JPH0357970A - Pulse input apparatus - Google Patents

Abstract

PURPOSE:To prepare the order set suitable for the measurement of a time interval by taking in a plurality of input signals at the same time to once store the same in an input memory and reading the signals at every channel to measure event generating time. CONSTITUTION:At least one register storing the measuring result of the order executed immediately before in an order execution part and a means adding at least one flag showing the measuring state of the order executed immediately before and having function calculating the generation time of an event and calculating the difference between the measured value and the value stored in the register are mounted. The first order measures the generation time of the first event and the second order measures the generation time of the second event and this measured value and the measured result of the first order are used to measure a time interval. The result of the event generated time mea sured is stored in the register and a flag is set corresponding to a measuring state by the first order and the difference between the measured value and the register memory value is calculated corresponding to the flag state.

Description

[Detailed Description of the Invention] [Objective of the Invention] (Industrial Field of Application) The present invention relates to a pulse input device that measures input pulses, and in particular, to central processing of pulse time intervals such as input pulse change times and pulse periods. This invention relates to a pulse input device that can perform calculations without depending on the device.

(Prior Art) Pulse input devices for measuring input pulses and processing the results in a central processing unit are widely used as state detection devices for various devices, together with control devices for these devices. The function of these devices is to measure time intervals such as the time when an input signal changes, pulse width, period, etc. As this type of pulse input device, for example, there is a device disclosed in Japanese Patent Application No. 132215/1982. Similar techniques are known.

FIG. 4 shows the configuration of the pulse input device according to the above-mentioned prior art. In the same figure, the block indicated by the dotted line is the pulse input device 4.
01 and is connected to a CPU (central processing unit) 403 via a bus 402.

The pulse input device receives a plurality of input signals 404 from the outside and proceeds with processing based on instructions from the CP0403 written in the instruction memory 405.

Then, by writing the result into the instruction memory, the CPU is informed of the measurement result of the input pulse. The pulse input device includes a timer counter 40 that generates reference time information as a basic element for executing instructions.
6. Input circuit 407 that samples the input signal periodically;
An input memory 408 for storing sampled input signals for a while, and an instruction execution unit 40 that sequentially executes processes necessary for executing instructions to control the entire pulse input device.
It is composed of 9.

The input circuit 407 samples the input signal at a predetermined period and transmits it to the input memory 408. The predetermined period usually coincides with the period at which the timer counter 406 is updated, and the input signal is sampled by a sample clock created by frequency-dividing the clock.

Further, the input memory 407 has a function similar to that of a FIFO, and can hold new input signals and erase old input signals, that is, update information. Therefore, the period during which the input signal is held is constant. The held data is read out from the instruction execution unit 409 in units of channels.

That is, when data is input, it is read out and an event specified by a command is detected.

For example, to find the time of change in the rise of the input signal of channel 3, the input history of channel 3 stored in the input memory 408 is read out all at once, and the portion where the change in the rise has occurred is examined. Furthermore, the input memory 407
It has a FO (first in, first out) configuration and is updated at regular intervals, so by checking the part where a change occurred, it is possible to know how long ago the change occurred. Using the change value and the current timer value, the time at which the event occurred can be determined.

The instruction memory 405 is configured so that the CPU 403 is connected to the pulse input device 4
A plurality of commands can be stored in which the input channel number of the event to be detected, the polarity of signal change (rising, falling), etc. are specified as the measurement content required for 01. Additionally, time information of signal changes, which is the result of executing the command, is similarly written to a predetermined location in the command memory 405. On the other hand, the CPU 403 generates an interrupt in the pulse input device 401 and periodically monitors the contents of the command memory to accurately know the times of multiple events that occur in an unknown order. I can do it.

The instruction execution unit 409 sequentially reads and executes instructions from the beginning of the instruction memory 405, and after finishing executing the instruction at the end of the instruction memory, executes the instruction at the beginning of the instruction memory again. In this way, the instruction execution unit 409 repeatedly executes the instructions in the instruction memory in order. A NOP instruction, which will be described later, is written in a free area of the instruction memory.

Further, the specified event measurement is performed, and instructions that do not require re-execution of measurement are also executed in order. In this case, the instruction is processed similarly to a NOP instruction.

Note that each instruction is normally designed so that the time from when it is executed once to when it is executed again is constant, and that time is designed to be the same as the input signal retention time of the input memory 408. ing.

(Problem to be Solved by the Invention) By the way, when measuring time intervals such as the cycle, pulse width, and phase difference of an input signal, the host CPU calculates the desired time based on the change time measured by the pulse input device. Although it is possible to determine the interval, processing speed and host CPU
This is undesirable because it increases the load on the machine.

