IT. 
   
  
  
   
    
  
  
  
  
  
  
  
  
    
   
  
   
   
    
  
  
   
   
    
   
  
   
  
  
“D 
The conduction of the laboratory tests was based on a survey, analysis, genera- 
lisation and the subsequent classification of all significant error sources. 
Then the respective partial tests have been conducted according to the principle 
of maximum response of the system to the relevant errors. For the instrumental 
errors which are physically interrelated, the partial tests have to be combined 
and arranged according to a prescribed sequence. Errors tested later should not 
affect the results of the preceeding testse 
The laboratory tests provide the initial performance level and the initial ca- 
libration data for the subsequent routine testse The task of the routine tests 
is to bridge the time periods between the laboratory tests, with the purpose of 
inspecting the less stable instrumental errors, This, however, requires a clas- 
sification of error sources according to their stability. In accordance with 
this classification, the optimum frequencies for the routine tests can be esti- 
mateds For this purpose, much experience is required — the estimates made so far 
may be regarded as a first approximation. This holds to some extent also for the e e» 
conduction of the individual partial tests, Hence, the proposals made should be 
regarded as a convenient initial approach rather than an ultimate optimum solu- 
tion, 
À compromise between the detection power of tests and the effort involved is 
necessary — regardless of how intelligible the conduction may bee 
After some general considerations on tests, the partial tests will be conducted 
first for the laboratory applications. The routine tests will be derived from 
the laboratory ones in a later stage, 
GENERAL CONSIDERATIONS 
The objectives of the tests will be restricted only to those which are import- 
ant for the users of comparators. In the case of laboratory tests, these objec- e o 
tives are: 
a) to define the state of adjustment of the measuring unit with all its compo- 
nents — thus to detect and calibrate the individual instrumental errors, 
b) to estimate the overall accuracy of observations performed with a compara- 
tore 
A further assumption has been made that the tests are based on Standard inputs 
(eg test grids), making use of the measuring system of the comparator. 
For the realisation of the objective a), the detective tests have to be app- 
liede This, however, requires creation and control of specific environmental 
conditions, adequate input structures and intelligible testing procedures. 
The creation of adequate environmental conditions may require Special re- 
Search. This applies to the external environments (e.g. allowable change in 
temperature, vibrations, shocks, etc.), internal environments (such as opti- 
mum optical magnification, measuring mark, illumination, "warming up" period