Patent Application
20060115280
In optical systems where data is transmitted through optical fibers, a demand for higher data transfer rate requires that optical receiver systems be designed with a wide bandwidth. In order to increase bandwidth in an optical receiver system, efforts have been made to increase the bandwidth of a transimpedance amplifier (TIA) within the optical receiver system. The bandwidth of the TIA is increased, for example, by use of more expensive, high performance process technology, by reducing the input resistance to the TIA and/or by employing new architectures, such as common base architecture, for the implementation of TIA in order to avoid the Miller effect. Bandwidth of the optical receiver system can also be increased by reducing parasitic capacitance in a photo-detector that is used to convert light signals to electrical signals.
Each of the above-discussed ways used to increase bandwidth in an optical receiver system has disadvantages and/or tradeoffs. For example, using more expensive, high performance technology to implement the TIA can significantly increase the cost of the optical receiver system. Reducing the input resistance to the TIA requires use of a bigger transistor size resulting in increased power consumption. Use of architectures, such as common base amplifier architecture, to avoid the Miller effect results in noise degradation in the TIA. Reducing parasitic capacitance in the photo-detector requires a decreased photo-detector size resulting in alignment issues and other bonding issues.
Additionally, as data transfer rate increases, additional issues arise which can result in degradation of reception quality. For example, the intrinsic non-linear characteristics of light-emitting devices used to transmit signals through the optical fiber can result in eye quality degradation of the established optical link. The intrinsic non-linear characteristics include relaxation oscillation or slow tail of falling signal. Such behaviors of light-emitting device characteristics are not simple to compensate for in the optical receiver system. Traditionally, low pass filtering techniques are used to filter out the relaxation oscillation frequency component of light emitting diodes. However, such low pass filtering techniques conflict with efforts to increase bandwidth in receiver systems. Designers have also tried to design light-emitting devices so as to speed up the falling edge of optical signal to reduce the effect of slow tail. However, designing light-emitting devices with these two contradicting design constraints requires a lot of effort, increases development time and provides diminishing returns in signal quality.
In accordance with an embodiment of the present invention, an optical receiver system includes an amplifier circuit and a compensation circuit. The amplifier circuit includes a light detector, and a transimpedance amplifier. The transimpedance amplifier produces an amplified signal. The compensation circuit includes at least one pole compensation stage that performs pole compensation on the amplified signal.
Amplifier circuit 11 includes a transimpedance amplifier (TIA) 16 and a feedback resistor 15. Amplifier circuit 11 also includes a photo-detector 14 connected to VCC 13 as shown. Photo-detector 14 detects light signals 12. The resulting electrical signal produced by photo-detector 14 is amplified by TIA 16.
Compensation circuit 21 provides pole compensation outside of TIA 16. Since TIA 16 typically consists of three poles, three compensating stages can be used. However, compensation of major two poles (input related and output related), requiring only two compensating stages, is often enough to achieve a desired bandwidth of operation.
In the implementation of compensation circuit 21 shown in
Variable control input 22, variable control input 23 and variable control input 24 can be used for weight adaptation. This is useful, for example, because TIA 16 can have more complex AC responses by having more than 3 poles. For example, an additional pole or poles can result when the emitter parasitic capacitance of input TIA transistors generates a zero in TIA bandwidth performance creating somewhat unpredictable behavior of TIA 16. Weight adaptation performed with variable control input 22, variable control input 23 and variable control input 24 can also be used to compensate for conventional interface issues such as intersymbol interference (ISI) or skin effect that occur in additional circuitry of optical receiver system 10.
Arranging compensation stages in series results in less output parasitic of each stage. Alternatively, to make it easier to control weights, the compensation can be arranged in a parallel configuration, as shown in
Amplifier circuit 31 includes a transimpedance amplifier (TIA) 36 and a feedback resistor 35. Amplifier circuit 31 also includes a photo-detector 34 connected to VCC 33 as shown. Photo-detector 34 detects light signals 12. The resulting electrical signal produced by photo-detector 34 is amplified by TIA 36.
Compensation circuit 41 provides pole compensation outside of TIA 36. Since TIA 36 typically consists of three poles, three compensating stages can be used. However, compensation of major two poles (input related and output related), requiring only two compensating stages, is often enough to achieve a desired bandwidth of operation.
In the implementation of compensation circuit 41 shown in
An output signal quality monitoring feature, which can be implemented in CDR and decision block 40 is used to control variable control input 42, variable control input 43 and variable control input 44.
The use of a compensation circuit as illustrated in
The technology used to implement the compensation circuit typically is the same used to implement the TIA based amplifier circuit, provided the bandwidth of each compensation stage covers maximum operational frequency required by equalization. Also, power consumption required for the compensation circuit is much less than power consumption resulting from an increase of TIA stage bandwidth. Additionally, the simple structure of the compensation stages leads to less time required for the development of the optical receiver system. Also, the compensation circuitry can be designed without affecting an existing TIA amplifier design, so compensation circuit can be added to an existing design of a TIA based amplifier circuit without affecting stability of the TIA based amplifier circuit. When the compensation circuit is used to compensate for transmitter light-emitting device characteristics, the compensation circuit improves the intrinsic relaxation oscillation of light-emitting diode without deteriorating rising or falling time of the signal.
In the compensation stage shown in
In the compensation stage shown in
For example, when a TIA amplifier has a major pole contributed by parasitic of a photo detector at 5 GHz and output associated pole at approximately 9 GHz with a 27 GHz buffer stage, the overall bandwidth of the TIA stage is approximately 4.3 GHz, which is not enough to process an optical signal operating at 10 Gbps. A compensation circuit with two frequency compensation stages provides sufficient compensation to allow accurate detection of signals.
Frequency compensation for the light-emitting device on the transmitter side is accomplished by extracting the impulse response of the light-emitting device. Since the impulse response of the light-emitting device contains relative information, the impulse response is used as a reference for the compensation process. From this impulse response, the matching filter that puts out identical impulse response is created using currently available optimization tools. The typical impulse response of the light-emitting diode is composed of three poles. Two of the poles are conjugated poles and control the light-emitting devices intrinsic relaxation oscillation frequency by having a real part and an imaginary part. The third pole contributes to the adjustment of rising and falling time of transient response from the light-emitting device. A final compensation filter required for the actual compensation of light-emitting device characteristic is the inverse function of the matching filter implemented using the impulse response.
The foregoing discussion discloses and describes merely exemplary methods and embodiments of the present invention. As will be understood by those familiar with the art, the invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. Accordingly, the disclosure of the present invention is intended to be illustrative, but not limiting, of the scope of the invention, which is set forth in the following claims.
