JPH0767110B2 - Signal synchronization method - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a signal synchronization method, and in particular, a token passing bus is used to instantly detect voltage, current, etc. at a plurality of points separated from each other. The present invention relates to a signal synchronization method suitable for use in generating a sampling signal or the like for simultaneously obtaining a value.

[Prior Art] Local Area Network (LAN) has been rapidly spreading in recent years, and one of the technologies that should be especially noted is MAP.
An industrial LAN known as (Manufacturing Automation Protocol) is known.

This MAP is aimed at facilitating easy and inexpensive interconnection of equipment such as robots and computers installed in factories, offices, etc. for FA, OA, etc., and standardization thereof is in progress. . This standardization is being advanced by filling each layer of the OSI (Open Systems Interconnection) hierarchical model of ISO (International Organization for Standardization), and the lower two layers (physical layer, data link layer) excluding the theoretical link control sublayer. IEEE (US Conference on Electrical and Electronic Engineers)
802.4 The Commission's token passing bus is used.

The token passing bus and signal transmission will be described below with reference to the drawings.

FIG. 3 is a diagram showing an example of a transmission signal of a token passing bus for signal synchronization, FIG. 4 is a diagram showing an example of a bus type network using a coaxial cable, and FIG. 6 is a subscriber station to the network. It is a figure which shows the example of the transmission signal in case the number is small. In FIG. 4, 1 ₁ to 1 _n are subscriber stations, 2 is a coaxial cable, and 3 _{1 to} 3 _n are branching devices.

In FIG. 4, the subscriber stations 1 ₁ to 1 _n forming the network
Are connected to the coaxial cable 2 via branching devices 3 _{1 to} 3 _n called taps. This kind of bus network also has a method (CSMA / CD) in which each connected subscriber station arbitrarily sends data, but this method is traditionally efficient when the amount of transmitted data increases and the load increases. Has the disadvantage that it drops rapidly. On the other hand, the token passing bus to which the present invention is applied can avoid drawbacks by the deterministic access method. That is, in this token passing bus, information indicating a transmission right called a token is sequentially transmitted between the subscriber stations that have subscribed to the token passing bus, and a plurality of subscriber stations simultaneously transmit at the same time. To prevent that. The token is passed, for example, from a station having a larger address to a station having a smaller address in turn, and each subscriber station stores a subsequent subscriber station to which the token is to be passed, whereby a token circulation logic is used. The upper ring (logical ring) is constructed. However, if this logical ring is fixed,
If a new subscriber cannot join the bus, he will not be able to join, and if any of the subscribers fails, the logical ring will be broken and communication will stop. In order to avoid this, the token passing bus has the following rig maintenance function.

(1) When the token is temporarily lost due to noise or the like, the subscriber station holding the previous token reissues the token.

(2) If the subsequent subscriber that should pass the token fails, the token-issuing subscriber gives the token to the successor of the failed subscriber and removes the failed subscriber from the logical ring.

(3) A new subscriber station is solicited at an almost constant cycle and a subscriber station wishing to join the logical ring is permitted to join.

On the other hand, there are application fields in which high-speed control of equipment is performed using a LAN including this token passing bus.
There is a system that requires each subscriber station to repeatedly generate a specific signal at the same time. For example, voltage, current at multiple points,
For example, generation of sampling signals for simultaneously obtaining instantaneous values such as displacement and acceleration. Therefore, if a token passing bus using LAN can be used to perform signal synchronization of such a specific signal, for example, a sampling signal, the token passing bus originally has immediacy, self-healing property, and expansion. In addition to the excellent characteristics, the usage of the token passing bus can be further widened.

As a method of performing signal synchronization as described above using a token passing bus, a specific subscriber station that constitutes the network is defined as a master station, and the master station is used by using a data frame periodically issued by this master station. There is a method in which a subscriber station (slave station) other than the station synchronizes with the master station to generate a specific signal such as a sampling signal.

