Realization of an active inductance for a low power high bandwidth DC power line communication network transceiver

https://doi.org/10.1016/j.aeue.2009.07.010Get rights and content

Abstract

An active inductor based on an improved gyrator circuit is proposed. The active inductor is developed to be implemented in a high impedance transceiver for a wearable DC power line communication network where requirements such as low power consumption, high bandwidth and numerous nodes support are prioritized. A load isolation step is introduced to ensure the stability of the active inductance's size on different load currents. The proposed gyrator circuit is analyzed and optimized by means of theoretical calculations. The theoretical results are then verified by simulations and experiments in the frequency range up to 10 MHz.

Introduction

Power line communication (PLC) is a technology for distributing data on the power supply line. A DC-PLC network (see Fig. 1) is a PLC network deployed in systems where DC power supply is used, e.g., automotive systems, robotic systems, wearable systems, etc. A number of N nodes are connected, without any restraint on placement, in parallel over a DC power supply unit powering the nodes. The communication signal is generated from either of the attached nodes and is superimposed on the DC voltage.

Every node interfaces the power line through a DC-PLC transceiver. Power is a scarce resource in battery powered systems, which puts hard requirements on the transceivers. On one hand, the transceiver has to have low impedance on low frequencies to minimize losses when powering the node. On the other hand, the transceiver's impedance on higher frequency has to be high to minimize power consumption when communicating. The impedance of the power line is determined by the number of nodes connected to the network. The lower the power line impedance, the higher current is required to establish the communication signal. A straightforward solution is to connect an inductor in series with the power stage of each node thus isolating the low impedance path. However, due to DC current flowing through the inductor, achieving an adequate high inductance with a passive inductor element is not possible. High inductance and low DC resistance are usually two contradictory requirements when designing an inductor with physical dimension constraints. Moreover, increasing the size of the inductance gives rise to an increase in the winding capacitance [1]. This results in a lower resonance frequency and consequently a reduction in the bandwidth of the DC-PLC transceiver.

There are several techniques to implement active inductors used in the domain of active power filtering. These techniques consist of feed forward filters [2], [3], feedback filters [4], [5] or hybrid filters [6]. The working principle for these filters is that a sense element measures the ripple component of the signal which is then actively reduced by, for example, injecting its inverse. These types of solutions are, however, not appropriate for DC-PLC transceivers due to either the need of a bias voltage that is higher than the voltage present on the DC power line or comprising capacitive load (used as the sense element), resulting in a low impedance path.

Three additional techniques to realize active inductors are presented and analyzed in [7], [8]: Gm-C gyrator, inductor multiplier and operational amplifier C (OA-C) gyrator, see Fig. 2. The Gm-C gyrator suffers from high power losses within the transconductance amplifiers [7], which is not suitable in a DC-PLC transceiver where a load current is always running through the inductor. While the inductor multiplier shows good efficiency on high load current, it still shows low efficiency on low load currents [7] and therefore discarded from further consideration. The OA-C gyrator is characterized by high inductance and high efficiency for low load current (current flowing through RL), which are key characteristics for DC-PLC transceiver. However, the operative bandwidth of the OA-C gyrator is poor [8] which limits the attainable data rate transfer on the network. Therefore this study is made to investigate an active solution to replace the passive inductor element based on the architecture of the OA-C gyrator and meet the requirements on high inductance, high bandwidth and low power consumption.

Section snippets

DC-PLC transceiver requirements

Designing a DC-PLC network has the main advantage over PLC network on other types of power supply lines, e.g., AC home power line, that the load conditions on the DC power line are known and controllable. This is the case when developing new DC systems where the loads and noise sources can be restricted through specifications and filtering measures thus minimizing interference with the DC-PLC network. This is not possible in AC-PLC networks where random nodes can be connected to the AC power

Modification of the OA-C gyrator

Using Laplace transform techniques, the impedance of the OA-C gyrator, ZOA-C, can be described as a function of the complex variable s, as shown below [8]:ZOA-C=RL1+s/ωz11+s/ωp1,where ωz1 is the location of the zero z1 that defines the lower frequency limit, ωlf, seeωlf=ωz1=(R0C0)-1and ωp1 is the location of the pole p1 that defines the upper frequency limit, ωhf, seeωhf=ωp1=(RLC0)-1.Having R0 much higher than RL results in an active inductance LOA-C=RLR0C0 and a bandwidth BOA-C=ωhf-ωlfωhf.

Realization of the LI-OA-LC gyrator

The LI-OA-LC gyrator is realized according to the schematic shown in Fig. 5. When choosing the components, a decision has to be made concerning the OA. There are two types of OAs: voltage feedback (VFB) and current feedback (CFB). In general, CFB OA has better power efficiency for higher bandwidth [12]. In this work, low power consumption and high bandwidth are prioritized and therefore, a CFB OA is chosen for the gyrator.

