Received noise on powerline communications where the in-building wiring acts as an antenna

IEEE Africon 2011 - The Falls Resort and Conference Centre, Livingstone, Zambia, 13 - 15 September 2011 Received Noise on Powerline Communications where the In-Building Wiring Acts as an Antenna † A. Emleh , H.C. Ferreira † , A. J. Han Vinck‡ and A. Snyders† † Department of Electrical and Electronic Engineering Science University of Johannesburg, P.O. Box 524, Auckland Park, 2006, South Africa {aemleh, hcferreira, asnyders}@uj.ac.za ‡ Institute for Experimental Mathematics Duisburg-Essen University Ellernstr. 29, D-45326, Essen Germany vinck@iem.uni-due.de Abstract – The powerline channel is considered as a challenging channel due to the heavy load and noise impairments. In this paper we investigate an important type of measurement with regards to the powerline communications channel, where the low voltage wiring inside a house is considered as a receiving antenna for interference. Several types of measurements are presented using the corresponding broadband PLC channel as a communication medium. Impulsive and interference noise play a major role in the contamination factors on a powerline channel. The results include the different measurements of noise in the frequency domain. Extensive tests and measurements have been used to ascertain which specific external interference, after filtering by the coupling circuit, does affect the entire PLC system being used in a house. Most of the residential powerline channels contain several types of noise because of the connected electric devices to the ordinary home power sockets [1]. The level of the disturbance that the electrical devices produce may vary from one building to another or a network in a certain area to another network in a different area. This paper will focus on a subset of the results that were obtained during the extensive tests that took place in two different residential houses. The results will mainly show the differences between the contamination factors and the similarities, if there are any, on the same network or on both networks in the two houses. To study this scenario, we have focused on the effects of interference and noise when using the house wiring as an antenna. Keywords: Powerline Communications, Coupler, Antenna, House Wiring, Interference and Noise. In fact, power lines were mainly designed and fabricated and installed for the purpose of providing and supplying electric power with the frequency of 50-60 Hz. However, in the frequency range between 3 kHz and 35 MHz [2] for narrowband and broadband PLC channels, the electromagnetically coupled noise and noise impairments play a major role in the distortion of transmitting and receiving communications signals or data bits. I. INTRODUCTION For a long time the powerline channel has been considered as one of the most challenging communications channels due to the noise impairments and impedance mismatching occurring during the transmitting and receiving of bits or pulses in the communication channel itself. In the frequency range mentioned above, AM and HF transmissions might cause interference if they 978-1-61284-993-5/11/$26.00 ©2011 IEEE IEEE Africon 2011 - The Falls Resort and Conference Centre, Livingstone, Zambia, 13 - 15 September 2011 transmit in close proximity to the ellectrical house wiring. We now present the results of an expeerimental pilot project. The main investigation that this paper focuses on, is the very serious externnal interference disturbance that occurs when an approvved ICASA [3] remotely controlled toy car is in a funnctioning status and transmits noise, which is receivedd by the wiring system inside the house. Many tests were done to demonstrate this interference and how w it affects the communications signal in the powerlinee channel. The paper outline is as follows: In Section II we introduce the design of the coupling circuits which were used [4], the different probbes, and the measuring devices that were used inn the practical measurements tests. Section III will deescribe the way the measurements were done and analyze the wiring system in the house where the relationship between the powerline channel and the outputt noise will be explained and classified. In Section IV V we will show the results of the measurements and ccompare them. We conclude this paper in Section V. Fig. 1. Simple coupling design. TX: Traansmitter, RX: Receiver A 400VAC transformer was desig gned for the purpose of building a coupling device. Currents C of 0.5A and 1.0A were also included in n the calculations, depending on the ratio of the transformer. For the capacitors, a 275VAC cap pacitor with 0.47nF, 1.0µF, 2.2µF, 3.3µF, and 4.7µF values v were selected for the same design of the couplerrs. Capacitors are widely used in powerline commun nications to couple the signal to the powerline while blocking b the lowfrequency power signal [7], [8]. y used to block dc This series capacitor is mainly voltages but passes ac signal voltages v due to the frequency-dependent impedancee of a capacitor as shown in the formula below: Zc = II. COUPLING DESIGN AND DIIFFERENTIAL PROBES Couplers with transformers were dessigned for the purpose of detecting the received signal on the powerline channel. The design of thee couplers was made to serve the class of broadbandd PLC signals that uses the frequency range of 1 MH Hz to 35 MHz [2]. Another design was made to comply with CENELEC narrowband rules and reguulations for the band 3 kHz – 148.5 kHz [2]. For thee design of the transformers, we specifically followed the method in [5]. Another design method foor broadband transformers [6] was studied andd taken into consideration and its general rules havve been applied in the design. C = C − 90 9 0 (1) A 75Ω coaxial cable was used d in the design for higher frequencies, and a BNC C terminal for end connection to the receiving modem. When the measurements were tak ken, we made use of different types of differential pro obes and a UPS AC output power supply to ensurre isolation of the measurement equipment from the t PLC channel to prohibit possible contamination of the PLC channel by the instrument’s power suppliees. To get the best results on the reeceiving side, all the extensive measurements and exp periments were done using one receiving instrument, which is the 8591E Hewlett Packard spectrum analyzer [9]. Fig. 2 illustrates the specific device thatt was used during the comprehensive measurements tak ken. The abovementioned methods relied onn using a backto-back combination of 1:1, 1:2, 1:3 and 1:4 transformer designs. Fig. 1 shows the design of one of the used couplers for the different measurements taken. 