Now, in FIG. 4, the subscriber station ₁₁ is the main station and the other subscriber stations 1 ₂
Assume that 1 to 1 _n are slave stations, and the subscriber stations ₁₁ to 1 _n including the master station transmit data. In this case, as shown in FIG. 3, first, after the subscriber station 1 which is the main station transmits the data frame D ₁ ,
Send token T ₁ . A slave station which has been kicked not a logically located position after the main station subscriber, e.g., subscriber station 1 _2, that this detects the token T ₁ of the from the main station, the transmission right to itself was Reality Confirming that, the data frame D ₂ is transmitted, and the token T ₂ is transmitted in order to transfer the transmission right to the succeeding subscribing station. When such an operation is performed by all the subscriber stations, the transmission right is transferred to the subscriber station ₁₁ which is the main station again, and the above-described operation is repeated.

Incidentally, it is assumed that the subscriber station 1 ₁ as the master station is occurring synchronizing signals generated for a particular signal, such as a sampling signal (hereinafter, referred to as the main synchronization signal) S ₁ at period T, the this main station subscriber 1 _1, if it is possible to transmit the data frames D ₁ in synchronization with the main synchronization signal S _1, and the subscriber station 1 ₂ to 1 _n comprising another slave station, based on the reception time of the data frame D ₁ Te, the main time of occurrence of the synchronization signals S ₁ can know, it is possible to generate the primary synchronization signal own station is synchronized with the S ₁ synchronous signals S _2, S ₃ .... That is, in the token passing bus, the data frame D ₁ is received by the previous slave station almost at the same time by appropriately setting the destination address.

To realize the above-mentioned signal synchronization, the main station is a subscriber station 1 ₁
There is a problem in how to transmit the data frame synchronized with the main synchronization signal S ₁ .

Originally, the tokens in the token passing bus circulate according to an algorithm for maintaining the logical ring, and the circulation cycle is different from the cycle T of the main synchronization signal S ₁ , as well as solicitation of new stations and token passing. It always fluctuates because retries and the like are performed stochastically. Therefore, in order to perform the signal synchronization as described above using the token passing bus, the means for enabling the subscriber station _{11 serving} as the main station to transmit the data frame D ₁ in synchronization with the main synchronization signal S _1. Is essential.
One method, to wait after the subscriber station 1 ₁ to be the master station to obtain a transmission right by receiving the token, the transmission of data frames D ₁ to be sent immediately, until the occurrence time of the main synchronization signals S ₁ A method can be considered. However, this method will be no signal state continues for a long time, the subscriber station preceding that sent the token to the subscriber station 1 ₁ to be the master station is, to retransmit the token is regarded as a token Patsu single fails, Furthermore there is a problem that removed from the logical ring is regarded as a failure of the subscriber station 1 ₁ to be master station.

As a method for solving such a problem, the master station transmits an empty data frame D _s prior to the transmission of the data frame D ₁ , and the length of this empty data frame D _s is set as the token reception time of the master station. There is a method of adjusting the transmission time of the data frame D ₁ by changing the transmission time of the data frame D ₁ . However, this method causes the following problems when the number of subscribers to the network is small.

FIG. 6 is a diagram showing a transmission signal when the number of subscribers to the network is two, including the main station. In this figure,
Subscriber 1 ₁ as the master station is, after transmission data frame synchronization to the main synchronization signal S _1, and send the token T _1, slave station detects the token T _1, after transmitting a data frame D _1,
Send token T ₂ . The number of subscribers to the network is 2
Since as a station, the subscriber station 1 ₁ to be the master station, by the token T ₂ will get a transmission right, immediately it is necessary to start sending null data frames D _s1. But,
In this case, from the transmission of the token T ₁ in the master station, time again t _INT until the token T ₂ of the own station arrives is short, the subscriber station 1 ₁ as the master station, to be carried out during this period In some cases, the process of calculating the length of an empty data frame and adding it to the head of the transmission queue may not be in time. In this case, the above method cannot maintain synchronization.

