OSSD Safety Output Explained for Light Curtains and Machine Safety
OSSD means Output Signal Switching Device. Safety light curtains typically provide two monitored outputs, OSSD1 and OSSD2. A field interruption normally switches them OFF; brief diagnostic pulses can also occur while logically ON. Compatible safety inputs distinguish those tests from a protective demand.
What makes an OSSD different from an ordinary sensor output?
An OSSD is the safety-related interface between the protective device and the machine control system. It is not the optical beam itself, a relay coil or a command to start production. Pilz’s OSSD definition connects the term to electro-sensitive protective equipment (ESPE), such as safety light curtains.
A normal PNP sensor and a semiconductor OSSD may both present an ON voltage near the 24 V supply. The voltage alone does not make them interchangeable. With an OSSD system, the safety-rated device, output diagnostics, channel evaluation and documented fault reaction work together. Two ordinary outputs wired to ordinary PLC inputs do not reproduce that function.
PNP describes the electrical behavior; OSSD describes a safety interface
Many light curtains use PNP, current-sourcing semiconductor OSSDs. That tells you how current reaches the receiving input, not which faults are detected or which safety performance is achieved. A non-safety auxiliary output may follow the same beam state but still be intended only for an indicator or standard PLC status signal. BERNSTEIN’s explanation describes the role of self-monitoring in electronic OSSD outputs.
For dry-contact safety devices, test signals may instead originate in the receiving controller and pass through the contacts. Do not copy that input configuration to a light curtain that generates its own output tests. The approved connection example must match the actual device type.
Integration boundary: this guide explains signals and verification questions. Wiring changes and safety-function validation require qualified personnel, the exact device manuals and the machine’s safety requirements. Do not join OSSD1 and OSSD2 or substitute a status output for the safety channels.
How OSSD1 and OSSD2 behave during operation
The two channels give the safety system separate signal paths. In a typical equivalent-switching pair, both are logically ON when the curtain is ready to permit operation. A protective-field interruption normally takes both OFF within the specified response time. A wiring or internal fault can produce another invalid state or a lockout.
“ON” in this description is a logical operating state, not a promise of uninterrupted DC voltage. The receiving safety system must distinguish permitted diagnostic pulses from an OFF demand and react to an invalid channel combination.
| Observed condition | Typical channel behavior | What it means |
|---|---|---|
| Field clear, device ready | Both logically ON; brief test pulses may occur. | The curtain permits operation. Other interlocks, reset logic or machine controls may still prevent starting. |
| Field interrupted | Both normally switch OFF. | A protective demand. The downstream safety function must produce the defined safe reaction. |
| One short LOW event | One channel may briefly differ from the other. | Possibly a documented test pulse. Duration and pattern matter; do not diagnose from one sampled bit. |
| Sustained disagreement | One channel stays ON while the other stays OFF. | An invalid enable condition, not permission to run on the remaining channel. |
| Clear field, outputs OFF | Both remain OFF despite no visible obstruction. | Check alignment, interlock, feedback, configuration, supply and fault codes. Clear beams alone are not a complete diagnosis. |
The exact voltage thresholds, transition timing and recovery sequence belong to the selected model and configuration. The table is a functional explanation, not a universal pinout or fault-code list.
Why a healthy OSSD can briefly pulse LOW
A semiconductor output can be tested by briefly commanding it OFF and checking the result. This helps the device detect relevant output-stage and wiring faults. If an output is held HIGH by a short to supply, for example, the expected LOW transition may not occur.
The downstream safety input must tolerate the curtain’s documented test pattern without overlooking a real protective demand. Pulse width, repetition and the relationship between the two channels are product-specific. A valid test on one model is not a universal OSSD timing specification.
Read the maximum pulse width, not only the typical value
Documented model example: the SICK deTec4 manual, revision 8021645/1SWC dated February 4, 2026, lists a test-pulse width of 130 µs minimum, 150 µs typical and 300 µs maximum in Table 88. These are deTec4 receiver data, not ratings for every curtain or for an xsz sensor model. See the manufacturer’s table and associated conditions.
