Transcript Field trial in the city of Arnhem

Experiments related to network issues
1.PLC Field trial in Arnhem
The Netherlands
2. Radiating characteristics of
Low Voltage Distribution
Networks
PLC Field trial in the city of Arnhem, NL
•
•
•
•
Start 2002, end May 2003
150 houses;
337 modems (included repeaters)
2 Mbit/s backbone via 6 medium
voltage transformer stations. (35 users
per transformer station)
• Modem power density -50 dBm/Hz;
• DSSS Technology
Operator results (Arnhem)
•
•
•
•
•
Users very satisfied with service,
Technology works,
Commercially not yet viable!
Practical 1,4 Mbit/s too slow for future,
Connecting 10 kV Transformers to
infrastructure a logistic/economical
problem, (40.000 Transformer stations
to be connected).
Results measurements in Arnhem
• Large variations in generated fields (30 to 40 dB)
• Generated Fields:
– Table A (JWG Questionnaire Option 2): pass
– Table B (JWG Questionnaire Option 3) NB 30: fail
• A few hundreds of meters from the PLC- area the
generated fields are below ‘ITU rural’ noise level.
• How many measurements are needed to assess the
network compliance?
2. Additional measurement
campaign
Radiating characteristics of Low
Voltage Distribution Networks
LVDN’s
Radiating characteristics of LVDN’s
• Voltage injection on mains wires (sym. and asym)
• E and H–field measurements in and around
40 different houses:
– Indoor: At 1 m and 3 m from injection point
– Outdoor: At 3 m, 10 m, 100 m and 1000 m from
building
• ‘Antenna gain’ measurements of LVDN’s.
Results
• Large variations in generated fields (>40 dB)
because of large variations in mains geometry
• The injection of voltages equal to the CISPR-22
class B mains voltage limits can result in
magnetic field levels equal to the ‘table A’ field
limits at 3 m distance.
However, in most cases the generated fields will
be lower than the ‘table A’ limits.
General conclusions
• Local effects arise at locations in and around the
PLC area. Further outside the PLC area fields will
be lower than the existing noisefloor.
BUT, the following phenomena is underestimated:
• If in a large part of the world PLC networks are
deployed with a large penetration rate, cumulative
effects cause a rise in the noise floor in a large
area (not near PLC systems!) (Applies << 50 MHz)
Local PLC- effects are well known and can
Conclusions
be controlled in some manner
Near field ‘coupling’
Groundwave
Direct wave
Near LVDN wires
In and around
PLC-area
Above PLCarea
Mondial PLC effects depend on deployment
and penetration rate!
Skywave
Rise of noise floor
Measured equivalent antenna gain (for sky wave) of
LVDN’s in the frequency range 5 MHz- 20 MHz amounts
on average -30 dBi.