An even more serious problem arises when measuring time intervals continuously. That is, for example, consider the case where the positive pulse width of the input signal shown in FIG. 5 is measured. The CPU is
The positive pulse width is calculated from the measurement results of event 1 and event 2. Part A in the figure shows the period during which the pulse input device holds time data necessary for its calculation. If the period during which the input signal is '0' becomes short, it becomes impossible for the host CPU to calculate the positive pulse width.

Further, it takes a certain amount of time from reading out the measurement results of event 1 to reading out the measurement results of event 2 from the host CPU. In the meantime, there is a possibility that a new measurement will be performed and the measurement result of event 3 will be updated. In this case, the host CPU uses a value larger than the actual time interval as the measured value of the time interval. In order to avoid such a situation, it is necessary to create a program for the host CPU with due consideration given to the nature of the input signal to be controlled, the measurement timing of the pulse input device, etc., which places a heavy burden on the user. Also, depending on the nature of the input signal,
In some cases, the above problems cannot be resolved. In order to solve this problem, a function for calculating time intervals may be added to the pulse input device.

When commands for measuring time intervals such as period, pulse width, and phase difference are added to a conventional pulse input device, the following three problems arise.

The first problem is that the bit width required by one instruction becomes wide. For example, a command to measure the change time of an input signal only needs to have one field indicating the channel to be measured and one field indicating the polarity. but,
An instruction that instructs to measure a time interval requires two sets. If the encoding is advanced to shorten the field length, the circuit for decoding the instructions becomes complicated and the instruction execution time increases.

The second problem is that the time interval measurement command requires memory to store the time when the first event occurs. For example, when creating an instruction to measure the width of a positive pulse, it is necessary to prepare a memory that stores the time at which the first event (rising edge change) occurs. Since the time between the occurrence of the first event and the occurrence of the second event (falling transition) is unpredictable, the storage memory described above must be able to store the event for an arbitrary amount of time.

Therefore, it is necessary to provide a memory for storing the first event occurrence time. Therefore, in order to provide a general-purpose pulse input device, it is necessary to set the same number of storage memories as the number of commands that can be set in the command memory. However, since this storage memory is not used by an instruction that only measures change times, it becomes a completely useless memory for users who do not measure time intervals.

The third problem is that the execution time of the time interval measurement command becomes long. For example, consider the number of memory accesses. A simple change time measurement command requires accessing the command memory twice and the input memory once in order to read the command and store the measurement result. Therefore, the time interval measurement instruction requires two instruction memory accesses and two input memory accesses to detect two events. Even in the processing performed by the instruction execution unit, the number of time interval measurement instructions clearly increases.

Therefore, the present invention provides an instruction set suitable for measuring time intervals, that is, an instruction set that has a short instruction code length and is easy to decode, and also enables effective use of the existing memory of the pulse input device. It is an object of the present invention to provide a pulse input device that can easily process each command.

[Structure of the Invention] (Means for Solving the Problems) In order to solve the above-mentioned problems, the present invention provides two instructions stored at addresses on an instruction memory that are successively executed to measure a time interval. The configuration is implemented using . The first instruction then measures the time of occurrence of the first event. The second instruction measures the time of occurrence of the second event, and uses the measured value and the measurement result of the first instruction to measure the time interval.

In order to perform this function, the present invention adds at least one register that stores the measurement result of the most recently executed instruction and at least one flag that indicates the measurement status of the most recently executed instruction to the instruction execution unit. In addition, it has a function of calculating the time when an event occurs, and a means of calculating the difference between the measured value and the value stored in the register.

The result of the event occurrence time measured in this way is stored in the above register, the first instruction sets the above flag according to the measurement situation, and the second instruction sets the above flag according to the state of the flag. The difference between the measured value and the value stored in the register is determined.

(Function) With the structure of the present invention, the memory for storing the measured first event can use the memory area of the instruction memory,
This memory area can be used to store measurement results of change times when time intervals are not measured, so a special memory area is not required.

The first and second instructions used for time interval measurement each have a field for specifying one event and a field for specifying whether to measure the first event or a later event in the time interval to be measured. It would be nice if there was a field to do so. Therefore, the code length of each instruction can be made equal to that of the conventional time measurement instruction, so it is possible to realize an instruction set with a short code length and easy decoding.

(Example) The following example of the present invention will be described with reference to the attached drawings.

FIG. 1 shows a command execution unit 40 of a pulse input device according to the present invention.
This shows the structure of 9. The other components constituting the pulse input device are the same as those shown in FIG. 4, and therefore will be omitted.

In FIG. 1, the instruction execution unit 409 includes the first subtracter 11
1, a second subtracter 1, a verification circuit 113, a register 114, and an R register 115. Further, although not shown, a V flag and an R flag are provided in association with the two registers.

