A relay alarm output is a voltage-free dry contact inside the switch that opens or closes when something goes wrong, so an external siren, PLC or SCADA system can react even when the network itself is down. Rayin (Shenzhen Rayin Technology) is a manufacturer of GPON OLTs and industrial Ethernet switches. It is the last line of reporting that still works when SNMP traps, syslog and the web interface have all gone quiet with the network they were supposed to describe.

"Dry contact" means the switch supplies a switchable pair of terminals and no voltage. The detection current has to come from the PLC or SCADA side — that single fact causes most failed installations.
Normally-closed wiring is the fail-safe choice: a cut cable, a lost power supply and a genuine fault all produce the same open circuit, so the alarm cannot be silently defeated.
Contact ratings are small. Industry-typical figures are around 1 A at 24 VDC or 0.5 A at 125 VAC into a resistive load — enough to drive a relay coil or a digital input, not enough to drive a siren directly.
Inside the switch is a small electromechanical or solid-state relay whose contacts are galvanically isolated from the switch electronics. The switch decides when to change the contact state; it does not put any voltage on the terminals. That is what "dry" means, and it is why the same output can be wired into a 24 VDC PLC input, a 48 VDC telecom rack or a 110 VAC panel without the switch caring.
What gets reported varies by vendor, but the common triggers are: loss of one of the redundant power inputs, a configured port going link-down, a ring protection protocol switching to its backup path, and an internal temperature threshold being crossed. On managed industrial designs these are usually selectable, so the alarm means something specific rather than "something somewhere".
The isolation matters. Because the contact is separate from the data plane, a fault that takes the switch's CPU or uplink down can still be reported — provided the relay is energised from the power input side, which is the usual arrangement.
This is the decision that decides whether the alarm actually survives a bad day.
| Wiring | Contact state when healthy | Behaviour on fault | Behaviour if the cable is cut |
|---|---|---|---|
| Normally open (NO) | Open | Closes, energising the alarm circuit | Stays open — fault never reported |
| Normally closed (NC) | Closed | Opens, de-energising the circuit | Opens — alarm fires |
| NC with supervision | Closed | Opens | Opens, and the monitoring input sees line break too |
With NO wiring, a severed alarm cable looks exactly like a healthy switch. The site stays silent until someone physically checks. With NC wiring, the loop is broken and the SCADA input drops out, which is the correct response to both a real fault and a damaged cable.
The pattern is simple: one terminal of the relay to the PLC digital input, the other to the common of the input bank, with the input's own supply providing the wetting voltage. The switch contributes only the closure.
| Alarm source | Typical trigger | Where it belongs |
|---|---|---|
| Dual power input | One feed lost | Critical alarm, siren or SMS gateway |
| Port link-down | Configured port drops | Warning, logged to SCADA |
| Ring / redundancy switch | Protection path activated | Warning — service survives, topology does not |
| Temperature threshold | Internal limit exceeded | Maintenance ticket |
The recurring mistakes are consistent across sites:
Expecting the switch to supply voltage. It will not. Size the PLC input supply yourself.
Driving a load beyond the contact rating. A siren or beacon usually needs an intermediate relay; the switch contact drives the relay coil.
Mixing alarm and power wiring in one conduit. Long parallel runs next to motor cables induce enough noise to look like contact chatter.
Never testing the loop. An output that has never been exercised is an untested assumption. Trip a power feed during commissioning and confirm the SCADA tag moves.

Does the switch supply voltage on the alarm terminals? No. The relay presents an isolated pair with no source. Wetting voltage comes from the PLC or SCADA input, so connect the input's own supply to one side of the loop.
Why is normally-closed wiring recommended? Because a broken cable then produces the same open circuit as a real fault. With normally-open wiring, a cut cable is indistinguishable from a healthy switch and the alarm is silently lost.
Can the relay drive a siren directly? Usually not. Industry-typical contact ratings are around 1 A at 24 VDC or 0.5 A at 125 VAC resistive. Use the contact to drive an interposing relay coil and let that switch the load.
What events should trigger the alarm? On managed industrial switches the usual set is loss of a redundant power feed, link-down on a configured port, ring protection switching, and an internal temperature threshold. Select them rather than accepting defaults.
How often should the alarm loop be tested? At commissioning, and then at every maintenance window. Trip one power feed and confirm the SCADA tag changes state — an untested relay output is an assumption, not a safety function.
Why every managed industrial switch still needs an RS232 port
Switch ports explained: RJ45, SFP, PoE and console in one pass
Deployment checklist: from console access to configuration backup
Bypass function on industrial switches: principles and use cases
About the author: Sara is a Customer Manager at Rayin with over 10 years of experience in the communications field. In her free time, she enjoys badminton and swimming.
About Rayin: Shenzhen Rayin Technology Co., Ltd. — company profile