Smart camera strobe output: enable it and wire the light

A smart camera that fires its own flash solves two problems in one exposure: motion stops because the light is on for microseconds, and the room's lighting stops mattering because the pulse swamps it. The missing piece on many benches is not theory but one setting and a cable. This guide walks through enabling the strobe output on a smart camera with its supported configuration tools, wiring that pulse into a light controller, and choosing a pulse width that matches the exposure. A worked timing example shows the arithmetic, and the bench checks prove the flash lands inside the frame before the camera goes on the line.

What the strobe output does

Most smart cameras and vision sensors expose a general-purpose I/O line that can be switched from input to output and driven by an internal signal. Set the source to exposure-active and the line rises with the exposure and falls when it ends, on every frame, whatever started the frame. The camera does not switch the lamp; it timestamps the exposure with a voltage, and the controller does the switching.

That timestamp earns its wire. A continuously lit lamp must stay dim enough to work beside, so the camera runs long exposures and the room's lights print themselves into every frame. A strobed lamp burns bright only while the sensor integrates: a few hundred microseconds of intense light, motion frozen, ambient contribution close to nothing. The lamp's average power falls by the duty factor:

At the pulse widths in this guide duty stays below one percent, which is also the regime where a strobe-rated lamp may be overdriven for extra peak brightness. The overdrive limits and their arithmetic are in strobe machine vision light: overdrive, duty cycle and timing.

What you need

• A smart camera with an output-capable I/O line. If the I/O table lists inputs only, Step 1 will confirm it in a minute.
• A light controller with an external trigger or strobe input, one channel per lamp, and a channel type that matches each lamp: constant-current for current-rated lights, 24 V constant-voltage otherwise.
• The camera's I/O cable or pigtail, plus shielded cable from the camera output to the controller's trigger input.
• The lamps themselves, and a 24 V supply sized for the controller's total output.
• For the bench check: a part you can move at line speed, and ideally an oscilloscope.

Step 1: enable the strobe output in the camera

1. Open the camera's supported configuration interface: its built-in web page, its configuration utility, or a GenICam browser on the network or USB connection.
2. Open the digital I/O section and select the general-purpose line you plan to use.
3. Switch the line from input to output and set its source to the exposure-active signal. This is the same line-settings page that governs trigger inputs, and hardware trigger source not detected: check GenICam line settings covers its usual failure modes.
4. Where the camera exposes a strobe width and delay, enter the width from the worked example below and leave the delay at zero to start.
5. Save the configuration to a user set and make that set the power-on default. A setting left only in the live session disappears at the next power cycle.
6. Software-trigger a frame and watch the line on the camera's I/O status page or the controller's input LED: one pulse per frame is what you want.

These settings apply across the smart camera families. A global-shutter color model with an 8 mm M12-mount lens covers most single-station freeze-motion checks:

Step 2: wire the controller and the light

1. Put the controller channel into external trigger or strobe mode, not continuous.
2. Choose the pulse mode. Follow-input repeats whatever width the camera sends and is the case the worked example assumes; a fixed-width mode lets the controller time the pulse itself from a short trigger edge.
3. Match polarity between the camera output and the controller input, sinking or sourcing, per the controller's printed pin-out. Wrong polarity shows up as a lamp that never fires, or never goes dark.
4. Run shielded cable from the camera's strobe line to the controller's trigger input and join the commons as the pin-out shows.
5. Set the channel output for the lamp: the rated current on a constant-current channel, or voltage and wattage headroom on a 24 V constant-voltage channel.
6. Connect the lamp to the channel output; the connector and current notes in light controller to light cable adapters: connectors and current prevent the usual mating mistakes.

