WO2018228773A1

WO2018228773A1 - A wide dynamic range optical receiver

Info

Publication number: WO2018228773A1
Application number: PCT/EP2018/062919
Authority: WO
Inventors: Colm Donovan; Dylan LOUGHNAN; Ciaran Cahill; Patrick Murphy; John D. LAMBKIN
Original assignee: Firecomms Ltd
Current assignee: Firecomms Ltd
Priority date: 2017-06-13
Filing date: 2018-05-17
Publication date: 2018-12-20
Anticipated expiration: 2019-12-13

Abstract

An optical receiver circuit (100) has a first photodiode (102) and a second diode (103) and an amplifier circuit (104-108) connected to the diodes. The photodiodes are arranged so that the second photodiode receives substantially the same optical signal as the first photodiode but reduced in magnitude. There is a TIA for each channel. The first photodiode is coupled to the input of a first TIA (104), and the second photodiode is connected to the input of a second TIA (105), and the receiver comprises an output (107) linked to the outputs of said TIAs. A switch (106) is connected by separate inputs to the TIAs, and a signal level detection circuit sends a control signal to the switch.

Description

"A Wide Dynamic Range Optical Receiver"

INTRODUCTION Field of the Invention

The invention relates to optical receiver circuits, particularly for use in communications applications.

Prior Art Discussion

In many industrial optical communications applications, receivers must be designed to accept a large range of optical input powers. Receivers must provide a transimpedance amplifier (TIA) with large transimpedance gain to amplify small signals to adequate levels. Receivers must also accept high power light signals without distorting the signals. A standard approach to addressing these competing requirements is to use an automatic gain control (AGC) circuit to reduce the transimpedance gain when large input levels of light are detected, and to increase the transimpedance gain when low levels of light are detected. While this works reasonably well it has some drawbacks. Fig. 1 serves to highlight some of the drawbacks of using AGC circuitry. An optical receiver comprises a photodiode Dl linked to a TIA, in turn linked with a filter and gain circuit. There is an AGC in a feedback loop to the TIA. The first disadvantage of an AGC circuit is the TIA circuit must remain stable over all conditions, and some compromise from the 'optimum' design for low light levels or high light levels may be required. The introduction of an active transistor to form part of the AGC introduces the thermal and shot noise from that transistor to the receiver. The AGC control circuitry connected to the gate of the MOS device if not carefully designed may present noise to the MOS device which will couple into the receiver. In addition, a MOS device such as a PMOS is constructed in an N-Well that needs to be biased with a power supply voltage. This power supply voltage therefore has a capacitive path to couple into the input of the TIA through the PMOS. A further disadvantage is when the AGC reduces the gain of the TIA, the linearity of the receiver may be distorted, which, in a differential TIA, compromises common-mode noise rejection, and may increase pulse- width distortion and jitter. These elements may reduce the sensitivity of the receiver, and reduce the robustness of the receiver to power supply noise - an essential property of robust industrial receivers. It is therefore desirable to design a TIA that can receive a wide range of input light powers without any gain adjustment circuitry which could compromise performance and robustness.

The TIA in [1] uses an AGC circuit to increase the dynamic range of the receiver.

An approach used in [2] describes the use of a signal receiving photodiode and a photometry detecting photodiode. The photometry detecting photodiode is to facilitate the detection of the levels of input light and adjust the operating currents or voltages of an amplifier connected to the signal receiving photodiode.

An approach used in [3] describes the use of an optical power detector, and gain adjusting amplifiers, to facilitate the detection of light power and adjust the gain accordingly.

An invention described in [4] makes use of a dual sensitivity photodetector that comprises of two photodetector types of different sensitivities connected.

An invention described in [5] makes use of concentric photodiodes in a photodetector.

An invention described in [6] makes use of multiple feedback paths to create a form of gain control.

A technique presented in [7] uses either an electrical circuit to simulate a photodiode in an illumination-free state, or uses a photodiode in an illumination-free state. [7] also describes the importance of its illumination-free element in creating an identical differential path. The second receiver element is always illumination-free in [7].

References

[1] Paper: A low noise, wide dynamic range, transimpedance amplifier with automatic gain control for SDH/SONET (STM16/OC48) in a 30GHz f_T

BiCMOS process

Author: M. A. T. Sanduleanu; P. Manteman

Publication: Proceedings of the 27th European Solid-State Circuits Conference.

Year: 2001

Pages: 190 - 193 [2] Patent: US6858830 B2 Feb 22, 2005

Author: Takayuki Suzuki et al

Assignee: Hamamatsu Photonics K.K.

Patent: US7605358 B2 Oct 20, 2009

Author: Takayuki Suzuki et al

Assignee: Hamamatsu Photonics K.K.

Patent: US4366377, Dec 28, 1982

Author: Notthoff et al

Assignee: McDonell Douglas Corporation

Patent: US4724313, Feb 9, 1988

Author: French et al.

The Boeing Company

Patent: US6462327, Oct 8, 2002

Author: Ezell et al.

