Cross resistance amplifier circuit structure supporting fast fine adjustment and high bandwidth high gain flatness balanced photodetector

Abstract

This application provides a transimpedance amplifier circuit structure supporting rapid and fine adjustment, and a high-bandwidth, high-gain flatness balanced photodetector, belonging to the field of transimpedance amplifiers. The feedback structure of this transimpedance amplifier introduces an additional resistor-capacitor network into the traditional single-resistor, single-capacitor transimpedance amplifier circuit structure, forming a transimpedance amplifier circuit structure with double-pole, single-zero characteristics. This increases the bandwidth and adds additional controllable poles and zeros, enabling precise shaping of the closed-loop frequency response of the transimpedance amplifier and improving gain flatness. It fundamentally reduces the dependence of high-bandwidth transimpedance amplifiers on parameters that are difficult to measure precisely, such as inherent parasitic capacitance, and also solves the problem of inaccurate matching of compensation networks due to parameter dispersion in traditional designs. This balanced photodetector can be applied to optoelectronic technologies such as coherent lidar, high-sensitivity laser spectroscopy, precision metrology and scientific research, and quantum technology.

Description

Technical Field

[0001] This application relates to the field of transimpedance amplifier technology, and in particular to a transimpedance amplifier circuit structure that supports fast and fine adjustment and a high-bandwidth, high-gain flatness balanced photodetector. Background Technology

[0002] As a key photoelectric conversion device, the balanced photodetector's core principle is to utilize two performance-matched photodiodes to differentially process the input optical signal, thereby significantly suppressing common-mode noise and effectively amplifying weak differential signals. This characteristic gives it an irreplaceable advantage in extracting weak optical signals, including but not limited to high-sensitivity lidar, precision spectral measurement, biomedical imaging, and quantum information processing. In lidar applications, especially coherent lidar systems (such as Doppler wind lidar and vibration measurement lidar), the balanced photodetector is crucial for achieving high-precision measurements. It acquires information such as target velocity and vibration by detecting the weak frequency or phase difference (Doppler shift) between the signal light and the local oscillator light. These frequency shifts are typically concentrated in the hundreds of megahertz range, with extremely weak signal strength. Therefore, the performance of the balanced photodetector directly determines the detection sensitivity, wind speed measurement accuracy, and vibration analysis capability of the entire system. In quantum information processing applications, especially in scenarios involving high-fidelity measurement of quantum states (such as quantum key distribution, quantum communication, and continuous-variable quantum computing), the balanced photodetector is essential for achieving quantum limit detection. Key information or features of a quantum state can be extracted by detecting the minute amplitude and phase differences (typically near the shot noise limit) between two quantum entangled optical fields or between a quantum state and its coherent local oscillator. These quantum signals exhibit extremely small fluctuations, often submerged in classical noise. Therefore, the common-mode rejection ratio, bandwidth, and intrinsic noise level of a balanced photodetector directly determine the secure key generation rate of quantum communication, the fidelity of quantum state measurements, and the ultimate performance limit of the entire quantum information system.

[0003] There are three main approaches to balanced detection technology. Chip-based solutions offer excellent performance but are expensive and lack flexibility. RF amplifier-based solutions offer high bandwidth (up to GHz) but low signal-to-noise ratio, and their input / output typically uses AC coupling. This filters out DC and low-frequency components, making it impossible to obtain complete information such as the average power of the optical signal, thus limiting its application in precision measurement scenarios requiring full-spectrum information (from DC to high frequencies). Operational amplifier-based balanced detection technology, however, exhibits unique advantages. This approach utilizes high-performance operational amplifiers to construct transimpedance amplifier circuits, enabling DC coupling and thus fully preserving the DC and low-frequency components of the signal. This allows it to not only demodulate high-frequency AC signals but also simultaneously and accurately measure the average power of optical signals, providing crucial technical support for precision measurement scenarios requiring full-spectrum information, and at a lower cost.

[0004] However, the bandwidth performance of high-performance balanced photodetectors is largely limited by their core signal conditioning circuit—the transimpedance amplifier circuit. The basic structure of a transimpedance amplifier circuit consists of an operational amplifier, a feedback resistor, and a feedback capacitor; its essence is a controlled source circuit that converts current into voltage. While the transimpedance amplifier circuit has a simple structure, its applications still face many limitations.

[0005] In the low-bandwidth, high-gain direction, there is an existing T-type resistor network structure, which achieves equivalent high transimpedance gain through the combination of three small-value resistors.

[0006] However, for bandwidth requirements exceeding 100 Mbps, the T-type resistor network structure is no longer suitable. Furthermore, the design of 100 Mbps bandwidth places higher technical demands on the transimpedance amplifier circuit. The design difficulty of the feedback capacitor shifts from the nanofarad and picofarad range at low bandwidths to the even lower femtofarad range, sometimes requiring capacitance values ​​far lower than those of current standard discrete components. Therefore, the design of high-bandwidth transimpedance amplifier circuits involves a precise trade-off between gain, bandwidth, and stability.

