COT architecture high-precision output error correction circuit

By designing ripple compensation module and error correction module in the COT architecture, the reference voltage is dynamically corrected, and the problems of output voltage accuracy difference and subharmonic oscillation in traditional COT converters are solved, thus achieving high-precision output and fast response system stability sex.

CN118939056APending Publication Date: 2024-11-12NANJING UNIV OF POSTS & TELECOMM
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410988885.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-23
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

The DC-DC converter of the traditional COT architecture has different output voltage accuracy and subharmonic oscillation, which leads to uncontrollable switching frequency and the error introduced by the ripple compensation circuit is difficult to effectively correct, affecting the system stability.

Method used

A high-precision output error correction circuit is designed, including a ripple compensation module, an error correction module and a summing comparator. The feedback voltage ripple is enhanced through the ripple compensation network, and adaptive adjustment is generated through the error correction module when the load changes. Error voltage, dynamically correct the reference voltage, eliminate the error between the feedback voltage and the reference voltage.

Benefits of technology

It realizes output voltage error correction under large load and wide range conditions, improves the output voltage accuracy and stability of the system, and combines fast transient response and high-precision output.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118939056A_ABST
    Figure CN118939056A_ABST
Patent Text Reader

Abstract

The invention discloses a COT architecture high-precision output error correction circuit, and relates to the technical field of integrated circuits, ripples of feedback voltage are enhanced through a proper ripple compensation network, and when output voltage fluctuates due to load change, the peak values of compensation ripples working under different duty ratios are different, so that the output error correction circuit can be used for correcting the output error. The error correction circuit adopting the subthreshold design can detect the output voltage in real time, quickly responds to generate the error voltage for self-adaptive adjustment, and eliminates the error between the feedback voltage and the reference voltage caused by ripple compensation by dynamically correcting the reference voltage through the error voltage in the summation comparator. Therefore, the working stability of the system is improved under the condition of wide input and wide load, and high-precision output voltage and rapid transient response are realized.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of integrated circuits, and in particular relates to a COT architecture high-precision output error correction circuit. Background Art

[0002] With the development of emerging technologies such as new energy vehicles, cloud computing, and the Internet of Things, high-performance electronic devices are increasingly favored by people, and the power management chip industry that provides them with stable power is thriving. The DC-DC converter with COT architecture has become one of the research hotspots due to its advantages of simple structure and fast response speed, but the stability of the traditional COT mode is heavily dependent on the ESR of the output capacitor, and there is a defect of poor output voltage accuracy. In line with the development trend of high precision, low EMI, and fast transient response, the present invention studies the ripple compensation circuit to enhance the feedback voltage ripple to achieve a stable output voltage, and generates an adaptive error voltage correction reference voltage through an error correction circuit to improve the output voltage accuracy. The topology diagram of the synchronous DC-DC step-down converter is shown in the figure. Figure 1 As shown, it mainly includes the main switch tube HS, the synchronous rectifier tube LS, the inductor L, and the output capacitor C OUT , capacitor equivalent series resistance R ESR , load resistance R L When the converter uses ceramic capacitors with smaller ESR, the key signal waveforms are as follows: Figure 2 As shown, the output voltage V OUT Ripple is caused by R ESR Voltage ripple V ESR and C OUT Charge and discharge ripple V COUT When V ESR Less than V COUT When the output voltage V OUT Ripple and feedback voltage V FB The ripple is dominated by the capacitor charging and discharging ripple. Since the capacitor voltage lags behind the current, when the inductor current rises, V OUT and V FB will not rise accordingly, the converter turns on T ON After time V FB Still less than the reference voltage V REF , so the converter will repeatedly trigger T ON On time until V FB Greater than V REF , the converter is continuously turned on for two or more cycles, which is called subharmonic oscillation. It will cause the switching frequency to be uncontrollable and the output voltage ripple to increase sharply. In order to ensure the stable operation of the system, it is necessary to introduce a ripple compensation circuit to enhance the feedback voltage ripple, but it will cause V FB With V REFThe error becomes larger, so a high-precision error correction module is needed to improve the output voltage accuracy by adaptively correcting the reference voltage.

