EuroWire May 2015

82 / 116

Technical article
May 2015 80
Temperature rise due to the Joule heating
effect is known to be proportional to I2R
losses[8] so, as current is fixed for each
measurement point, the resistance of
the cable pair under investigation will
differentiate temperature rise from one
cable to another.
Therefore, as expected, the cable with
highest DC resistance will have the most
temperature rise, and vice versa.
Discussion
Heating cables is known to increase
attenuation[9] which has a limiting effect
on cable reach. In relation to PoE, the
maximum temperature is likely to be in
the proximity of the energised conductors
which may be used for data transmission.
Therefore, the consequences of DC
powering on attenuation of the same pair
should be taken into consideration.
The results presented in this paper show
the temperature rise of one pair energised
with DC power using a cable located in a
controlled 20°C environment. Realistically,
the ambient temperature will vary from
site-to-site, and therefore, caution
should be taken when installing PoE
systems into uncontrolled and/or warmer
environments.
Further consideration should be given to
the correlation of simulated data and that
of the installation environment. On one
hand, the simulation may be based on a
worst case scenario. However, in reality,
the duty cycle may dictate that the power
is only supplied for a fraction of the time.
Good installation practices should be
implemented wherever possible, such
as minimising bundle sizes, accounting
for temperature rise for maximum cable
lengths, and keeping pathways and spaces
free of thermally insulating materials.
It is important to note that, while excellent
correlation was seen between simulated
and measured results for a single cable
setup, this work was not intended to
replicate the behaviour of cables in
bundles.
However, it is anticipated that good
correlation between theory and practice
will also apply to bundled configurations
in free air and in a variety of cable
containment systems, ie tray, trunking,
conduit, etc.
A comparison between the CCA 24 AWG
UTP and Cat6A 26 AWG U/FTP samples
show that it is possible for cables with
smaller conductors to radiate less heat
than those of larger conductors when
supplied with identical DC current values.
It is also known that the conductive foil
in screened cables act as a heat sink
which helps to reduce the amount of heat
radiated from the cable[10].
Therefore, it is important to take into
account the construction of cable, and
not only the conductor diameter for PoE
system deployment.
Individual Value Plot of Differences (with Ho and 95% t-confidence for the mean)
Differences
Δ Conductor temperature (ºC)
Current (A)
Cat6A-26AWG
CCA-24AWG
Cat5e-24AWG
Current (A)
Δ Conductor temperature (ºC)
Measurement
Simulation
Approximation
▲ ▲ Figure 4: Simulated, measured and approximated change in conductor temperature
▲
▲ Figure 5: Individual value plot of temperature differences
▼ ▼ Figure 6: Measured change in conductor temperature