A conveyor position that increases during motion but returns to zero at every stop does not prove the encoder scale is wrong. The display may show a relative position, a synchronization input may repeatedly redefine its origin, or the counter itself may lose its accumulated value. These faults require different tests.
This procedure is for commissioning and maintenance teams diagnosing failed counts-per-distance calibration. It separates counting, position latching and synchronization before addressing electrical connections. You will establish which value resets, identify any unwanted input event, and calculate the scale only after the count remains stable.
Identify which position actually returns to zero
Start by identifying the displayed variable from the controller's documented diagnostics. A label such as conveyor position does not establish whether it represents an accumulated encoder count or distance since the latest synchronization event.
Record these values together, where the equipment exposes them:
| Value or state | What it represents | What a reset suggests |
|---|---|---|
| Raw encoder count | Accumulated transitions accepted by the counter | Counter reset, reinitialization, power interruption or count-processing fault |
| Latched count | Encoder count captured at a particular event | Another latch event or replacement of the stored reference |
| Relative conveyor position | Distance calculated from a selected origin | New origin, display mode or synchronization-state change |
| Synchronization input state | Electrical or logical state presented to the synchronization function | Unwanted edge, persistent assertion or missing command |
| Conveyor tracking state | Whether position processing is active | Stop-related initialization or application sequencing |
Observe a stationary conveyor before motion, during an authorized test movement, and after stopping. Save the raw count immediately before and after the stop. Also record the synchronization event counter or timestamp if available.
If the raw count holds but the displayed distance returns to zero, investigate the reference and display logic first. Changing counts per distance cannot repair an origin that is being rewritten.
If several consumers use the same encoder, distinguish their counters and reference events. The related guide to conveyor encoder sharing for multi-robot vision picking explains the broader distribution problem; this procedure focuses on the reset itself.
Separate counting, latching and synchronization
An encoder channel measures movement. A latch captures the count associated with an event. Synchronization associates a captured position with an image, target list or tracking reference. Depending on the implementation, an input may perform more than one of these functions.
Draw a functional diagram using the installed equipment's documentation. Show the encoder channels, counter supply, reference conductor, synchronization input and every output connected to that input. Identify the configured active edge and input function without relying on wire color.
A camera strobe output and a manually commanded synchronization output are separate signal sources. Their names do not establish that they may share a terminal. One may remain asserted during a period when the other expects the input to be inactive.
Build a short event trace:
- Calibration mode becomes active.
- The manual synchronization command is issued.
- The physical synchronization input changes state.
- The counter or application records the event.
- The conveyor moves and stops.
- The raw count and displayed position are read again.
For every transition, ask which device generated it and which function consumed it. An indicator that does not flash is useful evidence, but a short pulse may be invisible or absent from a slow diagnostic display.
If the behavior occurs only after restarting the application, compare it with conveyor vision startup synchronization troubleshooting. Startup initialization and stop-related reference changes should not be treated as the same fault.
Run controlled tests in a repeatable order
Use the manufacturer's documented commissioning mode. Any necessary movement must remain under the machine's approved safeguarding and operating procedure. Do not change wiring on an energized installation.
-
Save the starting configuration. Record the calibration value, encoder mode, input assignment, edge selection and relevant application states. Capture the original wiring arrangement so that every authorized change is traceable.
-
Test stationary count retention. With motion stopped, observe the raw count and synchronization diagnostics. Unexpected count changes point toward signal integrity, mechanical movement or counter behavior. A changing relative display with an unchanged raw count points elsewhere.
-
Confirm command generation. Issue the documented manual synchronization command under its required operating conditions. Check the logical output state, any output diagnostic counter and the physical terminal signal. A command that never reaches the output is not an encoder scaling problem.
-
Confirm input reception. Compare the output transition with the receiving input diagnostic. Check terminal assignment, electrical compatibility, pulse duration and active edge. If a valid signal reaches the terminal but no input event is recorded, investigate the input configuration and documented timing requirements.
