9-3] EXTERNAL INPUTS: UNDESIRED AND DESIRED 477 



but it is an inherent characteristic of the desired target input. At shorter 

 ranges where the signal strength is greater, fading is less likely to occur.^ 

 Although fading is not a predictable phenomenon and there is no accurate 

 mathematical description of it, it may persist for periods greater than 1 sec 

 and possibly as great as 3 sec at long ranges.^ The tracking system must be 

 capable of maintaining the antenna position approximately on the target 

 during a fading period with an angular accuracy which will permit restora- 

 tion of tracking when the target return signal becomes strong enough to be 

 detected after the fade. 



Target Information Signal. Information about the angular space 

 error between the sight line and antenna tracking line is presented to the 

 antenna drive circuits as sidebands on a suppressed carrier signal. In a 

 conical scan radar, as shown in Fig. 5-13 for example, the carrier frequency 

 is obtained from a rotary motion of the antenna beam in space. This 

 motion is often provided by mechanically driving the antenna feedhorn 

 with a small motor. The spinning feedhorn provides a modulated pulse 

 return from the target when it is not on the tracking line. An unmodulated 

 carrier frequency is also obtained from a generator on the feedhorn to 

 provide a reference for extracting space error information from the mod- 

 ulated carrier. The suppressed carrier signal actually contains error 

 information for two antenna control channels simultaneously, and two 

 phase reference signals from the feedhorn generator are needed in a tracking 

 circuit demodulator to separate the information required for each channel. 

 The tolerances which apply to the carrier frequency amplitude and phase 

 are now discussed. 



The selection of the carrier (lobing) frequency is dictated by the following 

 considerations: 



(1) Natural modulating frequencies of the target (see Paragraph 4-8) 

 should be avoided. For example, a propeller-driven aircraft might produce 

 a 70-cps modulation on the signal return; if the lobing frequency was also 

 70 cps (or a harmonic of 70 cps) poor tracking performance would result. 



(2) The lobing frequency cannot be higher^ than ^ of the basic radar 

 sampling frequency; otherwise stability problems will arise incident to the 

 effective delays introduced by the sampling process. For example, if the 

 PRF of a pulse radar is 1000 pps, the lobing frequency should not be higher 

 than 100 cps. 



^Except in a pulsed doppler radar with high pulse repetition where fading is definite and 

 periodic because of eclipsing. 



^Because of flight and attack time limitations, longer fades would probably dictate a return 

 to the search mode of operation discussed in Chapter 8 with reference to the AGC design 

 problem. 



^To achieve almost complete protection from the effects of using sampled data in the usual 

 case, a 20 : 1 relationship should be used. 



