JPH03177396A - Temperature control device of semiconductor production device - Google Patents

Abstract

PURPOSE:To control temperature in high accuracy and flexibly in comparison with conventional method by independently controlling a heater operation amount and a cooling water valve operation amount by means of PID control by simultaneously regulating both the operation amounts by adopting a fuzzy control. CONSTITUTION:The title temperature control device is equipped with a pressure sensor 23 to detect pressure P in a chamber 1 and a temperature sensor 13 to detect temperature Ti of a specimen 5, a fussy controller 39 which outputs an operation amount H of heaters 20, 21 and 22 and an operation amount V of a cooling water valve 40 by using deviation T between specimen temperature Ti and target temperature Ts, d T as change thereof and pressure P as input information (calculated by a calculator 36) by fuzzy theory. By the above-mentioned controller, heater control and valve control are totally carried out by making fuzzy rule and membership function proper.

Description

DETAILED DESCRIPTION OF THE INVENTION (a) Industrial Application Field The present invention relates to a temperature control device for controlling the temperature of a reactor in a semiconductor manufacturing device such as a plasma CVD device.

(BL Conventional Technology Figure 4 shows a configuration diagram of a conventional semiconductor manufacturing device and a temperature control device. The figure shows a sputtering device as a semiconductor manufacturing device. A high-frequency electrode is placed inside a vacuum container (chamber) 2 is placed, and a thin film material 3 is applied to this high frequency electrode 2.A susceptor 4 is placed opposite the electrode 2.A sample 5 is placed on the susceptor 4.Sample 5
are made of glass substrates, ceramics, etc. A catalyst gas inlet 6 and an exhaust lower are attached to the chamber 1. Argon gas is often used as the gas. A cooling pipe 8 and a heater 9 are incorporated into the susceptor 4. The valve 10 is for adjusting the amount of cooling water, and the temperature controller 1) controls the heater power source 12. Also,
Inside the chamber l is a temperature sensor 1 that detects the temperature of the sample 5.
3 is attached, and a thermocouple or the like is used for this temperature sensor 13. 14 is a high frequency power source, 15 is an argon gas cylinder, and 16 is an exhaust pump.

The above sputtering device has a ceramic surface (sample 5).
This is a device for growing thin films of metals, semiconductors, and insulators. The operation of this device is generally as follows.

(1) After the chamber I is evacuated using the exhaust pump 16, it is evacuated while introducing a small amount of argon gas to maintain a constant degree of vacuum.

(2) High frequency voltage (13°56
MHz), a glow discharge occurs between the electrode 2 and the sample 5. Argon gas becomes a plasma.

(3) When a negative bias is applied to the electrode 2 or a DC magnetic field is applied, argon gas ions are attracted to the electrode side, collide with the thin film material 3, and eject atoms of the material.

(4) The ejected atoms are stacked on the sample 5 to form a thin film.

The properties of the thin film formed on the sample 5 are strongly influenced by the sample temperature. One of the factors that determines the sample temperature is heat radiation from the high-frequency electrode 2, and the power applied to the electrode 2 changes depending on the material of the film and the film formation speed. In other words, the applied power changes each time the work is performed.

(C) Problems to be Solved by the Invention Temperature control, which is a key point in forming a thin film, has conventionally been carried out by PID control on the heater side while flowing a constant amount of cooling water. However, this method has the following drawbacks.

■Since only the heater is controlled, convergence against disturbances is poor.

(2) It is the fate of PID control that the PID parameters that are optimally adjusted to the target value response when the system is started up cannot optimally control the disturbance response due to the instability of glow discharge. Conversely, if the disturbance response is optimized, a large overshoot will occur in the target value response at startup.

■ Even if the heater and cooling water volumes are controlled by two PID controllers, adjustment becomes difficult because the degree of mutual interference with temperature cannot be recognized.

An object of the present invention is to provide a temperature control device that can solve the above-mentioned drawbacks by performing fuzzy control. a pressure sensor for detecting the sample temperature Ti; a temperature sensor for detecting the sample temperature Ti; and a temperature sensor for detecting the sample temperature Ti.
Deviation Δ]′ between i and target temperature Ts and its change dΔT
and a fuzzy controller that outputs a heater operation amount and a cooling water valve operation amount by fuzzy reasoning using the pressure P as input information.

+8) Effect The temperature control device of the present invention outputs the heater operation amount and the cooling water valve operation amount by fuzzy inference using the deviation between the sample temperature and the target temperature, the change thereof, and the pressure as input information.

By making appropriate fuzzy rules and membership functions, heater control and valve control can be performed comprehensively.

That is, instead of performing heater control and cooling control independently as in conventional PID control, both controls can be performed comprehensively. Further, the fuzzy inference operation can be made extremely strong against disturbances, and the control accuracy can be increased.

(f) Embodiment FIG. 1 shows a block diagram of a temperature control device according to an embodiment of the present invention.

