ZA200108987B

ZA200108987B - Coupling stage for a data transmission system for low voltage networks.

Info

Publication number: ZA200108987B
Application number: ZA200108987A
Authority: ZA
Inventors: Jochen Mast; Lothar Heinemann; Miermann Rainer; Michael Arzberger
Original assignee: Abb Patent Gmbh
Priority date: 1999-08-26
Filing date: 2001-10-31
Publication date: 2002-07-10
Also published as: IL146376A0; PL353183A1; BR0013620A; CN1371555A; HUP0201602A2; NO20020402D0; SK2542002A3; DE19940544A1; NZ515475A; NO20020402L; EP1206846A1; WO2001015334A1; CZ2002597A3; HUP0201602A3; TR200200479T2

Description

¢ . t oT 99/657 ~1-

Input stage for a data transmission system for low- voltage power supply systems

Description

The invention relates to a device for inputting radio- frequency user signals for bidirectional data transmission on a low-voltage power supply system.

Deregulation of the electrical power market and the added-value electrical power services associated with this have resulted in a major increasé in inter&st in : direct transmission of data between the public electricity suppliers and the various end customers.

The bidirectional data transfer required for this purpose can in this case advantageously be carried by the low-voltage power supply system itself. This type - of data transmission is referred to as power line communication (PLC). The data transmission system can in this case be represented schematically by an arrangement in the form of a star. An intelligent control unit, which is referred to as an intelligent - network controller (INC) is located at the node of the star. At the end points, the signals for bidirectional data transfer are in each case produced and processed by a transmitter/receiver unit. This transmitter/receiver unit 1s also referred td as a transceiver (= TR). In practice, the star point normally corresponds to the 50 Hz distribution transformer station, and the end points are normally located in the vicinity of the building connection of the end customer.

In order to allow a low-voltage power supply system to be used as a transmission medium, a modulator/demodulator (= modem) is used in the INC and in each TR, where the digital user data are preprocessed in a suitable manner for the information channel. The modem itself essentially comprises a

\ p oo» /

Y 99/657 -2 - section for signal processing, where the modulation/demodulation is carried out, and a power supply system input stage, by means of which the analogue output signal is applied to the low-voltage power supply system, and is also received.

According to "Deutsches Institut fiir Normung: DIN EN - 50065-1: Signaliibertragung auf elektrischen

Niederspannungsnetzen 1m Frequenzbereich 3 kHz bis 148.5 kHz, VDE-Verlag, Berlin", [German Institute for

Standardization, DIN-EN 50065-1: Signal transmission on electrical low-voltage power supply systems in the frequency band from 3 kHz to 148.5 kHz, VDE Press,

Berlin], the entire frequency band which is authorised for communication on low-voltage distribution power supply systems for public electricity suppliers extends from 9 to 95 kHz. As is in this case stated in "Arz- berger, M, : Datenkommunikation auf elektrischen

Verteilnetzen fiir erwelterte Energiedienstleistungen,

Dissertation, Universitdt Karlsruhe, 1997" [Arzberger,

M.: Data communication on electrical power supply distribution systems for extended power supply services, Thesis Karlsruhe University, 1897], the frequency hopping modulation method (=FH modulation) with, for example 4 discrete frequencies within a frequency band which currently extends from 40 to 80 kHz is in this case used owing to the specific characteristics of the information channel. The INC periodically checks that the individual TRs can be reached and handles data traffic as required (= polling method) .

In principle, there are two different feasible versions for inputting and outputting signals onto and from the low-voltage power supply system, namely series and parallel inputting and outputting. Bidirectional data transmission in the frequency band from 40 to 80 kHz can be achieved more cost-effectively using parallel inputting and outputting, since the output stage for

Cw { 99/657 - 3 - outputting the signals is simply connected in parallel to the power supply system via a capacitor with a : suitable withstand voltage (in order to isolate the 230

V/50 Hz power supply system from the modem) and, possibly via a transformer for DC isolation. Inputting and outputting take place either between one phase and the neutral conductor or, for example, in the case of power supply systems without a neutral conductor, between two phases. Inputting and outputting between one phase and the neutral conductor is normally preferred for practical reasons, since the 50 Hz power supply system voltage, which represents interference from the transmission point of view, is only 230 V rather than 400 V for inputting and outputting between two phases. : Considerably more circuit complexity is required for inputting signals to the low-voltage power supply system than for outputting the signals. Greater importance must therefore also be attached to inputting of the signals since this is where there 1s the greater potential for further cost savings while at the same time achieving improved characteristics. For this reason, only the inputting of the signals is considered and discussed in the following text. The stage which is actually always provided for outputting the signals will be mentioned only when this appears to be necessary for the purposes of description of the invention.