[0007] Existing high-bandwidth balanced photodetectors typically use a circuit structure consisting of five core modules. The first module is a differential detection unit composed of two photodiodes connected in series, used to convert two light intensity signals into photocurrent and output the difference. The second module is an AC / DC signal separation module. The third module uses a transimpedance amplifier to convert the AC signal into a voltage signal. The fourth module uses a non-inverting or inverting voltage amplifier to perform secondary gain amplification of the AC signal. The fifth module is a DC signal amplification module.

[0008] In the field of high-speed photoelectric detection, the overall bandwidth performance of the detection system is mainly limited by the transimpedance amplifier in the third module. The bandwidth design of the transimpedance amplifier determines the final bandwidth of the photodetector. The basic structure of the transimpedance amplifier circuit consists of an operational amplifier, a feedback resistor, and a feedback capacitor. The core challenge in designing a high-bandwidth transimpedance amplifier lies in the fine adjustment of its feedback capacitor, specifically, ensuring stable operation of the transimpedance amplifier by adjusting the feedback capacitor. Currently, there are two main types of conventional compensation schemes in the industry, but both have significant technical defects that restrict the full realization of the transimpedance amplifier's performance.

[0009] The first approach uses discrete miniature capacitors for precise compensation. This approach forms a compensation network by connecting a precisely valued tiny capacitor (typically a few tenths of a picofarad) in parallel with a feedback resistor to suppress phase hysteresis caused by the photodiode junction capacitance and the op-amp input capacitance, ensuring loop stability. However, the bottleneck of this approach lies in the limited minimum resolution of the nominal values ​​of existing standard discrete capacitor components. Commonly available surface-mount capacitors typically have minimum nominal values ​​starting from 0.1 picofarads, with increments of 0.1 picofarads (e.g., 0.1 pF, 0.2 pF, 0.3 pF, etc.). This discrete minimum resolution inherently contradicts the continuous and precise compensation values ​​required by high-bandwidth transimpedance amplifiers. In actual circuit debugging, designers often cannot obtain the theoretically calculated ideal capacitance value (e.g., 0.07 pF) and are forced to choose adjacent, larger nominal values ​​(e.g., 0.1 pF). This "choosing the larger" directly leads to overcompensation. While overcompensation ensures circuit stability, it comes at the cost of artificially introducing an additional low-frequency pole, severely compressing the system's effective bandwidth. This prevents the operational amplifier chip's inherent high bandwidth potential from being fully realized, ultimately resulting in a significant reduction in the detector's bandwidth performance and making it difficult to meet design expectations. Simulation analysis reveals that the optimal compensation value for the capacitor, when using 20 femtofarads as the minimum resolution scanning parameter, is as follows: Figure 4 The required 205-fold capacitance value is often not available in actual designs. Therefore, using existing capacitance values ​​often results in undercompensation and overcompensation.

[0010] The second approach abandons discrete capacitors and instead utilizes the parasitic capacitance generated by the layout and routing of the printed circuit board (PCB) itself as the feedback capacitor. This approach forms a specific distributed capacitance by carefully designing the shape and spacing of the pads across the feedback resistor of the transimpedance amplifier and the back ground layer. While this method avoids the problem of insufficient resolution of discrete capacitors, it introduces new drawbacks such as poor repeatability and difficulty in controlling accuracy. The parasitic capacitance of the PCB is highly dependent on the dielectric constant of the substrate material, the thickness of the dielectric layer, and the line width and spacing. The manufacturing process parameters themselves have tolerances, and in actual sample production, the parasitic capacitance value will fluctuate due to different batches of PCB materials or slight layout differences, making it difficult to achieve precise, predictable, and repeatable control. The direct consequence is that the compensation amount deviates from the optimal value, which can easily lead to significant spectral peaking in the frequency response curve of the transimpedance amplifier within the gain range. This peaking means that the gain increases in certain frequency bands, disrupting the flatness of the frequency response. This not only introduces measurement errors and affects signal fidelity, but in severe cases, it can also cause circuit oscillations, resulting in unstable detector performance and reduced yield. Meanwhile, parasitic capacitance is an inherent parameter, its magnitude depending on manufacturing processes such as circuit board layout, lamination materials, and component packaging, making it difficult to precisely control and effectively adjust in the later stages of design. This contradiction leads to a significant design dilemma: to achieve system stability, designers are forced to passively limit the range of feedback resistor values ​​based on fixed parasitic capacitance values. This severely restricts the optimization space for the transimpedance amplifier's transimpedance gain, bandwidth, and noise performance, significantly limiting design freedom.

[0011] Finally, when it is necessary to maximize bandwidth while maintaining high gain, feedback capacitor compensation becomes more difficult for higher bandwidth transimpedance amplifier circuits, often requiring flyfarm-level capacitors for compensation. Therefore, for the design of high bandwidth transimpedance amplifier circuits, one approach is to sacrifice some gain, i.e., reduce the feedback resistor, to increase the required feedback capacitor value. Larger feedback capacitors are often readily available. Even so, because the nominal values ​​of standard discrete capacitor components vary discretely, the method of using feedback capacitors often easily leads to overcompensation or undercompensation, failing to maximize the performance of the operational amplifier. Summary of the Invention

[0012] To address the industry challenge of precisely matching the feedback resistor and feedback capacitor in the design of high-bandwidth transimpedance amplifier circuits, this application proposes a transimpedance amplifier circuit structure that supports rapid and precise adjustment, as well as a high-bandwidth, high-gain, flatness balanced photodetector.

