Your camera streams all day on software trigger. You wire a pulse generator to the multi-pin I/O connector, switch the acquisition tool to external trigger, press grab — and the call times out, with the tool sometimes flipping triggering back to off. Most of the time the camera and the software are both healthy: the pulse is landing on the wrong pin, or its return is on the wrong ground. This guide walks the fault in the order you meet it: which pin is the input, why the power ground does not work as its return, what the pulse must look like, how long the grab timeout must be, and a bench test that proves the wiring with a multimeter in ten minutes.
Why does the grab time out the moment I switch to external trigger?
Because the camera never sees the pulse: the timeout fires while it waits on an undriven input. The fix on the 12-pin connector below is GP Input 1 on pin 4 with Input GND on pin 7 — pin 1 is power ground and does not close the trigger circuit.
Software trigger working is the strongest clue you will get. It proves the interface link, the exposure settings and the grab path are all good; the only thing that changed is where the camera expects its start command. A hardware-triggered camera exposes only when a valid edge arrives on its input, and the grab call aborts when its deadline passes. When the tool then reopens the camera with triggering disabled, that is the recovery path after the abort, not a second fault.
Work the causes in this order:
| Check order | Cause | Fast test | Fix |
|---|---|---|---|
| 1 | Pulse return on power ground instead of Input GND | Meter between pin 4 and pin 7: no pulse swing | Move the return lead to Input GND |
| 2 | Pulse on an output pin or the wrong input | Ohm out each flying lead to its pin | Land on the GP input |
| 3 | Grab timeout shorter than the pulse interval | Raise the timeout to 10 s, trigger once by hand | Set timeout at or above 3 × interval |
| 4 | Activation edge mismatched to the generator's idle level | Meter the idle voltage | Set the matching edge |
| 5 | Opto-isolated input under-driven | Check the input type in the pin table | Drive it with the current its input stage asks for |
Which pin is the trigger input on a 12-pin I/O connector?
The pin named GP Input — GP Input 1 on pin 4 in the assignment below — and its return is Input GND on pin 7. This is a real assignment of the kind printed in a camera hardware guide:
| Pin | Name | What it is |
|---|---|---|
| 1 | External Ground | Return for the external power input |
| 2 | External Power (+8 ... +36 V DC) | Camera supply when the camera is not bus-powered |
| 3 | GP Output 4 | Free user output |
| 4 | GP Input 1 | Hardware trigger input — the pin you want |
| 5 | GP Output 3 (Default: Busy) | Readout-in-progress status |
| 6 | GP Output 1 (Default: IntEna) | Exposure-in-progress status |
| 7 | Input GND | Return for the GP inputs — the ground you want |
| 8 | RxD RS232 | Serial receive |
| 9 | TxD RS232 | Serial transmit |
| 10 | Output Power (max. 36 V DC) | Auxiliary supply for your trigger circuit |
| 11 | GP Input 2 | Second user input |
| 12 | GP Output 2 (Default: Follow GP Input 2) | Mirrors input 2 — a free test point |
Three patterns in that table save you an afternoon. Inputs, outputs, serial lines and power all share one shell, so a one-pin error lands you on an output, an RS-232 line or a power pin. The grounds are separate returns: External Ground serves the power input, Input GND serves the GP inputs. And the mirror output means whatever the camera registers on input 2 it drives back out on pin 12, which is a status point you can watch with a meter.
Pin numbers move between camera families and connector generations; the names do not. Read your camera's own I/O table once, anchor on the words GP Input, Trigger Input or Opto-in, and take the ground from that same table.
Ready-made leads from the I/O and power cables range cover both common connector families:
- Type
- Power and I/O cables
- Length
- 6-8 m
The LS-CBL-0211 is a 5 m, 12-pin I/O and power cable. Every pin of the connector arrives on its own conductor at your terminal block, so the input and its ground land exactly where you label them. The LS-CBL-0300 is an M8 8-pin power and I/O cable with an open end, 7 m, rated for drag-chain use, for the round M8 I/O connectors on current camera bodies.
Why doesn't the camera's power ground work as the trigger return?