Assignee: Microtune

Patent: US7406268, Jul 29, 2008

Author: Karl Schrodinger

Assignee: Avago Technologies Limited

Summary of the Invention

We describe an optical receiver circuit comprising:

a first photodiode,

a second photodiode,

wherein the photodiodes are arranged so that the second photodiode receives substantially the same optical signal as the first photodiode but reduced in magnitude, and

an amplifier circuit connected to the photodiodes. „

- 4 -

Preferably:

the amplifier circuit:

comprises at least one TIA coupled to said photodiodes, and

provides a sensitive first channel including the first photodiode and an overload second channel including the second photodiode, and

the receiver circuit comprises:

an output terminal coupled to outputs of said TIA or TIAs, and

a switch arranged to select a channel to provide an output at the output terminal. The amplifier circuit may comprise a TIA uniquely associated with each photodiode, and the outputs of the TIAs may be coupled to separate inputs of the switch.

The receiver circuit may comprise a single TIA which is shared between the first and second photodiodes and the switch is linked between the photodiodes and the TIA.

Preferably, the amplifier circuit comprises a second stage amplifier arranged to apply gain and filtering required by any subsequent stages.

Preferably, the amplifier circuit comprises a control circuit for controlling said switch. Preferably, the output of said second stage amplifier is coupled to an input of the control circuit.

Preferably, the output of said TIA or TIAs is coupled to an input of the control circuit. The control circuit may comprise a signal level detection circuit adapted to monitor signals in said channels. Preferably, the signal level detection circuit comprises a peak detector to provide signal level detection. Preferably, the control circuit comprises decision circuitry which, based on the received optical power, selects the appropriate photodiode for any given power. Preferably, the signal level detection circuit is configured to instruct the selection switch to choose a channel that remains undistorted. The control circuit may comprise a detection circuit arranged to determine a threshold at which the receiver circuit selects a channel based on comparing an output (e.g. RSSI) current with a reference current. Preferably, the reference current is programmable. Preferably, the detection circuit is arranged to select an alternative channel if the output current exceeds the reference current. The detection circuit may be adapted to choose a new reference current to implement a form of hysteresis which prevents oscillating the selection between the channels.

Preferably, the control circuit is configured to disable channel selection for a finite period of time after either a channel change or signal detection, which prevents unnecessary channel selection changes.

Preferably, the control circuit is configured to alter the reference current after a finite period of time after a channel change or signal detection, which prevents unnecessary channel selection changes.

Each channel may be adapted to perform differential signalling.

Preferably, said first channel includes said first photodiode and a first non-active photodiode, and the second channel comprises said second photodiode and a second non-active photodiode, each channel thereby having differential signalling. Preferably, within each channel the photodiodes are identical and are coupled to identical TIAs, by a direct physical connection. The receiver circuit may further comprise a control circuit linked to inputs of said TIA or TIAs to provide a feedback current to force a balanced differential output.

Preferably, the control circuit comprises an integrator arranged to provide said feedback current that is proportional to a low-frequency component of the current generated in the first or second photodiode in response to an input light signal. Preferably, a scaled copy of said feedback current from the integrator is used to provide signal level detection.

Preferably, the control circuit is adapted to couple the TIA outputs to inputs of a filter amplifier in order to receive two pairs of inputs and provide one pair of outputs, which connect a differential signal to the filter amplifier. Preferably, the non-active photodiodes are located out of range of an input optical signal, and so receive light only in the event of a light flash event, and an amplifier circuit is linked to said non- active photodiodes and is configured to perform light flash event compensation or electromagnetic interference (EMI) compensation. _r

- 6 -

The non-active photodiodes may be located close to the active photodiodes to receive the same light signal but at a lower intensity, to not compromise the differential signalling, and to improve light flash event compensation or to improve EMI compensation. Preferably, the first photodiode is at least partly surrounded by the second photodiode. The first photodiode may be approximately circular. The second photodiode may be annular, surrounding the first photodiode.

The receiver circuit may further comprise a guard-ring between the first and second photodiodes and configured to prevent formation of parasitic devices and to reduce the amount of any diffusion current generated outside of a depletion region of the photodiodes. Preferably, the guard-ring comprises a reverse biased diode consisting of an N-type implant in a grounded P- type substrate to form a depletion region guard-ring. The receiver circuit may further comprise a comparator arranged to provide a digital signal output.

Preferably, at least one photodiode and the amplifier circuit are integrated onto the same silicon substrate.

We also describe an optical receiver comprising a receiver circuit of any embodiment, and an optical element for directing light to the photodiodes. Preferably, the optical element comprises an optical fibre of numerical aperture of 0.3 or greater. Preferably, the optical fibre has a diameter of 0.5mm or greater. The optical element may include a plastic optical fibre.

The optical element may include a lens between the end of a fiber optic and a photodiode to focus light onto the photodiode. Preferably, the optical element is arranged to direct the light to illuminate an area larger than the first and second photodiodes.

The optical receiver circuit may receive more than two channels, in which the third and any subsequent channel comprises a photodiode arranged to receive less light than the preceding channels.