[0013] The technical solution adopted in this application is: a transimpedance amplifier circuit structure that supports fast and fine adjustment, including an operational amplifier and a feedback network structure connected between the input and output terminals of the operational amplifier. The feedback network structure is composed of multiple resistors and multiple capacitors, and a double-pole single-zero transfer function network is constructed by series and parallel connection.

[0014] Furthermore, the feedback network structure includes two parallel branches, one of which consists of capacitors C1 and C2 connected in series, and the other consists of resistors R2 and R3 connected in series. A resistor R1 is connected between the midpoints of the two parallel branches for finely controlling the relative positions of the zeros and poles.

[0015] Furthermore, a feedback network structure is connected between the inverting input and output of the operational amplifier.

[0016] Furthermore, there are three strategies for finely controlling the relative positions of the zeros and poles through resistor R1. The first is:

[0017] Strategy 1: Set R1 to a specific resistance value and fine-tune R1 to optimize the frequency response. This strategy is used in application scenarios where the flattest gain flatness is desired.

[0018] Furthermore, there are three strategies for finely controlling the relative positions of the zeros and poles through resistor R1, the second of which is:

[0019] Strategy 2: Shorting R1 facilitates quick location of zeros and poles. Under this strategy, the transfer function of the transimpedance amplifier circuit structure is simple, allowing for rapid optimization to the vicinity of the target bandwidth.

[0020] Furthermore, there are three strategies for finely controlling the relative positions of the zeros and poles through resistor R1, the third of which is:

[0021] Strategy 3: With R1 left floating, the feedback network degenerates into a simple parallel RC structure.

[0022] Furthermore, by leaving a capacitor's preset pad in a floating state, the circuit stability can be optimized using the equivalent parasitic capacitance of the floating pad.

[0023] Furthermore, the input of the transimpedance amplifier is connected to a sensing device.

[0024] Furthermore, the bandwidth of this transimpedance amplifier circuit structure is adjustable between 200MHz and 800MHz.

[0025] A high-bandwidth, high-gain, flatness balanced photodetector includes a photocurrent difference signal generation module, an AC / DC separation module, an in-phase proportional amplifier module, and a DC output module, as well as the transimpedance amplifier circuit structure that supports rapid and fine adjustment.

[0026] Furthermore, the optical signal acquisition module includes a first photodiode PD1 and a second photodiode PD2; the AC / DC separation module includes a DC blocking capacitor C3 and a resistor R9; the cathode of the first photodiode PD1 is connected to a positive bias voltage, the anode of the second photodiode PD2 is connected to a negative bias voltage, and the anode of the first photodiode PD1 is connected to the cathode of the second photodiode PD2, with node a;

[0027] Node a is connected to one end of DC blocking capacitor C3, and the other end of DC blocking capacitor C3 is connected to the inverting input port of operational amplifier. The output pin of operational amplifier is connected to non-inverting amplifier module, and the output of non-inverting amplifier module is AC signal.

[0028] Node a is also connected to one end of resistor R9 and the input terminal of the DC output module, which outputs a DC signal.

[0029] The beneficial effects of this application compared to the prior art are as follows: The transimpedance amplifier circuit structure that supports fast and fine adjustment proposed in this application can greatly simplify the design difficulty of transimpedance amplifier circuits under high bandwidth and high gain flatness conditions, and facilitate the design of high bandwidth transimpedance amplifier circuits with required gain flatness.

[0030] The advantages of this circuit structure are:

[0031] 1. The transimpedance amplifier circuit structure with feedback network can further improve bandwidth and support fast and fine adjustment of gain flatness, solving the industry problem that the feedback capacitor is difficult to finely adjust in traditional feedback schemes for high-bandwidth transimpedance amplifiers.

[0032] 2. The balanced photodetector built with a transimpedance amplifier circuit structure with a feedback network has ultra-high gain flatness and a highly consistent response to signals of various frequencies within the bandwidth, reducing the error caused by inconsistent frequency response during use.

[0033] 3. Adjusting the bandwidth flatness of the balanced photodetector under the transimpedance amplifier circuit structure with feedback network structure: Due to the use of the newly constructed transimpedance amplifier circuit structure, the compensation of the transimpedance amplifier no longer relies solely on a single feedback capacitor, but rather on a double-pole single-zero compensation method constructed by the transimpedance amplifier circuit structure and its feedback network structure as a whole. The values ​​used in this feedback network structure are all standard discrete components. It is worth noting that capacitors with larger capacitance values ​​are readily available in existing standardized discrete components, and their selection is flexible and cost controllable. This characteristic significantly reduces the complexity and design burden of high-bandwidth transimpedance amplifier circuits and balanced photodetectors in the implementation process.

