How Slide Gate Photo Eyes Work, and Where the Usual Advice Falls Apart

A slide gate photo eye is an infrared beam sensor that holds one line of light across the gate's path and reports to the operator's control board, continuously, whether that line is intact. An emitter sends a modulated infrared beam either to a matching receiver across the opening (through-beam) or to a reflector that returns it to the same housing (retroreflective). Anything opaque crossing the line drops the signal, and the sensor changes state within its rated response time, published by EMX Industries as 250 ms for the IRB-RET2 and under 300 ms for the IRB-MON2. The operator then stops the gate, reverses it, or refuses to move. The beam knows about one line at one height, and nothing about the rest of the driveway.

A photo eye does not detect vehicles

The receiver has no concept of a vehicle. It reports whether light arrived. Everything else — the car, the dog, the kid on a scooter — is inference by whoever placed the beam.

The consequence starts with mounting height. The EMX IRB-MON2 instruction manual tells the installer to pick a mounting location 4 to 25 inches above the ground, a 21-inch band of legitimate choices above finished grade. A beam at the bottom of that band runs under many vehicles, so a car can straddle it between the axles once the bumper has cleared. A beam near the top clears a low trailer tongue. Both pass inspection. What settles which is right for you is the underbody clearance of the vehicle in your driveway, and that takes a tape measure.

LiftMaster's UL 325 and ASTM F2200 site planning checklist states that a slide gate entrapment zone exists wherever, at any point during travel, the gap between the moving gate and a fixed counter-opposing edge or surface is less than 406 mm (16 in), in a location up to 1.8 m (6 ft) above grade. That is a volume six feet tall. Your beam is a line inside it.

Nineteen years at a detector manufacturer's warranty returns desk taught me the pattern at work here. Most of what came back worked on the bench. The failure sat in the space between the device and wherever somebody put it.

Two architectures, and the failure modes are not the same

Both types emit infrared. That is where the similarity ends, and it is the confusion I would most like to kill.

A through-beam pair puts a transmitter on one side and a receiver on the other, a one-way trip for the light. The EMX IRB-MON2 is listed at 5 to 115 ft (35 m); LiftMaster lists 90 ft for its LMTBUL through-beam pair. A retroreflective unit puts emitter and receiver in one housing and sends the light to a passive reflector, a round trip it pays for in range. The EMX IRB-RET2 is listed at 5 ft (1.5 m) to 60 ft (18.3 m); LiftMaster lists 50 ft for its LMRRUL. Roughly half the reach, in exchange for not trenching power to the far post.

| Property | Through-beam photo eye | Retroreflective photo eye | Inductive loop detector | |---|---|---|---| | What it senses | Interruption of a one-way beam | Interruption of a beam returned by a reflector | Change in inductance from metal entering a loop field | | Published range | 5–115 ft, EMX IRB-MON2 | 5–60 ft, EMX IRB-RET2 | Detection height about 70% of the loop's shortest side, EMX LP D-TEK | | Response time | Under 300 ms in NC or 10K monitoring, EMX IRB-MON2 | 250 ms, EMX IRB-RET2 | Set by sensitivity, filter and frequency settings; EMX publishes no figure | | Alignment | Receiver LED plus sensitivity potentiometer | Reflector centered in a detection pattern roughly 2 ft in diameter | No optical alignment; loop resistance 0.5–5 ohms | | Environmental rating | NEMA 4X | Weatherproof housing, gasket and cover screws seated | Sealed switch contacts, anodized aluminum housing | | Characteristic false state | Sunlight or cross-talk into the receiver | A shiny vehicle returning enough signal to look "clear" | A high-bed vehicle passing over undetected |

Alignment tolerance is where retroreflective units earn their reputation. The IRB-RET2 manual describes a detection pattern about 2 feet in diameter at the reflector, and tells the installer to move the reflector up, down, left and right to find its center before fixing it. Two feet sounds generous until a post shifts in a wet winter. Its troubleshooting table lists a failure a through-beam pair cannot produce the same way: signal bouncing off a shiny vehicle back into the sensor, reading as a clear beam while something solid sits in the gap.

Neither type works at close quarters: both EMX models state a minimum detection range of 5 feet, so a pedestrian-width gap falls outside the envelope of the devices most people reach for.