The cables from the distribution transformer station to the building connections of the end users are in some cases in the form of underground cables, and in some cases overhead lines, in which case there may be a number of Junctions between underground cables and overhead lines, and vice versa, en route from the distribution transformer station to the end user.

Furthermore, there are junction points, since the end users are not all connected to the distribution

Lo. { } hi 99/657 - 4 - transformer station via their own cable connection.

Underground cables have a higher capacitance and a lower inductance than overhead lines, which means that the underground «cable has a considerably lower characteristic impedance. For this reason, voltage division takes place at the Junction between an overhead line and an underground cable, and this makes a major contribution to the attenuation levels, which are very high overall. These interface points are also the reason why the distribution system does not have a reciprocal response as an information channel, so that the response depends on the communication direction.

The attenuation characteristics of the information channel vary over the course of the day depending on the load applied to the low-voltage power supply system by the connected loads (in particular appliances with : EMC filters on the input side, such as primary-pulsed power supplies for televisions etc.). In order to keep the negative effects of these characteristics on the reliability of the information transmission system low, it is desirable for the maximum permissible signal : amplitude always to be available over the entire usable frequency band at the power supply system feed point, irrespective of the load state at that time.

The input impedance, or access impedance, which is generally highly inductive, but may also be capacitive, at the power supply system feed point of the INC or the

TR may vary within wide limits. The magnitude of the impedance is frequency-dependent and, depending on the type of cabling, the load on the power supply system and the particular discrete transmission frequency, is between less than 1 ohm up to 100 ohms. This impedance forms the load on the output amplifier on the INC and the TR. The lower the impedance, the greater the volt- amperes required to apply a specific signal amplitude to the existing power supply system voltage. These volt-amperes must be provided as real power by the output amplifier power supply, with the majority being

¥ 99/657 - 5 - converted to power losses in the output amplifier. The reason for this problem is the mismatch between the access impedance and the source impedance of the coupling network. It would be desirable for the output power to be produced by the output amplifier to be pure real power. This is the only way to make it possible to achieve the optimum design of the amplifier and its power supply, with as low a rating as possible.

The waveform of the FH-modulated signal is distorted due, on the one hand, to the frequency dependency of the input impedance and, on the other hand, the impedance of the particular input network. This leads to switch-on and switch-off transient responses when changing between the discrete frequencies, and must be avoided as far as possible for interference-free data : transmission.

In commercially available systems at this time, the majority of the costs for the entire PLC system are incurred for the output amplifier together with the power supply, so that optimum design of the power supply system input stage allows the overall system costs, which until now have still been high, to be considerably reduced. Furthermore, commercially available systems do not comply with the elementary technical requirements mentioned above.

The invention is thus based on the object of specifying an input stage by means of which the overall system complies with the elementary technical requirements stated above, and which allows a considerable . efficiency improvement to be achieved, while having low production costs at the same time.

This object is achieved by an input network having the features specified in Claim 1. Advantageous refinements are specified in further claims.

bi 99/657 - 6 -

A further description of the invention and its advantages is contained in the following .text with reference to exemplary embodiments which are illustrated in drawing figures, in which:

Figure 1 shows a block diagram of a normal stage according to the prior art.

Figure 2 shows a block diagram of the input stage according to the invention. :

Figure 3 shows a circuit diagram of one implementation of the normal input stage according to the prior art.

Figure 4 shows a circuit diagram of a first possible implementation of the input stage according to the invention.

Figure 5 shows a circuit diagram of a second possible implementation of the input stage according to the invention.