[0034] 4. Adjusting the bandwidth flatness of the balanced detector within a transimpedance amplifier circuit with a feedback network structure and making reasonable use of the system's inherent parasitic capacitance is an effective strategy for optimizing circuit stability. This transimpedance amplifier circuit structure, containing multiple resistors and capacitors, behaves as a double-pole, single-zero system in the frequency domain. If one of the capacitors has its preset pad floating (i.e., no actual capacitor is soldered), then that node only retains its equivalent parasitic capacitance. Typically, the parasitic capacitance is small; therefore, to meet the total compensation capacitance required for system stability, other capacitors need to have larger capacitance values. This characteristic can also reduce the design difficulty of high-bandwidth transimpedance amplifier circuits and balanced photodetectors.

[0035] The transimpedance amplifier circuit structure with feedback network in this application offers significant economic benefits in adjusting bandwidth and gain flatness. It greatly reduces the compensation difficulty of high-bandwidth transimpedance amplifier circuits and solves the problem of accurately matching feedback capacitors with distributed capacitors. This provides a practical and efficient implementation path for amplifier optimization in applications such as high-speed optoelectronic reception.

[0036] 5. The technical value of the transimpedance amplifier circuit structure with feedback network in this application lies in its highly flexible adjustment of bandwidth expansion. By adjusting the values ​​of multiple components, the positions of zeros and poles can be flexibly placed, thereby achieving a better trade-off between expanding bandwidth and maintaining stability than a single resistor and single capacitor structure, while also achieving a near-ideal flat frequency response. Attached Figure Description

[0037] The following description, in conjunction with the accompanying drawings, further illustrates this application:

[0038] Figure 1 This is a circuit diagram of an existing single-resistor-single-capacitor transimpedance amplifier.

[0039] Figure 2 This is a circuit diagram of a transimpedance amplifier with a feedback network structure provided in an embodiment of this application;

[0040] Figure 3 A schematic diagram of a high-bandwidth, high-gain, flatness balanced photodetector circuit provided in the embodiments of this application;

[0041] Figure 4 Figure 1 shows the bandwidth simulation results of the transimpedance amplifier circuit with a total gain of 3k ohms. Figure 2(a) shows the bandwidth simulation structure of the traditional single resistor-single capacitor transimpedance amplifier circuit, and Figure 3(b) shows the bandwidth simulation results of the transimpedance amplifier circuit with feedback network structure proposed in this application at the same gain.

[0042] Figure 5The simulation results show the bandwidth of the gain flatness when the R1 is adjusted using strategy one.

[0043] Figure 6 The graph shows the measurement results of shot noise rise and linearity test of the 180M balanced photodetector with the transimpedance amplifier circuit structure with feedback network structure proposed in this application in different frequency bands.

[0044] Figure 7 The graph shows the measurement results of shot noise rise and linearity test of the 240M balanced photodetector with the transimpedance amplifier circuit structure with feedback network structure proposed in this application in different frequency bands.

[0045] Figure 8 The graph shows the measurement results of shot noise rise and linearity test of the 360M balanced photodetector with the transimpedance amplifier circuit structure with feedback network structure proposed in this application in different frequency bands.

[0046] Figure 9 The graph shows the measurement results of shot noise rise and linearity test of the 420M balanced photodetector with the transimpedance amplifier circuit structure with feedback network structure proposed in this application in different frequency bands.

[0047] Figure 3 In the middle: 1 is the photocurrent difference signal generation module, 2 is the AC-DC separation module, 3 is the transimpedance amplifier circuit structure, 4 is the in-phase proportional amplifier module, and 5 is the DC output module. Detailed Implementation

[0048] like Figures 1 to 9 As shown, this application provides a transimpedance amplifier circuit structure that supports fast and fine adjustment, hereinafter referred to as a transimpedance amplifier. A novel double-pole single-zero transimpedance amplifier feedback structure is designed, which transforms the traditional... Figure 1 The single-pole feedback scheme of the transimpedance amplifier with a single resistor and a single capacitor in parallel, as shown, becomes as follows: Figure 2 This paper presents a novel, finely controlled feedback scheme based on a feedback network structure with two poles and a single zero. The core idea of ​​this structure is to overcome the design bottleneck caused by discrete, discontinuous, small compensation capacitors through innovative circuit design. Furthermore, when the feedback capacitor needs to be smaller than its nominal value, an equivalent continuously adjustable compensation capacitor with a value much smaller than the minimum nominal value of a standard discrete component can be achieved. This allows designers to independently optimize the feedback resistor (to set the gain) and the equivalent feedback capacitor (to precisely optimize bandwidth and stability) over a wide bandwidth, thereby breaking through the limitations of existing technologies and achieving flattening of the gain curve, reliable stability control, and performance optimization of high-bandwidth transimpedance amplifiers.

[0049] Specifically Figure 1 The single resistor R in f and a single capacitor C f The parallel feedback scheme has an output voltage With input current The ratio is the transimpedance gain. , This is the response of a first-order low-pass filter. Its DC gain is... The -3dB bandwidth is: The frequency response curve is a single pole. The curve.