For a bench with one or two 24 V lamps, a four-channel digital box with a strobe feature is the workhorse choice:

Set the pulse width: a worked example

The exposure comes from the blur budget, the width from the exposure plus the lamp's rise time, and the duty from the width times the frame rate:

Take a belt at 300 mm/s, an image scale of 0.05 mm per pixel, one pixel of blur allowed, LED rise and settle of 30–50 µs (the microsecond rise-time behaviour of LED strobes is why the margin exists), a camera running at up to 60 fps, production presenting 30 parts per minute, and a lamp drawing 0.5 A at 24 V:

• Exposure limit: 0.05 mm ÷ 300 mm/s = 0.00017 s ≈ 0.17 ms. Choose 0.1 ms (100 µs). Actual blur = 300 mm/s × 0.0001 s = 0.03 mm, or 0.6 pixel.
• Pulse width: 100 µs + 50 µs margin = 150 µs, delay at 0.
• Duty at the 60 fps ceiling: 150 µs × 60 = 9 ms of light per second = 0.9%.
• Duty at the production rate: 150 µs × 30 = 4.5 ms of light per 60 s = 0.0075%.
• Power: peak = 24 V × 0.5 A = 12 W. Average at 0.9% duty ≈ 12 W × 0.009 ≈ 0.11 W. A 200 W four-channel box shares out to about 50 W per channel, so one 12 W lamp sits far inside the budget.

If the first rows of a global-shutter frame look dimmer than the rest, add 20–30 µs of delay so the flash peak lands mid-exposure, and stretch the width by the same amount so the light still covers the whole exposure.

A rolling-shutter sensor changes the rule: the pulse must cover the whole scan, not one exposure. At 6 MP full frame the scan runs several milliseconds or more, which pushes the width, and the duty with it, up by two orders of magnitude. Crop the region of interest to bring the scan time back down before you recompute. Rolling-shutter smart cameras strobe well once the width is set for the scan:

Verify the flash before production

1. Turn automatic exposure and gain off first, or every reading that follows moves on its own; the drift mechanism is in camera image brightness drifts over time: turn every auto mode off.
2. Trigger frames with the room lights on, then off, strobing both times. Mean brightness should barely move; if it moves, the pulse or the current is too small to dominate the integration.
3. Watch for black or half-lit frames at speed; the timing fixes are in strobe light on but camera image is black: fix the timing.
4. Run the part at line speed and check edge sharpness at the exposure you chose.
5. With a scope across the controller's trigger input, confirm one pulse per frame, the width you set, and no noise edges between frames.

Common mistakes

• Controller left in continuous mode: the lamp ignores the camera pulse and the strobed look never appears.
• Pulse shorter than the exposure: a global shutter loses light on every row, and a rolling shutter paints bright bands down the frame.
• Polarity mismatch: the lamp sits at full output, or never lights.
• Settings not saved to the power-on user set: perfect on the bench, gone after the morning power-up.
• Overdrive without a duty calculation: the larger the overdrive factor, the smaller the fraction of time the lamp may run. Compute the duty first, then take the overdrive factor from the lamp's strobe rating.

When the camera output is not the trigger

Some benches should not time the lamp from the camera at all. If the flash must fire before the camera knows a part is there, a photoelectric sensor can trigger the light while the PLC stages the frame, and trigger strobe lighting from robot IO digital outputs covers the robot and PLC side. Line scan is its own regime, where the controller follows an encoder rather than a camera pulse; lighting controllers for line scan cameras: what they fix explains that case. If the pulse source is a 5 V instrument or data-acquisition line rather than a camera, level matching decides whether the input ever sees it; triggering a camera with a 5 V output walks through that wiring.

Before you order

• Camera: an output-capable I/O line on the exact model, not just somewhere in the series.
• Controller: one channel per lamp, channel type matched to each lamp, and an external trigger or strobe mode in the feature list.
• Trigger input: a voltage range that covers the camera's output level and a polarity you can wire either way.
• Cables: the camera's I/O pigtail and a controller-to-light adapter with the right connector and current rating.
• Timing: your exposure, width and duty worked through like the example above, before anything is ordered.

Size the whole chain — camera lines, controller channels, lamp current — in the configurator:

If the camera's I/O options and a controller's trigger input do not obviously meet, send both model numbers through the contact page and we will check the wiring match with you.