On a guard locking safety switch the safety outputs are enabled only when the locking bolt is sensed in its extended position, an action that requires the guard to be both closed and locked [S1]. That single sentence is the engineering reason two independent signals appear on every compliant interlock: a door position contact and a lock state contact, both routed into the safety relay or safety PLC [S1][S4].
The split is not a marketing distinction. ISO 14119 makes it explicit: an interlocking device associated with a guard must provide a guard position monitoring function that detects whether the guard is closed, and a separate interlocking function that holds the guard closed until a defined safe condition is met [S7]. A door that is closed but not locked is a different safety state from a door that is closed and locked, and the standard treats them as such [S4][S7].
What Each Output Actually Reports
On a coded-magnet or RFID interlock such as the DAIDISIKE DX-R1, one OSSD pair reports actuator presence at the switch (the door position fact) and a second OSSD pair reports the internal lock element state (the lock fact) [S3]. A safety door lock such as the DX-W2, DX-W3 or DX-W5 adds a solenoid-driven bolt to that pair, with a published 1300 N holding force across the mechanical-contact family [S3]. The compact DX-D6 raises that figure to 2000 N while keeping the same two-channel split, in a 30 × 30 mm housing with dual NPN or PNP outputs [S3].
For non-locking monitoring only, the DX-D2 and DX-D3 mechanical switches publish NC/NO contact blocks with IP67 sealing and a 1,000,000-cycle mechanical life, and they deliberately omit the lock signal because no bolt is present [S3]. The DX-C1 coded-magnet version publishes a 17 mm sensing range for the same door-position-only role [S3]. SICK's portfolio carries the equivalent split: RE1 and RE2 magnetic switches, STR1 RFID-coded switches, and flexLock safety locking devices each treat the two facts as separate signal paths into the safety logic [S5].
Why ISO 14119 Forbids Treating Door Position as Lock Status
ISO 14119 requires systematic monitoring of both the guard position and the lock state; it is not enough to know that the door is closed [S4]. The standard's reason is mechanical: a door can register closed while the locking bolt has failed to extend, has been defeated, or has never been commanded to engage. Wiring only the door contact into the safety relay leaves the lock failure path open, even on a switch that is otherwise well specified [S4][S7].
The same rule applies in reverse. A door position switch is a discrete device that reports open or closed, while a latch or bolt monitor reports a locking element's position; these are two different sensor outputs and they answer two different engineering questions [S9]. On access-controlled industrial doors, adding a door position switch on top of the lock sensor enables a third check (door-forced, door-propped, door-held-open-too-long) that the lock output cannot give on its own [S2]. Conflating them removes diagnostic value and removes one of the two independent channels the safety standard is asking for [S4][S7].
Selection Criteria: When You Need Two Signals vs One
Specify two independent signals (door + lock) on any guard where a stop command cannot be confirmed as effective before a person can reach the hazard. This covers all process-generated hazards with run-down time (rotating inertia, hot surfaces, stored pneumatic or hydraulic energy) and all robot cells where the risk assessment requires the guard to remain closed until a defined safe state is reached [S4][S7]. The 2000 N DX-D6 and 1300 N DX-W2/W3/W5 families are sized for this class, with redundant dual outputs that the safety controller can evaluate channel-by-channel [S3].
Specify a single door-position signal only when the risk assessment confirms the hazard ceases before the guard can be opened, which is the published scope of the non-locking DX-D2/D3 mechanical and DX-C1 magnetic families [S3]. SICK's MLP1 magnetic safety locking device is the boundary case: a high offset tolerance locking device for process protection where nuisance trips are unacceptable, but it still exposes the lock state as a separate output rather than hiding it inside the door contact [S5]. When a non-locking door switch is the only device on the guard, the safety relay must see only the door contact, and the lock output is not present because no bolt exists [S1][S3].
OSSD Wiring, Dual Channels, and the Safety Relay Side
On a guard locking device wired to a safety relay, two OSSD safety signals carry the door position fact and two carry the lock fact, and the relay cross-checks them in pairs [S6]. A common installer question is whether the lock signal can be tied to the same input pair as the door signal to save wiring; the answer in compliant designs is no, because a short across one channel would then mask the other, defeating the redundancy the dual-channel architecture exists to provide [S6]. The relay sees four wires from the lock device, not two, even when the device is mechanically small [S3][S6].