Signal 102 is a signal from input memory 407 shown in FIG. 4 and indicates how long ago the event to be measured occurred, if detected. Signal 101 is a signal from timer counter 406.

A signal 104 sends a measured value due to instruction execution to the instruction memory 40.
This is a signal for storing in a predetermined memory area of No. 5.

The command execution unit 409 of the pulse input device according to the present invention includes:
The event occurrence time of the input signal is calculated using the subtracter 111. As previously mentioned, signal 102 indicates how long ago the event occurred. Therefore, signal 101
The time at which the event occurs is calculated by subtracting the value of the signal 102 from the timer counter value. Subtractor 1
11 as a signal 103, and further subtracter 1 second
is passed through and output as a signal 104. This through function can be easily realized by inputting zero as one of the second operands of the subtracter 1. In this way, the command execution unit 409 of the pulse input device according to the present invention realizes the function of measuring the event occurrence time.

In the pulse input device according to the present invention, the measurement result of the first command is used in the second command. For this purpose, the R register 11 is used as a register to store the measurement result according to the first instruction.
5 and V register 114 are provided in the instruction execution unit 409.

When the first instruction is executed, the memory area for storing the measurement results is read out and stored in the R register at the same time as the instruction is read out. Therefore, if the first instruction has been measured before that point, the measurement result will be stored in the R register. Further, at the end of execution of the first instruction, the measurement result is stored in a predetermined field of the instruction memory 405 and also stored in the ■register 114.

Next, the operation of the flag will be explained. As described above, this pulse input device has two types of flags, the R flag and the V flag, and the flags are set only when the 1st instruction is executed. Execution of other instructions always resets the flag on completion as well.

These flags are used to indicate whether the measured value of the first instruction is stored in the R register and the V register, respectively. When the first instruction is executed, the instruction memory 405
If the measured value is stored in the memory area that stores the measured result read from the , the R flag is set. This flag will not be set if it has not been stored. This determination can be easily made by providing a bit in the specific field of the instruction that indicates whether or not the measurement value is stored in the memory area for storing the measurement result. If the occurrence time of the first event is measured when the first instruction is executed,
■A flag is set. Since the flag of this pulse input device is set as described above, in the second instruction executed immediately after, the R register 115 and the V register 114 are
It can be determined whether the measurement result of the first instruction is stored in the first instruction.

The first instruction may measure the time of occurrence of the first event, and the subsequently executed second instruction may detect the second event. Since the pulse input device of the present invention detects an event by checking the input signal stored in the input memory, there is a possibility that a second event occurring before the first event will be detected. Therefore, it is necessary to know whether the second event occurred after the first event. Therefore, a verification circuit l13 is provided for this purpose. If the event detected by the second instruction being executed is new, 1 degree is output as the output signal 106.

The commands of the pulse input device shown in this embodiment are roughly divided into the following four types. That is, NOP, EDGE, WI
These are DTH and WIDTHP. The NOP instruction is an instruction that does nothing, and is written into an empty area of the instruction memory 405. The EDGE command is a command for measuring the change time of an event, and is also used as the first command for time interval measurement. In this embodiment, since no particular inconvenience occurs, the simple change time measurement command and the first time interval measurement command are set as the same command. WI DTH and WIDTHP are used as the second command for time interval measurement. WIDTHP is used as the second command for period measurement. WIDTH is used as the second command for pulse width and phase difference measurements.

The basic format of these instructions is shown in FIG. The instructions are of fixed length, and the lower bits are dedicated to fields that store measurement results. Since field 201 specifies each of the four types of instructions, field 201 only needs to have 2 bits. Field 202 specifies the input channel to measure. field 203
indicates the polarity (rising or falling) of the change being measured. Field 204 is a field that indicates whether the specified measurement has been performed. host CPU40
3 can determine whether a measured value is stored in the measurement result storage field by checking this field. Field 206 specifies whether or not to generate an interrupt pulse when a measurement result is obtained, and whether or not to perform measurements continuously.

Measurement of time intervals using flags will be explained.

The time interval is measured by setting the first command and the second command in the command memory 405 to be executed consecutively. The EDGE instruction is used as the first instruction.

When the execution of the EDGE instruction ends, the measurement results are stored in the R register 115 and the ■ register 11 in the instruction execution unit 409.
4, and its measurement status is set in the R flag and V flag. Since the first instruction and the second instruction are executed consecutively, at the start of execution of the second instruction,
R register 115, ■ register 114, R flag and V
The flag holds the state at the end of the first instruction. When a certain event is detected during execution of the second instruction, the time of occurrence is calculated using the subtracter 111, and the R register 115 and V register 1 are calculated using the R flag and the V flag.
14, and calculate the time interval using subtracter 1 and 2. The above is an outline of time interval measurement. Depending on the contents of the two flags, the time interval may not be calculated even if the second event is detected.