-
Investigate multiple signal sources. Determine whether the manufacturer permits the connected outputs to share the input. If an isolation test is necessary, plan it under the documented commissioning procedure, de-energize before changes, and maintain all required safety functions. Never treat removing a wire as a permanent repair without understanding its production role.
-
Repeat the movement and stop. Use the same operating sequence and record the same variables. Change only one configuration or connection condition between tests. Otherwise, an apparent improvement cannot be assigned to a cause.
-
Restore and verify the intended system. Before automatic operation, restore required image synchronization, covers, connections and approved settings. A calibration that works only with a required production signal absent is not a completed repair.
When an output command exists but the receiver does not recognize it, use the checks in camera trigger common reference: diagnose missing responses. Apply the installed equipment's own interface requirements rather than assuming all digital inputs share the same reference scheme.
Check shared-input electrical compatibility
Two outputs connected to one input can produce unexpected behavior even when both devices work correctly by themselves. Possible effects include an input held active, a missing inactive interval, extra edges or output contention.
Do not assume that a shared connection is acceptable because it previously appeared to work. The controlling evidence is the output topology and the receiving input specification.
| Check | What to verify | Why it matters |
|---|---|---|
| Output topology | Push-pull, sourcing, sinking, open-collector or isolated contact | Determines whether sharing is permitted at all |
| Input circuit | Voltage range, thresholds, reference and isolation | Establishes whether the receiver can recognize the signal safely |
| Idle state | Electrical level between commands | A valid event may require a return to the inactive state |
| Pulse timing | Minimum active and inactive durations, filtering and edge selection | Short or merged pulses may not create separate events |
| Power sequencing | Output behavior during startup and shutdown | An unintended transition can redefine the reference |
| Input function | Latch, synchronize, reset, enable or another assigned operation | A correct electrical edge can still invoke the wrong action |
Never directly parallel push-pull outputs unless the manufacturer explicitly documents that arrangement. An open-collector designation alone also does not establish compatibility: voltage, current, reference and pull-up requirements still matter.
Where the design needs multiple event sources, use the manufacturer's supported input arrangement or a reviewed interface design. Separate inputs or suitable interface logic may be appropriate, but the required timing relationship must remain intact.
Measure signals with instruments and probes suitable for the circuit. An oscilloscope ground connection can unintentionally bridge isolated references. Follow the equipment and instrument documentation rather than attaching a probe ground by habit.
Calculate counts per distance only after retention passes
A scale error normally changes the relationship between distance and count. By itself, it does not explain why an accumulated count disappears at a stop.
Once the raw count holds and synchronization behaves as documented, measure a known conveyor displacement. Mark or reference the actual belt movement rather than relying solely on a commanded motor distance. Slippage between the drive and belt can otherwise be mistaken for an encoder error.
For example, suppose a measured movement of 500 mm produces 37,000 accepted counts:
- Counts per millimeter = 37,000 / 500 = 74 counts/mm.
- Counts per meter = 74 × 1000 = 74,000 counts/m.
- A later change of 18,500 counts represents 18,500 / 74 = 250 mm.
These are illustrative application values, not specifications for any listed product. Use the count reported by the configured counter. Do not multiply again for quadrature decoding if that counter already reports decoded counts.
Record direction separately. The absolute value is useful for calculating scale, but the application still needs the correct sign convention. If a counter wraps during measurement, use the controller's documented rollover handling or choose an approved test interval that avoids ambiguity.
Repeat the measurement at least three times. Compare the calculated scales and repeat the test after a stop. A consistent scale with inconsistent reference retention indicates that the calibration arithmetic is not the remaining problem.
Recognize when the apparent reset has another cause
A position display can legitimately return to zero when configured to show displacement from the latest reference. Confirm the intended semantics before trying to make that display behave like a continuously accumulated counter.