Additional Statements

According to the invention, there is provided an optical receiver circuit comprising a first photodiode and a second diode and an amplifier circuit connected to the diodes, wherein the photodiodes are arranged so that the second photodiode receives substantially the same optical signal as the first photodiode but reduced in magnitude.

In one embodiment, the receiver circuit comprises a sensitive channel linked with the first photodiode and an overload channel linked with the second photodiode.

In one embodiment, the amplifier circuit comprises a TIA uniquely associated with each photodiode. Preferably, the first photodiode is coupled to the input of a first TIA, and the second photodiode is connected to the input of a second TIA, and the receiver comprises an output linked to the outputs of said TIAs. In one embodiment, the receiver further comprises a switch, and the outputs of the TIAs are connected to separate inputs of the switch, and the receiver further comprises a signal level detection circuit adapted to send a control signal to the switch. In one embodiment, the switch comprises at least one single-pole double-throw type (SP-DT) switch. In one embodiment, the output of said switch is coupled to the input of a second stage amplifier arranged to apply gain and filtering required by any subsequent stages.

In one embodiment, the receiver further comprises a signal level detection circuit. This may for example be coupled to the output of said second stage amplifier, or coupled to the TIA. In one embodiment, said signal level detection circuit has a digital output that controls the switch to choose which photodiode channel is coupled to the second stage amplifier.

Preferably, the signal level detection circuit is configured to choose a channel with the highest level of output signal that remains undistorted. In one embodiment, the amplifier circuit comprises a single physical TIA and a multiplex switch for routing signals from said first and second photodiodes into said single TIA, whereby only a single TIA is associated with a photodiode at any one time.

In one embodiment, the receiver further comprises a signal level detection circuit coupled to the output of said second stage amplifier and to said multiplex switch.

In one embodiment, the receiver circuit is adapted to perform differential signalling, in which said first photodiode is in the sensitive channel with a non-active photodiode, and the second photodiode is in the overload channel with a non-active photodiode, each channel thereby having differential signalling. - o -

In one embodiment, within each channel the photodiodes are identical and are coupled to identical TIAs, by a direct physical connection.

In one embodiment, the circuit further comprises a control circuit connected to the TIA or filter outputs and linked back to TIA inputs to provide a feedback current to force a balanced differential output. In one embodiment, the control circuit is also linked to a switch to provide a switch control signal for delivering signals from the TIAs of both channels to a filter and gain circuit. In one embodiment, the control circuit comprises decision circuitry which, based on the received optical power, selects the optimum photodiode for any given power. Preferably, the control circuit is adapted to couple the TIA outputs to the filter amplifier inputs in order to receive two pairs of inputs and provide one pair of outputs, which connect a differential signal to the filter amplifier.

In one embodiment, the non-active photodiode of each channel is located out of range of an input optical signal, and so receives light only in the event of a light flash event, and an amplifier circuit is linked to said additional photodiode and is configured to perform light flash event compensation or electro-magnetic interference (EMI) compensation.

In one embodiment, the non-active photodiode is located close to the active photodiode so as to receive the same light signal but at a lower intensity, so as to not compromise the differential signalling, and to improve light flash event compensation or to improve EMI compensation.

In one embodiment, the circuit further comprises an integrator arranged to provide a current that is proportional to a low-frequency component of the current generated in the input photodiode in response to an input light signal.

Preferably, the circuit further comprises a detection circuit arranged to determine a threshold at which the receiver circuit selects an overload TIA rather than a sensitive TIA, based on comparing an output (e.g. RSSI) current with a reference current.

In one embodiment, the detection circuit is arranged to select the overload TIA if the output current exceeds the reference current. In one embodiment, the detection circuit is adapted to choose a new reference current to implement a form of hysteresis which prevents oscillating the selection between the TIAs. In one embodiment, the circuit comprises a guard-ring, such as a reverse biased diode consisting of an N-type implant in a grounded P-type substrate to form a depletion region guard-ring.

In one embodiment, the control circuit performs peak detection to provide the feedback current.

In one embodiment, the circuit further comprises an integrator arranged to provide a current that is proportional to a low-frequency component of the current generated in the input photodiode in response to an input light signal. In one embodiment, the reference current is programmable. In one embodiment, the circuit further comprises a comparator providing a digital signal to drive a differential driver output circuit.

DETAILED DESCRIPTION OF THE INVENTION

Brief Description of the Drawings

The invention will be more clearly understood from the following description of some embodiments thereof, given by way of example only with reference to the accompanying drawings in which :-

Fig. 1 is a diagram showing a prior art AGC of a receiver;

Fig. 2 is a diagram showing a dual photodiode arrangement of a receiver of the invention;

Fig. 3 is a plot showing TIA ranges and hysteresis versus light power received from Plastic Optical Fibre (POF);

Figs. 4, 5 and 6 are circuit diagrams of optical receivers of three embodiments; and

Fig. 7 is a diagram showing EMI rejection. Description of the Embodiments

We describe an optical receiver including optics and a receiver circuit which has at least two photodiodes, arranged to receive different levels of input signal light, possibly generated by modulating a light-emitting diode with an optical transmitter circuit. One photodiode is arranged to receive less light intensity or magnitude than the other. The receiver circuit amplifies the photodiode currents. Because the intensity of light received could be either very small or very large it is difficult to provide an amplifier that will cover all ranges. As previously mentioned an AGC is not desired because of various drawbacks. Switching between the photodiode inputs provides a solution to achieve a range of optical powers.