From interruption to reversal, in the operator's own numbers

A monitored photo eye does not simply open a contact. The EMX manuals describe a 300 Hz "heartbeat" on the supply lines when the beam is clear, 0 Hz when it is obstructed, and in three-frequency mode 2 Hz obstructed and 0 Hz for a fault. The board listens for that signal every cycle.

What happens next is set by the standard and the operator's firmware. DASMA's Technical Data Sheet 353, which reprints UL 325's Section 32 and its tables, states that reversing must begin within 2 seconds of sensing an obstruction, and that after the first contact the gate must reverse a minimum of 2 inches. On a second sequential contact, a Type A inherent system stops the gate without reversing, while a Type B1 photo eye or B2 edge stops it and may also reverse.

Manufacturer behavior is more specific than the standard's floor. LiftMaster's SL595 installation manual documents that an obstruction on the close-direction photo eye input drives the gate to the full open position and resets the timer-to-close, while an obstruction on the open-direction input reverses the gate for 4 seconds and then stops.

Set those numbers beside the speed limit. TDS 353 notes that Class I and Class II slide gate operators shall not exceed 1 foot per second while pulling 75 pounds or more. At that speed a 250 ms sensor response is roughly 3 inches of gate travel before the board hears about it, and the 2-second bound allows about 2 feet more before motion turns around. That arithmetic is mine, from two published figures. It is the worst case your installation is allowed to have, not a measurement of your gate.

The monitored circuit, and the jumper that deletes it

Every external entrapment protection device on a modern operator is supervised. TDS 353 puts it plainly: the operator verifies every device at least once during each open and close cycle, and should a device go missing or a fault prevent the sensing of an obstruction, including an open or short in the wiring, the operator falls back to constant-pressure operation or manual movement. Your remote stops working. That is the system behaving correctly.

The same document rules out field modification to bypass, interfere with or defeat the monitoring function by adding, suppressing or changing wires, terminals, switches, jumpers or components. A jumper across a monitored input is not a repair. It converts a gate that told you something into a gate that cannot.

This is the part I know from the returns desk. A smoke alarm at end of life chirps once a minute, and it arrives in a padded envelope with a note calling it defective. It was announcing its own expiry, and nobody decoded the chirp. A monitored photo eye lockout is the same kind of message, read the same wrong way. My brother still sends me photographs of ceilings, and a photograph shows a housing, never an LED state, the only part carrying information. A periodic signal is diagnostic only if something is listening and knows what the period means. The fan above this desk ticks once per revolution, and it took me a week to stop hearing a fault in it.

Where one beam fails: mapping beams to zones

UL 325's Table 32.2, effective 1 August 2018 and reprinted in TDS 353, sets the count for a horizontal slide gate at two independent entrapment protection means in the opening direction and two in the closing direction. Table 32.1 lists the acceptable types for a slide gate as A (inherent), B1 (non-contact sensor), B2 (contact edge) and D (continuous-pressure control), with a note that the same type shall not serve both means. Two photo eyes do not satisfy that table. A photo eye plus the operator's inherent system does, and so does a photo eye plus an edge.

A slide gate presents more than one hazard, and they are in different places:

• The leading edge closing into the receiver post or catch.
• The trailing edge, which pinches against the fence line and the operator as the gate retracts.
• The gap between the gate frame and a fixed object nearest the roadway, which ASTM F2200 section 6.1.4 holds to 2¼ inches at both the fully open and fully closed positions, with an exception for objects more than 16 inches away.
• Reach-through openings in the gate and in the fence it covers when open, which UL 325 section 56.8.4.a.2 and ASTM F2200 section 6.1.2 address by screening or picket spacing that stops a 2¼-inch sphere, up to 6 feet above the ground.

LiftMaster's site planning sheet draws the answer: the slide gate layout shows photo eyes in four locations, plus edge sensors on the leading and trailing edges. Its SL595 manual adds the constraint installers miss most often. Where the gate opening is wider than the maximum separation distance of the photo eyes, edge sensors go in instead.

A loop detector answers a different question

The inductive loop is the device that actually does what people believe a photo eye does.

