NETWORKING
35
Fig. 2
Frequency spectrum of a typical PLC system.
Fig. 3
Interference masking the intended PLC signal.
Fig. 4
Interference spectrum (with the PLC signal removed).
Fig. 5
Interference spectrum with the main interferer removed.
interference can easily be measured. If interference
is overlapping the intended receiver signal, it will be
relatively obvious on the spectrum analysers display.
Signals on a typical PLC system are shown in
Figure 2. The actual signal is the large saddle shape
in the middle of the display.
Interference can be identified because every
interferer has a frequency ‘fingerprint’; different kinds
of interferers have different fingerprints. Figure 3
shows a spectrum measured from a power line feed
at a transformer station. The saddle shape of the PLC
hub trying to communicate with the meters can be
seen, but so can a large amount of interference.
The interferer with the highest amplitude is right
on top of the PLC signal at marker 5. Figure 4 shows
the same spectrum after stripping out the PLC
signal. This shows that the strongest interference is a
single frequency peak: this suggests that the source
is probably a switch-mode power supply, and that,
because of the large amplitude, it is close to the test
site.
In this case, the interference was tracked to an
adjacent building that housed a mobile telephone
base station run by a large power supply. The power
supply was turned off in order to verify the signal
source, and the result was a somewhat cleaner
spectrum (see Figure 5). The analyser also found that,
beneath the main interferer, there was additional
interference at a slightly lower frequency. A separate
diagnostics process could then be carried to find this
other source of interference.
Appropriate Test Equipment
The most useful and accurate tool for the kind of
interference-hunting process described above is a
hand-held broadband spectrum analyser, such as the
www.networkseuropemagazine.com