What actually happens the instant a single wire fault occurs — and why one wiring topology loses devices while the other doesn't.
Every NFPA 72 initiating device circuit, signaling line circuit, or notification appliance circuit is wired one of two ways. Class B runs out to each device and back to the panel along the same physical path — a simple "home run." Class A runs out to the last device and then keeps going, via a second, separately-routed conductor, all the way back to the panel — entering through its own dedicated connection and forming a genuine loop.
On paper that sounds like a minor wiring detail. In practice, it's the difference between a system that keeps working through a single break and one that doesn't. The two diagrams below show the identical fault — one break, in the identical location — wired each way.
A Class B circuit is a dead end — one conductor run out to the farthest device and back along the same path. Break it anywhere, and everything past the break has no way back to the panel; the panel notices the open circuit (that's what "trouble" annunciation is for) but simply cannot talk to those devices anymore. A Class A circuit is a closed loop: the same outgoing conductor continues past the last device and returns to the panel through a second, independent connection. When a single break occurs, the panel keeps talking to the "upstream" devices exactly as before, and it reaches the "downstream" devices by going the other way around the loop — out through the return conductor. Every device stays reachable from one direction or the other. The panel still knows precisely where the break is and reports it as trouble, but no device drops off the system because of it.
This undersells what Class A actually buys. It isn't a spare conductor sitting idle until the primary one fails, and it isn't just "better odds of eventually restoring service." It's a fundamentally different topology — a true loop with a genuine, separately-routed return path into the panel — that keeps every device fully functional during the fault itself, not just recoverable afterward. As the diagrams above show, a Class B circuit with a single break loses two of four devices immediately and stays that way until someone repairs the wiring. The same break on a Class A circuit loses zero devices, for the entire time the fault condition exists. That's exactly why NFPA 72 and many AHJs specifically require Class A wiring for certain applications — high-rises, systems serving large occupant loads, and other cases where uninterrupted operation of every device on a circuit during a single fault is treated as a life-safety requirement — rather than treating it as an optional redundancy upgrade.
Explains what actually happens during a single wire fault on a fire alarm initiating device circuit, signaling line circuit, or notification appliance circuit, depending on whether it's wired Class A or Class B per NFPA 72 — using a side-by-side illustrated comparison of the identical fault on each topology.
A Class B circuit is wired out from the panel to each device in sequence and back to the panel along the same physical conductor path — it is not a loop back to the panel from the far end. If a single break (open circuit) occurs anywhere along the run, every device farther from the panel than the break point is disconnected and loses communication with the panel. The panel does still detect the open circuit and annunciates a trouble condition — detection isn't the failure point, continued function is.
A Class A circuit runs out to the last device just like Class B, but then a separate, independently-routed return conductor carries the circuit all the way back to the panel, entering through a second, dedicated connection. This forms a genuine loop rather than a dead-end run. If a single break occurs anywhere in that loop, the panel can still reach every device on the circuit — the devices on one side of the break are reached normally, and the devices on the other side are reached by routing signal the opposite direction around the loop through the intact return conductor. Full functionality survives a single fault; the panel still identifies and annunciates the break location as a trouble condition, but no device is lost.
The practical difference is stark: Class B tolerates a single fault with partial, downstream loss of devices, while Class A is specifically engineered so that every device keeps functioning through that same single fault. That reliability comes at a real cost — a Class A circuit needs roughly double the wire of the equivalent Class B circuit, since a full separately-routed return path has to be pulled all the way back to the panel. NFPA 72 and the adopting building/fire codes require Class A wiring in specific occupancies and applications — commonly high-rise buildings, systems serving large or high-risk occupant loads, and other cases where uninterrupted device function during a single fault is treated as a life-safety requirement — rather than leaving the choice purely to cost.
No — a break is still detected and annunciated as a trouble condition on a Class A circuit, and NFPA 72 requires the panel to be able to identify its approximate location. What changes is that devices keep functioning during the fault; on Class B, the same break both trips trouble AND disconnects every downstream device.
Not necessarily. Class A is engineered to maintain full functionality through a single fault condition on the circuit. A second, independent break — especially one that isolates a segment on both sides — can still disconnect the devices between the two breaks, since the loop is now broken in two places rather than one.
No, and this is the most common misunderstanding. It isn't a spare conductor that sits unused until something breaks — it's a genuine loop topology where the panel can address every device from either direction simultaneously. That's what lets it maintain full functionality through a fault rather than just improving the odds of eventually restoring service.
NFPA 72 and the adopting codes require Class A circuits where the AHJ or code considers uninterrupted operation of every device on a circuit during a single fault to be a life-safety necessity — commonly high-rise buildings and occupancies with large or higher-risk populations. Where that level of survivability isn't mandated, Class B is permitted as the lower-cost option.
Wire quantity is roughly double, since a full separately-routed return conductor has to be run all the way back to the panel in addition to the outgoing path. Total installed cost is usually somewhat more than double once labor, conduit fill, and the panel's second circuit connection are factored in, though the multiplier varies by project.
Try our Fire Alarm Studio
More calculators, simulators, and guides for this discipline.