FIGS. 3(a) and 3(b) show how the second command is processed based on the contents of the flag and the value of the output signal 106 of the verification circuit when the second event is detected.

The signal 106 from the verification circuit 113 outputs ゜1゜ if the detected event has occurred after the first event detected by the first instruction executed immediately before. The measurement result of the first instruction executed immediately before is stored in the register 114.
is stored in. Therefore, when the signal 106 is 0', the register 114 is set as the time of occurrence of the first event.
The value stored in is never used.

FIG. 3(a) shows the second command for period measurement, WIDT.
It shows the processing content of the HP command. In period measurement, the event specified by the first command and the event specified by the second command are of exactly the same type (same channel, same locality event). In this case, the event detected when WIDTHP was executed was also detected in the first instruction executed immediately before, and its occurrence time is recorded in the V register 1.
As the occurrence time of the first event, the content of the R register 115, which stores the occurrence time of an event older than the event detected by WIDTHP, is used.

FIG. 3(b) shows the processing contents of the WIDTH command. ■
When the flag is 1°, it indicates that the first event was detected in the execution of the immediately preceding first instruction, and its measured value is stored in the V register 114. In this case, it is determined based on the signal 106 whether the event stored in the register 114 has occurred before the detected second event. When the event occurs before the second event, that is, when the signal 106 is 1, the content of the register 114 is used to calculate the time interval. When the signal 106 is "0" and the R flag is "1", the time interval is calculated using the contents of the R register 115.

Further, when only the R flag is "1", the time interval is calculated using the contents of the R register 115.

■When the flag is 1° and the signal 106 is 1°, the contents of the register 114 are used.

〔Effect of the invention〕

The embodiments of the present invention have been described above, and when measuring the time interval of input pulses, the pulse input device according to the present invention simultaneously captures a plurality of input signals, stores them once in the human memory, and stores the input signals for each channel. Since the system uses a configuration in which the data is read out and the time at which the event occurs is measured, even if multiple events occur in a concentrated manner within a short period of time, no event will be missed.

Further, in the pulse input device according to the present invention, the memory for storing the measurement results of the first event necessary for time interval measurement can use the memory area for storing the measurement results of the command memory, so that the memory area can be If time interval measurement is not required, it can also be used to store measurement results of change times. Therefore, there is no need for a new memory area for time interval measurement, and memory can be used flexibly.

Furthermore, since the instruction set used in the present invention has a relatively simple bit pattern, the decoding circuit for decoding the instructions can also be small-scale. Furthermore, in the present invention, memory accesses by the first instruction for measuring the first event and the second instruction for measuring the second event are performed twice to the instruction memory and once to the input memory. Since the number of accesses is small, the instruction execution time is short and constant.

[Brief explanation of drawings]

FIG. 1 is a diagram showing the configuration of an instruction execution unit in a pulse input device according to the present invention, FIG. 2 is an instruction format diagram used in the device according to the present invention, and FIG. 3(a) is a diagram showing the second instruction WIDTHP. Figure 3(b) is a diagram showing the processing content of the second command, WI DTH command, Figure 4 is an overall configuration diagram of a conventional pulse input device, and Figure 5 is a diagram showing the processing content of the command. Figures illustrating disadvantageous states of pulse input signals when measured with a conventional pulse input device are shown, respectively. 401...Pulse input device, 402...Bus, Φ403...CPU (central processing unit), 405...
・Instruction memory, 406...Timer counter, 407...Input circuit, 408...Input memory, 409...Instruction execution unit, 111 and 1 second...Subtractor, 113...Verification circuit , 114...V register, 115...R register. Figure 3 (a) Note: Don't cure Figure 3 (b)

Claims (1)

[Claims] A timer counter that counts a system clock and outputs reference time information; an input circuit that samples input signals from a plurality of channels at a predetermined period in synchronization with the time information; an input memory that stores the input signals for the number of times; an instruction memory that mainly stores a plurality of instructions; and an instruction memory that searches the instruction memory to sequentially read and execute the plurality of instructions, and controls the overall operation of the device. In a pulse input device having an instruction execution unit, the instruction execution unit includes at least one register that stores a measurement result of a change time of an instruction executed immediately before, and a flag that indicates the measurement status at the time of execution of the first instruction. and a subtraction means for calculating the time difference between the contents of the register and the measured event occurrence time, a first instruction that sets a flag at the time of execution and measures the occurrence time of the state change of the input signal; Prepare in advance a second instruction that instructs to take the difference between the contents of the register and the measured state change time value according to the state of the flag, and store this instruction in the instruction memory. A pulse input device characterized in that a first command and a second command are executed continuously and a time interval of input signals is measured.