[0003] The existing COT architecture uses an error amplifier EA to amplify V REF With V FB The error voltage between V C , V FB With V C By comparing with the loop comparator, the output voltage accuracy drop caused by small ESR can be improved under the regulation of EA, but the offset and delay of the loop comparator will introduce new output errors. There are also some circuits that improve and optimize the ripple compensation circuit, so that the ripple voltage with a valley voltage of 0 generated by the ripple compensation circuit is consistent with V FB Superposition, the valley voltage is V FB The voltage and V REF By comparison, the output error introduced by the ripple compensation circuit is eliminated. However, when the duty cycle changes significantly, the peak-to-peak value of the compensated ripple will change, causing the valley value of the ripple voltage to deviate, and the effect of error correction will be greatly reduced. Summary of the invention

[0004] To achieve the above-mentioned purpose, the technical solution of the present invention is as follows: a COT architecture high-precision output error correction circuit, the circuit includes a ripple compensation module, an error correction module and a summing comparator, the ripple of the feedback voltage is enhanced by a suitable ripple compensation network, and the output voltage is detected by the error correction circuit when the load changes, to generate a fast adaptively adjusted error voltage, and the reference voltage is dynamically corrected in the summing comparator to eliminate the error between the feedback voltage and the reference voltage after the ripple compensation is introduced, so as to realize the error correction of the output voltage under large load and wide range conditions.

[0005] As an improvement of the present invention, the ripple compensation module generates a square wave signal V in phase with the switch node SW. P and the inverted square wave signal V N , through the filter network to generate the inductor current I L The same phase ripple voltage V RIPPLEP and the reverse ripple voltage V RIPPLEN The differential mode signal is subtracted through the transconductance amplifier to eliminate the common mode component and obtain a ripple current with a DC value of zero.

[0006] As an improvement of the present invention, the ripple current I with a DC value of zero is obtained. RIPPLE , the output is converted into a voltage by the summing comparator and superimposed on the reference voltage V FB The ripple enhancement voltage V FBN .

[0007] As an improvement of the present invention, when the load changes and the output voltage fluctuates, the compensation ripple changes with the duty cycle, and the ripple enhancement voltage V FBN1 With V FBN2 With different peak-to-peak V PP1 With V PP2 The error correction module generates an adaptive error voltage V as the duty cycle changes. OFFSET1 With V OFFSET2 , superimposed on V REF The reference voltage is dynamically corrected under different duty cycles, so that V FB With V REF Still equal, eliminating the half-ripple error between the feedback voltage and the reference voltage.

[0008] As an improvement of the present invention, the summing comparator converts V FB Ripple Enhancement and V REF After error correction, the operational amplifier compares V FBN and V REFN , when V FBN Drop to V REFN When the output signal V COMP The logic control module controls the conduction of the main switch tube to maintain stable operation of the system and improve the output accuracy of the system.

[0009] As an improvement of the present invention, the ripple compensation module generates a ripple compensation voltage by using a voltage divider network and a filter network, converts the differential mode voltage into a ripple current output with a DC value of zero through a transconductance amplifier, and outputs the ripple current through a resistor R a1 , R a2 The switch node voltage V SW Voltage division, voltage division resistor ratio To ensure that the common-mode input level of the subsequent transconductance amplifier is within the normal operating range, a wide input range is achieved, and the input voltage V IN Generated by LDO structure Set the resistor ratio so that V LDO =aV IN Power the inverter and transmission gate, and generate the same voltage as V through two-stage inverters. SW In phase, amplitude is aV IN The square wave signal V P , the ripple voltage V is obtained through the filter network R1C1 RIPPLEP , its DC voltage:

[0010] V DC =aV IN *D=aV OUT

[0011] When the main switch is turned on, the voltage rise slope of C1 is:

[0012]

[0013] When the synchronous rectifier is turned on, the voltage drop slope of C1 is:

[0014]

