Mount a light curtain 300 mm from the hazard and it looks safe. Reach through, trip the curtain, and your hand is still inside the press before the machine stops. The curtain works. The mounting distance doesn’t.
The minimum distance formula
Safety standards define a minimum distance from the light curtain to the hazard: S = K × T + C. K is the hand speed, typically 2000 mm/s. T is the total stop time, including the curtain response, the safety relay, and the machine itself. C is an additional inset factor based on the beam spacing.
On a press with a 50 ms machine stop and a 30 ms curtain response, T is 80 ms. K × T = 160 mm. Add C for finger detection (typically 8 × (beam spacing − 50), capped), and you might land at 200 to 300 mm. That’s the minimum. Anything closer and you’re asking for a reach-in injury.
People plug in the sensor response time and forget the machine stop time. The curtain stops in 20 ms, but the press ram doesn’t. If the ram travels another 100 mm after the stop signal, that distance has to be accounted for. Measure the actual stop time with a tachometer or a safety evaluator; don’t use the datasheet number.
Beam spacing is not just resolution
Beam spacing determines what the curtain detects. A 14 mm spacing detects fingers. A 30 mm spacing detects hands. A 70 mm spacing detects arms or torso. Pick the wrong resolution and you’ve built a curtain that won’t see a finger reaching through.
The temptation is to buy the coarsest (cheapest) curtain that “seems” fine. But if the operator can slip a finger between beams and reach the hazard, the curtain is decorative. For point-of-operation guarding on a press or a robot cell, use finger detection where hands enter, not hand detection.
Mounting height matters
A curtain mounted too high leaves a gap underneath. Mounted too low, the bottom beam sits at floor level and debris trips it constantly. The standard approach is to align the bottom beam 100 to 150 mm above the floor, and size the curtain height so the top beam covers the maximum reach height. Don’t forget to account for the operator standing on a platform or a stool; that shifts the whole reach envelope up.
On tall machines, use two cascaded curtains or a floor-to-ceiling set. A single 900 mm curtain mounted in the middle leaves the operator to reach over the top or under the bottom.
Bypassing is the real failure mode
Operators defeat light curtains. They prop a piece of tape in front of the receiver, leave the curtain muted during setup, or stand off to the side where the beam doesn’t cover them. Safety curtains that get bypassed are worse than no curtain, because management believes the risk is controlled.
Design the cell so bypassing is hard. Use muting sensors only for actual material pass-through, not for setup. Require a held-down enable switch for jog modes. Mount the curtain so the operator can’t stand beside it without breaking the beam. And audit the installation: if the fault log shows the curtain triggered 50 times in a shift and nobody investigated, the curtain is being treated as an annoyance.
Blankings and overrides
Fixed blanking lets you mask a set of beams that the machine structure always occupies. This is legitimate: a conveyor passing through the curtain shouldn’t stop the press every cycle. But accidental blanking disables a zone you thought was guarded. Document every blanking configuration, and use device-aware blanking where the system detects when a masked area changes.
For setup and jogging, use a three-position enable switch (hold-to-run, release to stop, squeeze hard to stop). Don’t use a simple override key that latches the curtain off. Latched overrides get left on.
Bottom line
A light curtain is only as good as its distance calculation and its mounting geometry. Measure the real machine stop time, pick the right beam resolution, and mount it so there’s no reach-around. If operators are bypassing it, the installation is the problem, not the people.