A trans-impedance amplifier (TIA) converts a current input signal to a voltage output signal. There could equally be an individual TIA with tuned gain for each photodiode, or if it is feasible for the application, one physical TIA and a switch is used to determine which photodiode provides the input. Equally if an individual TIA is used for each photodiode an individual gain or filter circuit or comparator circuit could follow for each photodiode. Or for optimal circuit reuse some other combination of individual channel circuits and common circuits could be used. Preferably a filter or gain circuit as well as a comparator would constitute shared circuitry rather than having a filter, gain or comparator circuit uniquely associated with each photodiode.

In one embodiment shown in Fig. 2, an optical receiver 1 has a first photodiode 2 of an approximately circular or disc shape, and a second photodiode 3 is an annular photodiode surrounding the first photodiode. These are fabricated on a semiconductor substrate 4. A plastic optic fiber (POF) 6 emits light via a focusing lens 5.

The photodiodes 2 and 3 are arranged beneath a lens 5, which focuses light from the plastic optic fibre 6. The signal light from the plastic optic fibre 6 illuminates an area larger than the photodiodes 2 and 3. The second photodiode 3 surrounding the first photodiode 2 receives substantially less signal light than the first photodiode, but still receives enough signal light in high light power situations, for example greater than -20 dBm of light, to provide an adequate output current signal. The first photodiode is coupled to a TIA 8 that does not have an AGC. The TIA 8 is designed for optimal sensitivity. This TIA 8 can amplify very low light levels with adequate signal-to-noise ratios, for example signals levels such as -30dBm. However, this TIA may overload for higher light powers, and may fail to correctly amplify the input signal. This TIA may not be suitable on its own for applications that need to receive light power with a wide dynamic range. The second photodiode 3 is also coupled to a TIA 9 that does not have an AGC. The second photodiode will output less current, for example a lOdB reduction in current, compared to the first photodiode for the same light signal transmitted through the POF. If, for example, similar TIAs are connected to both photodiodes, then the total range of light powers that the receiver can process is extended by approximately lOdB from the case with only the first TIA. The TIAs and the transimpedance gains may vary between TIAs depending on the optimisation required.

The term "channel" is used to describe the signal path, differential or single-ended, from the photodiodes to the output of the receiver. A signal level detection circuit monitors the output of the amplifier circuit and controls via the switch which photodiode signal is transmitted to the output.

For example, in Fig. 4 there are two single-ended channels each consisting of a photodiode and dedicated TIA, a switch and a filter and gain circuit common to either channel. In Fig. 5 each channel constitutes a photodiode together with a shared TIA and filter and gain circuit.

Furthermore, the terms sensitive and overload channels are used descriptively to differentiate between the first and second channel. The overload channel is the channel which is envisioned to operate with high levels of input light to the overall receiver. In Fig. 2 the annular outer photodiode (3) is the input to the overload channel. The sensitive channel is the channel which is envisioned to operate with low levels of input light to the overall receiver. In Fig. 2 the circular inner photodiode (2) is the input to the sensitive channel. When the sensitive channel receives too much light to function without distortion and would 'overload', then the overload channel would be functional and would be used. It is preferred that the sensitive channel photodiode be arranged to receive more light than the overload channel photodiode.

As a further illustration of the terms channel, sensitive, and overload, Fig 6 shows an example of two differential channels, one sensitive channel, one overload channel. The sensitive channel constitutes two photodiodes (141, 142) and two TIAs (145, 146). Similarly, the overload channel constitutes two photodiodes (143, 144) and two TIAs (147, 148).

An example is shown in Fig. 3, where the sensitive TIA channel (the inner photodiode) is functional for the range of powers from -30dBm to -lOdBm, and the overload TIA channel (the outer photodiode) is functional for the range of powers from -20dBm to OdBm. There is significant overlap where both channels are potentially functional. This overlap serves to help the decision circuitry avoid oscillations between the two channels in a case where the average input light power experiences fluctuations due to signal content, or changes in the physical transmission medium, such as movement of the fiber or inconsistencies in the transmitter. This overlap can be reduced or removed if necessary to extend the operational range. It is preferred that there is some overlap between the operational light levels of the sensitive and overload channels. This overlap can be used to form a type of hysteresis between the selection of one channel or the other. For example, in Fig. 3, the sensitive channel may receive increasing light levels up until the light exceeds lOdBm, at which point the overload channel will be selected. The overload channel may then receive reducing light levels until less than -20dBm of light is received, at which point the sensitive channel will be selected. Thus, a hysteresis range from - lOdBm to -20dBm is formed.