[0050] Figure 2 The transimpedance amplifier circuit structure proposed in this application includes a sensing device (such as a photodiode) for converting physical signals (such as optical signals) into current signals, an operational amplifier (U1), and a feedback network structure (Z) connected between the inverting input terminal and the output terminal of the operational amplifier. Figure 2 The paper provides an exemplary circuit using a photodiode, which is represented by an equivalent circuit. The photodiode equivalent circuit includes a current source and a junction capacitance C. j and parallel resistor R s Description and notation of the feedback network structure: The bottom branch (from left to right) consists of series resistors R2 and R3, with the midpoint denoted as V. mr The top branch (from left to right) consists of capacitors C1 and C2 connected in series, with their midpoint denoted as V. mc There is a resistor R1 between the midpoints (connected to V). mr With V mc The left input of the feedback network structure is connected to the inverting input port of the operational amplifier. The right output of the feedback network structure is connected to the output port of the operational amplifier. It should be understood that the components in the system can be electrically connected to each other using any suitable electrical connection method (e.g., metal wires, printed circuit board traces, etc.) to transmit electrical signals.

[0051] In a specific implementation case, the parameters of the feedback network structure (Z) are determined as follows: First, the desired gain (e.g., 5kΩ) and desired bandwidth (e.g., 200MHz) of the transimpedance amplifier are determined according to application requirements, and then the bandwidth is calculated using the bandwidth formula (…). Calculate the required capacitance, which is approximately 0.15pF. Select a standard discrete capacitor component close to the calculated capacitance, such as 0.2pF, and allocate it to C1 and C2 (e.g., C1=C2=0.2pF). Then, distribute the desired gain of 5kΩ to R2 and R3 in an appropriate ratio, such as an average distribution strategy, where R2 equals 2.5kΩ and R3 equals 2.5kΩ. Select 0 ohms for R1. Then, test and debug, adjusting the values ​​of R1, R2, and R3 to flatten the bandwidth.

[0052] refer to Figure 2 In one exemplary embodiment, the signal source is a photodiode device configured to receive an optical signal and convert it into an input current. In other exemplary embodiments, any sensing device capable of outputting a current signal, such as an ion sensor, a photoconductive device, etc., can be used in this feedback network.

[0053] The transimpedance amplifier circuit structure with feedback network in this application, compared to... Figure 1 The principle behind the bandwidth enhancement of a single-resistor-single-capacitor transimpedance amplifier circuit is as follows:

[0054] The overall complex impedance of the single-resistor-single-capacitor transimpedance amplifier circuit structure is: ,in For capacitor C f The impedance.

[0055] Simplifying the derivation, assuming R1 is 0, the overall impedance of the transimpedance amplifier circuit structure with feedback network in this application is:

[0056] ,in Let C1 be the impedance. The impedance of capacitor C2;

[0057] In a transimpedance amplifier circuit structure with a single resistor and a single capacitor, and in a transimpedance amplifier circuit structure with a feedback network, both have the same compensation capacitor. ;

[0058] To ensure that the total gain of both is the same, then we have ;

[0059] For ease of verification, let's assume a... The situation;

[0060] at this time ;

[0061] Substituting the transfer function of the transimpedance amplifier circuit structure with a single resistor and a single capacitor, and the transfer function of the transimpedance amplifier circuit structure with a feedback network structure, we get:

[0062] ;

[0063] A comparison reveals that when the denominator of the transfer function of the transimpedance amplifier circuit structure with feedback network structure proposed in this application is reduced, the corresponding poles increase and the bandwidth increases. In other words, the bandwidth limiting effect caused by a single feedback resistor and feedback capacitor can be mitigated by dividing the resistor into two components.

[0064] The complete theoretical formula derivation process in the above principle is as follows: The transfer function of the complete feedback network is complex, and directly deriving the transfer function becomes very complicated and physically unclear. The derivation of the complete transfer function will not be detailed here; only the expression of the complete transfer function is given.

[0065] ;

[0066] in .

[0067] This is a rational function with a second-order denominator and a first-order numerator. It contains two poles and one zero. R2 and R3 are the main resistors that provide the gain, and they also play a decisive role in the distribution of the poles and zeros. R1 plays a minor adjustment role.

[0068] To demonstrate the principle and illustrate the bandwidth improvement effect, a simplified derivation is performed. We first assume R1 is a fixed value, for example, R1 is 0 ohms. Then we have: V mr =V mc The subsequent derivation is uniformly V. m V m =V mr =V mc .

[0069] R2 and C1 are connected in parallel. One end of this part is connected to the inverting input port of the operational amplifier, and the other end is connected to node V. m R3 and C2 are connected in parallel. One end of this part is connected to the output of the operational amplifier, and the other end is connected to node V. m ;

[0070] The derivation of the transfer function is as follows:

[0071] Part 1: R2 and C1 are connected in parallel. ;

[0072] Part Two: R3 and C2 are connected in parallel. ;

[0073] The entire feedback impedance Z f It is a series connection of Z1 and Z2: ;

[0074] The transfer function (transimpedance gain) of a circuit. for: ;

[0075] Will By finding a common denominator and simplifying, we can obtain:

[0076] ;

[0077] Expand the molecules:

[0078] ;

[0079] Increase DC gain ;

[0080] Zero-point time constant ;

[0081] Time constants of the two poles ;

[0082] The transfer function can be rewritten in a concise standard form:

[0083] .

