Field of the Invention

The present disclosure relates to a distributed amplifier, a method of operating a distributed amplifier, a method of biasing a distributed amplifier, a transmitter comprising a distributed amplifier, and an optical communication system comprising a transmitter.

Background of the Invention

Distributed amplifiers are used for amplifying electrical signals. For example, distributed amplifiers are used in transmitters of optical communication systems for increasing the power level of an electrical signal supplied to an electro-optical modulator.

A distributed amplifier typically comprises an input line, formed of series coupled inductive elements of transmission line sections, for connection to a source, an output transmission line, similarly formed of a series of inductive elements or transmission line sections, for connection to a load, and a plurality of trans conductive gain devices, for example vacuum tubes or more commonly transistors, coupled in parallel between the input line and the output line. A signal propagating down the input line excites each of the gain devices successively, in response to which each gain device injects a current into the output line. The currents injected by the gain devices add in the output line to provide the amplified signal at the output.

The input and output lines should ideally be terminated by an impedance equal to the characteristic impedance of the respective line to minimise reflection and ripple phenomena and improve power transfer from the distributed amplifier to the load. The impedance should ideally be constant over the entire frequency range of the amplifier. In view of the desire to maintain a matched input and/or output impedance of the amplifier, biasing of the input and output lines can be problematic. It is known to supply a direct current bias voltage to the input and/or output line via the impedance termination. However, considering in particular the output line of the amplifier circuit, which may typically require a relatively high bias current, it can be difficult to maintain a constant output impedance termination whilst supplying a relatively high DC bias current to the output line. As an alternative, is known to supply the DC bias current to the input and/or output lines of the amplifier circuit via a reverse termination resistor. However, again, considering in particular biasing of the output line, supplying the bias current across the termination resistor can undesirably increase DC power consumption, as a portion of the DC power is dissipated across the resistor as heat. Summary of the Invention

An object of the present disclosure is to provide a distributed amplifier in which a DC bias current may be supplied to the output line of the amplifier with relatively low power dissipation whilst still achieving good output impedance matching across a wideband range.

The foregoing and other objects are achieved by the features of the independent claims. Further implementation forms are apparent from the dependent claims, the description and the Figures.

A first aspect of the present invention provides a distributed amplifier comprising an amplifier circuit comprising an input line for coupling to an AC voltage source, an output line for coupling to a load, and a plurality of gain devices between the input line and the output line, and a bias circuit for supplying a DC voltage to the output line of the amplifier circuit, the bias circuit comprising a lossy low-pass filter having an input for coupling to a DC voltage source.

The plurality of gain devices are coupled between, i.e. logically disposed between, the input and output lines. In essence, the function of the gain devices is to conduct the signal from the input line to the output line with a degree of gain, to thereby amplify the signal. By way of example, the plurality of gain devices could comprise a plurality of transistors, for example, field-effect transistors or bipolar junction transistors.

The distributed amplifier may thus be used for increasing the power level of an electrical signal supplied to the input line. For example, the distributed amplifier could be used in an optical communications system for increasing the power level of an input radio -frequency (RF) electrical signal and supplying the amplified electrical signal to an electro-optical modulator for generation of an optical signal for onward transmission via an optical fibre.

In operation, the lossy low-pass filter may filter high-frequency RF energy to thereby isolate the DC voltage source from RF energy in the output line, whilst providing a path for direct current from the DC source to the output line of the amplifier. Because the filter is lossy, i.e. because it absorbs, rather than merely reflects, stopband frequencies, it still provides output impedance matching at (high) frequencies in the stopband of the filter. This may advantageously allow for the DC voltage source to be connected directly to an RF active point of the output line, rather than, for example, through a termination or matching resistance. Consequently, dissipation of the DC bias current may be reduced/avoided, and thus the overall DC power consumption of the amplifier may be reduced. In an implementation, the lossy low -pass filter may comprise an output for coupling to the output line of the amplifier circuit. The output line may thereby be conveniently coupled to the lossy low-pass filter.

