Low-Loss Coaxial Cable
December 1966 QST

When it comes to low loss transmission media, it's hard to beat waveguide and open wire. Open wire can exhibit less a couple tenths of a decibel per hundred feet at low frequencies, but it is very susceptible to perturbations from nearby objects, wind and moisture. Waveguide exhibits a few tenths of a decibel per 100 feet at very high frequencies, but it is expensive and difficult to work with. In the middle is coaxial cable, which for a good quality product of appropriate size, you can get very low attenuation. As with most things, you get what you pay for in coax cable. I once used really expensive Andrew (now Commscope) Heliax coax cable on an S-band radar (2.8 GHz) system that had only a little more than 1 dB/100 ft, which was necessary from a receiver noise figure requirement rather than for transmitter power efficiency. This article from QST covers some of the basics of low loss cable.

Low-Loss Coax

Reduced Attenuation Through Improvements in Construction

By Freeman F. Desmond and Richard Tuttle (W1QMR)

c/o Times Wire & Cable, Wallingford, Conn.

We have all used RG-8/U or similar cables for radio-frequency transmission. Very often we have wished for a cable with lower losses and improved power-handling capability - one that is also relatively easy to install and reasonable in cost. Such a cable, with compatible connectors, has been designed and is now available.

To see how this improvement is accomplished, we should first look closely at various coaxial constructions and examine the factors which cause losses. The ideal coaxial cable - an inner conductor suspended, by air alone, concentrically inside an outer conductor - is for all practical purposes impossible; actual coaxial cable must be manufactured with supporting material or dielectric between the two conductors. The loss in percent of total contributed by each of the cable components may be seen in Fig. 1.

At 100 Mc. approximately 80 percent of the total loss is copper loss in the center conductor. At lower frequencies this percentage contribution to loss is even greater. Therefore, the center conductor is the most important factor to consider in reducing cable losses in a given frequency band.

The Center Conductor

It would be most desirable, to achieve lower attenuation, to increase the size of the center conductor. We do not, however, wish to change the characteristic impedance nor contribute substantially to size or weight of the cable. Since impedance is dependent upon the geometry of the cable,

where Z_o = Characteristic impedance

           K = Dielectric constant

           D = Dielectric diameter

           d = Center conductor diameter

merely increasing the size of the center conductor would change either the characteristic impedance or add substantially to the overall size and weight of the cable, neither of which is desirable. The dielectric constant is the area that we can change, if a material of lower dielectric constant can be used practically.

Solid polyethylene (dielectric constant K = 2.3) is the dielectric material of most coaxial cables. By changing to foamed polyethylene (K = 1.5) we may increase the center conductor size and lower the attenuation without changing the overall diameter or the impedance. In this dielectric minute air bubbles are encapsulated in the polyethylene during manufacture. Incorporating air bubbles brings the finished product closer to the dielectric constant of air (1.0), which is the goal, and achieves the lowest attenuation characteristic. With cellular polyethylene, costly pressurization of the cable is not necessary, as it is in the case of disk-supported, helical-supported, or spline-supported semiflexible coaxial cables.

Figs. 2 and 3 show the attenuation and power capacity vs. frequency of RG-8A/U, Times T-4-50 and Times 1/2-inch Alumifoam. The lower attenuation is evident in the latter two because of increased size of center conductor and changed dielectric constant.¹

Fig. 2 also illustrates another point: When a stranded center conductor is used instead of a solid, smooth, center conductor the spiraling of the stranding results in a spiraling of the r.f. current along the conductor, creating a longer r.f. path length. Coupled with the higher resistivity of the center conductor because of the contact resistance between the strands, this contributes to higher attenuation in the finished cable. RG-8A/U has a stranded copper center conductor, while T-4-50 and 1/2-inch Alumifoam have solid-copper center conductors.

Jacket Material

Another factor which may affect attenuation is the jacketing material of the cable. Most flexible coaxial cables use polyvinylchloride (p.v.c.) as the jacketing compound. However, to use p.v.c. - a relatively hard, brittle substance - it is necessary to add plasticizers to make the compound pliable and flexible. The nonresinous plasticizers compounded with p.v.c. have a tendency, with sunlight and summer temperatures, to leach out of the p.v.c. and migrate into the polyethylene of the dielectric. The migration of the plasticizer through the braid into the dielectric causes the dielectric constant and power factor to rise, with a resulting rise in the v.s.w.r. and an increase in attenuation. A rise in attenuation of 1 or 2 db. per 100 feet is not uncommon, once contamination has begun. Also, with the migration of the plasticizer the p.v.c. becomes brittle and nonpliable, resulting in cracking and breaking of the jacket. RG-8/U, RG-11/U and RG-17/U are examples of coax cables with contaminating p.v.c. jackets.

The dangers of this condition have been recognized, and in many of the military cables, identical in every respect except for jacket material, the older styles have been replaced by new ones. Cables such as RG-8A/U, RG-11A/U, and RG-17 A/U use p.v.c. jackets with a resinous plasticizer which does not leach out or migrate, and thus does not contaminate the dielectric. Life expectancy of this type jacket is in excess of fifteen years.

High-molecular-weight, carbon-black-loaded, polyethylene jackets such as Xelon contain no plasticizers of any kind, consequently a useful life of 25 years or more can be expected. Because of this, polyethylene jackets permit direct burial and are usually specified for submersible applications.