978-1-61284-993-5/11/$26.00 ©2011 IEEE IEEE Africon 2011 - The Falls Resort and Conference Centre, Livingstone, Zambia, 13 - 15 September 2011 are connected to the powerline wiring system in the house. The coupler is connected to the spectrum analyzer through a coaxial cable and a BNC terminal on one end and to the powerline 3-prong plug on the other end. The spectrum analyzer is connected to the powerline socket in one of the experiments and to an external power supply unit in the other measurement. The UPS power supply unit is used to ensure that interference does not come from the instrumentation when the spectrum analyzer is connected to the powerline plug that provides 220V. Fig. 2. HP Spectrum Analyzer (8591E) The two different coupling circuits that are shown in Figs. 3(a) and 3(b) were specifically designed for narrowband and broadband channels respectively. The couplers have 3-prong plugs that are typical for South Africa and the United Kingdom. Fig. 3 (a) Narrowband coupler, and (b) broadband coupler Several tests were done on these couplers, and their performance has also been experimentally verified. In [10] it was shown that couplers pass through background noise depending on the variation of impedance of the transformers installed in couplers. For comparison purposes, two houses were used to conduct the tests, where the diameter of their wiring systems is 1.5mm and 2.5mm respectively. Their topology was nearly the same. The tests show clearly that there is not much difference on the level of the interference that can be noticed or recorded because of the diameter of the wires. Four types of couplers were connected separately to the powerline socket. On each occasion, the measurements were conducted using one of those couplers. The other end of the couplers was connected to the spectrum analyzer which is also connected to the powerline socket as it is depicted in Fig.4, where the spectrum analyzer is powered by the UPS. The toy car’s interference (noise) that was detected when using the house wiring as an antenna had a range thus limited to 27 MHz for this pilot experiment. The full details of the measurements will be discussed in the two following sections. III. WIRING SYSTEM AND THE PROCEDURE OF THE MEASUREMENTS Residential houses are similar to some extent in the way the wiring system is implemented in them. But they may be different in the type of cable and the topology of implementation. In this section we will focus on the scheme and the implementation part of our devices and the way they Fig. 4 Spectrum Analyzer connected to UPS Unit The TOP101 Home UPS power supply unit that was used has a clean sine-wave and was tested on different electrical devices such as a fridge, vacuum cleaner, personal computer and air conditioning unit. 978-1-61284-993-5/11/$26.00 ©2011 IEEE IEEE Africon 2011 - The Falls Resort and Conference Centre, Livingstone, Zambia, 13 - 15 September 2011 This is a 3000VAC unit that gives an output of 220V at 50Hz frequency. Fig. 5 shows the remote controlled car that has been used in the measurements. This car is designed to operate and respond to the remote control signal at 27 MHz frequency. Fig. 6 shows a simple sketch of the power circuit used in the measurements. The settings of the spectrum analyzer have been fixed at the range of 130 MHz and the reference level of the amplitude is at -20dBm where the attenuation is 10dB. A MAXHOLD setting on the instrument has been called and the toy car started running by a demand from the remote control unit. Fig. 5 The 27 MHz remote controlled car used in measurements. The tests were performed in one room where the measuring connection to the powerline wiring system was setup and the car was moved to another room where it was operated and switched on/off several times to verify repeatability of the measurements. The car is an ICASA [11] approved toy and can be bought from any toy store in South Africa, and that potentially makes the problem of interference more complicated. This toy produces noise and interference (see Section IV) when used inside a house or a powerline medium. Sending and/or receiving of data signals over the powerline channel can then be affected, where bits or packets could be lost. The manufacturers of such toys or devices are not aware of this type of problems that the users of powerline communications can easily face. Fig. 6 Power circuit used in the measurements. Spectrum analyzer connected to UPS. The car with the remote control was moved to another room inside the house before the operation started in order to illustrate that the interference is not suppressed sufficiently by the attenuation of the house wiring. Fig. 7 shows the first result obtained after a fraction of a second from the start of the operation. It is obvious that there is interference occuring at 27 MHz and in fact also other noise in the 3-24 MHz band. The following section will show the figures and the results that