The CFB OAs available on the market do not normally sink the high load

Measurement setup

The measurement setup is shown in Fig. 6 and consists of a signal generator (Rohde & Schwarz SMT03), a spectrum analyzer (HP E4402B) and an oscilloscope (Tektronix TDS3034B). The signal generator is coupled onto the DC voltage source (8 V lithium ion battery) using a 10μF capacitor. The input AC voltage signal, vin, on the gyrator input is measured by the oscilloscope. The AC current flowing through the gyrator, iin, is proportional to the AC voltage over the 1Ω sense resistor. Both battery and

Results

In order to evaluate the LI-OA-LC gyrator, Pspice simulations and experiments are conducted to analyze the performance of the gyrator in Fig. 5. Both Pspice simulations and experiments are made at two different power loads: 0.125 and 2 W and at frequencies ranging from 300 kHz to 10 MHz. The Pspice model was build according to the schematic in Fig. 5. The models for the different components are supplied by the corresponding manufacturer. For each load, several transient simulations are done where

Conclusions

An active inductor based on LI-OA-LC gyrator was proposed for a DC-PLC transceiver. The gyrator was optimized with regards to bandwidth by introducing a small inductor element in series with the load. This led to an 85% theoretical enhancement of the bandwidth compared to the commonly used OA-C gyrator. The load isolation step was introduced to provide an active inductance that is independent of the load. Both simulation and experimental data were in good agreement and demonstrated the

Acknowledgment

The authors acknowledge the financial support from the Swedish Knowledge Foundation through the Industrial Research School for Electronics Design.

Michel Chedid obtained his M.Sc. degree (Dipl.-Ing) in Electrical Engineering from the University of Linköping in 2002. He was then employed at Saab Training Systems AB as a system engineer. In 2004 he was enrolled as an industrial Ph.D. student in the University of Linköping in the Institute of Technology and Science, where, in 2006, obtained his degree of Licentiate of Engineering Science (Dipl. Lic.-Ing) in the domain of robust wearable electronics.

References (12)

  • W. McLyman

    Transformer and inductor design handbook

    (2004)
  • P. Cantillon-Murphy et al.

    An active ripple filtering technique for improving common-mode inductor performance

    IEEE Power Electronics Letters

    (2004)
  • Dumas J, Lanoue B, Tahhan B. Active analog power filters provide solutions for EMC and EMI. In: Proceedings of IEEE...
  • L. LaWhite et al.

    Active filters for 1 MHz power circuits with strict input/output ripple requirements

    IEEE Transactions on Power Electronics

    (1987)
  • N. Poon et al.

    Techniques for input ripple current cancellation: classification and implementation

    IEEE Transactions on Power Electronics

    (2000)
  • Chow A, Perreault D. Design and evaluation of an active ripple filter using voltage injection. In: Proceedings of the...
There are more references available in the full text version of this article.

Cited by (0)

Michel Chedid obtained his M.Sc. degree (Dipl.-Ing) in Electrical Engineering from the University of Linköping in 2002. He was then employed at Saab Training Systems AB as a system engineer. In 2004 he was enrolled as an industrial Ph.D. student in the University of Linköping in the Institute of Technology and Science, where, in 2006, obtained his degree of Licentiate of Engineering Science (Dipl. Lic.-Ing) in the domain of robust wearable electronics.

Hans Nilsson born in 1963, holds a degree from the upper secondary high school of engineering (Dipl.-Ing) in the domain of mathematics, measurement technology and EMC. He started working within production test and verification in telecom and vehicle industry. Since 1988, he has been employed at the Saab group with responsibilities within hardware construction and environmental test. His research interests focus on the analogue and digital design in general and vehicle igniting system, multi-chip bonding, DC/DC converters, gyro and resolver platform, microwave, camera CCD and laser detector in particular.

Alf Johansson holds a M.Sc. degree (Dipl.-Ing) in Engineering Physics, Chalmers University of Technology, since 1973. During 1993–1999 he worked as system engineer and R&D-manager at Combitech Traffic Systems AB. Since 1999, he is a senior lecturer in Electronics Design at the School of Engineering (Jönköping University), and since 2002, the head of master programme in Embedded Systems. He has a number of publications in the area of EMC and power quality.

Jan Welinder holds a degree of Licentiate of Engineering Science Dipl. Lic.-Ing in the domain of electrical measurement technology and is currently deputy manager at the Electronics Department at SP Technical Research Institute of Sweden. He has a long background in electronic engineering. At the beginning of his career Welinder was involved in the early work with ink jet technology at Lund University. His work at SP was in the beginning focused on measurement technology but from 1990 he worked with managing and the development of the large EMC facilities at SP and its connected research group. In 2008 he moved on to his current work with strategic development of the Electronics and ICT area and research project management. Current research activities deal with communications technology together with the telecom and vehicle industry.

View full text