In an implementation, the bias circuit may further comprise an inductor coupled to the input of the lossy low-pass filter for coupling in series with the DC voltage source, wherein the inductor has an isolation-band lower than an isolation band of the lossy low-pass filter. The inductor may thus provide further isolation of the DC voltage source from relatively lower-frequency RF energy, i.e. frequencies that are in the passband of the lossy low-pass filter. This may advantageously increase the operating range of the amplifier. The inductor may advantageously be a relatively simple cheap mechanism for filtering these relatively lower frequencies. However, providing an inductor that is capable of filtering the relatively higher frequencies may be impractical, as to be capable of filtering higher frequencies the inductor may be required to be relatively large in size. The inductor and the lossy low-pass filter may thus efficiently work together to filter different frequencies of RF energy, to thereby increase the operating range of the amplifier whilst maintaining an acceptably matched output impedance.

In an implementation, the inductor may be a ferrite bead. A ferrite bead may advantageously be relatively inexpensive and mechanically robust.

In an implementation, the lossy low-pass filter may comprise a series line for coupling in series between the DC voltage source and the amplifier circuit, and a parallel line for coupling the series line to electrical ground in parallel to the DC voltage source and the amplifier circuit, the series line comprising one or more inductive components, and the parallel line comprising one or more resistive components. This arrangement may represent a particularly efficient structure for the lossy low-pass filter. The series line may function to conduct bias current to the output line of the amplifier, whilst the parallel lines may serve to dissipate RF energy in the stopband of the filter. For example, the lossy low-pass filter could comprise standard discrete inductor, resistor and capacitor components, which may advantageously be relatively inexpensive and mechanically robust. In a simple implementation the filter may comprise only a single parallel line, but in alternative implementations the filter could comprise multiple parallel line, for example, two parallel lines. The resistive components on the parallel lines provide the ‘lossy’ function of the filter, inasmuch that they allow dissipation of RF energy in the stopband.

In an implementation, the one or more inductive components may comprise one or more transmission lines. In other words, the inductance could be provided by transmission line sections. Transmission lines may advantageously be implemented relatively easily in a monolithic microwave integrated circuit architecture, with a relatively higher DC current capability than conventional inductors. Accordingly, employing transmission line sections may advantageously facilitate the supply of relatively higher DC bias currents to the output line.

In an implementation, the parallel line of the lossy low-pass filter may further comprise one or more capacitive components coupled in series with the resistive components. In other words, each parallel line of the filter may further comprise a capacitive component in series with the resistive component. The capacitive components, which may, for example, comprise one or more capacitors, reduce DC power consumption on the resistive components. Consequently, dissipation of the DC bias current across the filter may be reduced/avoided.

In an implementation, the one or more capacitive components may comprise one or more distributed stubs, and/or one or more diodes. In other words, the capacitance could be provided by stubs and/or diodes. Diodes may advantageously serve the dual purpose of protecting the amplifier circuit from electrostatic discharge, thereby reducing the risk of damage to the amplifier circuit in use. The stubs may advantageously filter very high-frequency radio energy with very good simulation accuracy.

In an implementation, the distributed amplifier may further comprise a termination circuit for coupling the output line of the amplifier circuit to electrical ground via a resistive component in the termination circuit, wherein the termination circuit is coupled to an end of the output line opposite to the load. The termination circuit may advantageously function to dissipate the energy of reverse travelling waves on the output line, thereby reducing the risk of destructive interference on the output line, and so advantageously increasing the output power of the amplifier for a given input power. The resistive component may, for example, comprise one or more resistors.

In an implementation, the termination circuit may be coupled to the output line of the amplifier circuit by the lossy low-pass filter of the bias circuit, and the termination circuit is coupled between electrical ground and the lossy low -pass filter in parallel to the inductor of the bias circuit.

