Optical receiver
Field of the invention
This invention is generally related to an optical receiver. More specifically, the invention concerns a method and circuitry for enhancing the dynamic range of an input amplifier in an optical receiver.
Technical background
An optical transmission system includes (a) an optical transmitter, which turns the electric signal to be transmitted into an optical form, (b) an optical fibre working as the conductor for the optical signal, and (c) an optical receiver detecting the transmitted optical signal and turning it into electrical form.
A typical optical receiver in its input stage includes a photodetector and a transimpedance amplifier, the input of which is connected to the photodetector. The photodetector turns the received optical signal into an electric current, which is supplied to the transimpedance amplifier. The latter produces such a voltage at its output, which is proportional to the incoming current, which means that a voltage proportional to the photodetector's current is obtained from the amplifier's output. The photodetector is usually either an Avalanche Photo Diode (APD) or an optical PIN diode. Better performance values are achieved with an avalanche photo diode than with a PIN diode, but it is more expensive than a PIN diode and, in addition, it is clearly more difficult to use (due to the high bias voltage which it requires and which must be adjusted according to the temperature). Transimpedance amplifiers are generally used e.g. for the reason that with their aid relatively good sensitivity characteristics are achieved with a relatively simple structure.
A problem with optical receiver solutions is their inadequate dynamics: a good sensitivity often means poor power tolerance, whereas a good power tolerance for its part means a low sensitivity. Poor dynamics again will make the receiver less flexible in use; e.g. when beginning to use a shorter fibre an additional attenuator must be added between the transmitter and the receiver.
Since in practice the power level of the optical signal arriving at the receiver may vary even considerably (depending on how long a fibre is used), some kind of automatic gain control (AGC) is typically used in connection with
the transimpedance amplifier to keep the amplifier's output voltage essentially at a constant value, when the incoming signal is higher than a predetermined threshold value.
When a good sensitivity is desirable, the stray capacitances affect- ing the amplifier's input are essential; even a small capacitance will lower the sensitivity of the receiver. Thus, it is also essential that the parasitics affecting the amplifier input are made as small as possible.
Attempts have been made to extend the receiver's dynamic range by using an adjustable resistive element in front of the transimpedance amplifier. The element's resistance is adjusted in response to the strength of a signal arriving at the amplifier, so that the resistance is lowered at higher levels, whereby the current connected to the amplifier input is reduced (as a part of the current goes through the resistive element), and the amplifier will not be saturated. Several different variations are known of this basic solution, and they will be described briefly in the following.
A control circuit is presented in US Patent 5,012,202 and in EP Patent Publication 433 646-B1 , wherein a field-effect transistor (FET) is used as the resistive element. To prevent the drain capacitance of the field-effect transistor from reducing the amplifier's sensitivity, it must be compensated for with a feedback over the field-effect transistor. Such a feedback, however, makes the circuit even more complicated. In addition, the feedback makes it more difficult to design the receiver.
An alternative, which is better than the field-effect transistor, is to use a diode with a naturally low capacitance as the adjustable resistive element. There are several different solutions based on such a diode.
GB Patent Appplication 2 247 798-A presents a diode-based solution, wherein based on a voltage formed over a resistance (r, Figure 1) a switching transistor (TR1 , Figure 1) is used to control a voltage over a diode (D, Figure 1) and thus to control the dynamic resistance of the diode. To prevent the control circuit from interfering with the DC operating point of the transimpedance amplifier, a capacitor (C2, Figure 1) must be used to separate it from the amplifier. However, the use of a capacitor causes an additional time constant in the feedback loop, which complicates the design. In addition, the capacitors and the above-mentioned resistance make the circuit more complex, whereby more space than before is also required on the circuit
board. All additional components also cause parasitics at high frequencies, which reduces the sensitivity of the receiver.
US Patent 4,415,803 also describes a diode-based solution using such a peak-hold circuit in the control which monitors the peak value of the transimpedance amplifier output. Based on this peak value, the voltage over the diode is controlled so that a part of the amplifier's input current will pass through the diode. A drawback of this solution is that it is difficult to bring about an exact peak-hold circuit at high transmission rates. It is difficult to achieve a great precision in peak value measurements and, besides, expensive special components must be used.