It is possible to have a third photodiode which receives less light than the second photodiode, to form a third channel and further increase the receivable light range.

Plastic Optical Fiber (POF) is a class of optical fiber, generally made from Poly Methyl Methacrylate (PMMA), with a large core diameter (0.5mm to 2 mm) and high numerical aperture (approximately 0.3 to 0.6). Two characteristics of POF are taken advantage of in this invention. Firstly, the high NA results in relatively greater dispersion of light. Secondly, the large core is of a relatively large diameter. As a result of these characteristics POF tends to create an illuminated area over a receiver that is large relative to other types of optical fibers, such as glass. It will be appreciated that the invention in this embodiment uses the property of light dispersion from wide diameter Plastic Optical Fibre to utilize the over-illumination of a photodiode, located directly beneath the path of light from the POF, possibly through a lens, and uses a second photodiode to recover some of this 'over-illumination' light. It uses this property advantageously to improve the architecture of a photodiode receiver and remove the need for an AGC while providing an architecture capable of a receiving a wide range of powers.

As shown in Fig. 4. A receiver 100 compromises:

a first photodiode 102 arranged to receive signal light,

a second photodiode 103 arranged to receive proportionally less signal light, - a first TIA 104 to receive the output of the first photodiode,

- a second TIA 105 to receive the output of the second photodiode,

- a switching circuit 106 to receive as inputs the TIA outputs, and whose purpose is to

switch its output to select one of the TIA inputs,

- a filter and gain amplifier 107 to receive the output from the switching circuit,

- a control circuit 108 with a received signal strength indicator (RSSI) circuit along with some logic to detect the signal strength and provide the means to select the optimum photodiode-TIA channel for any given level of light.

Referring now to Fig. 5, a receiver 130 comprises:

- a first photodiode 131 arranged to receive signal light,

- a second photodiode 132 arranged to receive proportionally less signal light,

- a switching circuit 133 to receive as inputs the photodiode outputs, and whose purpose is to switch its output to select one of the photodiode inputs,

- a TIA 134 to receive the output of the switching circuit,

- a filter amplifier 135 to receive an input from the output of the transimpedance amplifier,

- a control circuit 137 with a received signal strength indicator (RSSI) circuit with a current or voltage output to determine the level of output light based on the voltage output from the filter amplifier.

The signal strength detection circuit provides information for a switching circuit to connect the optimal photodiode for a given light power to the TIA.

A benefit of the single-TIA architecture of Fig. 5 over the arrangement of Fig. 4 is using only one TIA rather than two TIAs.

The dual-TIA arrangement of Fig. 4 has two benefits over the arrangement of Fig. 5. Firstly, the switches 133 in the single-TIA arrangement can act as a source of noise and increased input impedance in the input signal path and these are removed. Secondly, the dual-TIA arrangement allows optimisation of the TIA and its gain for the specific photodiode to which it is connected. The designer can control the output signal voltage level as a function of both the transimpedance gain and the photodiode coupling efficiency in this arrangement, and may better control the range of powers for which the receiver will function correctly within specifications. _Λ„

- 14 -

In one embodiment, each channel comprises a photodiode, and a TIA uniquely associated with each photodiode, and a filter circuit, either active or passive, connected to the output of each TIA. In one embodiment the receiver circuit is designed to further compensate for levels of light received by each photodiode by using different TIAs with different gains and bandwidths.

Industrial receivers demand high levels of robustness to disturbances, such as power supply noise and electro-magnetic interference (EMI). One of the key techniques in the reduction of the effect of these disturbances is the use of a differential receiver. If a differential receiver is designed with passive transimpedance components, as in this invention, it will create transimpedance gain that does not vary with light input power, such as in an AGC receiver, and will offer greater robustness to disturbances such as EMI. Another advantage with passive transimpedance components, is that the linearity of the receiver is maintained, which can be important in pulse amplitude modulation (PAM) optical receivers.

Referring to Fig. 6, this is a simplified block diagram of a receiver 140 with a further enhancement to the architecture which uses differential signalling. Each channel consists of two identical photodiodes (141, 142 and 143, 144) coupled to two identical TIAs (145, 146 and 147, 148). One photodiode in each of these pairs consists of light receiving photodiodes (141, 143) as previously described, while the other non-active photodiode in each pair (142, 144) receives little or no signal light. Non-active in this situation simply means that very little light reaches the photodiodes, but they are otherwise identical to the light-receiving pair of photodiodes. There is an active photodiode 141 (circular) and non-active photodiode 142 (circular) for the sensitive channel. Similarly, there is an active photodiode 143 (annular) and non-active photodiode 144 (annular) for the overload channel. The photodiodes 141 to 144 are connected to the inputs of TIAs 145 to 148 respectively, the outputs of which are linked to a switch 149. The switch 149 is connected at its output to a differential filter and gain circuit 150, in turn connected to a comparator 151. A control circuit 152 is linked back to the TIA inputs to provide a feedback circuit to force the low-frequency average outputs from the filter to the same voltage to provide a balanced differential output. The control circuit 152 is also linked to the switch 149 to provide a switch control signal. The control circuit 152 contains decision circuitry which, based on the received optical power, will select the optimum photodiode for any given power. This architecture therefore consists of, four photodiodes, and four TIAs, organized in two pairs of differential signal channels. The switching to couple the TIA outputs to the filter amplifier inputs must be amended to receive two pairs of inputs and provide one pair of outputs, which connect a differential signal to the filter amplifier.