In other words, the termination circuit may be coupled to an input of the lossy low-pass filter in parallel with the DC voltage source, i.e. the termination circuit may be connected to a node of the bias circuit located between the inductor of the bias circuit and an input of the lossy low-pass filter. In this, parallel, arrangement, the path for DC current from the DC voltage source to the output line does not pass through the termination circuit resistor. Consequently, DC bias current is not dissipated on the resistor, and thus the power consumption of the amplifier may advantageously be reduced.

In an implementation, the termination circuit may further comprise a capacitive component coupled in series with the resistor. The capacitive component(s), which may, for example, comprise one or more capacitors, reduce DC power consumption on the resistive components of the termination circuit. Consequently, dissipation of the DC bias current by the termination circuit may be reduced, and thus the power consumption of the amplifier may advantageously be reduced.

In an implementation, the amplifier circuit and the lossy low-pass filter may be formed as an integrated circuit. For example, the amplifier circuit and the lossy low-pass filter may be formed as a Monolithic Microwave Integrated Circuit (MMIC). An MMIC construction may advantageously allow the amplifier and filter to be implemented in a relatively small size and mechanically robust circuit. Furthermore, MMIC may be relatively easy an inexpensive to manufacture.

A second aspect of the present invention provides a transmitter for an optical communications system, the transmitter comprising: a distributed amplifier as described in any one of the preceding statements, and an electro-optical modulator connected to the output line of the amplifier circuit. A distributed amplifier is particularly suited to use in an optical communication transmitter as it allows for output impedance matching over a relatively wideband range.

In an implementation, the transmitter may comprise a DC voltage source coupled to an input of the bias circuit, and an AC voltage source coupled to the input line of the amplifier circuit, wherein the distributed amplifier is operable to amplify a signal travelling between the AC voltage source and the electro-optical modulator. For example, the AC voltage source could be computing device which outputs a RF signal. The distributed amplifier may thus be operated to increase the power- level of the signal prior to modulation by the electro-optical modulator.

In an implementation, the transmitter may comprise a further amplifier circuit for amplifying an AC signal, wherein the further amplifier circuit is coupled between the AC voltage source and the input line of the amplifier circuit. In other words, the transmitter may comprise a ‘pre-amplifier’ coupled in series between the AC signal source and an input of the amplifier circuit. Whilst the distributed amplifier topology of the amplifier circuit may advantageously facilitate output, and further input, impedance matching over a wideband range, the additive, rather than multiplicative, gain characteristics of the plural gain devices of the distributed amplifier circuit makes it difficult to achieve a large gain. Accordingly, the pre-amplifier may be configured to provide a relatively high gain, for example, the pre-amplifier could be an open-collector type amplifier, whilst the distributed amplifier may provide good output impedance matching with the load, e.g. the electro-optical modulator. The synergistic effect of the pre -amplifier and distributed amplifier circuit combined is thus acceptably high gain characteristics combined with acceptable output impedance matching of amplifier stage to load.

A third aspect of the present invention provides an optical communications system comprising: a transmitter as described in any one of the preceding three statements, a receiver for location at a position remote from the transmitter, and an optical fibre coupled between an output of the transmitter and an input of the receiver for carrying an optical signal between the transmitter and the receiver, wherein the receiver comprises a further electro-optical modulator for transducing the received optical signal into an electrical signal. The optical communication system is thus operable to communicate systems coupled to the transmitter and receiver respectively.

A fourth aspect of the present invention further provides a method of operating a distributed amplifier as described in any one of the preceding statements, comprising coupling the input line to an AC voltage source, coupling the output line to a load, and coupling the input of the lossy low-pass filter to a DC voltage source.

A fifth aspect of the present invention further provides a method of biasing a distributed amplifier, the method comprising supplying a DC voltage to an output line of the distributed amplifier via a lossy low-pass filter.