[0015] Therefore, when R1C1 matches the inductor L, a ripple V in phase with the inductor current can be obtained. RIPPLEP , its peak-to-peak value is:

[0016]

[0017] The adaptive constant on-time T ON =DR ON C ON , R ON and C ON is determined by the converter T ON The resistance and capacitance match the R1C1 in the filter network and R1C1=N1R ON C ON , we can get:

[0018]

[0019] Through the inverter and transmission gate to get V P Inverted square wave signal V N The transmission gate ensures that the two are in opposite phases. After the filter network R0C0 equal to R1C1, a ripple voltage V inversely proportional to the inductor current is obtained. RIPPLEN , differential signal V RIPPLEP and V RIPPLEN The common mode component is eliminated by the transconductance amplifier. Get the ripple current with zero DC value:

[0020]

[0021] As an improvement of the present invention, when the output voltage fluctuates due to load changes, the peak-to-peak value of the compensation ripple voltage changes with the converter working at different duty cycles, so that the reference voltage minus the fixed voltage cannot guarantee that the converter can achieve high-precision output in all cases. Therefore, it is necessary to generate an adaptive adjustment error voltage to dynamically correct the reference voltage. The differential mode voltage V generated by the ripple compensation module RIPPLEN and V RIPPLEP Through the resistor R c1 , R c2 Get the DC common mode voltage V DC =aV OUT , input voltage V IN Through the voltage divider resistor R b1 , R b2The voltage is sampled:

[0022]

[0023] Op amps OP2 and OP3 use negative feedback loops to convert aV OUT and bV IN Converted into current:

[0024]

[0025] PM1, PM2, PM3, PM4 have the same width to length ratio, PM5, PM6 have the same width to length ratio, NM9, NM10 have the same width to length ratio, and the current can be replicated in equal proportion through the current mirror. Using subthreshold design, that is, NM2, NM3, NM4, and NM5 work in the subthreshold region, the relationship between the drain-source current and the gate-source voltage can be expressed as:

[0026]

[0027] Where I0 is the subthreshold current pre-exponential factor, which is proportional to W / L, m>1, is the subthreshold slope factor, k is the Boltzmann constant, T is the temperature, q is the electron charge, the drain-source current of the MOS tube working in the subthreshold region is much smaller than that in the saturation region, which can reduce static power consumption. The relationship between the gate-source voltage and the drain-source current of the MOS tube in the subthreshold region is:

[0028]

[0029] According to Kirchhoff's voltage law, the gate-source voltage relationship of NM2, NM3, NM4, and NM5 is:

[0030]

[0031] Right now:

[0032]

[0033] Simplified, the drain-source current of NM5 is:

[0034]

[0035] The width-to-length ratio of NM7 is N2 times that of NM6, and the width-to-length ratio of PM8 is N3 times that of PM7. The current mirror can be used to replicate the current in proportion. The error current is generated by subtracting the current through the subtraction circuit:

[0036]

[0037] when Right now When , we can get:

[0038]

[0039] When load changes cause the output voltage to fluctuate too high or too low, the error current is quickly adaptively adjusted as the duty cycle changes, and the reference voltage is dynamically corrected in the summing comparator, thereby achieving output voltage error correction, allowing the converter to have both fast transient response and high output accuracy.

[0040] As an improvement of the present invention, the ripple current I output by the ripple compensation module RIPPLE Converted into a peak-to-peak ripple voltage V RIPPLE Then superimposed on V FB The error correction module outputs the error current I OFFSET Converted into error voltage V OFFSET Then superimposed on V REF The operational amplifier compares the ripple enhancement voltage V FBN and the new reference voltage V after error correction REFN , output signal V COMP The logic control module determines whether the main switch is turned on to achieve V FB With V REF The half-ripple error is eliminated, the output voltage accuracy is improved, and the ripple current I RIPPLE Flows through R5 to regenerate the ripple compensation voltage V RIPPLE =I RIPPLE R5, its peak-to-peak value:

[0041]

[0042] Generate V on R5 FB Ripple enhancement voltage:

[0043]

