Table Of Content
Table Of Content
Power over Ethernet has changed Pro AV, industrial automation, security, machine vision, and control-system design. One Ethernet cable can now carry data and power to cameras, touch panels, wireless access points, AV-over-IP endpoints, intercoms, sensors, and other edge devices.
A common mistake is checking whether a switch supports PoE, PoE+, or PoE++ without calculating whether its total power budget can support the complete system.
A switch may support 60 W or 90 W on one port while its total PoE budget is far lower than the sum of all ports at maximum load. In industrial networks, the available budget may also change with DC input voltage.
Start With the Difference Between PSE and PD Power
The PSE, or Power Sourcing Equipment, is normally the Ethernet switch. The PD, or Powered Device, is the endpoint receiving power.
IEEE PoE specifications distinguish between the power supplied by the PSE and the minimum power available at the PD because some energy is lost in the cable. The commonly referenced maximum PSE values are 15.4 W for IEEE 802.3af, 30 W for 802.3at, 60 W for 802.3bt Type 3, and 90 W for 802.3bt Type 4. The corresponding maximum power available at the PD is lower: approximately 13 W, 25.5 W, 51 W, and 71.3 W.
First rule: do not confuse the switch-side rating with usable endpoint power.
Calculate the Real Device Load
The most useful starting formula is simple:
Required PoE load = sum of the maximum power requirement of every connected PD.
Use the manufacturer's maximum draw rather than average consumption. PTZ cameras, wireless access points, AV endpoints, and other devices can have short periods of higher demand.
For example, imagine a Pro AV and control-room network with eight AV-over-IP endpoints at 18 W each, four PTZ cameras at 25 W each, two touch panels at 15 W each, and two wireless access points at 22 W each. The calculated PD load is:
8 × 18 W + 4 × 25 W + 2 × 15 W + 2 × 22 W = 318 W.
A 370 W PoE switch may appear sufficient, but 318 W is not yet the number I would use to approve the design.
Add Engineering Headroom
A PoE system should not be designed to run continuously at 100 percent of its rated budget. Real installations face cable loss, temperature variation, power-up events, device replacement, and expansion.
For many projects, I prefer to add roughly 20 to 30 percent engineering headroom unless the equipment vendor provides a different design rule.
Using a 25 percent margin in the previous example:
318 W × 1.25 = 397.5 W.
I would therefore look for a switch and power-supply architecture that can reliably provide at least about 400 W of usable system budget under the actual operating conditions.
The margin is a design decision, not an IEEE requirement. Critical, hot, remote, or difficult-to-service systems usually justify more headroom.
Check the Total Budget, Not Just the Per-Port Rating
This is where many specifications are misunderstood.
A 24-port switch advertised as “90 W PoE++ per port” does not necessarily provide 24 × 90 W. That would require 2,160 W before switch losses. The actual chassis budget may be only a fraction of that.
Always check three specifications together:
- Maximum PoE power per port
- Total PoE system budget
- Number of ports that can simultaneously operate at the required class
This matters in Pro AV and surveillance systems where many endpoints operate simultaneously.
In Industrial Networks, Input Voltage Can Change the Budget
Industrial Ethernet switches often accept wide-range DC input, but the available PoE output can depend on input voltage and power architecture.
One industrial switch design, for example, may provide a 120 W total PD budget at 48 VDC input, 90 W at 24 VDC, and only 45 W at 12 VDC. This means the sentence “the switch supports PoE+” is not enough to approve the system.
Confirm the field power source, voltage under load, redundant-input behavior, power-supply capacity, and PoE budget at that voltage.
Treat Startup, Priority, and Failure Modes as Part of the Budget
PoE planning must also cover startup after an outage, simultaneous device power-up, and loss of a redundant power supply.
Managed industrial PoE switches can provide useful controls such as per-port power monitoring, PoE scheduling, port priority, current and voltage reporting, and automatic restart of an unresponsive powered device.
Port priority matters when power becomes constrained. A control processor, emergency camera, or critical AV decoder may deserve priority over a noncritical endpoint.
In other words, PoE budget should be designed together with the system's failure strategy.
Remember That Heat Is Part of the Power Problem
Higher PoE power means more current and heat in the switch and cabling, especially in dense bundles, sealed cabinets, hot environments, and fanless installations.
Cable category, conductor size, bundle size, ambient temperature, cabinet ventilation, DIN-rail spacing, and equipment derating all influence thermal performance.
A design that is electrically valid at room temperature may not be the right design inside a hot enclosure beside drives, PLCs, or power equipment.
A Better PoE Budget Formula
For practical engineering, I use this structure:
Minimum switch PoE budget ≥ Σ(maximum PD power) × design margin
Then verify five additional constraints:
- Every device is compatible with the required IEEE PoE type and class.
- Each individual port can supply the device's maximum requirement.
- The switch can provide the total budget at the actual input voltage.
- The external power supply can support the switch itself plus the PoE load.
- The system remains acceptable after a single power-supply or uplink failure if redundancy is required.
This is safer than selecting by port count and the largest PoE number in the datasheet.
Why This Matters More in Pro AV
Pro AV is increasingly network-based. AV over IP moves video, audio, control, and management onto Ethernet, while PoE simplifies deployment of encoders, decoders, touch panels, cameras, and wireless devices.
That convergence creates a new engineering reality: the Ethernet switch is no longer only a data-forwarding device. In many systems it is also part of the power-distribution architecture.
AV-over-IP designers already consider multicast, IGMP snooping, VLANs, QoS, uplink bandwidth, and redundancy. PoE budget deserves equal attention: perfect packet performance is useless if endpoints lose power.
My Practical Checklist
Before approving an industrial PoE switch for a Pro AV or automation project, I recommend checking:
- Number of powered devices now and after expansion
- Maximum wattage of every PD, not only typical consumption
- IEEE 802.3af / at / bt compatibility
- Per-port power limit
- Total PoE budget
- Budget available at the actual DC input voltage
- External power-supply capacity
- Redundant power behavior
- Port priority and power-management features
- Cable length, category, conductor size, and bundle conditions
- Ambient temperature and enclosure thermal design
- Monitoring, alarm, and remote-restart requirements
If these items are known, PoE selection becomes an engineering decision instead of a guess.
Final Thought
The best PoE design is not the one with the highest advertised wattage. It is the one that can power every required endpoint, under the real voltage, temperature, cable, redundancy, and expansion conditions of the project.
Across Pro AV, industrial automation, machine vision, and network infrastructure, one principle holds: reliability comes from designing for the system, not the headline specification.
At STAR FIRE TECH, our work across Pro AV connectivity, optical fiber communication, structured cabling, and industrial communication leads us to view the switch as part of the complete signal, network, and power path.
If you are designing an AV-over-IP network, industrial control system, machine-vision network, or other PoE-powered infrastructure, calculating the power budget early can prevent expensive redesigns later.