As noted previously, the lossy low-pass filter may filter high-frequency RF energy to thereby isolate the DC voltage source from RF energy in the output line, whilst providing a path for direct current from the DC source to the output line of the amplifier. Because the filter is lossy, i.e. because it absorbs, rather than merely reflects, stopband frequencies, it still provides output impedance matching at (high) frequencies in the stopband of the filter. This may advantageously allow for the DC voltage source to be connected directly to an RF active point of the output line, rather than, for example, through a termination or matching resistance. Consequently, dissipation of the DC bias current may be reduced/avoided, and thus the overall DC power consumption of the amplifier may be reduced. In an implementation, the method of biasing a distributed amplifier may comprise supplying the DC voltage via the lossy low-pass fdter directly to an RF active node of the output line. In other words, the lossy low-pass fdter could be coupled directly to the output line of the distributed amplifier without any resistive or inductive component being located between the lossy filter and the output line. Coupling the filter directly to a RF active node of the output line, rather than via an inductive or resistive component, may advantageously minimise dissipation of the DC bias current, and thereby reduce the DC power consumption of the amplifier.

These and other aspects of the invention will be apparent from the embodiment(s) described below.

Brief Description of the Drawings

Embodiments of the invention will now be described, by way of example, with reference to the accompanying drawings, in which:

Figure 1 shows schematically an optical communications system embodying an aspect of the invention;

Figure 2 shows schematically a transmitter of the optical communication system;

Figure 3 shows schematically stages of a driver amplifier of the transmitter shown in Figure 2,

Figure 4 shows schematically a distributed amplifier assembly of the driver amplifier shown in Figure 3,

Figure 5 shows schematically an amplifier circuit of the distributed amplifier assembly shown in Figure 4,

Figure 6 shows schematically a first embodiment of a lossy low-pass filter for biasing an output line of the amplifier circuit shown in Figure 5,

Figure 7 shows schematically a low frequency equivalent scheme of the lossy low-pass filter shown in Figure 6,

Figure 8 shows schematically a high frequency equivalent scheme of the lossy low-pass filter shown in Figure 6, Figure 9 shows schematically a second embodiment of a lossy low-pass fdter for biasing an output line of the amplifier circuit shown in Figure 5,

Figure 10 shows schematically a third embodiment of a lossy low-pass filter for biasing an output line of the amplifier circuit shown in Figure 5,

Figure 11 shows schematically a fourth embodiment of a lossy low-pass filter for biasing an output line of the amplifier circuit shown in Figure 5, and

Figure 12 shows schematically a receiver of the optical communication system shown in Figure 1

Detailed Description

An example of an optical communication system 101 embodying aspects of the present invention is depicted schematically in the Figures.

Referring in particular to Figure l, The optical communication system 101 comprises a transmitter 102, a receiver 103 located at a position remote from the transmitter 102, and an optical fibre 104 for carrying an optical signal from the transmitter 102 to the receiver 103.

The transmitter 102 is operable, as will be described in further detail with reference to later Figures, to encode a message onto an optical signal, which may be transmitted via the optical fibre 104 to the receiver 103, whereby the message may be decoded from the optical signal by the receiver 103. The system 101 may thus be operated to transmit a data signal between the transmitter location and the receiver location. For example, the system 101 could function as a long-distance broadband capable data link in a computer network for communicating remotely located computing devices. Transmitter 102 will be described in further detail with particular reference to Figures 2 to 11. Receiver 103 will be described in further detail with particular reference to Figure 12.

Referring to Figures 2 and 3, the transmitter 102 comprises an electrical signal source 201, a driver amplifier 202 and an electro-optical modulator 203.

The source 201 is controllable to generate a radio-frequency (RF) AC electrical signal. In the example, source 201 is a modem in communication with a computing device, and operable to modulate a carrier wave signal to encode digital information output by the computing device onto the signal. It will be understood by the skilled person however that the disclosure of the present specification has far broader utility than being specifically for coupling computing devices, and thus it will be appreciated that the source 201 could, in alternative embodiments, be an alternative signal source.