Incidentally, as described in the foregoing, the synchronous data frame D ₁ and main synchronization signals S ₁ to subscriber station 1 ₁ to be the master station transmits, FIG. 3,
As shown in FIG. 6, the end of transmission of the data frame D ₁ is controlled so as to be the synchronization position of the main synchronization signal S ₁ .

[Problems to be Solved by the Invention] As described above, in the above-mentioned conventional technique, when the number of subscribers to the network is small, the token is issued before the calculation of the empty data frame length is completed in the main station. There is a case where a call arrives, and in that case, there is a problem that the signal synchronization cannot be maintained.

SUMMARY OF THE INVENTION An object of the present invention is to solve the above-mentioned problems of the prior art and to provide a signal synchronization system capable of reliably maintaining signal synchronization even when the number of subscribers to the network is small.

[Means for Solving the Problem] According to the present invention, the above object is achieved by transmitting an empty data frame also in a slave station when the number of subscriber stations to the network is small.

[Operation] By transmitting an empty frame in the slave station, the master station transmits the token, and after that, the time until the token is received and the next transmission right is obtained becomes longer, and the calculation and transmission of the frame length of the empty data frame are performed. It is possible to secure sufficient time for performing the preparation process. Therefore, signal synchronization can always be reliably maintained.

[Embodiment] An embodiment of a signal synchronization system according to the present invention will be described in detail below with reference to the drawings.

FIG. 1 is a diagram for explaining signal transmission according to an embodiment of the present invention when the number of subscribers to the network is two, and FIG. 2 is the present invention when the number of subscribers is the maximum number of subscribers of the network. 6 is a diagram for explaining signal transmission according to the embodiment of FIG.

The network to which the present invention is applied is, as a conventional technique,
It has the same configuration as described with reference to Figure 4, in the following description, a main station with respect to the signal synchronizing subscriber stations 1 _1, and the slave station to other subscriber stations 1 ₂ to 1 _n. And subscriber stations 1 ₂
Addresses given to ~ 1 _n are arbitrarily added and deducted regardless of the containing position, and tokens are sequentially moved according to this address.

In FIG. 1, the subscribing station _{11 serving} as the main station has a main synchronization signal S ₁
A data frame (hereinafter, a data frame synchronized with the main synchronization signal in this main station is referred to as a synchronization data frame) D ₁ is transmitted in synchronism with, and a token T ₁ is subsequently transmitted.
The synchronous data frame D ₁ has a fixed frame length, its data pattern is known in advance by slave stations that use all synchronization signals, and, as its destination address, all slave stations that use synchronization signals. A group address including is added. Then, the transmission time of the synchronous data frame D ₁ is controlled so that the transmission end time is the generation time of the main synchronization signal S ₁ . Subscriber station 1 ₂ as the slave station, the receives the token T _1, and prior to transmission of the data frame D ₂ of the own station transmits a fixed-length null data frame D _s2 of, subsequently, of the own station data frame D ₂ And send token T ₂ . In the example of FIG. 1, since the number of subscriber stations accommodated in the network is two,
Subscriber station 1 ₁ to be the master station, by the receiving token T ₂ of the from the subscriber station 1 ₂ to be the slave station knows that the transmission right to itself has moved. The main subscriber 1 ₁ has a synchronization data frame D ₁
In order to synchronize the transmission of the data with the next main synchronization signal, it is necessary to transmit an empty data frame D _s1 during that period. Token Thus, subscriber 1 ₁ as the master station, which the own station transmits
Token T ₂ from the subscribing station 1 ₂ that becomes a slave after the end of transmission of T ₁
Empty data frame D _s1 during the time t _INT until the reception of
The frame length, the frame length of the previous null data frame, calculated from the error of the synchronization frame transmission end time and the main synchronization signal S _1, transmit queue empty data frame D _s1 frame length obtained by the result Perform processing to add to the beginning of. Frame length of an empty data frame D _s2 transmitted from subscriber station 1 ₂ as the slave station, previously set to be longer than the processing time for performing the above-described processing at the subscriber station 1 ₁ time t _INT described above becomes master station To be done.

