108 



RADAR SCOPES 



direct visual studies on phosphor materials. 

 So far this has not been done, at least in any 

 systematic fashion. The current method of 

 evaluating phosphor materials is by photo- 

 cell or multiplier phototube, the material 

 being excited by light of short wavelength. 

 Good descriptions of the method are found in 

 Hopkinson (23) and in Hardy (20) . A simi- 

 lar method is used in evaluating completed 

 CRT screens, except that here the excitation 

 is by electrons. Measures are made of spot 

 diameter, spot brightness, decay times, build- 

 up on successive excitations and others which, 

 together, constitute a fairly complete battery 

 of quantitative tests which ought to predict 

 operational performance. Good accounts of 

 the method are given by Johnson and Hardy 

 (26), by Garhck, Henderson, and Puleston 

 (14) , and by Nottingham (38, 39, 41) . The 

 scan pattern, of course, is not a radar one, 

 but a small test raster. Scientific caution 

 dictates that photocell measures ought prop- 

 erly to be regarded as only a set of pre-tests 

 which may not always predict very accu- 

 rately the luminances which are critical for 

 effective visual use. That they actually do 

 fail is witnessed by Nottingham (41), who 

 concludes that, "It has not been possible to 

 bridge the gap between objective measure- 

 ments and radar tube applications and it is 

 unlikely that this gap will be closed in the 

 near future" (41, p. 15). He was referring 

 specifically to a comparison of the "cycUc ex- 

 citation" method of the British — a photocell 

 index of phosphor decay — and the "build-up 

 ratio" method of the Americans — an index 

 of cumulative brightness with successive ex- 

 citations. Neither method measured di- 

 rectly a signal-background contrast of the 

 sort found in operational radar, mainly be- 

 cause they were not effectively simulating 

 noise excitation. Consequently no com- 

 pletely satisfactory test of phosphor perform- 

 ance has been developed to use in manufac- 

 turing control. The ultimate test is human 

 vision; but though this is admittedly cum- 

 bersome as an acceptance test in a tube 

 plant, certainly it might be possible to con- 



struct a visually vaUd photocell "aptitude" 

 index. Nottingham's problem may be solu- 

 ble by the statistical methods familiar in 

 aptitude testing. 



The measurement of luminescence during 

 operational scan is almost too difficult to 

 accomplish, although a partially successful 

 attempt has been made at Systems Research 

 Laboratory by Hamburger and King (17) 

 whose records proved extremely useful in the 

 interpretation of human visibility data (54, 

 55). 



The properties of an efficient screen are 

 difficult to specify. Screen thickness, for 

 example, goes through a definite optimum, 

 but the optimum in turn depends on amount 

 of binder and method of processing, as well 

 as on the working voltage (23). The junc- 

 tion of the phosphor and the glass and the 

 elimination of metallic impurities are ex- 

 tremely important. Some appreciation of 

 how tubes are made is almost a requisite for 

 visual research. As an empirical guide, a 

 population study of manufacturer's tube 

 types, using visual perfoiTnance measures, 

 is urgently needed. 



A great contribution to phosphor design 

 might be made by psychology and physiol- 

 ogy. For example, manufacturers have 

 often assumed a constant of i^ second (30, 

 31) as the value for "visual persistence." 

 We know, of course, that no theory of tempo- 

 ral integration of the eye is as yet completely 

 founded; yet, some of the empirical data of 

 the flicker studies might be applicable, and 

 certainly the needs of radar should stimulate 

 new research. In England, Hopkinson (24) 

 felt it necessary to carry out experiments not 

 only on flicker but on movement of flashes 

 and their apparent brightness in order to 

 arrive at some rational guide to CRT con- 

 struction. Recently research has been un- 

 dertaken by Bartlett (3) and Sweet and 

 Bartlett (48) which has a direct bearing on 

 the problem. Reliable data on integration 

 time as a function of stimulus size, brightness 

 and movement and of background area, tex- 

 ture and brightness should be brought to 