were obtained during the entire measurements procedure. IV. RESULTSnOFnMEASUREMENTS, INTERFERENCE AND NOISE ACTIONS Fig. 7 Interference at 27 MHz The operation was repeated in another double storey house of two levels. The upper level was reserved for the remote controlled toy and the bottom level had 978-1-61284-993-5/11/$26.00 ©2011 IEEE IEEE Africon 2011 - The Falls Resort and Conference Centre, Livingstone, Zambia, 13 - 15 September 2011 the spectrum analyzer and the 1:1 coup upler connected to the 220V powerline sockets. Mostly same results were obtained. The very same measurements were reepeated several times, the difference of the setup betw ween them was that in one case the equipment was powered by the 220V mains and in the other case the ccoupler and the spectrum analyzer had a different ppower source, which was the external UPS unit. Wheen the test was repeated, both of them were connecteed to the UPS. No interference was detected. O On the next connection, the spectrum analyzer wass connected to the UPS source and the coupler to the powerline socket. Fig. 8 shows the detected interrference on the screen of the spectrum analyzer. Figure 10 is presented, where we w notice the drastic change in the attenuation of th he signal when the spectrum analyzer, the couplers, the 10 times cable, and the differential probe are all a connected to the powerline sockets. In this case, th he differential probe is put to 1/20 ratio. When the ratio is shifted to 1/200, no interference can be detected. 8 6 4 Interference 2 Level (dB) 0 Fig. 10 Levels of interference at 1-30 MH Hz A small comparison is made in Fig. 11 between the four broadband couplers and th he one narrowband coupler in order to get accurate reesults. In this test, all the parts are connected to the PLC C socket. Fig. 8 Detected interference from power source The same distance was used when the measurements were repeated using a differential probee (ML013) that allows a maximum output of 7V into 2 kΩ and a 10 times cable instead of the 4 sets of coouplers. Fig. 9 shows the result on this connection wheen the 10 times cable was used and makes the obbtained result comparable to the previous results in thhis paper. 7 6 5 4 Noise Level 3 2 (dB) 1 0 Fig. 11 A comparison between the couplin ng devices Fig. 9 The 10 times cable is connected to PLC soocket The UPS was used to supply thee used devices when all the measurements were repeaated. Every time the spectrum analyzer was connecteed to the powerline socket, the couplers were conneccted to the UPS unit and visa-verse. Whenever a cou upler or a probe is connected to the UPS unit, we w could detect no interference on the powerline wiring system. 978-1-61284-993-5/11/$26.00 ©2011 IEEE IEEE Africon 2011 - The Falls Resort and Conference Centre, Livingstone, Zambia, 13 - 15 September 2011 ……..IEEE Transactions on Power Delivery, Vol: 20, V. CONCLUSION [5] Coupling circuits were designed and different differential probes and cables were used to detect the noise and interference on the wiring system inside a residential house. Powerline behavior in the house has been studied and the effect of the wiring system has been tested. Interference occurs at 27 MHz inside the house when the wiring system acts as an antenna. Results were obtained and the interference problem was illustrated. The remote controlled cars operating at 27 MHz that are used inside the residential homes can be purchased at many outlets. Since some PLC modems operate at 27 MHz, a high probability for interference will occur. Therefore, it would be recommended that PLC modems do not transmit information at 27 MHz. [6] Issue: 1 Publication Year: 2005, PP. 64 – 70 Petrus A. Janse van Rensburg “Effective Coupling for Power-Line Communications”, Doctoral degree, University of Johannesburg, January 2008 C.L. Ruthroff “Some Broad-Band Transformers” Proceedings of the Institute of Radio Engineers, vol. 47, Issue 8, pp. 1337-1342, Aug. 1959. [7] H.-K Podszeck, Carrier Communication over Power Lines, 4th Edition, New York: Springer-Verlag, 1972. [8] IEEE Guide for Power-Line Carrier Applications, IEEE Standard 643-1980. [9] O. G. Hooijen, “A Channel model for the Residential Power Circuit Used as a Digital Communications Medium”, IEEE Transactions on Electromagnetic Compatibility, Vol. 40, Issue: 4, P.331, Nov. 1998 [10] H. M. Oh, S. Choi, J. J. Lee, S. Shon, H. S. Kim, D. S. In, J. Bae, “Coupler with transformer for impedance matching on mv power distribution line for BPLC”, IEEE International Symposium on Power Line Communications and Its Applications, ISPLC April 2008, p. 409 [11] ICASA Approval code “TA-2009/1460” Online: Available http://www.icasa.org.za/ For the first time to our knowledge, in-building wiring has been investigated as a receiving antenna for noise reception influencing PLC signals. If further work should be conducted, it might include but not be limited to a reconsideration of ICASA regulations, testing the interference in multi-level blocks and considering the time domain of the interference when occurring as a result of the 27 MHz noise level. REFERENCES [1] [2] [3] [4] M. Tlich, H. Chaouche, A. Zeddam, F. Gauthier, “Impulsive Noise Characterization at Source”, IEEE conferences Wireless Days, 2008. WD '08. 1st IFIP 24-27 Nov. 2008. H. C. Ferreira, L. Lampe, J. Newbury and T. G. Swart, Power Line Communications: Theory and Applications for Narrowband and Broadband Communications over Power Lines, Chichester, England: John Wiley & Sons, 2010. The Independent Communications Authority of South Africa (ICASA) Website: http://www.icasa.org.za/ (Last accessed on 14 April 2011) P. A. Janse van Rensburg and H. C. Ferreira “Design of a Bidirectional Impedance-Adapting Transformer Coupling Circuit for Low-Voltage Power-Line Communications” 978-1-61284-993-5/11/$26.00 ©2011 IEEE