Why Non-Woven Bag Lengths Drift: Root-Cause Tests for Feed Encoder and Web Slip

Bag-length errors should be diagnosed by their pattern. A constant offset suggests recipe scaling or measurement bias. Drift through the roll suggests tension or diameter compensation. Speed-related error suggests slip, acceleration, servo, or timing. Random variation suggests intermittent grip, feedback, material, or mechanical play. Printed registration errors require a separate mark-tracking check.

Measure a Sequence Before Touching the Settings

Operators often measure one long bag, correct the set value, and create a short bag five minutes later. Instead, collect consecutive samples in production order. Record target length, actual length, time, speed, roll diameter, dancer position if available, print-mark status, alarm, and any adjustment.

Use one defined measuring method. Flexible PP nonwoven changes when pulled, folded, warm, or stacked. Confirm the datum, instrument, specimen state, and inspector. A ruler method that varies by several millimeters cannot diagnose a smaller feed error.

Classify the Error Signature

Confirm the Recipe and Length-Control Mode

Identify whether the machine is running fixed-length feed, print-mark registration, or another mode. Record the actual recipe name and revision. Check target length, units, correction limits, acceleration profile, mark repeat, knife phase, and any compensation values against the approved Manufacturer documentation.

Do not restore unknown default parameters. Save the current PLC, HMI, drive, and recipe state before a controlled change. An OEM or ODM line may use model-specific scaling that cannot be copied from another machine.

Separate Web Stretch from Feed-Roller Slip

A stretched web and a slipping nip can both change finished length, but they leave different evidence. Map the web path and identify where length is commanded, measured, and physically gripped.

• Inspect unwind tension, dancer travel, brake or drive response, and roll winding.
• Check feed rollers for contamination, glazing, wear, flat spots, and unequal pressure.
• Verify roller parallelism and bearing condition.
• Look for dust, ink, treatment residue, or folds reducing contact.
• Compare actual surface travel with encoder or commanded travel using an approved safe method.

Siemens motion-control guidance includes slip compensation between a drive roller, web path, and measuring reference. The general lesson is that encoder counts represent true web travel only when the mechanical relationship is known and stable.

Inspect Encoder and Mechanical Transmission as One Chain

The feedback path may include encoder, measuring wheel, coupling, shaft, pulley, belt, gearbox, feed roller, cable, and drive input. A perfect encoder cannot detect slip that occurs after its measurement point. A loose coupling can create position loss even when the HMI shows a normal command.

Isolate energy before physical inspection. Electrical signal work belongs to qualified personnel.

Test Servo Response and Cut Timing Without Guessing

If the actual web feed is correct but finished bags are wrong, inspect the relationship between feed completion and cutting. Mechanical delay, knife phase, clutch response, sensor latency, or control sequence may shift the cut.

Trend commanded position, actual position, following error, speed, trigger, and cut event if the system exposes them. Step from low to normal speed. A growing following error suggests a different direction from a constant calibration offset. Do not raise servo gain or torque limits without the exact drive manual and Manufacturer approval; unstable motion can create equipment and safety risk.

Keep Print Registration as a Separate Diagnostic Branch

For printed fabric, compare actual artwork repeat with the approved value before blaming the sensor. Inspect mark contrast, size, quiet zone, treated or glossy background, sensor distance, teach-in, mounting, read window, and correction limit. Then review web tension and slip because the tracking system cannot correct unlimited mechanical variation.

If unprinted fixed-length output is stable but print-to-cut position moves, continue with the registration system. If both modes show length variation, return to feed, mechanics, tension, and measurement.

Close the Fault with a Reproduction and Validation Record

1. Reproduce the original pattern under recorded conditions.
2. Change one verified cause or component.
3. Run the same sample sequence at low, normal, and agreed high speed.
4. Repeat through a useful roll-diameter range.
5. Confirm bag dimensions, print position, seams, and scrap.
6. Save the approved recipe and maintenance evidence.

Send Zhengxin the machine serial number, recipe backup, ordered measurements, speed and roll-diameter history, web-path photos, roller condition, encoder or drive trends, and printed-mark data. This evidence allows the Supplier to distinguish a setting correction from a mechanical or control fault.