Because Input GND and External Ground are separate nodes inside the camera: the trigger threshold is measured between pin 4 and pin 7, so a pulse referenced to pin 1 never crosses it. The input conditioning circuit watches its own pin pair and nothing else.
External Ground exists to return the supply on pin 2, and it belongs to the power input stage. Input GND is the signal return the GP inputs are measured from. Wiring your generator between pin 4 and pin 1 drives the input against a return it is not referenced to, so no edge registers — and the grab times out exactly as if no pulse existed.
The meter makes this visible in one reading. With the generator running at 1 Hz, measure DC between pin 4 and pin 7 at the camera-end connector: the level kicks from 0 V to 5 V once per second, and a min/max capture shows it cleanly. Move only the black lead to pin 1 and the swing disappears or distorts. That missing swing is the whole fault you are chasing.
If the pulse source and the camera run from different supplies, the same symptoms follow the wiring rules in missing trigger responses caused by a shared reference.
Is a 0-5 V TTL pulse the right drive for the input?
Yes: a 0 V low and 5 V high swing is the standard drive for a logic-level GP input, and in the classic version of this fault the level was never the problem — the pin was. Set the generator to idle at 0 V and pulse to 5 V, then match the camera's activation setting to that idle state: idle low pairs with active high and a rising edge, while a generator that idles at 5 V needs the falling-edge setting.
Two input families behave differently at the connector. Direct logic inputs take the TTL swing as it comes. Opto-isolated inputs — usually named Opto-in or trigger input with optocoupler in the pin table — need LED current, so drive them through the series resistor the camera's I/O chapter specifies; a bare 5 V logic output can be too weak, and the symptom is the same silent timeout.
One generator output is meant to drive one input. To run several cameras off one pulse, buffer it first, as covered in fanning one TTL pulse out to multiple cameras.
If no generator is on the bench yet, a frame grabber can emit the pulse train itself, which also gives you a source whose idle level you control.
How long must the grab timeout be?
At least three times the longest interval between pulses — with a 2 Hz generator that is 3 × 500 ms = 1.5 s, so set 2 s. The factor of three covers the worst case where the grab arms just after a pulse has gone past.
Worked example, with the arithmetic. The generator runs at 2 Hz, so one pulse arrives every 1 / 2 Hz = 0.5 s = 500 ms. The rule gives 3 × 500 ms = 1500 ms, and you round that to a 2000 ms setting in the tool. At that rate you collect 2 frames per second no matter how fast the camera runs on software trigger — the pulse rate is the frame-rate ceiling, as frame rate with external trigger pulses works through in detail.
A hand-fed station is the one case that needs more, because parts arrive when a person presents them and gaps run to seconds. Give that grab a 10 s timeout and treat a timeout as the part-missed fault rather than a wiring fault. When a grab does have to be abandoned, interrupting the wait cleanly keeps the acquisition loop reusable instead of wedged.
How do I prove the wiring at the bench in ten minutes?
With a multimeter, the generator and a label maker, in this order. Power everything down first and unplug the I/O connector before any continuity measurement.
- Ohm out the cable with it off the camera: confirm which flying lead lands on which pin, and label the two you will use. Never probe continuity on a live circuit.
- Wire the generator signal to pin 4 (GP Input 1) and the generator ground to pin 7 (Input GND). Clamp the lead at the column of one of the camera stands within about 100 mm of the connector, so cable flex never works on the contacts.
- Set the generator: 1 Hz, 10 ms pulse width, 0 V low, 5 V high, output enabled, idle low.
- Meter across pin 4 and pin 7 at the camera end: steady near 0 V, with a min/max capture showing a 5 V hit once per second. No hit means the fault is upstream of the camera — source, cable or pin map.
- In the tool: trigger mode 0, activation high, exposure 10 ms, timeout 10 s. Start the grab; the frame counter climbs at one per second.
- Recreate the fault once on purpose: move the return lead to pin 1, restart the grab, and watch it time out again. That is the failure mode reproduced, understood and closed.
When is the wiring right and it still times out?
When the pulse reaches the correct pin pair and the grab still fails, work four things: the activation edge, the trigger mode, the tool's recovery behaviour and the input type.
