Remote plasma source reaction chamber temperature control circuit and control method

Abstract

The present invention belongs to the field of semiconductor manufacturing technology, and provides a remote plasma source reaction chamber temperature control circuit and control method. The circuit includes a sampling unit, a control unit, a PWM drive unit, a full-bridge inverter circuit and a resonant converter. The sampling unit is used to collect the output current, voltage and chamber temperature of the resonant converter; the control unit fits the chamber temperature at time k+2 and determines the optimal sequence executed by the PWM drive unit; the PWM drive unit outputs a pulse signal according to the optimal sequence; the full-bridge inverter circuit includes a switch tube S1-S4, and the switch tube is closed or disconnected according to the output signal of the PWM drive unit; the resonant converter is connected to the full-bridge inverter circuit, and an AC signal is output to act on the reaction chamber. The present invention can fit the reaction chamber temperature at time k+2 and compare it with a preset threshold, and then the control unit can further determine the adjustment value of the closed-loop control, ensure the control effect of the chamber temperature, and improve the stability of gas dissociation in the remote plasma source.

Description

Technical Field

[0001] The present invention relates to the technical field of semiconductor manufacturing, and in particular to a reaction cavity temperature prediction circuit and a control method of a remote plasma source. Background Art

[0002] A remote plasma source is a device that obtains plasma by artificial means and is often used in photovoltaic, chemical, semiconductor and other fields. The remote plasma source is mainly composed of a reaction chamber and a power conversion circuit. The power conversion circuit provides power to dissociate the gas in the reaction chamber to form plasma. Plasma and a large amount of heat are formed in the reaction chamber under the action of airflow, electric field, magnetic field, etc. Excessive temperature will lead to a shortened reaction chamber life, low gas dissociation rate, and high failure rate. The heat control effect of the reaction chamber directly affects the safety of equipment operation.

[0003] At present, the overheating monitoring of the reaction chamber of the remote plasma generator can only pre-set the threshold value, and the warning will be issued only when it is triggered. It has low adaptability, weak control accuracy and no temperature prediction ability. It is difficult to take into account all the factors at the same time, such as rapidity, overshoot, stability, etc., which directly affects the quality and efficiency of the industrial process.

[0004] Therefore, how to more accurately and quickly control the temperature in the reaction chamber during the process of monitoring the temperature rise has become an urgent problem to be solved by those skilled in the art. Summary of the invention

[0005] In view of the defects in the prior art, the present invention provides a remote plasma source reaction chamber temperature control circuit and control method to solve the problem that the reaction chamber of the current plasma generator cannot accurately control the temperature.

[0006] In a first aspect, the present invention provides a remote plasma source reaction chamber temperature control circuit, comprising a sampling unit, a control unit, a PWM drive unit, a full-bridge inverter circuit and a resonant converter.

[0007] The sampling unit is used to collect the output current of the resonant converter , output voltage and cavity temperature ;

[0008] The control unit is used to determine the cavity temperature at time k. , fitting the cavity temperature at time k+2 and determining an optimal sequence for execution of the PWM drive unit;

[0009] The PWM driving unit is used to output a pulse signal according to the optimal sequence;

[0010] The full-bridge inverter circuit includes switch tubes S1-S4, and the switch tubes are closed or opened according to the pulse signal output by the PWM drive unit;

[0011] The resonant converter is connected to the full-bridge inverter circuit and outputs an AC signal to act on the reaction cavity.

[0012] It can be seen from the above technical solution that the present invention collects the cavity temperature and the primary winding voltage of the resonant converter at the current moment. , Current , the reaction cavity temperature at time k+2 can be fitted and obtained; the cavity temperature at time k+2 can be compared with the preset empirical threshold, and then the control unit can further determine the adjustment value of the closed-loop control, adjust the pulse signal output by the PWM drive unit, so that the corresponding switch tubes in the switch tubes S1-S4 are closed or disconnected, thereby ensuring the control effect of the cavity temperature.