An output of the source 201 is coupled to an input of the driver amplifier 202. The driver amplifier 202 may thus be operated to increase the power level of the received electrical signal. Referring in particular to Figure 3, in the example, the driver amplifier 202 comprises an open-collector type pre-amplifier 301 and a distributed amplifier 302 connected in series between the source 201 and the electro-optical modulator 203. The source 201 is coupled to an input of the pre-amplifier

301, and an output of the pre-amplifier 301 is coupled to the distributed amplifier 302. The preamplifier 301 is operable to perform an initial amplification of the power level of the signal from the source 201 and output the pre-amplified signal to the distributed amplifier 302. The distributed amplifier 302 is operable to perform a subsequent amplification of the power level of the signal received from the pre-amplifier 301, and output the amplified signal to the electro-optical modulator 203.

In the example, the pre-amplifier 301 has a substantially conventional open-collector type amplifier construction. However, it will be appreciated, that the particular construction of the preamplifier 301 is relatively unimportant to embodiments of the invention. In essence, the preamplifier 301 should be capable of amplifying an AC signal, such as a RF signal, and many alternative suitable constructions of pre-amplifier will be known to the skilled person.

An input of the electro-optical modulator 203 is coupled to an output of the distributed amplifier

302. In the example, the electro-optical modulator 203 comprises an Electro-absorption Modulated Laser device, having an electrical output coupled to electrical ground and an optical output coupled to an input of the optical fibre 104. The electro-optical modulator 203 is thus operable to receive the amplified signal from the distributed amplifier 302, transduce the electrical signal into an optical signal, and inject the optical signal into the optical fibre 104.

Referring in particular to Figure 4, the distributed amplifier 302 of the driver amplifier 202 comprises an amplifier circuit 401 for amplifying a signal received from the pre-amplifier 301, a bias circuit, indicated generally at 402, for supplying a DC bias voltage to an output line of the amplifier circuit 401, an output line termination circuit, indicated generally at 403, for coupling the output line of the amplifier circuit 401 to electrical ground, and an input line termination circuit 408 for coupling an input line of the amplifier circuit 401 to electrical ground. The amplifier circuit 401 comprises a signal input terminal 404 electrically connected to an input line of the amplifier circuit, and a signal output terminal 405 electrically connected to an output line of the amplifier circuit. In an operable state, the input terminal 404 of the amplifier circuit

401 is connected to an output of the pre-amplifier 301, and the output terminal 405 of the amplifier circuit 401 is connected to an electrical input of the electro-optical modulator 203.

The amplifier circuit 401 further comprises an output line termination terminal 406 electrically connected to the output line of the amplifier circuit 401, and an input line termination terminal 407 electrically connected to the input line of the amplifier circuit. The output line termination terminal 406 is coupled to the bias circuit 402. The input line termination terminal 407 is coupled to electrical ground via the input line termination circuit 408.

The bias circuit 402 comprises principally of an electronic choke, in the form of ferrite bead 409, connected in series with a lossy low-pass filter 410. The bias circuit 402 is coupled between a DC voltage source VDD, which is in turn coupled to electrical ground 417, and the output line termination terminal 406 of the amplifier circuit 401, such that the bias circuit 402 may supply a DC voltage from the source VDD to the output line of the amplifier circuit. The operation of the bias circuit 402 will be described in further detail with reference to Figures 6 to 11.

The output line termination circuit 403 comprises principally a low-frequency matching resistor 411 connected in series with a low-frequency shunting capacitor 412. In the example, the termination circuit 403 is partly integrated with the bias circuit 402, such that the termination circuit 403 is electrically coupled between electrical ground 413 and a node 414 of the bias circuit

402 that is located between the lossy low-pass filter 410 and the ferrite bead 409. The termination circuit 403 thus electrically couples the output line of the amplifier circuit 401 to electrical ground 413 via the lossy -low pass filter 410, the resistor 411 and the capacitor 412.