- Activation edge: set it from the generator's measured idle level, not from memory. Idle 0 V goes with active high; idle 5 V goes with active low.
- Trigger mode: DCAM-era and most current cameras default to mode 0, one exposure per edge. Mode 1 ties exposure to the pulse level, so the pulse width must span the exposure you want — why trigger pulse width sets your exposure time shows that arithmetic. Mode 15 is a vendor-defined multi-trigger mode; mode 0 already does one frame per pulse.
- Tool recovery: after a timeout many acquisition tools reopen the camera with triggering disabled, so the mode you set appears to revert. Cure the timeout and the setting holds; the timeout mechanics are the same as in software-trigger one-shot timeouts.
- Input type: an opto input driven by a bare logic output behaves exactly like a dead one. Drive it with the current its input stage is built for.
When the station moves on to a current camera, the same pin discipline carries over to its round I/O connector:
- Sensor format
- 1/1.7"
- Lens mount
- C-mount
- Interface
- GigE
- Resolution
- Up to 3 MP
The MV-CU020-90GC is a 2.5 MP GigE area scan camera with a global shutter and a C-mount. On current GigE and USB3 bodies the I/O lives on M8 or M12 connectors, and the rule is unchanged: pulse on the input the pin table names, return on the ground it names. Browse GigE area scan cameras for the sensor and interface your station needs.
What should I check before calling the camera faulty?
Run this list top to bottom; any single failed line explains a timeout on its own:
- Pulse lands on the GP input pin — pin 4, GP Input 1, in the assignment above.
- Return lands on Input GND, pin 7 — not on External Ground, pin 1.
- A meter between those two pins shows the full 0 V to 5 V swing at the camera end.
- Activation edge matches the generator's idle level; idle 0 V means active high.
- Trigger mode 0 unless the pulse width must set the exposure.
- Grab timeout at least three times the pulse interval — 2000 ms at 2 Hz.
- I/O cable clamped, connector shell tight, leads labelled at the terminal block.
Every line that passes narrows the fault to the camera itself, which is where a warranty conversation starts from evidence instead of a guess. For the parts side, the accessories finder lists leads for both connector families, and the configurator below assembles the trigger chain:
The configurator matches a camera, lens, light, light controller and cables to each other, step by step, and lists only parts that fit what you already picked.
Bring your camera's pin table and we will check the lead assignment with you through the contact page.
Questions
Can I bridge Input GND to the power ground pin and save a wire?
No — run the separate return wire. The two grounds are different nodes: pin 1 returns the external supply, pin 7 returns the trigger inputs. Bridging them ties your signal return to the supply return, so supply current and its noise flow through the very path your trigger threshold is measured across. One extra conductor costs less than chasing intermittent edges for a day.
How do I find the pin assignment for my exact camera?
Open the camera's hardware installation guide or I/O chapter and read the pin table. Anchor on the names, not the numbers: the row you want is called GP Input, Trigger Input or Opto-in, and its return is Input GND or I/O GND. If the table names an optocoupler, note the series-resistor value it asks for before you choose a pulse source.
Why does my acquisition tool switch triggering back off after the error?
Because the grab timed out and aborted, and the tool then reopens the camera with its default configuration — triggering disabled — so the next grab succeeds. It looks like the setting will not stick; it is really the timeout repeating. Cure the wiring and the timeout, and the trigger mode you set survives. If it still reverts with pulses present, the tool is dropping the camera handle after every error.
What timeout should a manually loaded station use?
Ten seconds. A person presenting parts makes gaps of several seconds, and the rule of three times the longest gap still governs: 3 × 3 s = 9 s rounds up to 10 s. Keep the timeout rather than removing it — it is what turns a missing part into a clean fault report instead of a hung grab loop.
Can I drive two cameras from one pulse generator output?
Not directly. Two input loads on one logic output double the drive current and round the edge until neither camera registers it reliably. Put a buffer stage in — one driver channel per camera — and bring each camera its own return wire on the input ground. The same fan-out discipline applies whether the source is a generator, a PLC output or a microcontroller pin.