Error correction – HEIDENHAIN ND 1100 User Manual

Page 64

Advertising
background image

Chapter 4

Setup

4-22

Error Correction

Linear, segmented linear and optional non-linear error correction
methods can be used with the QC100. Linear and segmented
linear error corrections are provided with every QC100 system and
non-linear error correction is provided as an option. Each method
compensates for encoder and machine travel variations using er-
ror correction coefficients developed by comparing actual mea-
surements of a standard to the standard’s nominal values.

Linear error correction (LEC) uses one correction coefficient for
each axis to compensate for variations along the axis. A standard
is measured and the standard’s nominal and measured values are
specified during the setup process. The resulting correction coeffi-
cient is then applied to total measured displacements. For example,
a linear error correction coefficient of 0.0002 “/inch applied to a
6” measurement along the X axis would produce a measurement
result of 6.0012”.

Segmented linear error correction (SLEC) divides each axis into up
to 7 segments. Each segment uses one error correction coeffi-
cient to compensate for variations within the segment. Each axis
also includes a machine zero offset to refer the position of the
standard’s datum to the machine zero reference. The standard’s
nominal and measured segment values must be specified during
the setup process. The machine zero offset for each axis must
also be specified. When SLEC is used, the stage must be moved to
cross a reference mark/hard-stop on each axis upon system startup
to initialize SLEC.

Non-linear error correction (NLEC) uses a standard that divides
the entire measurement area into a grid of up to 30 X 30 cells.
Each cell uses error correction coefficient for each axis to compen-
sate for variations within a cell. The standard’s nominal and mea-
sured X and Y values must be specified for the positions of each
cell during the setup process. The machine zero offset to the grid
datum must also be specified. When NLEC is used, the stage must
have a repeatable machine zero defined on startup to initialize
NLEC.

Advertising