[0044] Then V FBN Minimum value of:

[0045]

[0046] The error current I from the error correction module OFFSET The error voltage generated by R4 is:

[0047]

[0048] Generates the correction V on R4 REF The new reference voltage after:

[0049]

[0050] Set R4=R5, make Can achieve V MIN,FBN = REFN , the right side op amp completes V FBN and V REFN When V FBN Drop to V REFN When the output signal V COMP The high level is triggered, and the main switch is turned on through the logic control circuit. At this time, the average value of the ripple enhancement voltage is V FB With V REF Equal, eliminating the half ripple error between the feedback voltage and the reference voltage, and achieving high-precision output voltage. PM8 and PM9 use a common source and common gate structure to make the output node V A The maximum value is reduced to NM3 uses a diode connection method. FBN Greater than V REFN When it is turned on, NM2 and NM3 add a current path to make the output node V A The minimum value increases to The response speed of the operational amplifier is improved by reducing the output node voltage variation range, thereby achieving a fast transient response.

[0051] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention generates an adaptively regulated error voltage according to different duty cycles, and eliminates the error between the feedback voltage and the reference voltage after the ripple compensation is introduced by dynamically correcting the reference voltage, thereby achieving accurate voltage output. When the load changes and the output voltage fluctuates too high or too low, the peak-to-peak value of the compensation ripple working under different duty cycles is different. Because the generated error voltage is strictly related to the duty cycle, the reference voltage can be quickly adaptively regulated, and the correction effect on the output voltage is not affected when the load changes, which is conducive to optimizing the dynamic characteristics of the converter in a large load and wide range under the continuous working mode, thereby ensuring the stable operation of the system under wide input and wide load conditions, and achieving high-precision output voltage and fast transient response; the present invention also includes a ripple compensation circuit, which generates an internal ripple compensation voltage through the SW terminal voltage to enhance the feedback voltage ripple. When RC / a is highly matched with L, the generated ripple voltage is strictly in phase with the inductor current, and the common mode component is eliminated by subtracting the differential mode signal, and a ripple compensation voltage with a DC value of zero is obtained to achieve feedback voltage ripple enhancement, so that the ripple compensation of the output voltage can be achieved when the output capacitor ESR is small, thereby improving the working stability of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 This is a diagram of the existing technical system architecture;

[0053] Figure 2It is a waveform diagram of key signals in the prior art;

[0054] Figure 3 Schematic diagram of the high-precision output error correction circuit in this embodiment;

[0055] Figure 4 This is a key signal waveform diagram of the high-precision output error correction circuit in this embodiment;

[0056] Figure 5 This is the circuit diagram of the ripple compensation module in this embodiment;

[0057] Figure 6 is a circuit diagram of the error correction module in this embodiment;

[0058] Figure 7 FIG. 4 is a circuit diagram of a summing comparator in this embodiment. DETAILED DESCRIPTION

[0059] The present invention will be further explained below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the following specific embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention.

[0060] Embodiment: This embodiment can be applied to a COT mode synchronous rectification DC-DC buck converter. The schematic diagram of the high-precision output error correction circuit proposed in this embodiment is as follows: Figure 3 As shown, it is mainly composed of three units, namely a ripple compensation module, an error correction module, and a summing comparator.