[0084] Its frequency response curve behavior can be described as follows: starting from zero Hz, the gain is flat, that is, the DC equivalent impedance is R2+R3. As the frequency increases, it encounters the first pole. ,pole It starts working, with the gain decreasing by -20dB / decibels before hitting zero. midnight It starts to work at zero frequency. At this point, the slope of the gain drop is "lifted," changing from -20dB / decibels to 0dB / decibels (i.e., the gain curve becomes flatter). This is key to achieving wide bandwidth and improving bandwidth flatness. High-frequency band: at the second pole frequency. At that point, the gain drops again at a slope of -20dB / decibels.

[0085] Simulation results verification as follows Figure 4 As shown, changing R2 and R3 significantly alters the pole-zero distribution, resulting in a marked increase in bandwidth. Simulation analysis reveals that in a traditional single-resistor-single-capacitor transimpedance amplifier circuit structure, with a total gain of 3k ohms, the optimal compensation value for the capacitor, using 20 femtofarads as the minimum resolution scanning parameter, is 205fF, achieving a -3dB bandwidth of 405MHz. Other parameter values, however, cause spectral peaking. In practice, achieving the same level of precision as the simulation is challenging.

[0086] With the same total gain, by changing the single-resistor-single-capacitor transimpedance amplifier to the transimpedance amplifier circuit structure with feedback network proposed in this application, and by changing the resistance distribution relationship between R2 and R3, different degrees of bandwidth improvement can be obtained, while still maintaining a flat response process, which has obvious advantages.

[0087] Simulation analysis shows that the -3dB frequency drop point changed from 405MHz to 542MHz, and the bandwidth increased by 33.8%. Therefore, this transimpedance amplifier circuit structure can not only simplify the design of high-bandwidth balanced photodetectors, but also further improve the bandwidth.

[0088] Introducing R1 introduces an additional control zero-pole method. Simulation software can then be used to quickly verify the correctness of the principle. Simulation analysis, such as... Figure 5 As shown, changing the value of R1 does not significantly increase the bandwidth, but it has a greater impact on the flatness of the bandwidth response range, acting as a fine-tuning of the zeros and poles, making the curve response flatter. Figure 4 Taking a set of ratios as an example, such as R2 being 600 ohms and R3 being 2.4k ohms, the effect of changing the value of R1 to fine-tune the gain flatness is demonstrated. It can be seen that changing R1 will cause a small change in the zero-pole distribution, thus resulting in a small change in bandwidth. Its more important role lies in its small adjustment of the gain flatness.

[0089] The advantage of this transimpedance amplifier circuit structure with a feedback network is that it distributes the resistance value required by a single resistor across multiple resistors. This reduces the resistance, thus increasing the bandwidth-limiting pole, thereby improving bandwidth. The feedback network structure consists of multiple capacitors and resistors, forming a two-pole, single-zero transfer function network through series and parallel connections. By appropriately setting the positions of the zeros and two poles, and using a fine-tuning resistor R1, the closed-loop frequency response can be further precisely shaped to increase gain flatness. By adjusting the resistor and capacitor values ​​in the feedback network structure, the zero and pole positions can be flexibly placed, achieving a better trade-off between bandwidth expansion and gain flatness compared to a single-resistor, single-capacitor transimpedance amplifier circuit structure, resulting in a near-ideal flat frequency response.

[0090] The control effect of R1 is to finely adjust the relative positions of the zeros and poles, and its specific implementation examples include three strategies:

[0091] Strategy 1: Fine-tune R1 to a specific resistance value to optimize frequency response.

[0092] Strategy 2: Short-circuit R1 (resistance value is 0). At this time, the transfer function of the feedback network is simpler and it is easier to quickly locate the zeros and poles.

[0093] Strategy 3: Leave R1 floating (open circuit). In this case, the feedback network degenerates into a simple parallel RC structure.

[0094] The core principle of this transimpedance amplifier circuit structure can be explained as follows: Analyze the components of the feedback network. R2 and R3 are resistors that provide gain and are also the dominant factors affecting the distribution of zeros and poles. R1 is a fine-tuning resistor. For ease of analysis and understanding, assume that R1 is a fixed value, set to 0 ohms here. As shown in Strategy 2 in the specific implementation case, Strategy 2: Short-circuit R1 (resistance value is 0). At this time, the transfer function of the feedback network is more concise and it is easier to quickly locate the zeros and poles.

[0095] like Figure 3 As shown, this application also proposes a high-bandwidth, high-gain flatness balanced photodetector, including a photocurrent difference signal generation module 1, an AC / DC separation module 2, a transimpedance amplifier circuit structure 3, a non-inverting proportional amplifier module 4, and a DC output module 5. The optical signal acquisition module 1 includes a first photodiode PD1 and a second photodiode PD2; the AC / DC separation module 2 includes a capacitor C3 and a resistor R9; the transimpedance amplifier circuit structure 3 includes an operational amplifier U1 and a feedback network structure Z.