The input line termination circuit 408 comprises a further low-frequency resistor 415 coupled in series between the input line termination terminal 407 and electrical ground 416. In the example, a DC bias voltage VGG is applied to an input line of the amplifier circuit 410 across the resistor 415.

In the example, the amplifier circuit 401, lossy low-pass filter 410, and resistor 411 are implemented as a monolithic microwave integrated circuit (MMIC), whilst the bead 409, capacitor 412, and resistor 415 are implemented as surface-mounted devices. The resistor 415 could additionally be relatively easily implemented in an MMIC construction. It will of course be appreciated however that whilst this MMIC construction confers certain advantages on the circuit, for example, reduced size, complexity, and cost, embodiments of the invention are not limited in its utility to an MMIC implementation, and could alternatively have a non-MMIC construction,

Referring next to Figure 5, the amplifier circuit 401 comprises an input line 501, an output line 502, and a plurality of trans conductive devices, in the form of transistors 503 (denoted Q1 - QN), electrically coupled in parallel between the input line 501 and the output line 502.

The input line 501 is connected between the input line signal input terminal 404 and the input line termination terminal 407, i.e. between the output of the pre-amplifier 301 and electrical ground 416. In the example, the input line 501 is an artificial transmission line formed by a plurality of series connected inductors LG1 - LGN+1 spaced apart along the length of the input line. As an alternative, the input line could be formed by actual transmission line segments coupled in series. The output line 502 is connected between the output line termination terminal 406 and the signal output terminal 405, i.e. between the bias/termination circuits 402, 403 and the electro-optical modulator 203. Similarly, the output line 502 is an artificial transmission line formed by a plurality of series connected inductors LD1 - LDN+1 spaced apart along the length of the line. As with the input line 501, the output line 502 could instead be formed by actual transmission line segments coupled in series.

The plurality of trans conductive devices comprise N transistors 503, which in the example are field-effect transistors. The transistors 503 are spaced apart, i.e. “distributed”, along the input line

501 and output line 502, and coupled in parallel between the input line 501 and the output line 502. Each of the transistors 503 comprises a source terminal 504, a drain terminal 505, and a control, or gate, terminal 506. The gate terminal 506 of each transistor 503 is coupled to the input line 501 at a node between adjacent gate inductors LGN, and the drain terminal 505 of each transistor 503 is coupled to the output line 502 at a node between adjacent drain inductors LDN. The source terminal 504 of each of the transistors 503 is coupled to electrical ground. The parasitic input and output capacitances of the transistors 503 are thereby absorbed into the artificial input and output transmission lines 501, 502 respectively.

The transistors 503 act to couple electromagnetic energy from the input line 501 to the output line

502 with a certain amount of gain. In operation, an input signal from the pre-amplifier 301, received at the signal input terminal 404, propagates down the input line 501. As the input signal propagates down the input line 501, it sequentially excites the transistors 503, and is finally dissipated in the resistor 415 of the input line termination circuit 408, thereby preventing destructive back reflections on the input line which may otherwise undesirably reduce the input signal. In response to the signal propagating on the input line 501, the transistors 503 each inject a current into the output line 502, such that each successive transistor 503 contributes to the output signal. The gain of the transistors 503 and thus the output power of the amplifier circuit 410 is a function of the size of the transistors, i.e. the source width or emitter area, and the applied bias currents.

The current injected into the output line 502 by each transistor 503 splits equally in the output line 502 between a forward travelling component, i.e. a component travelling towards the signal output terminal 405, and a reverse travelling component, i.e. a component travelling towards the output line termination terminal 406. Provided the signal delay between outputs of the transistors 503 matches the delay between the inputs of the transistors, the forward travelling waves add in phase. This coherent signal is the amplified output of the amplifier. Conversely, where the delays are matched, the reverse travelling waves add out of phase, i.e. incoherently, and are dissipated in the resistor 411 of the termination circuit 403, at low frequencies, or in the lossy low-pass fdter 410 of the bias circuit 402, at high frequencies.

