Tech Tips Blog

PoE Issues… How to Find and Avoid Common Problems

Network World cites (trib.al/9hLb9KS) a survey the Ethernet Alliance (https://ethernetalliance.org/)  conducted, in which a significant number of users reported issues with PoE including reliability, long repair times, and connection challenges. This should be of key concern to network operators, who – in addition to delivering voice and data – are now in the “power delivery business.”

Today, PoE is powering many mission critical devices, including VoIP systems, Wi-Fi networks, surveillance systems, digital lighting, building access, IoT devices, digital signage, etc. As with bandwidth, PoE has also increased, evolving from 12W to 25W. Now with IEEE 802.11bt delivering up to 71W to the device, a myriad of new applications and endpoints previously not possible are now available. Unfortunately, testing during deployment is typically conducted by simply plugging in a simple low powered device (PD) such as a phone or AP and watching to see if the LEDs come on. The problem with this method is it does not consider or test for possible issues in the network infrastructure or with the worst-case PD power requirement – most of which could be avoided and resolved before the network turn-up, if properly tested.

We would go so far as to say that the main reason so many users are finding problems with PoE, is precisely BECAUSE OF THE LACK OF APPROPRIATE TESTING!

Remember, your cable plant is likely being used in a manner it was never tested for at installation.

Some of the most common PoE issues include:
i. Cable Type is incorrect 802.3bt requires at least CAT5E

  1. High cable resistance, unbalanced cable pairs, or poor installation workmanship (i.e., bad punch-downs)
    ii. Bad Cabling (Opens, Shorts, Split Pair, Crossed) as a 100Mbps high powered PD requires 2 additional pairs
    iv. Misconfigured PSE (Power Sourcing Equipment) i.e., switches or injectors
  2. Incorrect port usage (PoE limited to specific ports)
    vi. Incompatible technologies
    vii. Vendor specific vs. standards compliant
    viii. PSE PoE budget limitations –  PSE’s have a limit on how much power they can provide

These issues must be identified during rollout of the PoE network so they can be resolved immediately, and this goes well beyond negotiating PoE and measuring voltage. To do so properly, it is important to validate that the expected power (wattage) is being supplied by the PSE. Technicians must be able to validate a PoE switch’s power budget and the cabling infrastructure’s ability to deliver full power to the endpoint – across ANY distance. This is especially critical when pushing the IEEE maximum length spec of 100m (in many cases, users are forced to push beyond the spec due to site constraints.)

Both LinkRunner G2 and EtherScope nXG feature NetAlly’s proprietary TruePower™ technology, to test loaded PoE performance by drawing actual power up to 802.3bt 90W across all four pairs. TruePower validates that the Switch (Power Sourcing Equipment) and cabling can provide the requested power under load by applying a load equivalent to the selected class to mimic a Powered Device (PD). This testing loads the circuit to stress switches, cabling, and patch panels, all while measuring the voltage drop, actual power delivered, and pairs being used. This test emulates any PoE client in terms of negotiating and drawing (!) any PoE client specification (Class 0 through 8, including UPOE and PoE injectors).

To simplify the process of validation, NetAlly’s AutoTest technology automatically changes the test warning threshold per class negotiated, depending on the class type. This ensures a proper “go/no go” test result.

This is the only kind of thorough testing that validates power delivery when installing IoT connected devices to ensure a smooth deployment and avoid future problems.