Camera trigger delay timing for offset inspection sensors

An upstream part sensor can detect every object correctly while the camera still photographs the wrong position or misses closely spaced parts. The problem is often the timing design, not the trigger wiring. For irregularly spaced conveyor inspections, you need to align exposure with part travel and establish what happens when another event arrives during the delay. This guide shows how to calculate the delay, choose where to implement it, and verify the complete sequence before ordering hardware.

Choose the timing strategy before setting a delay

Use a fixed time delay when conveyor speed remains sufficiently constant throughout the sensor-to-camera journey. Use position-based scheduling when acceleration, stops or speed variation would move the part outside your inspection tolerance.

A third condition matters just as much: the system must retain every required part event. A camera that accepts only one pending trigger may become unavailable while its programmed delay runs. That can make an otherwise adequate camera miss parts long before its advertised acquisition rate becomes relevant.

Hardware triggering generally provides a more controllable timing path than a host application issuing commands through an operating system. However, it does not remove sensor latency, input filtering or camera response time.

For the separate question of how quickly a camera can accept external pulses, use the guide to camera frame rate with external triggers. Here, the additional concern is how far each part travels before exposure actually occurs.

Define the reference points and timing budget

Measure distance along the part's travel direction, using the same physical feature at both locations. A sensor detecting a leading edge does not automatically locate a barcode centered elsewhere on the part. Include that feature offset in the required travel distance.

Choose whether the target position should coincide with exposure start or exposure midpoint. Midpoint alignment is often useful for moving objects because it centers their movement during integration. The following model assumes constant speed and a global exposure interval.

Here:

• L is the feature's travel distance from physical sensor detection to its desired image position.
• v is conveyor speed during that travel.
• Tsensor is the delay from physical detection to the sensor's electrical output edge.
• Tcontrol is the delay from that output edge to the camera input edge.
• Tcamera is the camera input-to-exposure latency, excluding the programmed delay.
• Texposure is the integration time.
• D is the additional programmed delay.

For exposure-start alignment, omit the half-exposure term. For rolling-shutter imaging, different rows have different exposure times, so a single midpoint does not describe the entire image.

Check the manufacturer's documented timing reference before applying the equation. A delay attached to frame start may not have the same reference as one attached to exposure start. Avoid subtracting a latency twice if a documented measurement already includes another term.

Work through a constant-speed example

Consider a hypothetical inspection with these application measurements:

These values illustrate the calculation; they are not specifications for any catalog product.

First calculate travel time:

Convert exposure time before combining terms. An exposure of 40 µs is 0.040 ms, making its midpoint offset 0.020 ms.

The full timing chain is therefore:

This is an initial setting, not a finished calibration. Read back the actual accepted setting because supported increments may round the requested value. Verify the image position with a measured target, then investigate any discrepancy rather than hiding every error inside a larger delay correction.

Speed sensitivity can dominate the timing budget. If speed rises to 808 mm/s while the midpoint remains scheduled at 225 ms, travel becomes:

A speed increase of only 1% has shifted the feature by 1.8 mm. If the inspection cannot tolerate that movement, a fixed delay requires tighter speed control or replacement with position-based scheduling.

Budget exposure blur and timing jitter separately

Short exposure limits motion blur. It does not correct uncertain exposure timing. Conversely, precise triggering does not freeze a moving part during a long exposure.

Start with the required object-space sampling. Suppose a hypothetical inspection covers 120 mm horizontally and needs four pixels across a 0.20 mm feature:

The camera selection finder helps narrow the resulting camera choices. Use the resolution calculator to check the field width and feature-sampling requirement; motion blur still requires the separate calculation below.

At 800 mm/s, an exposure of 40 µs produces estimated object movement of:

Whether that is acceptable depends on edge contrast, feature size and the inspection algorithm. The guide to exposure time versus shutter speed explains the exposure decision in more detail.

Now allocate a separate placement-error budget. If allowable timing-related position error is 0.10 mm at the same speed:

That allowance covers the combined variable timing errors, not a separate allowance for every component. Include sensor repeatability, control scheduling, input response variation and delay quantization. Account separately for mechanical position variation and changing conveyor speed.

For a conservative bound, add bounded contributions. Do not combine them statistically unless you have evidence supporting their distributions and independence.

Check event acceptance during the waiting period

Non-cyclic inspection must be designed around the shortest credible event spacing, not average throughput. Random arrivals can produce two closely spaced parts even when the average production rate is low.

For leading-edge spacing p and constant speed v:

If the minimum pitch is 100 mm at 800 mm/s:

Now suppose testing establishes that a particular configuration accepts no new event throughout its programmed delay, exposure and subsequent recovery. If recovery after exposure takes 22 ms, its unavailable interval would be:

The 125 ms incoming interval is shorter than the 246.120 ms unavailable interval. Some part events will therefore arrive too early. Under this simplified behavior, the theoretical pitch corresponding to readiness is:

This calculation is deliberately conditional. Other cameras or configurations may support pending events, overlapping acquisition or different readiness behavior. Do not assume either rejection or queuing from a product's interface or frame-rate specification.

Moving the delay into a controller can help if that controller stores independent records for all parts traveling toward the camera. A single timer that restarts on every sensor edge does not provide that function. The controller must also deliver properly spaced camera pulses when records become due.