Reach past the switch into the actuator side. An RFID-coded actuator such as the uniquely coded variant in the DX-R1 family rejects a common spare key or magnet, raising the manipulation resistance of the door position fact itself rather than relying on lock force alone [S3][S5]. Mechanical variants such as the DX-W5 use six gold-plated contacts in four published configurations, giving integrators a direct path to four independent signals (door NO, door NC, lock NO, lock NC) without an external safety relay doing the multiplexing [S3].
Comparison: Door-Only vs Door-Plus-Lock Families
Across the DAIDISIKE 9-series the door-only group (DX-C1 magnetic, DX-D2/D3 mechanical, DX-R1 RFID) reports a single safety fact at published ratings of 17 mm sensing distance, IP67, 1,000,000-cycle mechanical life, or dual-channel PLe/Type 4 output depending on the variant [S3]. The door-plus-lock group (DX-W2, DX-W3, DX-W5 solenoid locks and DX-D6 electronic lock) adds a second fact and publishes holding force values of 1300 N for the mechanical-contact trio and 2000 N for the DX-D6 [S3]. Holding force is the spec to track on the lock channel, while sensing distance and cycle life are the specs to track on the door channel [S3].
SICK's portfolio maps the same way: RE1/RE2 magnetic and STR1 RFID handle door position only, while flexLock and the safety locking devices add the second channel for process protection [S5]. For high-manipulation-risk guards, an RFID-coded flexLock is the equivalent upgrade over a magnetic RE-series device, with the same two-signal wiring on the controller side [S5]. The decision criterion is not brand but channel count: how many independent safety facts the device publishes into the relay.
Limits, Failure Modes, and Common Wiring Mistakes
The most common failure on a guard locking switch is mechanical, not electrical: a door that appears closed but whose bolt has not reached the extended position because of door sag, misaligned hinges, or worn actuator keys [S1]. The switch's two independent outputs are designed to catch this, but only if both are wired to the relay; tying the lock input to the door input discards that protection [S1][S6]. A second common failure is using a coded-magnet switch in a high-vibration location where the published sensing range of 17 mm is consumed by door flex, leaving the door channel in an undefined state [S3].
A third failure mode is diagnostic blindness. Without the door position contact, an access control system cannot distinguish a propped door from a forced door from a held-open door, and the SOC loses the alarm set that door position switches normally generate (door open too long, door forced open) [S2]. This is the access-control argument for keeping the door channel even when the lock channel is the safety-critical one, and it is the reason DAIDISIKE's product map keeps the two facts on two contacts rather than collapsing them [S2][S3].
Standards, Sourcing, and What to Verify on the Datasheet
Two standards govern the architecture above: ISO 14119 for the interlocking device and its dual-monitoring requirement, and ISO 13849-1 (referenced through ISO 14119's performance level expectations) for the safety relay side [S4][S7]. On the datasheet, verify four numbers before specifying: holding force in newtons, mechanical life in cycles, sensing distance in millimetres for non-contact variants, and the safety rating (PLe/Type 4, SIL3, or the equivalent category under ISO 13849-1) [S3][S5]. A device that publishes only one of those numbers is signalling which role it is built for, and a device that publishes them all is signalling that both signals can be relied on independently [S3].
Track one regulatory signal: the December 2025 update to the Machinery Safety 101 guidance on ISO 14119, which restates the dual-monitoring requirement in the same terms and is the most current public restatement of the rule [S4]. For a deeper look at how stop categories and safety wiring decisions flow downstream of these interlock choices, the IEC 60204-1 selection rules cover the controller side of the same circuit, while the POM Homopolymer vs Copolymer Rod and CSA Z94.3 Class 1 vs Class 2 comparisons sit in the same spec-by-decision-criterion frame this article uses for interlock selection.
Detailed specification references: switches, displacement position, and lock nut.