JPH0577017B2 - - Google Patents

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a method for managing the condition of a clean room, and more particularly, to a method for processing data measured by a particulate detector for managing the cleanliness of a clean room. be.

[Background Art] Recently, as the integration density of patterns in semiconductors has become more sophisticated, dust control in clean rooms, which are manufacturing environments, has become stricter. is per cubic foot
It is said to be 10 or less. In order to effectively maintain such good cleanliness, attention must be paid to cleaning the air with filters, the amount of clean air input, and the distribution route, but at the same time, there are Therefore, it is necessary to accurately measure the ever-changing dust particles and make effective use of this data.

To this end, it has been conventional practice to place particulate detector sensors at key points within a clean room to ascertain the spatial distribution of particulates and provide feedback for improving distribution routes.

On the other hand, the conventional data processing method regarding temporal fluctuations is a simple method that calculates and outputs the cumulative number of particles measured during a certain period of time, such as 3 minutes or 1 hour, depending on the situation. It's just a thing. Figure 4 shows this, where Figure a is a configuration diagram of the data processing device, Figure b is the measurement time,
It is a graph showing the relationship between measured values and output data.
Since the particle detection signal detected by the particle detector 1 corresponds to each particle, its number indicates the number of particles, and this is determined by the microcomputer 2 as the cumulative number n of particles during the time Ts from time t _p to ts. is counted and outputted to the output device 3.

Now, the temporal distribution of dust in a clean room generally fluctuates quite non-uniformly, and in recent highly cleaned conditions, the above-mentioned fixed time Ts has to be quite long, for example over an hour. Without it, it is a fact that valid measurement values cannot be obtained. Therefore, the output of the measurement result cannot be obtained without waiting for a long time, and the measures taken accordingly are delayed. Furthermore, when a large amount of dust is generated temporarily due to special circumstances and the cumulative value becomes excessive, the temporal distribution of the dust generation is unknown and the cause of the dust generation cannot often be ascertained. Moreover, with such data, it is difficult to judge whether it is appropriate to use feedback to control the amount of clean air. As described above, the conventional method for processing measurement data of a particulate detector in a clean room has been incomplete as far as time is concerned.

[Purpose of the Invention] This invention eliminates the drawbacks of the above-mentioned conventional method of processing measurement data of a particulate detector, and in order to effectively utilize this method for dust management in a clean room, the present invention aims to reduce the number of particulates that fluctuate over time. The purpose of this invention is to provide a method for managing the state of a clean room that can perform calculations in a rational manner and output immediately.

[Means for Solving the Problems] In the present invention, the number of particles continuously detected by a particle detector is determined by a moving average value within a certain period of time or a moving cumulative value within a certain period of time (hereinafter referred to as (provisionally referred to as a moving cumulative value). That is, the feature of the clean room condition management method of the present invention is that in the clean room,
The number of particles Pr (however,
r is the number of detections), and calculates and outputs the cumulative value P of n items (n is an integer greater than or equal to 2) from the previously stored number Pi (where i<r) to the number Pr. In addition, when the cumulative value P exceeds a predetermined limit value, an alarm is output or the amount of clean air to the clean room is controlled.

The concept of moving average is defined in statistics. According to this, the interval between the start and end points is a constant interval, and this interval moves as time passes. This is called a moving section, and the moving average value is the average of the number of phenomena that occur within this section per unit time. In this case, instead of taking the average value, if you take the total number of phenomena within the section, that is, the cumulative value over time, you will get the above-mentioned moving cumulative value.

Figure 1 explains the movement cumulative value performed in this invention, where the unit time is ta, nta = T where this continues n times is the movement section, and the number of particles detected in each unit time is P1, P2. ,...Pn,..., the moving cumulative value P is P=Σpr, r: 1 to n. When the time changes in unit time ta, the start and end points of T move by ta and become ti, ti + n-1 (however, a is 1, 2, etc.), and the value of i is updated by 1 every time ta changes in unit time. The value of P for this is newly calculated. The moving cumulative value can be obtained by following the same procedure.