Camera acquisition mode remains a separate check. If capture stops after the first image, follow the acquisition-mode troubleshooting guide rather than adjusting transport delay.

Select camera and cable candidates without assuming features

Begin with area scan cameras, then verify the timing functions in the exact candidate's manufacturer documentation. A suitable resolution and interface do not establish support for delayed triggers, event queues or readiness outputs.

The MV-CU020-90GC is a GigE, global-shutter, C-mount candidate. Its catalog interface and shutter information make it relevant to an initial shortlist, but its trigger behavior still needs separate confirmation.

The MV-CH120-11UC is a USB3, global-shutter, C-mount candidate. Evaluate host connectivity and acquisition requirements independently from the sensor-to-exposure timing design.

For each candidate, request or verify:

• The supported trigger selector, source and activation edge.
• Delay availability, range, increment and timing reference.
• Minimum input pulse width and any configurable input filtering.
• Minimum accepted trigger interval in the intended exposure and readout configuration.
• Behavior when a trigger arrives during a pending delay or unavailable period.
• Whether readiness, exposure-active signals or useful event records are available.
• Whether image size, pixel format or acquisition overlap changes those limits.

Select cables by their actual role. Data connectivity and trigger I/O are separate unless the specific documented arrangement combines them. The accessories finder is useful for narrowing cable types, but connector appearance alone does not establish electrical compatibility.

The LS-CBL-0199 is a high-flex GigE camera cable with screw locks. The LS-CBL-0219 is a high-flex USB3.0 camera cable. Neither description establishes suitability as a camera trigger I/O cable.

For the trigger connection, confirm pin assignments, permitted voltage, input current, signal reference and output type. Follow the camera GPIO wiring checks before connecting an unfamiliar sensor or controller output.

Validate timing with measurements and part identities

Validate the entire chain under representative operating conditions. An image reaching the host proves acquisition and transfer occurred, but does not identify when exposure happened.

1. Establish a geometric reference. Use a target whose detected feature and required image position are unambiguous. Record the actual travel distance and feature offset.
2. Begin with widely separated parts. Verify one accepted event produces the intended image before introducing close spacing.
3. Observe the electrical timing. Using appropriately rated measurement equipment, compare sensor output, camera trigger input and documented exposure-active output when available.
4. Separate response latency from programmed delay. Check which measured intervals include filtering, controller execution and camera latency. Confirm how accurately any exposure-active output represents actual exposure.
5. Measure placement over repeated passes. As an initial screening test, record 100 passes at each intended steady speed. Treat this as a screening sample, not proof of rare-event reliability.
6. Test the shortest allowed spacing and bursts. Count detected parts, issued trigger pulses, camera acceptance events when available, and received images. Preserve part identity through the comparison.
7. Exercise operating transitions. Include speed changes, conveyor stops, acquisition restart and the configured maximum processing load.

Read back settings after startup and after any configuration load. If writes do not take effect in the expected sequence, use the guide to verifying trigger-related startup settings.

Keep host arrival timestamps separate from exposure timestamps. Network delivery, buffering and application scheduling can change arrival order or latency without changing the physical exposure time. For asynchronous processing, match frames and results using explicit identifiers and timestamps.

Know when a fixed delay is the wrong answer

Use encoder-based scheduling when travel speed changes enough to consume the placement tolerance. Store the encoder count at detection, then schedule capture after the required additional displacement. If encoder scaling is 20 counts/mm and required travel is 180 mm, the displacement is 3600 counts.

Account for control and camera response during the final approach. An encoder establishes traveled distance; it does not make the camera's response latency disappear. Also verify that encoder motion represents part motion. Belt slip or a part sliding relative to the belt can invalidate the relationship.

A downstream sensor closer to the imaging position may be simpler when part slip is unavoidable. At very low speeds, stationary-part inspection may make a software trigger entirely adequate after measured latency testing.

Do not replace a part-specific event with an arbitrary free-running acquisition rate unless continuous coverage and part association are explicitly designed. Matching average frame rate to average part rate does not guarantee one usable image per irregularly spaced part.

Inspection triggering is not a personnel-protection function. Keep machine safety functions independent of this acquisition timing design.

Check the design before ordering

Use this final checklist to turn the timing calculation into a verifiable specification:

• Geometry: The sensor event and inspected feature have a defined relationship.
• Speed envelope: Maximum speed, acceleration and stopping behavior are documented.
• Image requirement: Sampling and exposure satisfy feature size and blur limits.
• Placement tolerance: Timing uncertainty and mechanical uncertainty fit the allocated error budget.
• Delay implementation: The camera or controller implements the required reference and range.
• Event capacity: All parts between sensor and camera retain independent identities and scheduled captures.
• Readiness: Minimum event spacing exceeds the verified acceptance limit with justified margin.
• Electrical compatibility: Trigger wiring is checked against the actual input and output documentation.
• Verification: Counts and identities agree from detection through received images, including burst tests.

Start a camera configuration once these requirements are defined. If acceptance during a pending delay remains undocumented, contact the applications team with your distance, speed range, minimum pitch, exposure requirement and preferred interface before selecting the final parts.