[0061] The key signal waveforms of the high-precision output error correction circuit are as follows: Figure 4 As shown, the ripple compensation module generates a square wave signal V in phase with the switch node SW. P and the inverted square wave signal V N , through the filter network to generate the inductor current I L The same phase ripple voltage V RIPPLEP and the reverse ripple voltage V RIPPLEN The differential mode signal is subtracted through the transconductance amplifier to eliminate the common mode component and obtain a ripple current I with a DC value of zero. RIPPLE , the output is converted into a voltage by the summing comparator and superimposed on the reference voltage V FB The ripple enhancement voltage V FBN The present invention adopts the valley voltage control mode and introduces the ripple-compensated V FBN Average V FB Always higher than the reference voltage V REF High, the difference is the peak-to-peak value of the ripple V PP The average output voltage will also be higher than the target value, so an error correction module is required to generate an adaptive error current I OFFSET , which is converted into an error voltage V in the summing comparatorOFFSET After V REF Add together to get a new reference voltage V REFN When the load changes and the output voltage fluctuates, the compensation ripple changes with the duty cycle, and the ripple enhancement voltage V FBN1 With V FBN2 With different peak-to-peak V PP1 With V PP2 The error correction module generates an adaptive error voltage V as the duty cycle changes. OFFSET1 With V OFFSET2 , superimposed on V REF The reference voltage is dynamically corrected under different duty cycles, so that V FB With V REF Still equal, the half ripple error between the feedback voltage and the reference voltage is eliminated, significantly improving the accuracy of the output voltage. FB Ripple Enhancement and V REF After error correction, the operational amplifier compares V FBN and V REFN , when V FBN Drop to V REFN When the output signal V COMP The logic control module controls the conduction of the main switch tube to maintain stable operation of the system and improve the output accuracy of the system.

[0062] The following will explain the working principle of the entire system in detail by explaining the specific working principles of the system submodules:

[0063] Ripple compensation module such as Figure 5 As shown, the ripple compensation module generates a ripple compensation voltage by using a voltage divider network and a filter network, converts the differential mode voltage into a ripple current output with a zero DC value through a transconductance amplifier, and outputs the ripple current through a resistor R a1 , R a2 The switch node voltage V SW Voltage division, voltage division resistor ratio To ensure that the common-mode input level of the subsequent transconductance amplifier is within the normal operating range, a wide input range is achieved, and the input voltage V IN Generated by LDO structure Set the resistor ratio so that V LDO =aV IN Power the inverter and transmission gate, and generate the same voltage as V through two-stage inverters. SW In phase, amplitude is aV IN The square wave signal V P , the ripple voltage V is obtained through the filter network R1C1 RIPPLEP , its DC voltage:

[0064] V DC=aV IN *D=aV OUT

[0065] When the main switch is turned on, the voltage rise slope of C1 is:

[0066]

[0067] When the synchronous rectifier is turned on, the voltage drop slope of C1 is:

[0068]

[0069] Therefore, when R1C1 matches the inductor L, a ripple V in phase with the inductor current can be obtained. RIPPLEP , its peak-to-peak value is:

[0070]

[0071] The adaptive constant on-time T ON =DR ON C ON , R ON and C ON is determined by the converter T ON The resistance and capacitance match the R1C1 in the filter network and R1C1=N1R ON C ON , we can get:

[0072]

[0073] Through the inverter and transmission gate to get V P Inverted square wave signal V N The transmission gate ensures that the two are in opposite phases. After the filter network R0C0 equal to R1C1, a ripple voltage V inversely proportional to the inductor current is obtained. RIPPLEN , differential signal V RIPPLEP and V RIPPLEN The common mode component is eliminated by the transconductance amplifier. Get the ripple current with zero DC value:

[0074]

[0075] Compared with the traditional two-stage filtering network to obtain DC voltage, the resistance value of the filter capacitor is reduced and the system integration is improved. RIPPLE The output is converted into a voltage by the summing comparator and superimposed on the reference voltage V FB The ripple enhancement voltage V FBN , can optimize the stability of the system when the ESR is small, but using V FBN Valley value and reference voltage V REF When compared, the average value VFB With V REF The existence of errors reduces the accuracy of the output voltage, so it is necessary to introduce an error correction module to lower the reference voltage to achieve the purpose of improving the output voltage accuracy.

