312 BELL SYSTEM TECHNICAL JOURNAL 



At low frequencies, conductor loss is often the principal cause of attenu- 

 ation. At high frequency, this loss may be still more important^ and in addi- 

 tion there may be losses in the medium around the two conductors. The 

 latter is particularly true when the conductors are supported on insulators 

 or are embedded in insulating material. There may also be losses due to 

 lines of force that detach themselves from the wires and float off into the 

 surrounding space (radiation). All three lead to attenuation and may be 

 expressed in terms of an equivalent resistance. They are amenable to cal- 

 culation for certain special cases. 



According to one view of electricity, the individual charges to which 

 lines of force attach themselves are unable to flow through the conductor 

 with the velocity of light If this is true, lines of force snap along from one 

 charge to the next in a rather mysterious fashion which we will not attempt 

 to picture at this time. This view, like others mentioned previously, tends to 

 relegate the charges and hence the currents to a secondary position. 



Although infinitely long transmission lines cannot be constructed in prac- 

 tice, it is possible, by a variety of methods, to approximate this result. In 

 general, a resistance connected across the open end of a short transmission 

 line, of the kind here assumed, absorbs a portion of the arriving wavepower 

 and reflects the remainder. If the resistance is either very large or very small, 

 the reflected power may be very substantial but, by a suitable choice of inter- 

 mediate values of resistance, the reflected part may be made very small in- 

 deed. In the ideal case, the arriving wavepower is completely absorbed. A 

 line connected to this particular value of resistance appears to a generator 

 at the sending end as though it were infinitely long. The particular resistance 

 that can replace an infinite line at any point, without causing reflections, is 

 known as the characteristic impedance of the line. This quantity depends on 

 the dimensions and spacings of the two conductors as well as the nature of 

 the medium between. A parallel-wire line, in air, usually has a characteristic 

 impedance of several hundred ohms. A coaxial line filled with rubber often has 

 a characteristic impedance of a few tens of ohms. A line having characteristic 

 impedance connected at its receiving end is said to be match-terminated. 



Reflections on Transmission Lines 



If the transmission line ends in a termination other than characteristic 

 impedance, or if there are discontinuities, due to impedances connected 

 either in series or in shunt with the line, reflections of various kinds will 

 occur.^ Much of the practical side of microwaves has to do with these re- 

 flections. 



"■ The losses in most conductors increase with ihe square rod of the frequency. 



•> At the higher fre(|ucncies, rellcclions may also occur at points where the wire spacing 

 changes al)rui)liy. In some instances al)rupt changes in wire diameter may be sulVicient 

 to cause reflection. These discontinuities may be regarded as changes in characteristic 

 impedance. 