This embodiment may also employ the previously described technique of the receiver shown in Fig. 5, in which the inner and outer photodiode may be coupled through switches to a single TIA for each side of the differential channels. The channel selection switches require an input to select which input channel to use. This may be as simple as an external user digital input, which is simply set depending on the application. An integrated circuit with the present invention may use a programmable fuse to make the TIA selection. Another option is a wire bond from a pad that selects a particular digital input depending on the intended application.

However, the present invention preferably uses an internal control circuit that automatically chooses an input channel depending on the light received. For a control circuit to perform this task it would need some information about the input signal based on the output of a TIA or a subsequent amplification stage. One method of determining the input power is a Received Signal Strength Indicator (RSSI) circuit. There are many variations of such a circuit. The RSSI generation circuit (137) of Fig. 5 could be generated for example with a peak detector.

The RSSI generation circuit (152) of Fig. 6 could be generated for example with an integrator circuit that forces the low frequency components of the differential outputs of the differential amplifier (150) to be equal.

The output of the integrator may be a current that is proportional to the low-frequency component of the current generated in the input photodiode in response to an input light signal manner. A copy of this current would be representative of the received signal strength and could serve as an RSSI. An integrator circuit in this situation advantageously performs two tasks at once. It provides a feedback current to the TIA inputs to create a balanced differential output from the second stage amplifier. It also provides a scaled copy of this feedback current as an RSSI current for the decision circuit. The feedback connections in Fig. 6 are shown connecting to the active TIAs but could equally connect to the non-active TIAs (with a change in polarity). In order to determine the threshold at which the circuit selects the overload TIA rather than the sensitive TIA, a copy of this RSSI current could be compared to a reference current. When the RSSI exceeds this reference current the overload TIA would be selected. The RSSI from this overload TIA will be lower than that from the sensitive TIA, for example, by lOdB. A new reference current must be used to decide when to switch from the overload TIA back to the sensitive TIA. In this situation the switching occurs if the output current drops below the new reference current rather than exceeding it (i.e. it exceeds it in a negative direction). This new reference current can be chosen to implement a form of hysteresis which prevents oscillating the selection between the TIA. Preferable, both reference currents would be programmable. It will be noted that for a short period of time after a channel change by the decision circuitry some pulse-width-distortion of the incoming signal may occur. This is undesirable, therefore unnecessary channel changes should be avoided. A timer circuit can be used to disable the decision circuitry for a finite period of time after changing between TIAs. This ensures that no further TIA selection changes occur to give the circuit adequate time to settle before the decision circuitry again becomes enabled. Another refinement could be that when light is initially detected, one particular threshold current level be used to choose the overload level. The danger is that the signal level could be very close to this overload level, and on the verge of exceeding the overload level and may do so due to small perturbations in the signal. Therefore, after a finite period of time, if the decision circuitry determines that no further TIA change is necessary, a new threshold level can be chosen, with the purpose of further preventing unnecessary TIA changeovers due to the signal initially having been close to the threshold.

The non-active photodiode is a light-receiving photodiode located close to the active photodiode on the substrate. The non-active photodiode is identical in construction to the active photodiode. Because of its location, it will receive a small amount of the incoming signal light. It is therefore an illuminated photodiode that receives a much smaller amplitude signal than the active photodiode. However, because the signal light the non-active photodiode receives is significantly less than that of the active photodiode, the differential signal architecture functions with only minimal reduction of differential signal which will have little or no practical impact to performance. A benefit of the non-active photodiode being a light receiving photodiode is twofold. Firstly, the alternative to an identical photodiode is a 'dark' photodiode, which may have metal or an alternative material covering the photodiode. This metal forms a parasitic capacitive load to the photodiode. In a differential architecture, it is important that both differential channels are identical in order to maximise common mode rejection. The 'dark' photodiode compromises this because it does not present an identical electrical load to the active photodiode.