Figures 6a, b, c show a comparative illustration of the transient response of the transmission signal according to the prior art and with the input stage according to the invention.

According to the block diagram illustrated in Figure 1, the normal configuration according to the prior art for feeding in a transmission signal urs in principle comprises a module for signal processing 101, in which the modulation/demodulation is carried out, a module which contains an output amplifier 102, a module 103 for supplying power to the output amplifier 102 and the circuit for signal processing 101, a module for inputting the signals 104 into the low-voltage power supply system, and a module for outputting the received signal 105. The modules 104 and 105 normally also each contain a transformer for DC isolation and for signal amplitude matching. The analogue output variable from the module 101 is a voltage ugrr, which is amplified by the output amplifier 102. The output received voltage uzy may be regarded as the input variable for the module

). 99/657 -7 - 101. Furthermore, the module 101 also handles the digital data transfer for the rest of the circuit. The output variable from the overall circuit is the voltage u; which is then effectively applied across the power supply system impedance Zp. u;, ui, uz denote input and output voltages of the power supply module 103. uw denotes the AC voltage of the low-voltage power supply system. The output voltage from the amplifier 102 is annotated uUamp-

In the block diagram of the power supply system input stage according to the invention shown in Figure 2, there are three further modules 206, 207 and 208 in : addition to those in the block diagram illustrated in

Figure 1, which are used to comply with the fundamental technical requirements mentioned initially. The modules annotated 201 to 205 may in this case be constructed in the same way as the prior art modules 101 to 105 illustrated in Figure 1. Module 206 is arranged between the module for the output amplifier 202 and the module for inputting the signals 204. This network, which is preferably composed of passive components, provides impedance matching to the impedance load on the output . amplifier 202 (matching network). The network 206 is in this case expediently designed such that, in the worst case to be expected (this is the case in which the power to be produced by 202 and hence by 203 as well is otherwise the maximum), there is a resistive load on the output of the amplifier 202. This allows the power drawn from the output amplifier 202, and hence the output power to be provided by the power supply unit 203 as well, to be reduced further. The regulator module 207 contains a circuit which allows the output voltage ur to be regulated without any knowledge of the input impedance Z.. This regulation process can be carried out either directly (by measuring the output voltages u; while the input stage is at the power supply system potential) or indirectly by measuring a current (for example the output current from the output b 99/657 - 8 - amplifier iy or from the series-connected network for impedance matching). Figure 2 shows an example of how the output voltage ur can be regulated indirectly. The input variable for the regulator module 207 is the current ix (which is mapped by the instrument transformer 208 onto a voltage umeas) ©f the network 206 for impedance matching. Uy denotes the output voltage from the module 206. The output variable usp from the regulator module 207 is a voltage which is equivalent to the current in the network for impedance matching, and is thus equivalent to the output voltage U.. As is normal in control engineering, this voltage is compared with the AC voltage urr supplied from the module 201.

The difference voltage uaisf formed by means of an addition point 209 is then applied to the input of the output amplifier, having been set as a function of the impedance of the power supply system Zp . It should also be mentioned that the module 206 is located within the illustrated control loop together with the network for impedance matching.

The arrangement illustrated in Figure 2 in consequence produces an output voltage ur of virtually constant amplitude, irrespective of frequency and of widely differing impedances of the power supply system, while the power drawn by the output amplifier 202 is hence low, and the power drawn from the power supply 203 is thus also low.

Figure 3 shows a circuit diagram of a conventional implementation according to the prior art. The fundamental problem of the known circuit is that the output amplifier 102 emits a signal voltage at a constant amplitude, irrespective of the input impedance

Z.. The dynamically low output impedance of the amplifier 102 is raised to a major extent by the impedance. Zo formed by the coupling capacitors Cxi1 and

Cx; and the transformer T (whose short-circuit impedance and stray inductance are frequency-dependant), which

N

1 99/657 - 9 - has parasitic characteristics. The coupling capacitor

Cx1 is required in order to isolate the 230V/50Hz power supply system voltage uy from the output amplifier. The choice of capacitance values of Cy; and Cy: is in this case a compromise. On the one hand, the value should be as low as possible in order to limit the 50 Hz current through the output amplifier. On the other hand, the impedance of the capacitors Cx: and Cy» must not be too high at the lowest signal frequency to be transmitted since, together with the input impedance Z., it forms a frequency-dependent voltage divider and thus reduces the input signal amplitude. Particularly at low signal frequencies, where the particularly low magnitude of the input impedance means that comparatively high input power levels are required, the high-pass characteristic of the arrangement results in noticeable interference.

