7.4 Fine Tuning the System

• Several subsequent fine adjustments procedures rely on the field being correctly scaled, and undistorted.

Skip this step if marking-on-the-fly is not used.

Method 1:

• Stop the belt, then read the encoder position Enc₀.

• Advance the belt by a measured distance D and again read the encoder position Enc₁.

• The encoder’s distance per count can be calculated as:

  • DistancePerCount = D / ( Enc₁ – Enc₀)

  Note: The precision improves as the distance between encoder readings increases.

Method 2:

• Compile a CommandList consisting of two parallel lines L1 and L2, perpendicular to the axis of the belt’s movement, and separated by a MOTF_WAIT_DISTANCE(D) command, where D is equal to the difference between the two lines’ ordinates on the axis of belt movement.

  For example, if the belt moves in the negative X direction (horizontally), then we would use two vertical lines.

  Hint: set up the two lines with different lengths so you can easily distinguish them.

• Execute the list.

• If the encoder's current distance per count is correct, the two lines will overlap.

  Otherwise, measure the distance delta between the two marked lines along the axis of belt movement, which in our example is the X axis:

  • delta = X_L1 – X_L2

  (This means delta is positive if L1 is further to the right than L2.)

  Now the corrected distance per count can be determined as follows:

  • Corrected distance per count = ( 1 – delta / D ) * Current distance per count

Skip this step if marking-on-the-fly is not used.

• Compile a CommandList consisting of two lines L1 and L2 of equal length L, which are co-linear with themselves and with the axis of belt movement, and separated by a MOTF_WAIT_DISTANCE(L) command.

  For example, if the belt moves in the negative X direction (horizontally), we would use two co-linear horizontal lines.

• Execute the list.

• If the scanner coordinate system is perfectly co-linear with the belt’s coordinate system, the end point of L1 and the start point of L2 will coincide.

  Otherwise, measure the distance between those two points perpendicular to the axis of belt movement. In our example, this is the Y direction:

  • delta = YEnd_L2 – YStart_L1

  (This means delta is positive if L2 is further towards the top than L1).

  Now the angle by which the current field transformation matrix should be rotated can be calculated as follows:

  • angle = arcsin( delta / L )

Tracking Error (Method 1):

• Run a bi-directional pattern, for example two horizontal lines separated by a very small delta in the vertical direction.

• Adjust the Tracking Error until the end points of both lines are horizontally aligned.

• Rotate the example pattern by 90 degrees in order to calibrate the Tracking Error for the Y axis.

Tracking Error (Method 2), or Tracking Error & Laser Trigger Delay:

• If supported by the scan head, stream both the scanner command and feedback channels of each axis, and calculate the Tracking Error based on their temporal difference.

• Optionally, for Laser Trigger Delay:

  • Run a bi-directional pattern, for example two horizontal lines separated by a very small delta in the vertical direction.

  • Adjust the Laser Trigger Delay until the end points of both lines are horizontally aligned.

• If your application requires high jump and mark speeds, these delays should all be adjusted according to the suggestions in 7.1.6 Synchronising the Laser and Scanners.

• The velocities for the fine adjustment should exceed the application’s maximum speeds by an appreciable margin.

  Any errors are far less noticeable at lower speeds.

• We recommend leaving the Laser On & Off delays at 0 during this step.

• Fine tune these two parameters at each distinct mark velocity, and if necessary individually for different objects, until the desired marking quality is achieved.