MHz.
It is priced (1992) at $54.95 (+ $3 shipping) for the balun alone or $69.95 (+ $4 shipping) for an antenna kit with 41 feet of wire, support rope and insulator (but no coaxial feedline).

There is a less expensive alternative, although it has received less publicity.
It is a 10:1 broadband matching transformer built and sold by Steven Lare N8KDV.
It was originally designed for use with Beverage antennas and required an earth ground at the connection point to the antenna. Steven has now redesigned the transformer so that it requires only that the coax shield be grounded at the receiver, making the transformer suitable for use with random wires.
It comes in a 4 x 2 x I inch blue plastic box and has two binding posts for antenna/ground connections and a S0-239 coax connector. Its price (1992) is $17 00 plus $2 for shipping.

In the laboratory

After four rainy weekends in a row, I was finally able to get out and evaluate these transformers.
The test antenna was a 70' length of wire {at least that's how long Radio Shack said it was!) suspended between trees with one end at about 30 feet and the other at about 7 feet. Seventy feet is equal to one wavelength at 13.4 MHz and equals one quarter wavelength at 3.34 MHz.
The actual measured 114 wavelength point was lower, at about 3.1 MHz, probably due to nearby trees and foliage.
I attached either of the transformers or a short wire at the low end of the wire and used a Heathkit antenna noise bridge to measure the antenna impedance directly at the feedpoint.
I made measurements at 2 MHz and at the midpoints of each of the international broadcast and amateur bands.
I ended up with two sets of curves -- the first was just the resistance reading given by the bridge, ignoring the fact that the antenna was non-resonant at many of the data points.
The second set was of the calculated feedpoint impedance based on both the resistance and the reactance.

I also did A-B comparisons between each transformer, between each transformer and a coaxial feedline and each transformer and a single-wire lead-in connected to the high impedance input of the Kenwood R-l 000 used as the test receiver. The receiver's "S" meter was used to compare signal levels.

The results

The "resistance" curve for the MLB was pretty well centered on 45 ohms with variation between 40 and 60 ohms throughout the test frequencies except for a dip to 20 ohms at 6 MHz (the 1/2 wavelength point which should be a high impedance point) and a peak to 75 ohms at 11.8 MHz.

The N8KDV transformer was closer to 50 ohms at the data points with the same variation except at 6 MHz where the reading was 190 ohms and 11.8 MHz where it was 70 ohms. The 70' wire alone gave readings between 60 and 80 ohms between 2 and 3.75 MHz, rose to 110 ohms at 4 MHz and was unmeasurable (greater than 220 ohms -the limit of my bridge- between 4 and 7.15 MHz).
At 7.15 and 9.75 MHz, I got readings of 200 and 190 ohms. There was another off-the-scale reading at I 1.8 MHz and above that it varied between 140 and 220 ohms.
For the calculated impedance values the curves were similar, but with wider variations in the figures: the MLB centered about 60 ohms, the N8KDV about 70 ohms, but with more variation than the MLB. The random wire itself varied about a central value of 180 ohms. Remember that these last values also reflect the contributions of a non-resonant antenna length to overall impedance.

A-B testing was very revealing. Comparing either transformer (with coax feeding the 50 ohm receiver input) to a single wire lead-in connected to the high impedance receiver input yielded slightly higher signal levels for the wire but slightly more noise as well.
I would expect the noise level to rise in a "real world"-situation, as I was testing out in the yard, away from the usual indoor interference sources.
Using the transformers versus coax feeding the random wire alone gave signal strengths about 10 dB higher when the MLB was used and 10-15 dB higher with the N8KDV unit.
Comparing the MLB vs. the N8KDV showed that the N8KDV gave signals 1/2 to 1"'S" unit higher than the MLB at most of the test points.

Recommenation

How essential is a perfect match? For receiving, anything less than a 3:1 match is reasonable, practically speaking.
For casual listening, feeding a random wire with 50 ohm coax with no baton will probably work just fine.
For those who are sticklers for optimization, I recommend these baluns, although the $55 (+ $3 shipping) price for the Magnetic Longwire Balun seems a bit high.

Based on price and performance, the N8KDV 10:1 transformer seems to be the best buy.
However, I must qualify my recommendation of this unit by pointing out that the quality of the components used might limit its useful lifespan if additional precautions are not taken during installation. All components, including the enclosure, are Radio Shack quality. I doubt that the plastic enclosure is very resistant to sunlight, although I have not done any long term testing.
The binding posts and coax connector are plated and will quickly rust if left exposed. Also, the N8KDV balun should not be used as an end insulator for the random wire it is connected to, as l don't think the plastic enclosure will take much strain.
All that said, the price (1992: $17 + $2 shipping) is certainly right if you don't mind expending a little creative energy in installing the unit. I think I'll install mine suspended from the antenna end insulator and enclosed in a plastic freezer container.

The MLB definitely wins in the convenience category.

The Magnetic Longwire Balun is available (1992) from Universal Radio, Inc., 6830 Americana Pkwy., Reynoldsburg, OH 43068, 0rders 800-431-3939 Info 614-866-4267.
The N8KDV 10:1 matching transformer is available (1992) from Steven R. Lare, 824 West 25th St., Holland, MI 49423