Impedance Uniformity

Attenuation is also increased by substantial v.s.w.r. Since v.s.w.r. is a function of the impedance of a cable, it follows that the more uniform the impedance the lower will be the v.s.w.r. (for a given termination). Because coaxial cable is manufactured of plastic materials by means of bulky extruders, it cannot be held to the tolerances of machined parts, especially in lengths of many hundreds of feet. Each individual extruder has its own peculiar eccentricities that cause variations in the cable during manufacture.

These variations in dimensions are very small but, unfortunately, sum up electrically along a length of cable and, at specific frequencies, may result in a v.s.w.r. as high as 4:1 even though the cable is properly terminated. In cable constructions where impedance uniformity and low v.s.w.r. are critical, the impedance can be held to tight tolerances by close control of the extrusion processes.

Cable Construction

Taking the foregoing into account, let us look at RG-8A/U, shown in cross-section in Fig. 4. The center conductor is stranded copper and the dielectric is solid polyethylene. The attenuation of RG-8A/U could be improved by 25 percent if we could increase the center conductor size and change to foamed polyethylene. This has been done in cable such as Times T-4-50, now available at about the same cost as RG-8A/U. Note that the overall diameter is the same, Fig.  4, but the attenuation is substantially improved (Fig. 2) and the cable weight is improved (99 lbs./1000 ft. for RG-8A/U, 94 lbs./1000 ft. for T-4-50).

However, for longest life and most carefree installation, even further improvements have been made. The largest factor contributing to degradation of attenuation in foamed polyethylene flexible coaxial cables, especially above 100 Mc., is moisture. Since moisture affects the power factor, the effect of moisture in the cable becomes significant as we increase frequency. This can be seen from the formula for attenuation:

          where, 

As frequency is increased, the power factor becomes a more significant figure. Moisture has been known to degrade the power factor by as much as ten times.

But how does moisture get into a cable? It enters flexible cables as water vapor, which is a very penetrating gas. This vapor condenses to water or moisture, changes the power factor and consequently raises the attenuation. For this reason, a solid, seamless, pinhole-free, metallic barrier or shield which positively excludes water vapor gives the longest-lived cable. In addition, with a solid metallic sheath the radiation into and out from the cable is eliminated, and isolation the order of 100 db. is achieved.

In cables such as the Times Alumifoam series moisture is precluded during manufacture by a completely dry core, and with the addition of the aluminum tube the foamed polyethylene is under constant pressure. Moisture traps and vapor paths are designed out, and the user has a self-sealing cable.

For above-ground applications, the seamless shield serves the dual function of electrical shield and protective cover. It eliminates the necessity for an outer jacket and thus represents the most economical use of weight and space to achieve desired electrical characteristics. To approximate the electrical characteristics of 1/2-inch Alumifoam in an RG cable, it would be necessary to use RG-17A/U (attenuation, 0.85 db. at 100 Mc.; power handling, 3.6 kw. at 100 Mc.: cost, approximately 30 percent higher). Cross sections of the two types are shown in Fig. 4.

System Installation Using Semiflexible Coaxial Cable

Fig. 5 illustrates a typical system installation employing 1/2-inch Alumifoam. The cable is simple to install, and connectors are readily available for it.

It is generally most convenient to run from the transmitter to the wall of the shack with flexible coax (RG-8A/U or T-4-50), although to eliminate losses, this run should be kept as short as possible. One end of this short run should be terminated in a connector that will mate with the transmitter, and the other end may terminate either in a type N or go directly into a splice connector. Splice connectors to accept flexible coax in one side and solid-sheath coax in the other are also available.

The main feeder run should be cable similar to 1/2-inch Alumifoam because of its low-loss characteristics. Since the cable is designed to be bent upon installation, there is no electrical or physical damage in bending it, even on a radius as small as ten times the o.d. of the cable. The cable should be terminated to match with the transmitter cable connector (type N or splice). The type N connector is better than the PL-259 because of its lower v.s.w.r., greater power-handling capability and improved radiation characteristics. For still lower system v.s.w.r., the conversion splice is the wiser choice.

The antenna end of the main feed line should be terminated by the same procedure as the transmitter end. Jumping from the solid-sheath cable to the antenna is accomplished by means of a short length of RG-8A/U or T-4-50. The cable loops once around the rotator and is terminated as you now terminate in your antenna.

In conclusion, solid-aluminum-sheath cable can be reasonably expected to deliver more power to the antenna because of its low v.s.w.r. and lower attenuation characteristics. Combining solid-sheathed, foamed-polyethylene dielectric, aluminum-shielded coaxial cable and low-v.s.w.r. connectors gives a feeder connecting system which, at v.h.f. and u.h.f., may deliver as much as 70 per cent more power from transmitter to antenna in a comparatively short run.

¹ T-4-50 is a flexible cable with foamed dielectric and braid outer conductor; Alumifoam is similar but uses seamless aluminum tubing as the outer conductor. Both are made by Times Wire & Cable Div., International Silver Co., 358 Hall Ave., Wallingford, Conn. 06492. -- Editor.

Posted April 7, 2020
(updated from original post on 3/15/2013)