[0013] Optionally, the resonant converter includes an LCL resonator and a transformer , the LCL resonator and transformer The LCL resonator is connected to the full-bridge inverter circuit.

[0014] In a second aspect, the present invention provides a remote plasma source reaction chamber temperature control method, which is used in a remote plasma source reaction chamber temperature control circuit of any possible implementation of the first aspect, comprising:

[0015] Collect the cavity temperature at time k ;

[0016] Based on the sampling temperature and preset thermal equilibrium stable temperature , get the temperature error at time k ;

[0017] Get the cavity temperature at time k+1 respectively , and the primary winding current at time k+2 , primary winding voltage ; Among them, the cavity temperature The temperature error at time k is calculated by the PID regulator Obtaining the opening and closing degree of the cold water flow valve and the duty cycle of the fan, and acting on the cavity temperature of the reaction cavity at time k+1 later;

[0018] Predict the cavity temperature at time k+2 and actual power ;

[0019] According to the absolute value of the temperature error at time k+2 and preset thresholds , and determining an optimal sequence based on a cost function; a PWM control signal output by the optimal sequence acts on the full-bridge inverter circuit.

[0020] Optionally, the primary winding current at time k+2 is The following prediction is obtained:

[0021] ;

[0022] Primary winding voltage The following prediction is obtained:

[0023] ;

[0024] Among them, the RMS value of the fundamental component of the input square wave voltage is , is the DC input of the full-bridge inverter circuit, the characteristic impedance , quality factors , resonant frequency , is the switching frequency at time k+1, is the inductance ratio.

[0025] Optionally, the cavity temperature at time k+1 is Fitting by the inverse tangent function yields:

[0026] ,

[0027] in, is the cumulative sampling time at time k+1, is the cumulative sampling time at time k, A is the time-varying temperature factor, and B is the influencing factor of the temperature change rate of the reaction chamber temperature.

[0028] Optionally, the cavity temperature at time k+2 and actual power The prediction model predicts that

[0029] Cavity temperature at time k+2 Determined according to the following formula:

[0030] ,in is the initial sampling time, is the initial temperature of the cavity during measurement;

[0031] Actual power at time k+2 Determined according to the following formula:

[0032] ,in For current With voltage The vector angle of .

[0033] Optionally, the temperature error at time k+2 and preset thresholds The relationship between , determines the optimal sequence based on the cost function, including:

[0034] when When , the optimal sequence is determined based on the cost function 1; the cost function 1 ;

[0035] when When , the optimal sequence is determined based on cost function 2; cost function 2 , , are the weight coefficients for adjusting the reaction chamber temperature and actual power respectively.

[0036] By adopting the above technical solution, the present application has the following beneficial effects:

[0037] (1) The present invention collects the current cavity temperature and the primary winding voltage of the resonant converter , Current , the reaction cavity temperature at time k+2 can be fitted and obtained; the cavity temperature at time k+2 can be compared with the preset empirical threshold, and then the control unit can further determine the adjustment value of the closed-loop control, adjust the pulse signal output by the PWM drive unit, so that the corresponding switch tubes in the switch tubes S1-S4 are closed or disconnected, so as to ensure the control effect of the cavity temperature; through the two compensation methods, the cavity temperature at time k can track the predicted value at time k+2, and the influence of control delay can be eliminated as much as possible to improve the control accuracy of the cavity temperature;

[0038] (2) The present invention subtracts the reaction chamber temperature at time k+2 predicted by the control unit from the temperature thermal equilibrium stable value, and determines in real time whether the temperature error value of the remote plasma source chamber is greater than a preset threshold value; different cost functions are substituted according to different situations to select the optimal sequence to control the plasma system, which is beneficial to reducing the amount of calculation in the stable temperature prediction process and alleviating the computing pressure and power consumption of the processor. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for the specific embodiments or the description of the prior art. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn according to the actual scale.