Referring next in particular to Figure 6, the lossy low-pass fdter 410 comprises an input terminal 602, an output terminal 603, a series line 604, and first and second parallel lines, indicated generally at 605, 606 respectively.

The series line 604 is coupled between the input terminal 602 and the output terminal 603 of the filter to conduct current therebetween. Series line 604 comprises a plurality, N, of inductors L0 - LN coupled in series in a spaced-apart arrangement along a conductor of the series line. The parallel lines 605, 606 are each connected at one end to the series line 604 at respective nodes located between adjacent inductors of the series line 604, and at the opposite end to electrical ground. Each of the parallel lines 605, 606 comprises a resistor 607 connected in series with a capacitor 608.

The input terminal 602 of the lossy low-pass filter 410 is coupled in series with the series line 604 of the filter 410. The lossy low-pass filter 410 thus communicates the DC voltage source VDD with the output line termination terminal 406 of the amplifier circuit via the ferrite bead 409 and the series line 604 of the lossy low-pass filter 410. The ferrite bead 409 and the low-pass filter 410 of the bias circuit 402 together act to filter the high frequency AC current, i.e. RF signals, applied to the bias circuit 402 from the reverse travelling wave in the output line 502 of the amplifier circuit 401. The ferrite bead 409 and the lossy low-pass filter 410 thereby serve to isolate the DC source VDD from the RF energy, while allowing a DC current path from the DC source VDD to the drain terminals 505 of the transistors 503 to thereby bias the transistors to place them in an amplifying state. The bead 409 of the bias circuit 402 is selected to have first stopband characteristics, such that lower frequencies of RF energy are effectively filtered by the bead 409, thereby isolating the DC source VDD from lower frequency RF energy. For example, the bead 409 may function to stop RF currents having frequencies up to approximately 30 GHz. The lossy low-pass filter 410 is configured, by selection of appropriate value for RCL components, to have stopband characteristics higher than the bead 409, such that the lossy low-pass filter 410 may effectively filter higher frequency RF energy emanating from the output line 502 of the amplifier circuit 401. Thus, in the example, the ferrite bead 409 and lossy low-pass filter 410 are configured to filter successive frequency bands. Because the ferrite bead 409 and the lossy low-pass filter 410 filter successive frequency bands, the operating range of the amplifier circuit is increased.

The termination circuit 403 is connected to the bias circuit 402 at the node 414 located between the bead 409 and DC source VDD and the input terminal 602 of the lossy -low pass filter 410. The termination circuit 403 thus communicates the output line termination terminal 406 of the amplifier circuit 401 with electrical ground 413 via the lossy low-pass filter 410. Similarly to the termination circuit 408 connected to the input line 501 of the amplifier circuit 401, the termination circuit 403 serves to dissipate, by the resistor 411, reverse travelling RF wave energy in the output line 502 of the amplifier circuit 401 to prevent back reflections of the output signal on the output line 502 to thereby increase the output signal level for a given input signal level. The termination circuit 403 thus serves as a RF signal path to ground. Because the resistor 411 is coupled in series with the low frequency capacitor 412 and is not in the direct current path between the bias voltage source VDD and the drain electrodes of the transistors 503, DC power dissipation by the resistor 411 is avoided in the course of biasing the output line 502 of the amplifier circuit 401.

Referring to Figure 7, at low frequencies the lossy-low pass filter 410 behaves as a short circuit. Thus, in this low frequency condition, the bias voltage is supplied from the DC source VDD via the bead 409, and the output matching impedance for the output line 502 of the amplifier circuit 401 is achieved by the resistor 411 and the low frequency capacitor 412 of the termination circuit 403.

In contrast, referring to Figure 8, at high frequencies the bead 409 drastically reduces its isolation capabilities to isolate the DC source VDD from RF energy in the output line 502 of the amplifier circuit 401. However, at high frequencies the lossy low-pass filter 410 functions to filter the high frequency energy, thereby isolating the amplifier circuit 401 from the bead 409 and DC source VDD. Additionally, at high frequencies, the lossy filter 410 provides the output matching impedance for the output line 502 of the amplifier circuit 401 by the ‘lossy’ nature of the resistive components 607 on the parallel lines 605, 606.