A solution is known from EP Patent Application 402 044-B1 , wherein the detector current passes through a resistance (R1 , Figures 4a and 4b) and causes a voltage over the diode (DO, Figures 4a and 4b). When this voltage is sufficiently high, the diode begins to conduct, whereby it functions as an attenuator. One drawback of this solution is that the resistance causes noise, which again reduces the sensitivity of the receiver. The feedback also makes the implementation more difficult, because it must be stable under all circumstances.
Attempts have also been made to enhance the receiver's dynamic range by making the transimpedance amplifier's feedback resistance adjustable. However, such a solution has the same problems as the solutions described above; the adjustment will easily cause stray capacitance at the amplifier input, and a compromise must be made in the adjustment between dynamics and sensitivity.
Brief summary of the invention
It is a purpose of this invention to eliminate the drawbacks described above and to bring about a solution with the aid of which the dynamic range of the receiver can be increased so that the circuit solution remains as simple as possible and, in addition, so that the receiver sensitivity will suffer only minimally.
This objective is achieved with the solution defined in the independent claims.
The idea of the invention is to perform the control entirely on the input side of the transimpedance amplifier by controlling the impedance of an impedance element connected to the amplifier input directly according to the
optical power arriving at the receiver. In other words, the idea is to perform the control without any feedback from the transimpedance amplifier output. Since no feedback is needed, less design and less components are needed in this regard. It is another advantage of the solution according to the invention that ordinary economically advantageous components may be used in the circuit.
Brief description of the drawings In the following the invention and its preferred embodiments are described in greater detail and by way of examples, referring to Figures 1-3 in the appended drawings, wherein
Figure 1 illustrates the solution in accordance with the invention at a general level;
Figure 2 illustrates the operation of a circuit in accordance with Figure 1 ; and Figure 3 shows in greater detail an embodiment of the circuit in accordance with Figure 1.
Detailed description of the invention
Figure 1 illustrates the principle in accordance with the invention at a general level by showing the front end of an optical receiver, which includes a transimpedance amplifier TA. As is known, a transimpedance amplifier is an inverting (phase shift of 180 degrees) voltage amplifier having an internal feedback resistance (not shown in the figure). An (optical) signal arriving from optical fibre OF is converted into a current in photodetector PD, which may be of any known type in principle. In this example it is assumed that the photodetector is an optical PIN diode. The anode of the PIN diode is connected to the input of the transimpedance amplifier and the catode to a biasing voltage +V1. A control unit CU in accordance with the invention measures the optical power arriving at the receiver, and based on this it adjusts the impedance of an impedance element ZU connected to input point P1 of the amplifier. Measurement of the input power takes place by measuring the current passing through the diode, which in this case is directly proportional to the optical mean power. Determination of the power is preferably performed from the photodetector's cathode, so that the measure-
ment will not have any harmful effect on the sensitivity of the transimpedance amplifier.
The operation of the circuit according to the invention is illustrated in Figure 2, wherein the vertical axis shows the optical input power mean while the horizontal axis shows the impedance seen at input point (P1) of the transimpedance amplifier. At small input levels, the impedance of impedance element ZU is high, and the impedance seen at the input point has a (high) value (Z1), which depends on the parallel connection of the resistance R2 and the transimpedance amplifier's input impedance. When the optical input power increases above a certain threshold value PTH, the impedance of impedance element ZU starts falling. The decrease is either linear (curve C1) or non-linear (curve C2), whereby the curve shape depends on the properties of the impedance element used.
Figure 3 shows in greater detail an embodiment of the circuit in accordance with Figures 1 and 2, wherein the impedance element is formed by successively connected PIN diodes D1 and D2 and the control unit by a differential amplifier circuit controlling the current passing through the impedance element. It should also be mentioned as a clarification that the diodes are RF PIN diodes. In the circuit shown in Figure 3, the anode of photodetector PD is connected through capacitor C4 to the input of transimpedance amplifier TA and through resistance R2 to earth. Capacitor C4 is used to perform DC separation in order to ensure that the control will not affect the DC operating point of the transimpedance amplifier. The capacitor is not essential from the viewpoint of the invention, and the need for it depends on the transimpedance amplifier used. Resistance R2 (which has a high value) together with the input impedance of the transimpedance amplifier determine the impedance of point P1 at low input levels (when no current passes through the impedance element), cf. Z1 in Figure 2. The cathode of the photodetector for its part is connected through capacitor C1 to earth and through resistance R1 to biasing voltage +V1. Capacitor C1 functions as a signal filtering component, which is used to obtain an averaged voltage for the photodetector's cathode (voltage U1). With the aid of resistance R1 , a voltage Urn is formed which is proportional to the optical input power and from which a control voltage (U2) is formed for the impedance element with the aid of the amplifier circuit.
