'
')XQFWLRQDO'HVFULSWLRQRIWKH&RQWUROOHU
signal and the pulse frequency at the counter are increased. The technical
limits in the direct counting mode are reached at a velocity of 10m/s, however
in the high resolution ranges this value decreases approximately by the multi-
plier (refer also to section 4.2.3, Measurement range).
'$ FFXUDF\
Due to its working principle, the displacement decoder attains its highest
accuracy in the direct counting mode. As long as the optical signal is free of
drop-outs, exactly one COUNT pulse is generated and accumulated for each
phase cycle of the interferometer. The accuracy of the equivalent increment of
316.4nm is solely determined by the wavelength stability of the helium neon
laser which is in the order of magnitude of 10
-5
. Helium neon lasers are a gen-
erally recognized standard for length measurements. Apart from the digital
residual error, the status of the digital fringe counter thus very accurately cor-
responds with the instantaneous position of the object, independent of the
influence from electronic components. The same applies for the extended
measurement ranges where only every nth COUNT pulse is accumulated.
Due to tolerances and drift of the analog components however, the subse-
quent digital-to-analog conversion and amplification cause an additional static
calibration error of maximum
±
1% of the measurement value. At higher fre-
quencies, the amplitude frequency response of the smoothing filter causes an
additional frequency dependent error of
±
0.5%.
In the high resolution ranges, the interpolation (phase multiplication) is an
additional potential source of error. As with every interpolation, additional lin-
earity errors can occur between the known values, which depend on the fre-
quency and acceleration of the object. Up to a characteristic frequency for
every measurement range, this dynamic amplitude error can practically be
ignored as it remains below 1%. Above this frequency which is to be taken
from the decoder specifications, the amplitude error can increase up to
approximately 10%. A precise error diagram can not be presented in a clear
way due to the dependency on acceleration.