Luma PST Position Error: Scaling Is Wrong, Not PID

Correct spindle-factor numerator and denominator values restored repeatable positioning after the backup battery failure. The 0.1 correction tolerance could not compensate for the wrong 1/1 scaling because every new target was converted through the same bad position relationship.

Stop applying the usual quick fixes

Do not keep correcting each position from the keypad and saving the result. That changes one taught point or offset, but it does not repair the conversion between encoder movement and linear travel. The next commanded dimension invokes the faulty conversion again, producing another error.

Get the conversion right before touching the loop. Otherwise every tuning change masks the calibration fault and adds another variable.

Identify scaling as the real fault

The Luma PST commands a blade through a motor, gearbox, ball screw, and incremental encoder. Manual +/- commands move the blade directly. Automatic operation converts a keyboard dimension into an encoder target, travels at high speed when distance permits, then uses low speed to refine the final position.

The backup battery failure erased stored functions or parameters. Replacing the battery restored power retention, but reloading the supplied parameter sheet with spindle numerator and denominator values of 1 and 1 left the axis incorrectly scaled. Correct machine-specific spindle values ultimately restored operation without replacing the positioner.

A spindle factor normally relates encoder increments to mechanical travel. The relationship can be expressed conceptually as:

linear distance per controller count = screw travel per revolution / effective encoder counts per screw revolution

The controller may apply its numerator and denominator in either direction, so derive the convention from its parameter documentation or a controlled displacement test. Do not infer the direction solely from the parameter names.

The SIKO IG09M encoder is mounted on the screw in this installation. That placement makes screw rotation the measured variable; motor gearbox ratio does not directly change encoder counts per screw revolution. Screw pitch, encoder resolution as interpreted by the controller, quadrature evaluation, electronic scaling, and any coupling between encoder and screw still determine the final distance conversion.

Separate spindle scaling from motion-loop tuning

Do not treat spindle numerator and denominator as proportional and integral gains. A wrong scale changes the relationship between requested distance and measured distance. Loop parameters instead change acceleration, correction strength, settling, overshoot, and the response near the target.

The working 200 mm axis normally passed the requested dimension, reversed, and advanced again to reach an error near one tenth. That sequence shows a functioning coarse/fine positioning strategy. Copying its dynamic parameters to a different axis is still unsafe: the two configurations had different loop values, and the 200 mm screw moved more mass.

Use the 200 mm axis as an A/B test

The known-good axis provides a strong functional test because it uses an identical encoder and the same stated motor and screw arrangement. Its spindle factors are 7 and 10, a numeric ratio of 0.7. When the positioner assigned to the 500 mm axis was connected to the 200 mm mechanism and loaded with 7/10, positioning worked correctly.

That test proves three useful points:

• The Luma PST can read the encoder and execute the positioning sequence.
• The positioner hardware is capable of accurate control when its scaling matches the mechanism.
• The 1/1 values supplied for the 500 mm configuration are the first parameters to challenge.

It does not prove that 7/10 belongs on the 500 mm axis. Course length alone does not determine spindle scaling. If both axes use the same screw pitch, encoder interpretation, and encoder mounting, they may share a conversion factor; if any of those differ, they require different values. Different moving mass mainly changes dynamic tuning, not the geometric encoder-to-distance ratio.

Recover the machine-specific parameter pair

Recover the original numerator and denominator from machine records, a configuration backup, another identical 500 mm axis, or the machine builder's commissioning data. The final repair used the correct spindle values supplied by a qualified contact; their actual numbers were not recorded here, so replacing them with calculated guesses would create another uncontrolled calibration.

If no valid record exists, collect the physical data needed to calculate and test the scale:

• Encoder resolution as evaluated by the Luma PST, including the controller's edge-counting convention.
• Ball-screw travel per revolution.
• Any mechanical transmission between the screw and encoder.
• The controller's documented numerator/denominator convention.
• The configured engineering unit used by the position display.

Stop here if encoder resolution at the controller or screw travel per revolution is unknown. Read the encoder configuration, machine drawings, screw identification, and positioner parameter description before moving the axis. A ratio derived from travel length or moving mass has no physical basis.

Program and test the restored configuration

1. Record every current parameter before making another change. Capture the spindle pair, reference values, direction, limits, coarse/fine thresholds, loop value, timing values, and correction tolerance.
2. Check the encoder coupling at the screw, connector seating, feedback direction, and mechanical freedom. The installation reported no mechanical abnormality, but these checks prevent a loose coupling from imitating a scale fault.
3. Load the recovered machine-specific spindle numerator and denominator for the 500 mm axis. Do not copy 7/10 unless the mechanical and encoder conversion data show that it applies.
4. Leave the already validated dynamic parameters unchanged during the first scaling test. Changing scale and loop response together makes the result difficult to interpret.
5. Reference the axis using the machine's established procedure. Confirm that positive manual motion increases the displayed position in the expected direction.
6. Command a short movement in a clear part of the stroke. Compare physical travel with displayed travel using a suitable measurement method.
7. Repeat at a longer distance. If the error grows in proportion to distance, revisit the spindle ratio. If the error remains nearly constant, check the datum or fixed offset.
8. Run automatic moves from both directions so the high-speed approach, low-speed refinement, reversal, and final correction are exercised.

Keep personnel and tooling outside the travel envelope. Wrong scaling can make the controller believe it has more or less remaining travel than the mechanism actually has.

Verify more than one taught dimension

A single accurate target does not validate the repair. Test several positions across the usable 500 mm stroke, including moves that start above and below each target. Record commanded position, measured position, approach direction, final displayed position, and any reversal before settling.

Use the pattern to decide the next action:

• Error increases with position: correct the spindle conversion slope.
• Error remains constant: correct the reference or datum offset.
• Position is accurate but repeatedly overshoots: review the dynamic loop and coarse-to-fine transition only after scaling passes.
• Repeated moves to one target produce different physical positions: inspect encoder coupling, backlash, electrical feedback integrity, and mechanical loading.

The repaired axis should return consistently to multiple requested dimensions, not merely accept a keypad correction at one point. After verification, export or photograph the complete configuration and store the machine-specific spindle pair with the maintenance records. Replace the backup battery under a controlled maintenance plan so parameters can be captured before another loss.

FAQ