The main consideration for the lossy low-pass filter 410 is that its isolation band starts from the frequency where the isolation band of the bead 409 reduces below an acceptable level. At the same time, the filter 410 has to provide good input/output return loss in its isolation band in order to provide impedance matching for the output line 502 of the amplifier circuit 401. The isolation band and input/output return loss characteristics of the filter 410 may be tuned by selecting appropriate values for the resistive, capacitive and inductive components of the filter.

In the example, the bead 409 is selected to have an isolation band up to approximately 30 GHz, this selection representing a commonly commercially available and suitably sized bead. Consequently, in the example the RCL components of the lossy low-pass filter 410 are selected to provide the filter with an isolation band starting at approximately 30 GHz and continuing up to approximately 50 GHz, 50GHz being the expected maximum operating frequency of the amplifier circuit 401 in the example embodiment. This relatively narrow stop-band requirement of the filter 410, i.e. 30GHz to 50GHz, may advantageously allow the filter 410 to be conveniently integrated with the MMIC architecture of the amplifier circuit 401.

As will be appreciated by the skilled person, the functionality of the lossy low-pass filter 410 could be readily achieved by a number of alternative circuits.

Referring to Figure 9, in a first exemplary alternative implementation of the lossy -low pass filter 410, the plural inductors L0 to LN of the series line 604 are substituted for transmission line segments TL0 - TLN. Transmission line segments may advantageously have a higher DC current capability than inductors in a MMIC implementation. Accordingly, employing transmission line segments in place of the inductors L0 - LN may advantageously facilitate reliable application of relatively higher DC bias currents to the output line 502 of the amplifier circuit 410.

Referring next to Figure 10, in a second exemplary alternative implementation of the lossy low- pass filter 410, the capacitors 608 of each of the parallel lines 605, 606 is substituted for a stub 1001. The use of stubs 1001 in place of the capacitors 608 in the parallel lines 605, 606 may advantageously improve the filtering of very high-frequency RF energy in the series line 604, whilst exhibiting improved simulation accuracy.

Referring next to Figure 11, in a third exemplary alternative implementation of the lossy low-pass filter 410, the capacitors 608 of each of the parallel lines 605, 606 is substituted for a diode 1101. The diodes 1101 may advantageously protect the amplifier circuit 401 from electrostatic discharge to electrical ground.

Referring finally to Figure 12, the receiver 103 comprises principally a second electro-optical modulator 1201 and a modem 1202.

An optical input of the second electro-optical modulator 1201 is coupled to an opposite end of the optical fibre 104 to the electro-optical modulator 203 of the transmitter 102, such that the optical signal propagating in the optical fibre 104 may be received by the second electro-optical modulator 1201 of the receiver 103. The electro-optical modulator 1201 is operable to transduce a received optical signal back into an electrical signal comprising the original message modulated onto a carrier signal. In the example, the second electro-optical modulator 1201 comprises a semiconductor photo detector. An electrical output of the second electro-optical modulator 1201 is coupled to an input of the modem 1202.

The modem 1202 is operable to receive an electrical signal from the second electro-optical modulator 1201, and demodulate the message from the carrier wave to thereby extract the data from the signal. An output of the modem 1202 may be coupled to further signal processing circuitry, for example, to a client computing device.

References in this specification to “electrical ground” or equivalent are references to a body capable of functioning as an infinite source or sink for electrical charge, and which can absorb an unlimited amount of current without changing its potential, i.e. to a reference voltage. Such a body may, for example, be earth.

Although embodiments of the present invention and its advantages have been described in detail, it should be understood that various changes, substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims. In the claims, the word “comprising” does not exclude other elements or steps, and the indefinite article “a” or “an” does not exclude a plurality.