If the number of stations subscribed to the network is large, the time t _INT from the token transmission to the reception of the next token at the master station is calculated even if the slave station does not transmit an empty data frame.
Since the time is sufficiently long, it is not necessary to transmit an empty data frame from the slave station. FIG. 2 shows the state of signal transmission in this case.

In FIG. 2, the subscriber station _{11 serving} as the main station has a main synchronization signal S ₁
After transmission of the sync data frame D ₁ synchronized with the token T ₁
To send. The slave station, which is a subordinate station receiving the token T ₁ , transmits the data frame D ₂ of its own station and then transmits the token T ₂ to shift the transmission right to the subordinate station. When the transmission of the data frame in all the subscriber stations is completed by such an operation, the subscriber station, which is the lowest station, transmits the talk T _n . As a result, the transmission right is transferred to the subscriber station ₁ which is the main station again. After sending the token T ₁ in subscriber station 1 ₁ as a master station, the time t _INT until the reception of the token T _n is
In this case, it is large enough that the empty data frame
Processing such as calculation of the frame length of D _s1 can be sufficiently executed during this period. Therefore, in this case, none of the subscribing stations that are subordinate stations need to transmit null data, and the number of subscribing stations to the network can be increased to the maximum number determined by the network capacity.

With such a configuration, the subscriber station 1 ₁ as the master station is always
Can send synchronous data frame D ₁ in synchronization with the main synchronization signal S _1, the total capable of receiving synchronous data frame D ₁ The slave station, the synchronization position the end of the synchronization data frame D ₁ self It becomes possible to generate a station synchronization signal.
Also, since the subordinate station that is a slave station has its own clock, it is possible to predict the time at which the main synchronization signal S ₁ from this master station should be detected, and if the detected main synchronization signal S 1
_{When 1} is significantly different from this predicted time, it is configured to ignore the detected main synchronization signal S ₁ as an abnormality in the network. As a result, it is possible to further improve the reliability of the synchronization signal at the subordinate station that is the slave station.

In the above-described embodiment of the present invention, one slave station transmits an empty data frame, but the slave station transmitting an empty data frame may be one slave station accommodated in the network. Also, a plurality may be provided. Further, whether or not the slave station should transmit the empty data frame may be instructed to each slave station by the master station according to the number of subscriber stations accommodated in the network, or each slave station may be instructed. In the above, the determination may be made according to the number of subscriber stations accommodated in the network, and the determination may be made using the frame length of the empty data frame from the main station.

In addition, the number of slave stations that transmit empty data frames may be variable according to the number of subscriber stations accommodated in the network, or the number of slave stations that transmit data may be variable depending on the number of subscriber stations accommodated in the network. It is also possible to make the frame length of the data frame variable. The empty data frame transmitted by the main station may be divided into two or more and transmitted as necessary.

In the embodiment of the present invention, an empty data frame transmitted from a slave station or a master station in a network does not need to be received by any subscriber station, and its destination address is not included in any network. It is a group address or individual address that does not instruct the subscriber station.

In the above-described embodiment of the present invention, when the slave station transmits an empty data frame, the slave station transmits the empty data frame and then transmits the own station data frame.
An empty data frame may be transmitted after transmitting the own station data frame.

FIG. 5 is a diagram showing the structure of another network to which the present invention is applied. In FIG. 5, 4 is an optical star coupler,
Reference numerals 5 _{1 to} 5 _n are optical fiber cables, and other reference numerals are the same as those in FIG.