[0040] Figure 1A schematic diagram of a remote plasma source reaction chamber temperature control circuit provided by an embodiment of the present invention is shown;

[0041] Figure 2 A flow chart showing a remote plasma source reaction chamber temperature control method provided by an embodiment of the present invention;

[0042] Figure 3 A logic block diagram of a remote plasma source reaction chamber temperature control method provided by an embodiment of the present invention is shown;

[0043] Figure 4 A schematic diagram showing the compensation of actual power and actual temperature over time provided by an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0044] The following embodiments of the technical solution of the present invention are described in detail in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and are therefore only used as examples, and cannot be used to limit the protection scope of the present invention.

[0045] It should be noted that, unless otherwise specified, the technical terms or scientific terms used in this application should have the common meanings understood by those skilled in the art to which the present invention belongs.

[0046] In one embodiment, Figure 1 As shown, a remote plasma source reaction chamber temperature control circuit is provided, including a sampling unit, a control unit, a PWM drive unit, a full-bridge inverter circuit and a resonant converter. The sampling unit is used to collect the primary winding current of the resonant converter. , primary winding voltage and cavity temperature ; The control unit is used to calculate the cavity temperature at time k , fitting the cavity temperature at time k+2 And determine the optimal sequence executed by the PWM drive unit; the PWM drive unit is used to output a pulse signal according to the optimal sequence; the full-bridge inverter circuit includes switch tubes S1-S4, and the switch tubes S1-S4 are closed or opened according to the pulse signal output by the PWM drive unit; the resonant converter is connected to the full-bridge inverter circuit to output an AC signal to act on the reaction cavity.

[0047] Using the above circuit, the current cavity temperature and the primary winding voltage of the resonant converter are collected. , Current , and the reaction cavity temperature at time k+2 can be fitted. The cavity temperature at time k+2 can be compared with the preset empirical threshold, and then the control unit can further determine the adjustment value of the closed-loop control, adjust the pulse signal output by the PWM drive unit, and make the corresponding switch tubes in the switch tubes S1-S4 closed or disconnected to ensure the control effect of the cavity temperature.

[0048] Specifically, the resonant converter also includes an LCL resonator and a transformer , LCL resonator and full-bridge inverter circuit connection, transformer Connect to the LCL resonator. The LCL resonator includes an inductor , and capacitor ,transformer It is a high-frequency transformer with a turns ratio of n:1. Through transformer coupling, it can achieve efficient energy transmission and reduce costs. The reaction chamber includes inlet and outlet ports and water channels. The gas load is .

[0049] Among them, the control unit can fit the cavity temperature at time k+2 through the prediction model. The prediction model is indispensable for the implementation of MPC. It is a model that describes the operating state of the object. Its purpose is to accurately predict the future state of the system output in the control time domain. It can use the state of the system at time k to predict the output state of the system k+1, so that the output value of the control quantity at the next moment is closer to the expected value, thereby obtaining a better control effect. As the name implies, the prediction model is to predict and control the state of the system, so it is not restricted to the form of the model, nor does it focus on the structure of the model. As long as the input and output relationship of the control quantity can be clearly expressed, it is feasible. There are two common prediction models, one is a parametric model represented by a state equation and a transfer function; the other is a non-parametric model represented by an impulse response or a step signal.

[0050] The prediction model selected in this embodiment is the state space equation, and the discretization form obtained according to the system transfer function is:

[0051] (1)

[0052] Where A, B, and C are the coefficient matrices of the controlled system; is the state quantity, is the output, is the input quantity, i.e. the control quantity; , are the state quantity and input quantity at time k+1 and k respectively.

[0053] The control model is predicted based on formula (1), and the process is as follows:

[0054] (2)

[0055] In formula (2), P is the prediction step size; y(k+j / k) is the output at the k+jth moment; and u(k+1 / k) is the input at the k+1th moment.

[0056] Rewrite formula (2) into matrix form:

[0057] (3)

[0058] in, ; ;

[0059] ; .

[0060] In one embodiment, a remote plasma source reaction chamber temperature control method is provided, which is used for a remote plasma source reaction chamber temperature control circuit provided in the above embodiment, such as Figure 2 As shown, including:

[0061] S1. Collect the cavity temperature at time k .