[0076] Furthermore, when the output voltage fluctuates due to load changes, the peak-to-peak value of the compensation ripple voltage changes with the converter operating at different duty cycles, so that the reference voltage minus the fixed voltage cannot guarantee that the converter can achieve high-precision output in all cases. Therefore, it is necessary to generate an adaptive error voltage to dynamically correct the reference voltage. The error correction module is as follows: Figure 6 As shown, the differential mode voltage V generated by the ripple compensation module RIPPLEN and V RIPPLEP Through the resistor R c1 , R c2 Get the DC common mode voltage V DC =aV OUT , input voltage V IN Through the voltage divider resistor R b1 , R b2 The voltage is sampled:

[0077]

[0078] Op amps OP2 and OP3 use negative feedback loops to convert aV OUT and bV IN Converted into current:

[0079]

[0080] PM1, PM2, PM3, PM4 have the same width to length ratio, PM5, PM6 have the same width to length ratio, NM9, NM10 have the same width to length ratio, and the current can be replicated in equal proportion through the current mirror. Using subthreshold design, that is, NM2, NM3, NM4, and NM5 work in the subthreshold region, the relationship between the drain-source current and the gate-source voltage can be expressed as:

[0081]

[0082] Where I0 is the subthreshold current pre-exponential factor, which is proportional to W / L, m>1, is the subthreshold slope factor, k is the Boltzmann constant, T is the temperature, q is the electron charge, the drain-source current of the MOS tube working in the subthreshold region is much smaller than that in the saturation region, which can reduce static power consumption. The relationship between the gate-source voltage and the drain-source current of the MOS tube in the subthreshold region is:

[0083]

[0084] According to Kirchhoff's voltage law, the gate-source voltage relationship of NM2, NM3, NM4, and NM5 is:

[0085]

[0086] Right now:

[0087]

[0088] Simplified, the drain-source current of NM5 is:

[0089]

[0090] The width-to-length ratio of NM7 is N2 times that of NM6, and the width-to-length ratio of PM8 is N3 times that of PM7. The current mirror can be used to replicate the current in proportion. The error current is generated by subtracting the current through the subtraction circuit:

[0091]

[0092] when Right now When , we can get:

[0093]

[0094] When load changes cause the output voltage to fluctuate too high or too low, the error current is quickly adaptively adjusted as the duty cycle changes, and the reference voltage is dynamically corrected in the summing comparator, thereby achieving output voltage error correction, allowing the converter to have both fast transient response and high output accuracy.

[0095] Furthermore, the summing comparator is as follows Figure 7 As shown, the ripple current I output by the ripple compensation module RIPPLE Converted into a peak-to-peak ripple voltage V RIPPLE Then superimposed on V FB The error correction module outputs the error current I OFFSET Converted into error voltage V OFFSET Then superimposed on V REF The operational amplifier compares the ripple enhancement voltage V FBN and the new reference voltage V after error correction REFN , output signal V COMP The logic control module determines whether the main switch is turned on to achieve V FB With V REF The half-ripple error is eliminated, the output voltage accuracy is improved, and the ripple current I RIPPLE Flows through R5 to regenerate the ripple compensation voltage V RIPPLE =I RIPPLE R5, its peak-to-peak value:

[0096]

[0097] Generate V on R5 FB Ripple enhancement voltage:

[0098]

[0099] Then V FBN Minimum value of:

[0100]

[0101] The error current I from the error correction module OFFSET The error voltage generated by R4 is:

[0102]

[0103] Generates the correction V on R4 REF The new reference voltage after:

[0104]

[0105] Set R4=R5, make Can achieve V M1N,FBN = REFN , the right side op amp completes V FBN and V REFN When V FBN Drop to V REFN When the output signal V COMP The high level is triggered, and the main switch is turned on through the logic control circuit. At this time, the average value of the ripple enhancement voltage is V FB With V REF Equal, eliminating the half ripple error between the feedback voltage and the reference voltage, and achieving high-precision output voltage. PM8 and PM9 use a common source and common gate structure to make the output node V A The maximum value is reduced to NM3 uses a diode connection method. FBN Greater than V REFN When it is turned on, NM2 and NM3 add a current path to make the output node V A The minimum value increases to By reducing the output node voltage variation range, the operational amplifier response speed is improved to achieve fast transient response. In addition, the Schmitt trigger and inverter cascade prevents V COMP False triggering due to noise affects the stable operation of the system.