Secondly the non-active light receiving photodiode can help immunity to Electro-Magnetic Interference (EMI) disturbances, in which electro-magnetic signals, including unwanted light, may reach the photodiodes through paths other than through the POF and focusing lens. Referring to Fig. 7 there are two inner photodiodes 141, 142, and corresponding annular outer photodiodes 143, 144, all on a substrate 163. Light is directed by a POF 164 via a lens 165. There are TIAs 145, 146, 147, and 148 for the photodiodes (respectively, 141, 142, 143, 144). In the case of a light- flash disturbance (173) with a darkened photodiode, only the active photodiode receives light signal, and thus there is no common mode rejection, and false signalling occurs. Similarly, in the event of an EMI disturbance the darkened photodiode, due to its different electrical construction, will experience a different disturbance to the active photodiode and compromise the common-mode rejection. In the present invention, the non- active photodiode is also a light receiving photodiode that is located near the active photodiode. In the event of an EMI disturbance a significant amount of unwanted signal reaches both the active and non-active photodiode. Thus, unlike the case where the non-active photodiode is darkened, the present invention will generate a common mode signal, and it will better reject such a common-mode disturbance. While the photodiodes have been described as approximately circular, integrated circuit lithography usually only allows a limited range of angles for drawing shapes. A square or octagon for example, may in some cases approximate the circle required. The annulus- shaped photodiode may similarly require approximation with the use of limited angles. There is a danger that a parasitic bipolar device can form between the inner and outer photodiodes in an integrated circuit. For example, the n-cathode of the inner photodiode, the n- cathode of the outer photodiode, and the p-anode of the area in between may form an n-p-n bipolar device. The layout of these photodiodes may advantageously contain guard-rings surrounding the photodiodes, and guard-rings in between the inner and outer photodiodes. The - 1δ - purpose of the guard-rings is to prevent the formation of possible parasitic devices. Because this invention is directed towards the case of the signal light over-illuminating the photodiode area, diffusion currents, currents that are generated outside the depletion region of the active photodiode area, are expected. This guard-ring also reduces the amount of diffusion current from reaching the active photodiodes that would otherwise result in an unwanted low bandwidth component in the signal received. There are many options for the construction of a guard-ring, such as a reverse biased diode consisting of an n-type implant in a grounded p-type substrate to form a depletion region guard-ring. There are many methods of constructing a guard-ring available to one skilled in the arts.

In one embodiment, each channel is differential. Such an embodiment is more robust to common mode disturbances such as the flash of a camera or an electrically generated spark. Each differential channel consists of one active photodiode and one non-active photodiode. The non-active photodiode provides light flash event compensation or electro-magnetic interference (EMI) compensation in the form of common-mode cancelation. An electrical spark is an example of an event that can cause both electromagnetic interference and optical light interference (a 'light flash event').

It may be that no filtering is required from the second stage. It may also be that the second stage gain requirements are sufficiently lax that no feedback is required in the second stage amplifier. This may be beneficial in creating a high-speed amplifier as stability requirements can be relaxed without the presence of feedback. It may be that this second stage amplifier is a limiting amplifier, or a comparator. It may be that this second stage amplifier simply acts as a buffer. It may be that this second stage amplifier has a second pair of inputs to act as an offset adjustment input.

In one embodiment there are three or more channels. This may facilitate increasing the range of receivable light levels. Where there are more than two channels, the third and any subsequent channel preferably comprises a photodiode arranged to receive less light than the preceding channels.

The invention is not limited to the embodiments described but may be varied in construction and detail.

Claims

An optical receiver circuit comprising:

a first photodiode (102, 131, 141),

a second photodiode (103, 132, 143),

wherein the photodiodes are arranged so that the second photodiode receives substantially the same optical signal as the first photodiode but reduced in magnitude,

an amplifier circuit connected to the photodiodes, wherein the amplifier circuit:

comprises at least one TIA (104, 105, 134, 145-148) coupled to said photodiodes, and

provides a sensitive first channel including the first photodiode and an overload second channel including the second photodiode,

an output terminal (109, 136, 153) coupled to outputs of said TIA or TIAs, and a switch (106, 133, 149) arranged to select a channel to provide an output at the output terminal.

An optical receiver circuit as claimed in claim 1, wherein the amplifier circuit comprises a TIA (104,105,145-148) uniquely associated with each photodiode, and the outputs of the TIAs (104,105,145-148) are coupled to separate inputs of the switch (106, 149).

An optical receiver circuit as claimed in claim 1, wherein the receiver circuit comprises a single TIA (134) which is shared between the first and second photodiodes (131, 132) and the switch (133) is linked between the photodiodes (131, 132) and the TIA (134).

An optical receiver circuit as claimed in any of claims 1 to 3, wherein the amplifier circuit comprises a second stage amplifier (107, 135, 150) arranged to apply gain and filtering required by any subsequent stages.

An optical receiver circuit as claimed in any preceding claim, wherein the amplifier circuit comprises a control circuit (108, 137, 152) for controlling said switch (106, 133, 149).

6. An optical receiver circuit as claimed in claim 5, wherein the output of said second stage amplifier (107, 135, 150) is coupled to an input of the control circuit (108, 137, 152).

7. An optical receiver circuit as in claims 5 or 6, wherein the output of said TIA or TIAs is coupled to an input of the control circuit (108, 137, 152).

8. An optical receiver circuit as claimed in claim 5 to 7, wherein and the control circuit (108, 137, 152) comprises a signal level detection circuit adapted to monitor signals in said channels.

9. An optical receiver circuit as claimed in claim 8, wherein the signal level detection circuit comprises a peak detector to provide signal level detection.

10. An optical receiver circuit as claimed in claims 5 to 9, wherein the control circuit (108, 137, 152) comprises decision circuitry which, based on the received optical power, selects the appropriate photodiode for any given power.