Illustrative compatibility check: assume a candidate safety input is documented to tolerate source test pulses only up to 200 µs. Comparing it with the 150 µs typical figure looks encouraging, but the 300 µs maximum exceeds that stated tolerance. The evidence does not yet support this pairing. Find a documented compatible configuration or input; do not assume that an arbitrary 300 µs filter makes the complete interface compatible.
Why a multimeter can show 24 V while a PLC logs OFF events
A steady meter display can miss short pulses because of its sampling and display behavior. Conversely, a sufficiently fast input may record the same pulses as transitions. Neither observation alone proves a bad curtain. Start with the light-curtain and safety-input diagnostics. Waveform capture, when necessary, requires qualified personnel, suitable instruments and a safe measurement procedure; the result must be compared with both manuals.
Can OSSD outputs connect directly to a safety PLC?
Yes, when the exact safety input and its configuration support the curtain’s OSSD interface. A separate safety relay is not always necessary. It can be useful for a straightforward, fixed-function circuit; a safety PLC or safety I/O may suit several zones or more complex logic. Neither choice avoids checking the actual connection.
Keep each OSSD on its own conductor and receiving input. Channel evaluation may reside in the input module or an approved safety function block, depending on the system. Do not assume that one universal “dual-channel” input setting applies to every controller.
Do not confuse sensor-generated tests with contact-input tests
Rockwell’s Light Curtain instruction documentation explains pulse handling and provides a specific input configuration. It also warns that added filtering increases safety reaction time. Use it as an example of why hardware settings and logic matter, not as a configuration recipe for another controller.
| Check | Evidence to compare | Why it changes the decision |
|---|---|---|
| Output and input type | Output polarity, specified supply/reference and the safety input’s supported semiconductor-device mode. | A matching 24 V label does not establish electrical or diagnostic compatibility. |
| HIGH and LOW recognition | Guaranteed output voltages, OFF leakage, input thresholds and input current. | The input must recognize both states under the stated supply and load conditions. |
| Pulse handling | Maximum pulse width, pattern, load dependency and documented input tolerance. | A normal self-test must not become a nuisance trip or justify masking a real OFF demand. |
| Cable and load | Pin assignment, cable limits, input capacitance and permitted output loading. | A matching connector or current rating alone does not prove the signal will behave correctly. |
| Evaluation and timing | Input mode, channel-discrepancy handling, safety logic, configured delays and response data. | The installed hardware and program must reject invalid states within the required time. |
What about a standard PLC or direct contactor connection?
A standard PLC can receive an approved non-safety status signal, but ordinary inputs and software must not replace the protective safety function. Prefer the device’s documented auxiliary or diagnostic interface for production status; do not add a parallel load to an OSSD without verifying that it cannot impair the safety path.
Directly driving final devices is a separate design question. It is acceptable only where the manufacturer permits that architecture and the complete safety function supports it. Check load behavior, suppression, feedback and required redundancy—not just whether a coil draws less than the OSSD’s rated current. The wiring diagram must come from the selected devices and engineered circuit.
Why a clear light curtain may still block restart
Field clearance, reset and downstream feedback answer different questions. Also check where each function is implemented: a restart interlock inside the curtain can keep its OSSDs OFF, while an interlock in the safety controller can keep the machine disabled even when the curtain’s OSSDs are ON.
Reset restores readiness; it is not the production-start command
Where manual reset is required, clearing the beams does not complete the restart sequence. Reset acknowledges that the conditions for rearming have been checked; the machine-start command remains separate. Banner’s EZ-SCREEN Point manual illustrates this distinction for its access-guarding system.
This matters when someone can walk through a perimeter curtain and remain inside the cell. The beams can become clear behind that person. Reset position, visibility of the hazard zone and any additional presence protection must address that possibility. A time delay alone does not establish that the area is empty.
EDM checks final-device feedback, not the optical field
External Device Monitoring (EDM) checks feedback from downstream switching devices, not physical stopping time. Suitable mechanically linked or mirror-contact arrangements must match the approved circuit. The deTec4 EDM example shows how a missing required contactor state prevents re-enabling.