The capacitor Ci: ensures that any DC voltage component in the output voltage from the amplifier which may be present due to amplifier offset does not magnetize the transformer in one direction, thus leading to transformer saturation.

The increased source impedance Zp leads to frequency- dependent and load-dependent voltage division between Zp and Zg, thus resulting in the amplitude of the output voltage up fluctuating to a very major extent. In an : arrangement such as this according to the prior art, the impedance Zo of the input stage must be low in comparison to the lowest power supply system impedance in order to avoid the described voltage division process becoming too severe. Even if this is achieved, it is impossible for this circuit to set the amplitude of the input transmission signal reproducibly over a relatively wide frequency band.

In contrast, Figure 4 shows a first possible implementation of the input stage according to the invention with indirect regulation of the output voltage ur. In this implementation version, the primary

Co 7 99/657 - 10 - current iy of the transformer T is used as a measure of the output voltage. This principle is based on the fact that, if the values of the coupling capacitors C,. and

Cx and the parameters of the transformer are Xknown : : 5 approximately, the voltage across the power supply system impedance Zp can be calculated from the integral of the current im. In this case, it is assumed for simplicity that the 50 Hz voltage uy - as in the circuit according to the prior art as well - is dropped across the coupling capacitor and that the corresponding network is closed on the power supply system side via the main inductance of the transformer. This means that, if the current iy in the primary of the transformer is known, the voltage ur. can be inferred approximately from the input impedance. The actual voltage signal can be reproduced from the current iy if the output signal umeas from a current measurement device 208 (which may, for example, be in the form of a resistor or a current transformer) is first of all amplified in the block 207 (indicated by the operational amplifier OP1l with the gain factor k) and is then subjected to high-pass filtering in order to eliminate the residual 50 Hz current component. By way of example, this is achieved in Figure 4 by the O0P2, which is connected as a second-order high-pass filter by means of Rupi, Rup2, Curis Cup2. Finally, the current signal is supplied to an integrator (OP3, Rint; Cint).

The image usp 0f the output voltage up, obtained in this way, and the output voltage uy from a matching network 206 are added, weighted by means of R; and R,, and are subtracted from the nominal output voltage urs in order to form the control difference ugirr. This is done in the input section (which is advantageously in the form of a differential amplifier) of the power amplifier 202.

The matching network 206 is generally represented in the form of a T equivalent circuit. In the simplest conceivable case, it may consist of only a single series inductance which, together with the series

Y 99/657 - 11 - circuit formed by the capacitors Ci; and Cy, in the module 204.1, forms a series tuned circuit, whose resonant frequency in the transmission band is close to the lowest frequency to be transmitted. Such a series inductance may be either a discrete component, or may alternatively be formed by the stray inductance of a transformer for level matching. In contrast to the prior art illustrated in Figure 3, the stray inductance of such a transformer is not a disadvantage with the concept proposed in Figure 4 and, in fact, a parasitic characteristic, which previously had a disturbing effect, of a component is advantageously made use of.

Essentially, this 1s done by including the matching network 206 in the inner loop of the double control loop (referred to as cascade regulation), which forms the arrangement proposed in Figure 4.

Figure 5 shows a second possible implementation of the input stage according to the invention. The entire input stage is in this case at the power supply system potential. If the transformer for DC isolation in the output circuit 1s dispensed with, this considerably simplifies the module for the power supply input 204.2, and this now contains only one coupling capacitor Cy:.

Not only does this save the complex and expensive transformer but, furthermore, it allows the regulator module 207 to reproduce the voltage up, more exactly. In this version, the load current iy is used to calculate the output voltage up directly. This is done, analogously to Figure 4, by using the current measurement device 208, in this case by way of example a simple shunt resistor Recs, to convert the current iy to a voltage Upeas. This voltage is amplified once again by 0OPl, is filtered by OP2 and is integrated by OP3.

