320 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1962 



a tuned grid circuit with the grid returned to the cathode through the 

 tuning coil. Grid coupling capacitance was then reduced to a min- 

 imum by using maximum inductance to capacitance ratio in the tuned 

 circuit, and loading the tuned circuit to the proper Q value with the 

 driving plate resistor. 



The second design requirement was to minimize the ring time of 

 tuned circuits from the transmitter-induced high signal level. This 

 was achieved by returning grids to cathodes without bias, thus limit- 

 mg the level to which the circuits could be driven by the transmitter. 



The third requirement was fast response to amplify the short pulse 

 echoes. This meant tailoring the Q values of all tuned circuits so that 

 the composite Q of the receiver would match the pulse length. This 

 was accomplished with the help of the appropriate equation published 

 by Mesny.* 



The fourth requirement was complete absence of regenerative feed- 

 back in the presence of high gain. A communication receiver of that 

 day was considered stable if it did not oscillate. Equivalent Q^ how- 

 ever, is a sensitive function of feed-back, and response characteristics 

 are readily altered by feed-back long before the point of oscillation is 

 reached. This requirement was met by using a superheterodyne re- 

 ceiver, limiting voltage gain on any one frequency to one thousand, 

 and changing intermediate frequency as required to accomplish an 

 overall voltage gain on the order of 10'. In addition, extreme pre- 

 cautions were taken in shielding, filtering, and common point 

 groimding. 



The receiver was intended for a 5-jusec. pulse. The overall response 

 was 90 percent of steady state in 5 /^sec. This characteristic was 

 independent of gain up to the point where thermal noise at the input 

 filled the cathode-ray screen. 



A new transmitter of the self-quenching or "squegging" type was 

 built to go with the new receiver. The transmitting antenna was a 

 4 X 4-wavelength curtain array with resonant reflector. The receiving 

 antenna was a single half-wave doublet with single resonant reflec- 

 tor. The frequency was 28.6 Mc, with pulse length of 5 /itsec. and 

 pulse recurrence rate of 3720/sec., giving a range scale of 25 statute 

 miles. The system went on the air in April 1936. The receiver 

 recovery to full sensitivity following the transmitted pulse appeared 

 to be instantaneous. Beautifully sharp echoes from aircraft were 

 observed almost at once, and within a few days they appeared all the 

 way to the 25-mile limit of the indicator. 



The spectacular success of the experiment was followed by a greatly 

 intensified effort. A primary objective was to reduce the size of the 

 equipment so it could be used on ships. The 28.6-Mc antenna was 



* Mesny, R., Time constants, build-up time and decrements. I'Onde Elec, vol. 13, 

 pp. 237-243, June 1934. 