The EMX LP D-TEK manual describes detection of metallic objects entering the field formed around an induction loop, with loop inductance of 20 to 2000 µH at a Q factor of 5 or better and a healthy loop measuring 0.5 to 5 ohms across the leads. Its coverage rule is explicit: detection height is approximately 70% of the shortest side of the loop, so a 4 ft by 8 ft loop yields roughly 33.6 inches. It also carries an automatic sensitivity boost whose stated purpose is continual detection of high-bed vehicles that would otherwise go undetected.

A loop covers an area and senses only metal. A photo eye covers a line and senses anything opaque. A child standing in a driveway trips neither a loop nor a badly aimed beam, which is why UL 325's Table 32.1 lists inherent systems, non-contact sensors, contact edges and continuous-pressure controls as entrapment protection types, and does not list loops. Loops manage traffic: hold-open, free exit, reversing on a vehicle. An installer who adds a loop and drops an eye has traded a safety device for a convenience.

Fault isolation without a bypass

When a gate refuses to close and the driveway looks empty, work the signal path in order. At the returns desk I logged every indicator state before touching anything, because the paperwork always arrived with the symptom already lost. I do not install gates; I read manuals and failure reports. Every number above comes from a document you can open yourself.

1. Read the operator's diagnostic display, write the code down, then read the photo eye's indicator. On the EMX IRB-MON2 a solid green receiver LED means aligned and flashing means blocked or misaligned. On the IRB-RET2 solid red means best alignment, slow flashing poor, fast flashing closer.
2. Clean the lens and the reflector face. Cobwebs, road film and frost break a beam exactly the way a leg does.
3. Confirm the geometry sits inside the listed envelope: more than 5 feet of separation, less than the published maximum, and a height matching the operator manual's zone drawing.
4. Realign. On a retroreflective unit, move the reflector to the center of the roughly 2-foot pattern. On a through-beam pair, turn receiver sensitivity down until the LED flashes, then back a quarter turn clockwise until it holds solid.
5. Check the monitoring method against the operator manual and set the DIP switches to match. Pulse monitoring on both EMX models runs on 6–35 VDC, and a mismatch shows up as a device the board cannot see.
6. Inspect the wiring and its entry point. Both manuals call for a UL listed watertight fitting, and the NEMA 4X rating holds only while the gasket is seated and cover screws snug.
7. Obstruct the beam at several distances along its length and confirm the board acknowledges each one.
8. If the sensor passes every step, the fault is downstream at the input terminal, the board or the wiring, and you have a documented result, not a guess.

Nothing on that list involves a jumper. The gate that will not close is running the only diagnostic it has, and each zone on a slide gate has its own heartbeat to lose.

Frequently asked questions

How does a slide gate work?

A slide gate rolls horizontally on a track or on cantilever rollers, driven by an operator turning a chain or a gear rack. Limit switches tell it where to stop. ASTM F2200 calls for positive stops at the fully open and closed positions, plus roller covers and guarded pinch points.

How do photo eyes work?

A photo eye emits a modulated infrared beam and monitors whether it reaches a receiver, either straight across the opening or via a reflector. Any opaque object breaking that line changes the sensor's output within its rated response time: 250 ms on the EMX IRB-RET2, under 300 ms on the IRB-MON2.

How do gate sensors work?

Gate sensors fall into distinct families. Photo eyes sense a broken optical line. Edge sensors sense contact along the leading or trailing edge. Inherent systems sense motor current or rotation stalling. Loops sense metal entering a buried coil's field. Only the first three count as entrapment protection under UL 325.

Why does my slide gate reverse when the driveway is empty?

Because something interrupted the beam or the monitoring signal, and neither has to be visible. Sun striking the receiver, a wet or shifted reflector, a spider web, a shiny vehicle bouncing the beam back, or a wiring fault all read as an obstruction. The diagnostic code names the input.

Which entrapment zones does a photo eye protect on a slide gate?

Only the line it is aimed along. UL 325's Table 32.2 sets two independent entrapment protection means for the opening direction and two for the closing direction of a horizontal slide gate, and Table 32.1 rules out using the same type twice. Leading edge, trailing edge and reach-through hazards each need their own device.

What are the most common problems with sliding gates?

Misalignment and contamination of photo eyes, monitored inputs locking the operator into constant-pressure mode, track debris and worn rollers, gaps exceeding the 2¼-inch limit in ASTM F2200 section 6.1.4, and layouts covering one zone while leaving the trailing edge or the fence-line gap unwatched.