If the raw count changes abruptly without a synchronization event, inspect counter power, communications status, initialization commands and encoder diagnostics. If the count becomes erratic rather than resetting cleanly, investigate signal integrity and mechanical coupling.
A wraparound can also resemble a discontinuity. The guide to conveyor target acknowledgment near tracking position rollover addresses boundary handling in downstream tracking logic. Do not hide a rollover problem by changing the calibration factor.
Finally, stable counts do not prove that image events carry the right position. When calibration passes but picks remain displaced, follow conveyor vision scene timing: troubleshoot stale pick data. Count retention, image association and target age are separate acceptance tests.
Check parts against the diagnosed function
Replacement hardware should follow the failed test, not the symptom alone. Compare area scan cameras through the camera finder only if the imaging device or its documented output behavior needs changing. Sensor resolution and data interface do not establish synchronization-output compatibility.
The LS-CMR-0053 is a GigE area scan camera candidate. The supplied catalog information does not establish its strobe-output electrical characteristics, so request those details before using it in a shared-input design.
- Sensor format
- 1/1.7"
- Lens mount
- C-mount
- Interface
- GigE
- Resolution
- 12 MP
Select cables using the accessories finder only after identifying the affected circuit. A camera data cable and an encoder synchronization lead perform different jobs.
The LS-CBL-0199 is a high-flex GigE camera cable with screw locks. It is relevant to camera data transport, not evidence that an encoder input connection is suitable.
For vision controllers, verify the exact installed input circuit and counter functions in the manufacturer's documentation. A port count alone does not establish encoder-input, latch or synchronization capabilities. If replacing a camera, start with the camera configuration step and include the required trigger and output functions in the selection brief.
Use a final acceptance checklist
Before accepting the calibration, confirm the following:
- Raw count retention: the accumulated count remains consistent through the approved move-and-stop sequence.
- Reference behavior: each origin change corresponds to an intended, documented event.
- Command delivery: manual synchronization produces the expected physical signal and receiving event.
- Shared-input approval: every connected output is compatible with the documented circuit arrangement.
- Scale repeatability: repeated measured movements produce consistent counts per distance.
- Direction: the sign convention matches actual conveyor travel.
- Production restoration: required camera synchronization and all safeguards are restored before automatic operation.
- Evidence saved: configuration, wiring revision, count readings and event traces are retained for maintenance.
For the next step, send the signal diagram, installed interface specifications and synchronized count trace through application support. That evidence is more useful than ordering a replacement encoder before identifying which state actually resets.
Questions
Why does conveyor position return to zero when the belt stops?
First identify whether the display shows accumulated encoder position or displacement from a synchronization reference. If the raw count remains unchanged, the reset is probably in the reference or display behavior. If the raw count also resets, investigate counter initialization, power, configuration and documented reset inputs before recalculating the scale.
Can camera strobe and manual synchronization outputs share one input?
Only use a shared connection when the equipment documentation explicitly permits the output types and receiving circuit involved. Matching nominal voltage is not sufficient. Output topology, reference, current limits, idle states and pulse timing all matter, and conflicting outputs may require separate inputs or an approved interface design.
How do you calculate conveyor encoder counts per distance?
Subtract the starting raw count from the ending raw count, then divide the absolute difference by measured belt travel. Express the distance in the units required by the controller. Confirm the configured decoding mode and avoid applying a quadrature multiplier twice.
Does an output indicator that stays dark prove synchronization failed?
No. A brief pulse may not be visible on an indicator or slow diagnostic display. Check the logical command, physical output signal and receiving event separately. Also verify the operating conditions required for the command, since application logic may prevent it from being issued.
Should you replace the encoder before repeating calibration?
Replace it only after evidence identifies a fault in the encoder or its mechanical coupling. A correctly accumulating count combined with a resetting relative position points toward reference logic instead. Confirm supply, signals, counter configuration and synchronization behavior before selecting replacement hardware.