[0062] S2. Based on sampling temperature and preset thermal equilibrium stable temperature , get the temperature error at time k .

[0063] S3. Obtain the cavity temperature at time k+1 respectively , and the primary winding current at time k+2 , primary winding voltage ; Among them, the cavity temperature The temperature error at time k is calculated by the PID regulator The opening and closing degree of the cold water flow valve and the duty cycle of the fan are obtained, and act on the cavity temperature at the time k+1 after the reaction cavity.

[0064] The current sampling temperature Compared with the thermal equilibrium stable temperature, the output temperature error After passing through the PID regulator, the cooling water flow valve opening and closing degree and the fan duty cycle are obtained and acted on the reaction chamber. After multi-step fitting, the chamber temperature at time k+1 is output. to the model prediction unit; similarly, the primary winding current ,Voltage Also fitted.

[0065] like Figure 3 As shown, the cavity temperature at time k+1 is Fitting by the inverse tangent function yields:

[0066] (4)

[0067] in, is the cumulative sampling time at time k+1, is the initial sampling time, is the initial temperature of the chamber during measurement; A is the time-varying temperature factor, and B is the influencing factor of the temperature change rate of the reaction chamber temperature.

[0068] Furthermore, the reaction chamber temperature at time k+1 is The cavity temperature at time k can be To express:

[0069] (5)

[0070] in, is the cumulative sampling time at time k+1, is the cumulative sampling time at time k.

[0071] For the LCL circuit, due to the resonance of the converter, the resonant frequency of the converter is defined as and switching frequency The normalized value of They are:

[0072] ; (6)

[0073] The RMS value of the equivalent resistance and the fundamental component of the input square wave voltage is expressed as:

[0074] ; (7)

[0075] Among them, fs is the output frequency; the inductance ratio and characteristic impedance of its equivalent circuit are and quality factor Q

[0076] ; (8)

[0077] (9)

[0078] It can be seen that the primary winding current Ipri and voltage Vloop are:

[0079] (10)

[0080] (11)

[0081] Primary winding current at time k+2 The following prediction is obtained:

[0082] (12)

[0083] Primary winding voltage at time k+2 The following prediction is obtained:

[0084] (13)

[0085] Corresponding to the LCL resonant network, the RMS value of the fundamental component of the input square wave voltage , is the DC input of the full-bridge inverter circuit, the characteristic impedance , quality factors , resonant frequency , is the switching frequency at time k+1, is the inductance ratio.

[0086] S4. Predict the cavity temperature at time k+2 and actual power .

[0087] Specifically, the cavity temperature at time k+2 is and actual power Predicted by the prediction model. Cavity temperature at time k+2 The cavity temperature at time k+1 can be And the cavity temperature sampled at time k To express, it is determined according to the following formula:

[0088] (14)

[0089] in is the initial sampling time, is the initial temperature of the cavity during measurement;

[0090] Actual power at time k+2 Determined according to the following formula:

[0091] (15)

[0092] in For current With voltage The vector angle of .

[0093] S5. Based on the absolute value of the temperature error at time k+2 and preset thresholds The optimal sequence is determined based on the cost function; the PWM control signal output by the optimal sequence acts on the full-bridge inverter circuit.

[0094] in, is the predicted cavity temperature at time k+2 Thermal equilibrium stability value The difference between .

[0095] Specifically, step S5 includes:

[0096] when When , the optimal sequence is determined based on the cost function 1; the cost function 1 ;

[0097] when When , the optimal sequence is determined based on cost function 2; cost function 2 , , are the weight coefficients for adjusting the reaction chamber temperature and actual power respectively.

[0098] The output sequence , , , Substitute the corresponding cost function for rolling optimization, and use the minimized cost function to output the optimal sequence to the PWM drive unit, which generates a pulse signal to control the corresponding full-bridge inverter circuit switch to open or close. Based on this, the closed-loop multi-step predictive control of the reaction chamber is achieved, and the influence of control delay is eliminated as much as possible to improve the precise control of the chamber temperature.