The controlled variable uy of the inner control loop is scaled by means of R,, and the feedback variable of the outer loop urs is scaled by Rz. This weighted sum is subtracted from the nominal output voltage urs in order to form the control difference uaisr, which is also once b) 99/657 - 12 - again advantageously done in the power amplifier 202.

The characteristics of cascade regulation are applied just as well in this version as in the arrangement shown in Figure 4. If DC isolation from the rest of the circuit 201 is required, this may be done, for example, at the output of the module 201, where urr is produced.

The advantage over the prior art (Figure 3) is in this case that the power levels that occur at this point are only low, so that it is possible to use a small-volume, low-cost transformer. Furthermore, its parasitic characteristics are of only secondary importance at the proposed installation location. It would also be possible to use an optocoupler instead of the transformer. :

Finally, by way of example, Figures 6a to 6c show a simulation result for the transient response of the transmission signal ur. The set of oscillations which is illustrated in Figure 6a and is at three different frequencies, but whose amplitudes are identical, is intended to be applied as accurately as possible with an amplitude of 2 V to the unknown power supply system impedance u;. Figure 6b shows the waveform of the output voltage u, when using the input stage according to the prior art (Figure 3): this shows severe transient distortion, and the amplitudes of the individual transmission frequencies are very different. The highest and the lowest transmission frequencies do not reach the desired level even at the feed point, thus restricting the range and the reliability of data transmission. As can likewise be seen well in Figure 6b, uncontrolled amplitude peaks also occur at certain frequencies, due to resonance effects. However, a response such as this is unacceptable in accordance with the Standard EN50065-1, which defines maximum amplitudes for such power line communication systems.

In some cases, the permissible transmission levels are considerably exceeded, depending on the impedance characteristics of the feed point. The only way to

Y 99/657 - 13 - prevent this is to select the transmission amplitude from the start to be so low that infringements are precluded. However, this drastically reduces the range and reliability of the PLC system.

The results illustrated in Figure 6c, when using the proposed novel input stage, have a different appearance: the desired constant and very highly reproducible amplitude of the transmission signal of 2V is achieved, and transient responses causing interference are largely suppressed. This arrangement allows the transmission levels specified in EN50065-1 actually to be fully utilized without having to accept any level reductions at the feed point or infringements occurring at certain frequencies. At the same time, the matching network 206 shown in Figures 2, 4 and 5 in this case makes it possible to minimize the power level to be supplied from the transmission amplifier 202.

From the above statements, it can be seen that the aims of: a) improving the reliability of the overall power line data transmission system, while b) at the same time reducing the power losses in the transmission amplifier and the power level required from the power supply unit, are achieved by using the proposed input device for power supply communication systems.

Claims

RE 1 99/657 - 14 - Patent Claims

1. A device for inputting a radio-frequency transmission signal (Urr) into a low-voltage power supply system (Ll, N), wherein a) an output amplifier (202) is provided, to which as the input voltage, a difference voltage (ugiff), which is formed at an addition point (209), between the transmission signal (ure) and a feedback signal (us) 1s supplied, b) the output amplifier (202) is followed by a network (206) for matching to the power supply system impedance (Z.), c) means (208) for detecting a voltage (Upeas) Which is proportional to the network output current (iy) is arranged at the output of the network (206), and d) the detected voltage (Umeas) 1s supplied to a regulator module (207) whose output signal is the feedback signal (us) .

2. The device as claimed in Claim 1, characterized in that the network (206) is formed from passive components.

3. The device as claimed in one of the preceding claims, characterized in that the regulator module (207) contains means for carrying out amplification, high-pass filtering and an integration function.

4. The device as claimed in one of the preceding claims, characterized in that the network (206) is followed by an input module (204.1), which contains a transformer (T) with coupling capacitors (Cui, Crz) on the primary and secondary sides.

Y 99/657 - 15 -

5. The device as claimed in one of Claims 1 to 3, characterized in that the network (206) is followed by an input module (204.2) which contains only one coupling capacitor (Cyi).