11. An optical receiver circuit as claimed in any of claims 8 to 10 wherein the signal level detection circuit is configured to instruct the selection switch to choose a channel that remains undistorted.

12. An optical receiver circuit as claimed in any of claims 5 to 11 claim, wherein the control circuit comprises a detection circuit arranged to determine a threshold at which the receiver circuit selects a channel based on comparing an output (e.g. RSSI) current with a reference current.

13. An optical receiver circuit as claimed in claim 12, wherein the reference current is programmable.

14. An optical receiver circuit as claimed in claims 12 or 13, wherein the detection circuit is arranged to select an alternative channel if the output current exceeds the reference current.

15. An optical receiver circuit as claimed in claim 14, wherein the detection circuit is adapted to choose a new reference current to implement a form of hysteresis which prevents oscillating the selection between the channels.

16. An optical receiver circuit as claimed in any of claims 5 to 15, wherein the control circuit is configured to disable channel selection for a finite period of time after either a channel change or signal detection, which prevents unnecessary channel selection changes.

17. An optical receiver circuit as claimed in any of claims 5 to 16, wherein the control circuit is configured to alter the reference current after a finite period of time after a channel change or signal detection, which prevents unnecessary channel selection changes.

18. An optical receiver circuit as claimed in any preceding claim, wherein each channel is adapted to perform differential signalling.

19. An optical receiver circuit as claimed claim 18, in which said first channel includes said first photodiode and a first non-active photodiode, and the second channel comprises said second photodiode and a second non-active photodiode, each channel thereby having differential signalling.

20. An optical receiver circuit as claimed in claim 19, wherein within each channel the photodiodes are identical and are coupled to identical TIAs, by a direct physical connection.

21. An optical receiver circuit as claimed in claims 18 to 20, further comprising a control circuit (152) linked to inputs of said TIA or TIAs to provide a feedback current to force a balanced differential output.

22. An optical receiver circuit as claimed in claim 21, where the control circuit comprises an integrator arranged to provide said feedback current that is proportional to a low- frequency component of the current generated in the first or second photodiode in response to an input light signal.

23. An optical receiver circuit as claimed in claim 22, wherein a scaled copy of said feedback current from the integrator is used to provide signal level detection.

24. An optical receiver circuit as claimed in any of claims 18 to 23, wherein the control circuit (152) is adapted to couple the TIA outputs to inputs of a filter amplifier (150) in order to receive two pairs of inputs and provide one pair of outputs, which connect a differential signal to the filter amplifier (150).

25. An optical receiver circuit as claimed in claims 19 to 24, wherein the non-active photodiodes (142, 144) are located out of range of an input optical signal, and so receive light only in the event of a light flash event (173), and an amplifier circuit is linked to said non-active photodiodes and is configured to perform light flash event compensation or electro -magnetic interference (EMI) compensation.

26. An optical receiver circuit as claimed in any of claims 19 to 24, wherein the non-active photodiodes are located close to the active photodiodes to receive the same light signal but at a lower intensity, to not compromise the differential signalling, and to improve light flash event compensation or to improve EMI compensation.

27. An optical receiver circuit as claimed in any preceding claim, wherein the first photodiode is at least partly surrounded by the second photodiode.

28. An optical receiver circuit as claimed in claim 27, wherein the first photodiode is approximately circular.

29. An optical receiver circuit as claimed in claim 27 or 28, wherein the second photodiode is annular, surrounding the first photodiode.

30. An optical receiver circuit as claimed in any preceding claim, further comprising a guard- ring between the first and second photodiodes and configured to prevent formation of parasitic devices and to reduce the amount of any diffusion current generated outside of a depletion region of the photodiodes.

31. An optical receiver circuit as claimed in claim 30, wherein the guard-ring comprises a reverse biased diode consisting of an N-type implant in a grounded P-type substrate to form a depletion region guard-ring.

32. An optical receiver circuit as claimed in any preceding claim, further comprising a comparator (151) arranged to provide a digital signal output. An optical receiver circuit as claimed in any preceding claim, wherein at least one photodiode and the amplifier circuit are integrated onto the same silicon substrate.

An optical receiver circuit as claimed in any preceding claim, comprising more than two channels, in which the third and any subsequent channel comprises a photodiode arranged to receive less light than the preceding channels.

An optical receiver comprising a receiver circuit of any preceding claim, and an optical element for directing light to the photodiodes.

An optical receiver as claimed in claim 35, wherein the optical element comprises an optical fibre of numerical aperture of 0.3 or greater.

An optical receiver as claimed in claim 35 or 36, wherein the optical fibre has a diameter 0.5mm or greater.

An optical receiver as claimed in any of claims 35 to 37, wherein the optical element includes a plastic optical fibre.

An optical receiver as claimed in any of claims 35 to 38, wherein the optical element includes a lens between the end of an fiber optic and a photodiode to focus light onto the photodiode. 40. An optical receiver as claimed in claims 35 to 39, wherein the optical element is arrang to direct the light to illuminate an area larger than the first and second photodiodes.