[0099] Understandably, Indicates that the switch tubes S1 and S3 are turned on; Indicates that the switch tubes S1 and S4 are turned on. , And so on.

[0100] like Figure 4 As shown, the power is tracked and controlled based on the cost functions G1 and G2, so the actual power Able to track the preset reference power at all times ; For cavity temperature control, the temperature rise of the reaction cavity is monitored, and a corrected control matching result is generated based on the monitoring result. The difference between the actual temperature and the preset thermal equilibrium temperature is compared with the preset threshold to facilitate the subsequent compensation adjustment of the cavity temperature and adjust the heat dissipation system.

[0101] The above embodiments are only used to introduce the technical solutions of the present application in detail, but the description of the above embodiments is only used to help understand the methods of the embodiments of the present invention and should not be understood as limiting the embodiments of the present invention. Any changes or substitutions that can be easily thought of by those skilled in the art should be included in the protection scope of the embodiments of the present invention.

Claims

1. A remote plasma source reaction chamber temperature control method, characterized in that: include: Collect the cavity temperature at time k ; Based on the sampling temperature and preset thermal equilibrium stable temperature , get the temperature error at time k ; Get the cavity temperature at time k+1 respectively , and the primary winding current at time k+2 , primary winding voltage ; Among them, the cavity temperature The temperature error at time k is calculated by the PID regulator Obtaining the opening and closing degree of the cold water flow valve and the duty cycle of the fan, and acting on the cavity temperature of the reaction cavity at time k+1 later; Predict the cavity temperature at time k+2 and actual power ; According to the absolute value of the temperature error at time k+2 and preset thresholds , determining an optimal sequence based on a cost function; a PWM control signal output by the optimal sequence acts on a full-bridge inverter circuit; Cavity temperature at time k+1 Fitting by the inverse tangent function yields: ， in, is the cumulative sampling time at time k+1, is the cumulative sampling time at time k, A is the time-varying temperature factor, and B is the influencing factor of the temperature change rate of the reaction chamber temperature; Cavity temperature at time k+2 and actual power The prediction model predicts that Cavity temperature at time k+2 Determined according to the following formula: ,in is the initial sampling time, is the initial temperature of the cavity during measurement; Actual power at time k+2 Determined according to the following formula: ,in For current With voltage The vector angle of The temperature error according to time k+2 and preset thresholds The relationship between , determines the optimal sequence based on the cost function, including: when When , the optimal sequence is determined based on the cost function 1; the cost function 1 ; when When , the optimal sequence is determined based on cost function 2; cost function 2 , , are the weight coefficients for adjusting the temperature of the reaction chamber and the actual power respectively; is the reference power.

2. The method according to claim 1, characterized in that Primary winding current at time k+2 The following prediction is obtained: ； Primary winding voltage The following prediction is obtained: ； Among them, the RMS value of the fundamental component of the input square wave voltage is , is the DC input of the full-bridge inverter circuit, the characteristic impedance , quality factors , resonant frequency , is the switching frequency at time k+1, is the inductance ratio.

3. A remote plasma source reaction chamber temperature control circuit, characterized in that: Based on the method described in any one of claims 1-2, comprising a sampling unit, a control unit, a PWM drive unit, a full-bridge inverter circuit and a resonant converter, The sampling unit is used to collect the primary winding current of the resonant converter , primary winding voltage and cavity temperature ; The control unit is used to determine the cavity temperature at time k. , fitting the cavity temperature at time k+2 and determining an optimal sequence for execution of the PWM drive unit; The PWM driving unit is used to output a pulse signal according to the optimal sequence; The full-bridge inverter circuit includes switch tubes S1-S4, and the switch tubes are closed or opened according to the pulse signal output by the PWM drive unit; The resonant converter is connected to the full-bridge inverter circuit and is used to output an AC signal to act on the reaction cavity.

4. The circuit according to claim 3, characterized in that The resonant converter includes an LCL resonator and a transformer , the LCL resonator and transformer The LCL resonator is connected to the full-bridge inverter circuit.

Images