Do not equate an EDM fault with a proven welded contactor. Incorrect feedback wiring, configuration or timing can also prevent the expected return. Use the code, circuit and recorded feedback behavior to distinguish them. Repeated resets or bridging the feedback loop do not correct the cause.
How to narrow down an OSSD fault without guessing
First preserve the light-curtain code, safety-input diagnostics and event timing, where these can be read safely. A reset or power cycle may clear useful evidence. Before changing wiring or inspecting hazardous areas, prevent unintended motion and follow the machine’s isolation procedure.
Separate three observations: what the curtain reports, what the receiving input sees, and what the downstream safety logic permits. A complaint that “the light curtain will not start” can originate in any of these layers.
| Symptom | First useful distinction | Next evidence to collect |
|---|---|---|
| Both channels OFF | Does the receiver report beam blockage, interlock or fault? | Alignment/beam indicators, exact code, supply status and configured restart conditions. |
| Channel disagreement | Brief documented pulse or sustained invalid state? | Time-stamped diagnostics; then qualified inspection of separate conductors, connector assignment and load. |
| Intermittent input events | Does the safety function trip, or only the status logger? | Pulse specifications, input configuration and matching timestamps. If events track motor/contactor switching, also investigate supply and electrical disturbance. |
| EDM blocks reset | Which required feedback state is missing or late? | Feedback circuit, expected contact behavior, configuration and allowed timing for the selected system. |
Example: a logger reports flicker, but the safety input stays enabled
Illustrative scenario—not a recorded customer test. Assume an installation has a manufacturer-permitted, non-interfering diagnostic connection. Its fast logger records short LOW events while the receiver reports a clear field and the OSSD-compatible safety input remains enabled.
The first hypothesis is that the logger sees normal test pulses. It is not yet a conclusion: compare event duration and channel pattern with the curtain’s specification. If they agree and the safety interface is documented, review how production diagnostics should represent those pulses. If the events are longer, irregular or accompanied by safety-input faults, continue investigating the supply, wiring, load and input configuration.
Transferable lesson: establish which layer is reporting the problem before changing the sensor. “The meter reads 24 V” and “the PLC logged LOW” can both be true. Neither justifies adding an arbitrary delay to the safety channel.
What a working OSSD interface still does not prove
Seeing the two channels switch correctly is an interface check, not acceptance of the complete machine safety function. The light curtain, safety logic, final switching elements, machine mechanics and safeguarding geometry must work together.
Output response time is not the complete stopping time
The timing assessment includes the selected curtain’s response, the configured safety-input and logic path, relevant communication and output delays, and the machine’s stopping behavior. Identify the start and end points of any measured time so that a delay is neither omitted nor counted twice. Changes to filtering, cascade, logic or final devices can change the result.
ISO 13855:2024 addresses safeguard positioning with respect to human approach. The installation must also account for detection capability and ways to reach over, under or around the field. Use the dedicated light-curtain installation-distance guide for that separate decision; a pulse-width comparison cannot establish a safe mounting distance.
Two OSSD wires do not automatically establish Type 4 or PL e
IEC 61496-1:2020 covers general ESPE requirements; IEC 61496-2:2020 adds requirements for AOPD technology. The declared Type must be supported for the actual product. The achieved Performance Level or SIL of the complete safety function is a different assessment.
ISO 13849-1:2023 provides the methodology for safety-related control-system design and integration; it does not assign one required PL to every machine. For the device-level distinction, see Type 2 versus Type 4 safety light curtains.
Finally, a protective stop is not hazardous-energy isolation. Servicing may require isolation and control of stored energy even when both OSSDs are OFF. The distinction is reflected in the U.S. OSHA hazardous-energy control requirements; use the procedures applicable to the actual machine and workplace.
The practical takeaway: read OSSD as a monitored safety interface—not simply “two 24 V outputs.” Match the pulse behavior and electrical limits to the receiving input, establish who owns reset and feedback, and validate the complete protective function.
When discussing an xsz sensor light-curtain option, identify the required safety function and exact receiving safety-input model, then request the proposed curtain’s OSSD specifications and covered configuration. Those details make an interface review meaningful.