In this embodiment, the heaters provided in the susceptor 4 are composed of three heaters 20, 21, and 22. In addition to the temperature sensor 13 that detects the temperature of the sample 5, a pressure sensor 23 that detects the pressure inside the chamber 1 is provided as a sensor.

The outputs of the pressure sensor 23 and temperature sensor 13 are transmitted to a transducer 31 and a transducer 32., respectively. The signal is guided to an arithmetic unit 36 through an amplifier 34 and an amplifier 35. A target temperature set by an operator using a target temperature setting device 30 is introduced to this calculator 36 through an amplifier 33. In the end, the data input to the calculator 36 are sample temperature Ti, pressure T
, target temperature Ts.

The switch changer 37 switches between the target temperature Ts and the sample 'IL.
Based on the degree Ti, the switch L of the switch unit
Select one of l-L3. This switch LL-L3
is a thyristor 5 for driving the heaters 20 to 22.
Turn on CRI-3CR3. For example, when switch Ll is selected, thyristor 5CRI is turned on and heater 20 is driven. Further, when the switches LL and 'L2 are selected, the thyristors 5CR1 and 5CR2 are turned on and the heaters 20 and 21 are driven. In this embodiment, three heaters 20 to 22 are used in this way, and the heater power can be switched in three stages.
As a result, even if the size of the high-frequency electrode changes and the temperature process changes, within the range of the heater's ability to adapt to the change,
Accurate control becomes possible.

The calculator 36 calculates the deviation ΔT between the sample temperature Ti and the target temperature Ts, and also calculates the change dΔT.

The output of this calculator 36 is the target temperature Ti, the above deviation ΔT
, its change dΔT, and pressure T. These outputs are given to the fuzzy controller 39 as input information for performing fuzzy inference operations. This fuzzy controller 39 performs a heater operation IH for controlling the power supplied to the heater selected by the switch unit 38, and a cooling water valve operation I for controlling the amount of cooling water.
V is output. Both manipulated variables are analog values. The cooling water valve operation amount V is determined by the control unit 41 of the flow rate control valve 40.
The flow rate control valve 40 is outputted here, and the valve adjustment of the flow rate control valve 40 is performed here.

FIG. 2 shows the temperature control rules set in the fuzzy controller 39, and FIG. 3 shows the membership functions.

As is well known, the fuzzy rule is expressed as if...then, and in this example, the variables of the antecedent part are ΔT, dΔT,
P is used. In addition, the variable of the consequent part is the heater operation I.
H and cooling water valve operation 1v are used. Total 13
Among the rules, for example, rule Th1O has the following meaning.

“If ΔT is a little smaller than the target temperature (NS),
If the change is quite large in the negative direction (NB) and the pressure is low (S), increase the heater operation amount (increase the power supplied to the heater) L (B), and leave the cooling water valve operation amount unchanged. Do it.”

When the fuzzy controller 39 performs inference operations based on the above fuzzy control rules, it uses the membership functions shown in FIG. Using the ship function, m
The calculation result based on the 1ni rule is passed to the calculation processing unit of the subsequent consequent part. In the computation of the consequent part, the membership function shown in the lower part of FIG. 3 is used to obtain an inferred definite value by truncation processing according to the max rule.

Through the above fuzzy calculation operation, it is possible to comprehensively judge the film forming conditions within the chamber 1 and simultaneously control the cooling water and the heater with high precision. Further, in the embodiment, since the heater power is switched in three stages, it is possible to set a wide range of temperature control widths in combination with cooling water temperature control. For example, according to experiments, it was possible to control the temperature within a range of approximately 70 to 800° C. by switching the heater power in three stages as described above. Further, since the fuzzy rules and membership functions can reflect the control know-how possessed by the operator, flexible control that cannot be obtained with conventional PID control becomes possible.

(0 Effects of the Invention According to this invention, in order to simultaneously control the heater operation amount and the cooling water valve operation amount by fuzzy reasoning, p
Compared to devices such as lDail control that control these independently, thin films with various properties can be controlled with high precision. Furthermore, since temperature control is performed using fuzzy inference, it is resistant to external disturbances, and since the operator's control know-how can be reflected in membership functions and fuzzy rules, highly accurate and flexible control is possible.

[Brief explanation of drawings]

FIG. 1 is a block diagram of a temperature control device according to an embodiment of the present invention;
FIG. 2 is a diagram showing fuzzy rules, and FIG. 3 is a diagram showing membership functions. Further, FIG. 4 is a schematic configuration diagram of a conventional temperature control device and semiconductor manufacturing device. 13-temperature sensor, 23-pressure sensor, 9-fuzzy controller.

Claims (1)

[Claims] (1) A pressure sensor that detects the pressure P in the chamber, a temperature sensor that detects the sample temperature Ti, a deviation ΔT between the sample temperature Ti and the target temperature Ts, a change dΔT thereof, and the pressure P as input information. A temperature control device for semiconductor manufacturing equipment, comprising: a fuzzy controller that outputs a heater operation amount and a cooling water valve operation amount by fuzzy reasoning;