[0096] The cathode of the first photodiode PD1 is connected to the positive bias voltage, the anode of the second photodiode PD2 is connected to the negative bias voltage, and the anode of the first photodiode PD1 is connected to the cathode of the second photodiode PD2. The node is a.

[0097] Node a is connected to one end of DC blocking capacitor C3, and the other end of DC blocking capacitor C3 is connected to the inverting input port of operational amplifier U1, and is also connected to feedback network structure Z. The other end of feedback network structure Z is also connected to the output pin of operational amplifier U1.

[0098] The output pin of the transimpedance amplifier circuit structure 3 is connected to the non-inverting amplifier module 4 to increase the additional gain. The output of the non-inverting amplifier module 4 is an AC signal.

[0099] Node a is also connected to one end of resistor R9 and the input terminal of DC output module 5, which outputs a DC signal.

[0100] Figure 6This graph shows the measurement results of shot noise rise and linearity tests of a 180MHz balanced photodetector with a transimpedance amplifier circuit structure featuring a feedback network, as proposed in this application, at different frequency bands. The yellow line represents the spectrometer noise floor, the red line represents electronic noise, and the purple line represents the shot noise curve of the detector at an input optical power of 1mW (500μW incident optical power from both channels). Similarly, the green line represents the noise curve at 2mW optical power, the blue line at 4mW optical power, and the black line at 8mW optical power. Its gain flatness is within 1dB.

[0101] Figure 7 This graph shows the measurement results of shot noise rise and linearity tests of a 240MHz balanced photodetector with a transimpedance amplifier circuit structure featuring a feedback network, as proposed in this application, at different frequency bands. The black line represents the spectrometer noise floor, the red line represents electronic noise, and the blue line represents the shot noise curve of the detector at an input optical power of 1mW (500μW incident optical power from both channels). Similarly, the green line represents the noise curve at 2mW optical power, the purple line at 4mW optical power, and the yellow line at 8mW optical power. Its gain flatness is within 1dB.

[0102] Figure 8 This graph shows the measurement results of shot noise rise and linearity tests of a 360MHz balanced photodetector with a transimpedance amplifier circuit structure featuring a feedback network, as proposed in this application, at different frequency bands. The black line represents the spectrometer noise floor, the red line represents electronic noise, the yellow line represents the shot noise curve of the detector at an input optical power of 1mW (500μW incident optical power from both channels), and so on. The purple line represents the noise curve at 2mW optical power, the green line at 4mW optical power, and the blue line at 8mW optical power. Its gain flatness is within 1dB.

[0103] Figure 9 This graph shows the measurement results of shot noise rise and linearity tests of a 420MHz balanced photodetector with a transimpedance amplifier circuit structure featuring a feedback network, as proposed in this application, at different frequency bands. The yellow line represents the spectrometer noise floor, the black line represents electronic noise, and the red line represents the shot noise curve of the detector at an input optical power of 1mW (500μW incident optical power from both channels). Similarly, the blue line represents the noise curve at 2mW optical power, the green line at 4mW optical power, and the purple line at 8mW optical power. Its gain flatness is within 1dB.

[0104] This application addresses the challenges faced by existing technologies in designing transimpedance amplifiers in the high-bandwidth domain of hundreds of megahertz (MHz). In this frequency band, reference... Figure 1The capacitance requirements for feedback capacitors have increased from the nanofarad (nF) and picofarad (pF) levels to the femtofarad (fF) level. Even a tiny change in parasitic capacitance can lead to a drastic deterioration in the system's frequency response, significantly increasing design complexity. Therefore, this application provides a transimpedance amplifier circuit structure supporting fine-tuning and its innovative feedback network structure. (Reference) Figure 4 and Figure 5 The transimpedance amplifier circuit structure 3 can not only effectively improve the bandwidth (for example, from 405MHz in the traditional structure to more than 542MHz), but also accurately shape the closed-loop frequency response and enhance the gain flatness by reasonably setting the pole and zero positions in the feedback network structure, while taking into account the speed and accuracy of gain adjustment.

[0105] This application proposes a transimpedance amplifier circuit structure with a feedback network. In the high bandwidth range of hundreds of MHz, the feedback network structure replaces the traditional single-resistor and single-capacitor scheme, increasing the overall impedance of the transimpedance amplifier circuit structure. This results in a larger bandwidth-limiting pole, thereby increasing the bandwidth.

[0106] The feedback network structure of this application consists of multiple capacitors and resistors, etc. It constructs a double-pole single-zero transfer function network through series and parallel connections. By reasonably setting the positions of the poles and zeros, the closed-loop frequency response can be further precisely shaped to increase the gain flatness, which reduces the design difficulties of traditional high-bandwidth transimpedance amplifiers and solves the problem that distributed capacitors cannot be accurately compensated, thus making it difficult to finely adjust the bandwidth and gain flatness of the balanced detector.