[Operation] Since such a movement cumulative value P is calculated and output every time the section ends, it is clear that the state of temporal fluctuations of particles in the clean room can be accurately grasped. In other words, if the cumulative movement value is calculated for clean room condition management as in the configuration described above, the cumulative value over a long period of time is short for a highly cleaned clean room, and the measurement time unit is short. Moreover, even if the state changes temporarily due to special circumstances, the numerical value will increase rapidly in units of measurement time, and this will have a large effect on the cumulative value, resulting in the obtained result. Therefore, by comparing this with the limit value, it becomes possible to detect an abnormality.

In short, by employing the statistical movement cumulative value, long-term management and short-term management can be achieved simultaneously, and the state of the clean room can be managed in detail. Furthermore, a limit value for P is determined in advance, and when the measured value exceeds this limit value, an alarm or a control signal for increasing the amount of clean air is output. These allow the cleanliness of the clean room to be maintained in a good condition.

The data processing including the calculation of the movement cumulative value described above is performed by a microcomputer, and the limit values for the unit time ta, n, which determines the section length, and P are all set as initial values.

[Embodiment] FIG. 2 shows an embodiment of the equipment configuration when the clean room condition management method of the present invention is applied. The detection signal from the particle detector 1 is sent to the microcomputer 2 every unit time ta. is counted and stored in the memory 4. The movement cumulative value P is calculated and outputted by the microcomputer 2, displayed on the display 5, and printed by the printer 6 at an appropriate time. When the value of P exceeds the limit value, the above-mentioned warning is displayed on the display 5. In addition, the control signal is output to the clean air controller 7, but in this case, the control signal is not output only once, but only when the limit value is exceeded several (N) times or more. All you have to do is use software technology to find a method that suits the actual situation.

FIG. 3 shows a general flowchart of data processing by the computer 2 in the embodiment shown in FIG. A limit value M and a number N for the number of times P exceeds the limit value M are set. Next, the number pr of particles is counted and stored every unit time based on the detection signal inputted from the particle detector 1 (11). The moving cumulative value P is calculated in (12) using the stored number pr, but the calculation method is a simple one using addition and subtraction as described above, and will not be described in detail. The value of P is displayed on the display 5 each time the calculation is performed, and is printed out at the appropriate time (13). In (14), the value of P is checked and when it exceeds the limit value M, a warning to that effect is displayed on the display 5 (15). Then, when the number of times exceeds N (16), a control signal is output to the clean air controller 7 (17). Here, as a method for determining the number N, in addition to cases in which P exceeds M continuously, it is possible to output a control signal depending on the actual situation even if it is not necessarily continuous, which can be easily implemented using a software method. It is.

[Effects of the Invention] As is clear from the above, by applying the clean room condition management method according to the present invention, it is no longer possible to clearly grasp the temporal fluctuations of dust particles in the clean room compared to the conventional method. The current drawbacks have been overcome, and the information can be accurately and immediately obtained, and the so-called limit value management method can be automated to output an alarm or a control signal for the amount of clean air when the number of particulates exceeds the limit value. It enables
It greatly contributes to maintaining the cleanliness of the clean room.

[Brief explanation of drawings]

FIG. 1 is a graph diagram illustrating cumulative movement values in a data processing method for a particle detector when the clean room condition management method of the present invention is applied;
FIG. 2 is a diagram of one embodiment of the equipment configuration when the clean room condition management method of the present invention is applied;
Figure 3 is a flowchart explaining the data processing procedure of the microcomputer in Figure 2, Figure 4 relates to the data processing method of a conventional particle detector, Figure a is a block diagram of the equipment configuration, Figure b
is a graph diagram showing measured values and output data. 1...Particle detector, 2...Microcomputer, 3...Output device, 4...Memory, 5...Display, 6...Printer, 7...Clean air controller.

Claims (1)

[Claims] 1. In a clean room, count the particle detection signals output from the particle detector every unit time to determine the number of particles Pr (where r is the number of detections)
, and the previous number Pi that has already been memorized
(However, when i<r) and the number Pr (n is an integer of 2 or more) of cumulative values P are calculated and output, and when the cumulative value P exceeds a predetermined limit value, A method for managing the condition of a clean room, characterized by outputting an alarm or controlling the flow rate of clean air to the clean room.