The terminal of resistance R1 on the photodetector side is connected through resistance R7 to the first (non-inverting) input of the differential amplifier A1 , whereas the terminal on the biasing voltage side is connected through resistance R5 to the second (inverting) input of the differential amplifier. This input terminal is also connected to earth through resistance R3 and to the amplifier output through feedback resistance R4.
PIN diodes D1 and D2 form a current-controlled impedance element ZU, the dynamic impedance of which is controlled in order to increase the dynamic range of the transimpedance amplifier. The diodes are connected in succession so that the common terminal formed by the anode of diode D1 and by the cathode of diode D2 forms the input point P1 of the transimpedance amplifier, which at the same time forms such a common terminal for resistance R2 and for the photodetector which is connected through capacitor C4 to the transimpedance amplifier input. The anode of diode D2 is connected to preset voltage Vi and to earth through capacitor C3. The cathode of diode D1 for its part is connected to the first terminal of resistance R6 and to earth through capacitor C2. The second terminal of resistance R6 is connected to the output of differential amplifier A1. The said capacitors are not essential to the inventive idea proper; C2 and C3 form a low impedance current path to earth, whereby the diodes from the viewpoint of circuit operation are located in parallel from point P1 to earth. Resistance R6 again functions as a limiting unit limiting the maximum current passing through the impedance element.
The circuit in accordance with Figure 3 operates in the following manner. When the optical input power is zero, diodes D1 and D2 are in reverse and an impedance is seen at point P1 , the value of which is determined by the parallel connection of resistance R2 and the input impedance of the transimpedance amplifier. When optical power arrives at the receiver, the voltage Urn affecting over resistance R1 is directly proportional to the mean value of the optical power arriving at photo diode PD. Thus, voltage U1 will fall as the optical input power increases. The output voltage U2 of the differential amplifier A1 changes as a function of voltage U1 , and when U1 is sufficiently low (compared to the threshold value set with the aid of the voltage divider formed by resistances R3 and R5), U2 too is sufficiently low compared to the preset voltage Vi, whereby diodes D1 and D2 will be forward biased and current starts flowing through them.
Thus, diodes D1 and D2 are reverse biased at low power levels of the input signal. At high power levels, the diodes are forward biased and a current passes through them, which is proportional to the (mean) input power. In other words, the dynamic impedance of diodes D1 and D2 (impedance element ZU) is reduced and the amplitude of the input signal of the transimpedance amplifier is reduced. As a result of this, the dynamic range of the transimpedance amplifier will extend.
Although the invention was described above by referring to the examples in accordance with the appended drawings, it is obvious that the invention is not limited to these, but it can be modified within the scope of the inventive idea defined in the appended claims. In practice, circuit solutions implementing a similar functionality may vary in many ways. Thus, the implementation of both the impedance element and of the control circuit may vary in many ways without deviating from the principle described in the foregoing. Some variations are described briefly in the following.
The impedance element may include one or more components, and the element may be implemented with components of different types. However, it is advantageous to use a PIN diode as the impedance element, because it has inherently a low capacitance. The embodiment presented above is advantageous in that two diode components are available as commercial components and it is easy for the component in question to implement a voltage control of the kind described above. The impedance element may be implemented by using only one diode, but it is then not easy to supply a control voltage to point P1 without causing any stray capacitance. The control circuit may also be implemented e.g. by a current mirror producing a current which is proportional to the current passing through the photo diode and which is used for controlling the impedance of the impedance element. However, the current mirror must be embodied in such a way that it has a certain threshold value, which must be exceeded by the photodetector current before the current mirror will generate any current at its output. The impedance element may also include a network formed by several components, but due to the reasons mentioned above, the less components are needed, the better. Neither does the preset voltage need to have any predetermined constant value, but it may be made controllable, e.g. based on voltage U2.