[0107] The broadband transimpedance amplifier based on the feedback network of this application has greatly improved bandwidth flexibility, and can be designed from 200MHz to 800MHz, all with high bandwidth gain flatness.

[0108] Based on the transimpedance amplifier circuit structure with feedback network of this application, circuit stability can also be optimized by making reasonable use of the inherent parasitic capacitance of the system. Since this feedback network structure includes multiple resistors and capacitors, it behaves as a double-pole single-zero system in the frequency domain. If one of the capacitors has its preset pad in a floating state (i.e., no actual capacitor is soldered), then that node only retains its equivalent parasitic capacitance. Typically, the value of parasitic capacitance is small; therefore, to meet the total compensation capacitance required for system stability, other capacitors need to have larger capacitance values. When the required capacitance is extremely small, this method can also be used to quickly level the bandwidth response. This characteristic can also reduce the design difficulty of high-bandwidth transimpedance amplifier circuits and balanced photodetectors.

[0109] The high-bandwidth, high-gain flatness balanced photodetector based on the transimpedance amplifier circuit structure with feedback network of this application can flatten the gain curve of the high-bandwidth balanced photodetector, avoiding overcompensation and undercompensation. Simultaneously, this structure makes the bandwidth of the high-bandwidth transimpedance amplifier easier to design and improves the debugging cycle of the balanced detector. This increases product yield and reduces the product production cycle from several weeks to one week.

[0110] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A transimpedance amplifier circuit structure supporting fast and fine adjustment, comprising an operational amplifier and a feedback network structure connected between the input and output terminals of the operational amplifier, characterized in that: The feedback network structure consists of multiple resistors and multiple capacitors, which are connected in series and parallel to form a double-pole single-zero transfer function network.

2. The transimpedance amplifier circuit structure supporting fast and fine adjustment according to claim 1, characterized in that: The feedback network structure includes two parallel branches. One branch consists of capacitors C1 and C2 connected in series, and the other branch consists of resistors R2 and R3 connected in series. A resistor R1 is connected between the midpoints of the two parallel branches to finely adjust the relative positions of the zeros and poles.

3. The transimpedance amplifier circuit structure supporting fast and fine adjustment according to claim 2, characterized in that: The feedback network structure is connected between the inverting input and output of the operational amplifier.

4. The transimpedance amplifier circuit structure supporting fast and fine adjustment according to claim 2, characterized in that: There are three strategies for finely controlling the relative positions of zeros and poles using resistor R1. The first is: Strategy 1: Set R1 to a specific resistance value and fine-tune R1 to optimize the frequency response. This strategy is used in application scenarios where the flattest gain flatness is desired.

5. The transimpedance amplifier circuit structure supporting fast and fine adjustment according to claim 2, characterized in that: There are three strategies for finely controlling the relative positions of zeros and poles using resistor R1. The second one is: Strategy 2: Shorting R1 facilitates quick location of zeros and poles. Under this strategy, the transfer function of the transimpedance amplifier circuit structure is simple, allowing for rapid optimization to the vicinity of the target bandwidth.

6. The transimpedance amplifier circuit structure supporting fast and fine adjustment according to claim 2, characterized in that: There are three strategies for finely controlling the relative positions of zeros and poles using resistor R1. The third one is: Strategy 3: With R1 left floating, the feedback network degenerates into a simple parallel RC structure.

7. A transimpedance amplifier circuit structure supporting fast and fine adjustment according to any one of claims 4-6, characterized in that: By leaving a capacitor's preset pad in a floating state, the circuit stability can be optimized using the equivalent parasitic capacitance of the floating pad.

8. The transimpedance amplifier circuit structure supporting fast and fine adjustment according to claim 7, characterized in that: The input terminal of the transimpedance amplifier is connected to the sensing device.

9. The transimpedance amplifier circuit structure supporting fast and fine adjustment according to claim 7, characterized in that: The bandwidth of this transimpedance amplifier circuit is adjustable between 200MHz and 800MHz.

10. A high-bandwidth, high-gain, flatness balanced photodetector, comprising a photocurrent difference signal generation module, an AC / DC separation module, an in-phase proportional amplifier module, and a DC output module, characterized in that: It also includes the transimpedance amplifier circuit structure that supports rapid fine adjustment as described in any one of claims 1-9.

11. A high-bandwidth, high-gain flatness balanced photodetector according to claim 10, characterized in that: The optical signal acquisition module includes a first photodiode PD1 and a second photodiode PD2; the AC / DC separation module includes a DC blocking capacitor C3 and a resistor R9; the cathode of the first photodiode PD1 is connected to the positive bias voltage, the anode of the second photodiode PD2 is connected to the negative bias voltage, and the anode of the first photodiode PD1 is connected to the cathode of the second photodiode PD2, with node a; Node a is connected to one end of DC blocking capacitor C3, and the other end of DC blocking capacitor C3 is connected to the inverting input port of operational amplifier. The output pin of operational amplifier is connected to non-inverting amplifier module, and the output of non-inverting amplifier module is AC signal. Node a is also connected to one end of resistor R9 and the input terminal of the DC output module, which outputs a DC signal.

Images