[0106] It should be noted that the above content only illustrates the technical idea of ​​the present invention and cannot be used to limit the protection scope of the present invention. For ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications all fall within the protection scope of the claims of the present invention.

Claims

1. A COT architecture high-precision output error correction circuit, characterized in that: The circuit includes a ripple compensation module, an error correction module and a summing comparator. The ripple of the feedback voltage is enhanced by a suitable ripple compensation network, and the output voltage is detected by the error correction circuit when the load changes, so as to generate a fast adaptively adjusted error voltage. The reference voltage is dynamically corrected in the summing comparator to eliminate the error between the feedback voltage and the reference voltage after the ripple compensation is introduced, so as to realize the error correction of the output voltage under the condition of large load and wide range.

2. A COT architecture high-precision output error correction circuit according to claim 1, characterized in that: The ripple compensation module generates a square wave signal V in phase with the switch node SW. P and the inverted square wave signal V N , through the filter network to generate the inductor current I L The same phase ripple voltage V RIPPLEP and the reverse ripple voltage V RIPPLEN The differential mode signal is subtracted through the transconductance amplifier to eliminate the common mode component and obtain a ripple current with a DC value of zero.

3. A COT architecture high-precision output error correction circuit according to claim 2, characterized in that: The ripple current I with a DC value of zero is obtained. RIPPLE , the output is converted into a voltage by the summing comparator and superimposed on the reference voltage V FB The ripple enhancement voltage V FBN .

4. A COT architecture high-precision output error correction circuit according to claim 3, characterized in that: When the load changes and the output voltage fluctuates, the compensation ripple changes with the duty cycle, and the ripple enhancement voltage V FBN1 With V FBN2 With different peak-to-peak V PP1 With V PP2 The error correction module generates an adaptive error voltage V as the duty cycle changes. OFFSET1 With V OFFSET2 , superimposed on V REF The reference voltage is dynamically corrected under different duty cycles, so that V FB With V REF Still equal, eliminating the half-ripple error between the feedback voltage and the reference voltage.

5. A COT architecture high-precision output error correction circuit according to claim 4, characterized in that: The summing comparator converts V FB Ripple Enhancement and V REF After error correction, the operational amplifier compares V FBN and V REFN , when V FBN Drop to V REFN When the output signal V COMP The logic control module controls the conduction of the main switch tube to maintain stable operation of the system and improve the output accuracy of the system.

6. The COT architecture high-precision output error correction circuit according to claim 1, characterized in that: The ripple compensation module generates a ripple compensation voltage by using a voltage divider network and a filter network, converts the differential mode voltage into a ripple current output with a DC value of zero through a transconductance amplifier, and transmits the ripple current to the output through a resistor R a1 , R a2 The switch node voltage V SW Voltage division, voltage division resistor ratio To ensure that the common-mode input level of the subsequent transconductance amplifier is within the normal operating range, a wide input range is achieved, and the input voltage V IN Generated by LDO structure Set the resistor ratio so that V LDO =aV IN Power the inverter and transmission gate, and generate the same voltage as V through two-stage inverters. SW A square wave signal with the same phase and amplitude V P , the ripple voltage V is obtained through the filter network R1C1 RIPPLEP , its DC voltage: V DC =aV IN *D=aV OUT When the main switch is turned on, the voltage rise slope of C1 is: When the synchronous rectifier is turned on, the voltage drop slope of C1 is: Therefore, when R1C1 matches the inductor L, a ripple V in phase with the inductor current can be obtained. RIPPLEP , its peak-to-peak value is: The adaptive constant on-time T ON =DR ON C ON , R ON and C ON is determined by the converter T ON The resistance and capacitance match the R1C1 in the filter network and R1C1=N1R ON C ON , we can get: Through the inverter and transmission gate to get V P Inverted square wave signal V N The transmission gate ensures that the two are in opposite phases. After the filter network R0C0 equal to R1C1, a ripple voltage V inversely proportional to the inductor current is obtained. RIPPLEN , differential signal V RIPPLEP and V RIPPLEN The common mode component is eliminated by the transconductance amplifier. Get the ripple current with zero DC value:

7. The COT architecture high-precision output error correction circuit according to claim 1, characterized in that: When the load changes and the output voltage fluctuates, the peak-to-peak value of the compensation ripple voltage changes with the converter working at different duty cycles, so the reference voltage minus the fixed voltage cannot guarantee that the converter can achieve high-precision output in all cases. Therefore, it is necessary to generate an adaptive error voltage to dynamically correct the reference voltage. The differential mode voltage V generated by the ripple compensation module RIPPLEN and V RIPPLEP Through the resistor R c1 , R c2 Get the DC common mode voltage V DC =aV OUT , input voltage V IN Through the voltage divider resistor R b1 , R b2 The voltage is sampled: Op amps OP2 and OP3 use negative feedback loops to convert aV OUT and bV IN Converted into current: PM1, PM2, PM3, PM4 have the same width to length ratio, PM5, PM6 have the same width to length ratio, NM9, NM10 have the same width to length ratio, and the current can be replicated in equal proportion through the current mirror. Using subthreshold design, that is, NM2, NM3, NM4, and NM5 work in the subthreshold region, the relationship between the drain-source current and the gate-source voltage can be expressed as: Where I0 is the subthreshold current pre-exponential factor, which is proportional to W / L, m>1, is the subthreshold slope factor, k is the Boltzmann constant, T is the temperature, q is the electron charge, the drain-source current of the MOS tube working in the subthreshold region is much smaller than that in the saturation region, which can reduce static power consumption. The relationship between the gate-source voltage and the drain-source current of the MOS tube in the subthreshold region is: According to Kirchhoff's voltage law, the gate-source voltage relationship of NM2, NM3, NM4, and NM5 is: Right now: Simplified, the drain-source current of NM5 is: The width-to-length ratio of NM7 is N2 times that of NM6, and the width-to-length ratio of PM8 is N3 times that of PM7. The current mirror can be used to replicate the current in proportion. The error current is generated by subtracting the current through the subtraction circuit: when Right now When , we can get: When load changes cause the output voltage to fluctuate too high or too low, the error current is quickly adaptively adjusted as the duty cycle changes, and the reference voltage is dynamically corrected in the summing comparator, thereby achieving output voltage error correction, allowing the converter to have both fast transient response and high output accuracy.

8. The COT architecture high-precision output error correction circuit according to claim 1, characterized in that: The ripple current I output by the ripple compensation module RIPPLE Converted into a peak-to-peak ripple voltage V RIPPLE Then superimposed on V FB The error correction module outputs the error current I OFFSET Converted into error voltage V OFFSET Then superimposed on V REF The operational amplifier compares the ripple enhancement voltage V FBN and the new reference voltage V after error correction REFN , output signal V COMP The logic control module determines whether the main switch is turned on to achieve V FB With V REF The half-ripple error is eliminated, the output voltage accuracy is improved, and the ripple current I RIPPLE Flows through R5 to regenerate the ripple compensation voltage V RIPPLE =I RIPPLE R5, its peak-to-peak value: Generate V on R5 FB Ripple enhancement voltage: Then V FBN Minimum value of: The error current I from the error correction module OFFSET Flowing through R4 generates an error voltage: Generates the correction V on R4 REF The new reference voltage after: Set R4=R5, make Can achieve V MIN,FBN =V REFN , the right side op amp completes V FBN and V REFN When V FBN Drop to V REFN When the output signal V COMP The high level is triggered, and the main switch is turned on through the logic control circuit. At this time, the average value of the ripple enhancement voltage is V FB With V REF Equal, eliminating the half ripple error between the feedback voltage and the reference voltage, and achieving high-precision output voltage. PM8 and PM9 use a common source and common gate structure to make the output node V A The maximum value is reduced to NM3 uses a diode connection method. FBN Greater than V REFN When it is turned on, NM2 and NM3 add a current path to make the output node V A The minimum value increases to The response speed of the operational amplifier is improved by reducing the output node voltage variation range, thereby achieving a fast transient response.