TRANSMISSION LINES: why 600V?

Jan 1, 2000 12:00 PM, Bill Whitlock

What is a transmission line, anyway? There are broad definitions, such as "the conductive connections between system elements which carry signal power" (ANSI/IEEE Std 100-1977 Dictionary of Electrical and Electronic Terms, page 739), under which every cable becomes a transmission line. Most engineers, myself included, would narrow the definition considerably. The following is the most concise, accurate and simple definition I have encountered: "A transmission line consists of an arrangement of electrical conductors by means of which electromagnetic energy is conveyed, over distances comparable with the wavelength of the electromagnetic waves, from one place to another. Transmission lines differ from simple electrical networks in that their inductance, capacitance, and resistance are not lumped but are distributed over distances such that the time required for electrical energy to travel from one part to another has to be taken into account. A uniform transmission line has what is called a `characteristic impedance'. This is the impedance that would be measured at the end of such a line if it were infinitely long. The importance of this characteristic impedance lies in the fact that if any length of line is terminated in an impedance of this value, then all the energy flowing along the line is absorbed at the termination and none is reflected back along the line." (Radiotron Designer's Handbook, F. Langford-Smith, Amalgamated Wireless Valve Company, Sydney, 1953, pages 890-891.)

The first widespread users of balanced transmission lines were the early telephone companies. The earliest systems had no electronic amplification yet needed to deliver maximum audio power from one telephone to another up to 20 miles (32 km) away. It is well known that, with a signal source of a given impedance, maximum power will be delivered to a load with the same, or matched, impedance. As stated above, it is also well known that reflection and standing wave effects will occur in a transmission line unless both ends are terminated in its characteristic impedance. Note that, as implied in the definition by "distances comparable with the wavelength," these effects occur only at frequencies where the time it takes the signal to propagate from one end of the transmission line to the other is significant. Because propagation time through 20 miles of line is significant even at voice frequencies, equipment at each end had to match the line impedance to avoid severe frequency-response errors because of reflections and standing waves. Early Bell engineers knew this, but the system impedance of the earliest telephone systems was dictated by the miles of open wire pair transmission lines strung along poles, which already existed for telegraph use. These lines typically used AWG#6 wires spaced 12 inches (305 mm) apart, which made their characteristic impedance exactly 600 V. For those who are interested, the formula for the characteristic impedance of such an open two-wire transmission line in air is Z = 276 log (2D/d), where d is the diameter; D is the spacing of the wires (in same units of measure), and Z is in ohms. Therefore 600 V became the standard impedance for these balanced telephone lines and all early telephone equipment in general.

Interference, mostly from the AC power lines frequently running parallel to the phone lines for miles, was largely eliminated through balanced operation of the lines. Balanced (for noise rejection) and impedance-matched (for power transfer) transmission lines were clearly necessary for acceptable operationof the early telephone systems, which had no amplifiers. Later, as the telephone network grew, amplifiers, filters and "hybrid" transformers were added to enable long-distance transmission. Proper operation of these components depended critically on rather precise 600 V impedances. Because other audio systems were rare, telephone equipment and practices were adopted by radio broadcasters and, later, by recording studios and audio professionals in general. Passive filters and EQs, step attenuators and amps having 600 V input and output impedances were widely used well into the 1960s, and some of this equipment is still used today.

In modern professional audio, the goal of signal transmission systems is to deliver maximum voltage, not maximum power. Thus, devices need low output impedances and high input impedances. This practice is the subject of a 1978 I.E.C. standard requiring output impedances to be 50 V or less and input impedances to be 10 kV or more. Sometimes called "voltage matching," it minimizes the effects of cable capacitances and allows an output to drive multiple inputs simultaneously with minimal level losses.

Nor do modern systems need to terminate audio cables. For a 20 kHz audio signal in typical cable, a wavelength is about 40,000 feet (12,200 m). As I have said in a previous column on long lines, these effects begin to become significant only when the physical length of the line becomes about 10% of a wavelength at the highest frequency of interest. For audio signals in typical cable, this is about 4,000 feet (1,220 m). Therefore, with rare exceptions, such as telephone equipment interfaces, the use of matched 600 V sources and loads in professional audio is generally unnecessary and usually degrades performance.

First, driving a 600 V terminated input rather than a 20 kV (i.e., high-impedance or bridging) input to a level of +22 dBu requires an additional 23 mA of peak current from the line driver. Because the line driver IC may already be taxed near its limit driving cable capacitance (see S&VC, October 1999, page 78), the termination may cause high-frequency current limiting, making cymbal crashes sound ugly. Second, if the driving equipment actually has a 600 V output impedance, driving a 600 V load will throw away half the signal, reducing headroom and S/N ratio by 6 dB. The notion that impedance matching and termination are required for modern audio equipment and ordinary audio cabling is simply an old idea whose time has long past.