3.6. The error theory of the Decca system has been covered in detail elsewhere's, 
   
   
  
  
  
  
   
  
  
   
  
  
  
   
  
  
  
   
general pi 
and here we need consider only the data provided by the contour diagrams and the giving thr 
all-important distinction between relative and absolute accuracy. The figures will be av 
assigned to the various contours refer to the relative accuracy or repeatability of 39. P 
the system; that is to say, everywhere within the contour labelled 50 feet there is an photograt 
approximately 67 per cent chance that an aircraft would be able to recover a position servo-driv 
at which a previous Decca reading had been taken, with this accuracy. In other of one DX 
words, the figures assume initially that the Decca grid is an arbitrary framework of Decomete 
reference, highly stable within itself but having no known position on the earth's good. The 
surface. It is worth noting that the repeatability offered by a long-wave system in the Flig 
such as Decca is in general unaffected by changes in meteorological conditions. are convei 
At very long ranges and great altitudes investigations may in fact show that some paras. 5.6 
changes with atmospheric conditions do exist on the Decca frequencies, but so far 
these have not been detected. The uniformity of tracking (for example) that the 4. USE 
relative accuracy of the system permits, is of a high order, as will be seen by reference 4.1 Fi 
to the Ordnance Survey work below, and appears to be wholly unaffected by meteoro- mitting st: 
logical conditions. 3806 pour 
3.7. It will already have become clear that the distinction between relative and test gear, 
absolute accuracy is bound up with the question of the effective speed of propa- station an 
gation. Over sea water, where the speed is now known within very small limits, the approxime 
added tho: 
ratio between the absolute accuracy and the relative approaches unity and the system 
is accordingly used in hydrography on a considerable scale for absolute fixation 
with very little visual control. As soon as overland transmission paths are used, 
supplies ai 
insulated 1 
however, the ratio increases; from the evidence so far available, the best possible in radius t 
absolute accuracy to be expected over homogeneous terrain or electrically good used, vary 
soil is about two-fifths that of the relative, i.e., the figures assigned to the various requiremet 
contours for a given chain layout should be multiplied by 2.5. As the terrain gets long opera 
more broken, with patches of moderately and very bad conductivity, the ratio engineer ci 
increases sharply and recent trials in exceptionally unfavourable geological con- 42. TI 
ditions have shown that some 16 visual control points would have been required manpower 
in the chain coverage to restore the ratio from about 7:1 to 2.5:1. when it is 
3.8. Equipment. The position-fixing equipment carried in each aircraft using equipment 
the system consists essentially of two Decometer meters, one for each hyperbolic Cross-coun 
pattern, driven by a radio receiving set. The set contains three channels for the North Afr 
master and slave stations, the tuning and other radio adjustments being preset. erection of 
Any number of receivers can, of course, be used simultaneously and the set itself two hours 
is capable to some degree of adaptation to installation requirements of particular up S re 
Is climable 
classes of aircraft. Similarly, the Flight Log can take various physical forms, each 
performing the basic function of displaying the aircraft's position and recording 
its track on a chart. The three ground stations constituting the chain are broadly 
gressively : 
erect this n 
similar in composition, the basic electrical function of the central *master' station 43. A 
being to radiate a wave of constant frequency which is picked up by the receivers m order to 
and also by receiving equipment at the ‘slave’ stations. At the latter, the phase of operation « 
the received master signal is compared with that of the local transmission and the tively, but 
two are rigidly held in a pre-determined angular relationship, so ensuring a stable due to the 
and correctly-located hyperbolic pattern. If the survey area is greater than can be that when 
accommodated by the three stations in a single siting, one or more of the stations on pattern 
is moved to a new position when the initial area has been covered and this ‘leap- although it 
frogging’ process continues until the whole region has been surveyed. Occasionally a 
$ 
a fourth station is provided to speed up this process; it is also possible, and is the 
8 
  
    
wa 
  
  
ww A 5 
ee SSL emm