PoE budget and the 100-metre limit: two common causes of unstable edge devices
A camera reboots, an access point drops out. Half the time the cause is not the network or the firmware but the power budget and the cable length.
Power over the network cable is the most practical part of low-voltage work. Here the model sits closest to day-to-day operations, and mistakes show up immediately.
The reference catalogue lists the Power over Ethernet classes with a nominal 48 V supply and typical per-port budgets. This is engineering reference data, not a magic calculation. The tool helps you see that the planned load is close to the limit.
What the problem looks like on site
A camera reboots at night. An access point drops out when many clients connect. A device runs for a month and then starts to misbehave.
All three symptoms read as network or software faults, so the investigation goes that way: firmware, configuration, switch logs.
The common cause is often simpler. The total planned load on the switch is close to its budget, and at peak one of the devices does not get power.
Why it is found late
Nobody likes adding up the budget by hand, and each device fits on its own.
The situation is typical. A cabinet holds a switch, access points and cameras. Each device sits in its own power class, each within the port limit. The question of whether the sum fits the switch’s total budget is rarely asked. It is almost never asked again when two more cameras go into the cabinet.
A model that records the switch, its total budget, the planned load and the PoE class answers that question on every recalculation.
100 metres as a planning limit
The second common source is link length. 100 metres for a copper Ethernet link is a planning limit.
The wording matters. It is not a physical threshold beyond which the link stops working; a 110-metre run often works. It works worse, and it degrades with heat, with a poor connector, with one more patch cord. In other words, it fails late and not reproducibly.
The tool flags a link over the planning length as a risk. Finding it in the description is cheaper than hunting for an intermittent fault on site.
What a cable record should contain
Endpoints and ports at both ends. Medium and cable type. Length. Label. Route. What the link shares a route with, and whether it is separated from power cabling. Load power.
When all of this is present, installation, audit and fault-finding go quickly. When something is missing, work on site turns into a manual investigation.
Why a shared route is a separate field
A low-voltage link next to a power cable with no declared separation means interference, and it is also a compliance question.
The tool cannot check physical separation: it does not see the site. It can check whether the description says anything about it and point out where it does not. For the Russian profile, the reminder about route separation is one of the main items.
What this gives operations
A quick win on large sites. An overloaded power budget and an unaccounted cable length are common causes of unstable edge devices, and both are found by checking the description, without a site visit.
It is dull work, done once when the site is described. It pays for itself the first time someone asks why the camera keeps rebooting.