Laserscale Stability |
| . |
. |
As we mentioned before, scales are superior to
interferometers.
The main reasons drift occurs are temperature,
humidity, the effects of stress on the
mounting, and long-term changes in the gauge
itself. While these also depend on the usage
environment, in general, the influence of
temperature is large. To minimize drift, a scale
made from materials with a small coefficient
of thermal expansion is selected. Also, the
detector head, which is a source of heat, is
designed for minimal power consumption.
When accuracy enters the nanometer range,
humidity becomes a problem. This is because
moisture can be absorbed if the optical and
other components in the gauge itself are
mounted with adhesives. Also, inadequate
hardness in the detector block can cause drift
over long time periods.
Laserscale stability is verified by observing
the signal drift in the usage environment.
The scale and head are amounted on a block.
As shown in figure 3, stability is within
±1 nm for measurements taken over a 40-day
period. Laserscale also provides a ±0.1 nm
stability over an 8-hour period. |
|
Figure 3 Laserscale Static Stability
When the ultrahigh accuracy 138 nm signal wavelength is used. |
Interpolator |
| . |
. |
The interpolator converts the sinusoidal
primary signal voltage acquired from the
detection head to position information by
finely interpolating (dividing) that signal.
The scale accuracy is determined by
the variations in the wavelength and the
interpolation error that occurs during
each primary signal period. Since there
are no variations in the wavelength in an
interferometer, only the interpolation error is a
problem.
In a scale, there are cases when the variations
in wavelength during recording can be
averaged and ignored if the average phase of
multiple lattices is detected. For interpolation
accuracy, variations occur when the two-phase
waveform, which is the primary signal,
is displaced relative to an ideal sine wave.
That is, both errors in the offset and phase of
the two-phase sine wave output and errors in
the amplitude of the signal cause interpolation
errors. Therefore, highly accurate signal
correction is required to achieve highly
accurate interpolation.
This signal distortion problem occurs when
the measurement light and the reference light
are not completely isolated, even in ordinary
optical interferometers. Therefore it is not easy to require sub-nanometer accuracy in
an optical interferometer, and the correction
function is critical. The interpolator performs
the primary signal correction described above.
The interpolator calculates the phase from
the four signals with phases that differ by 90
degrees each that are output by the head. If
the signal center and the zero calculated by
the interpolator do not agree, an interpolation
error with the same period as the primary
signal wavelength has occurred. If the phase
difference is not 90 degrees and if the primary
signal amplitude is not equal, an interpolation
error of 1/2 has occurred.
The magnitudes of these errors can be
expressed as follows. The signal amplitude is expressed as R, and the error caused by the
signal center phase displacement, d, is the DC
error. The error when the signal amplitudes
are not the same is the gain error, and the error
due to the deviation of the inter-signal phase
from 90 degrees, α, is the phase error.
When we compute the amount of error, it
becomes clear that the primary signal error
must be made extremely small. For example,
if we want to hold the error to 0.1% of the
primary signal wavelength, the DC error must
be held to be within 0.3% of the amplitude.
It is necessary to perform extremely
accurate correction at all times to increase
the interpolation accuracy. It is possible to
achieve a ±50 pm interpolation accuracy by
monitoring and correcting the primary signal
DC components for each primary signal and
the gain in the interpolator. |
|
Photograph 1 Interpolator |
 Future Developments
 |
See
all articles with figures and tables.  |
|
Vol.54 |
|