The present invention is not applicable only to the network provided with the coaxial cable shown in FIG. 4, and for example, as shown in FIG. 5, the network using the optical star coupler 4 and the optical fiber cables 5 _{1 to} 5 _n. Can also be applied to. Each of the optical fiber cables 5 _{1 to} 5 _n includes two optical fiber core wires. In this Nets work, for example, transmit signals from the subscriber station 1 _1, the optical fiber cable 5 ₁
Is sent to the optical star coupler 4 and is branched here to send the same signal to the subscriber stations 1 ₁ to 1 _n via the receiving fibers of the optical fiber cables 5 _{1 to} 5 _n . The topology of this network is physically star-shaped, but logically it can be handled in the same way as in the case of FIG. 4 as a bus-type. Although it is a CSMA / CD system and not a document of optical token bus, as a document of a network using a star coupler, signal total size No. 1780 (1984),
There is No. 177 (1984), etc.

[Effect of the Invention] As described above, according to the present invention, signal synchronization is achieved by adjusting the frame length of an empty data frame in an IEEE 802.4 talk passing bus or a network having a communication protocol similar to the bus. When the number of subscribers accommodated in the network is small, the slave station can also transmit a null data frame to achieve stable signal synchronization even when the number of subscribers is small. Also, if the number of subscribers accommodated in the network is large, it is possible to increase the number of subscribers to the maximum number of stations depending on the capacity of the network by stopping the transmission of empty data frames in the slave stations. Become.

[Brief description of drawings]

FIG. 1 is a diagram for explaining signal transmission according to an embodiment of the present invention when the number of subscribers to the network is two, and FIG. 2 is a book when the number of subscribers is the maximum number of subscribers of the network. FIG. 3 is a diagram for explaining signal transmission according to an embodiment of the invention, FIG. 3 is a diagram showing an example of a transmission signal of a token passing bus for signal synchronization, and FIG. 4 is an example of a bus type network using a coaxial cable. 5 and 5 are views showing an example of a network using an optical fiber cable, and FIG. 6 is a view for explaining signal transmission according to the prior art when the number of subscribers to the network is two. 1 ₁ to 1 _n ...... Subscriber station, 2 ...... Coaxial cable, 3 _{1 to} 3 _n ...... Branching device, 4 ...... Optical star coupler, 5 _{1 to} 5 _n ...... Optical fiber cable.

 ─────────────────────────────────────────────────── ─── Continuation of front page (71) Applicant 999999999 Hitachi Cable, Ltd. 1-2-2 Marunouchi, Chiyoda-ku, Tokyo (72) Inventor Yukito Sejima 1-1-1 Kokubuncho, Hitachi-shi, Ibaraki Hitachi Ltd. Factory Kokubun Factory (72) Inventor Mamoru Suzuki 1-3-1, Uchisaiwaicho, Chiyoda-ku, Tokyo Tokyo Electric Power Company (72) Inventor Yutaka Moriyama 1st, Toshiba Town Fuchu-shi, Tokyo Toshiba Fuchu Factory ( 72) Inventor Yasuhiko Terao 1-2-1, Wadasaki-cho, Hyogo-ku, Kobe-shi, Hyogo Mitsubishi Electric Corporation Control Works (72) Inventor Yasumasa Imai 5-1-1 Hidaka-cho, Hitachi-shi, Ibaraki Hitachi Cable Electric Cable Laboratory Co., Ltd. (56) Reference JP-A-63-263938 (JP, A) JP-A-63-318842 (JP, A)

Claims (8)

[Claims] 1. Token passing in which a plurality of subscribers form a logical ring, a token is circulated in the logical ring, and only a subscriber holding the token is given a right to transmit data. In a signal synchronization method using a bus or a network having a communication protocol similar to that, one of the plurality of subscriber stations is a master station for signal synchronization, and the other subscriber station is a slave station for signal synchronization, and The subscriber station serving as the main station sends an empty data frame for adjusting the transmission time of the synchronization data frame so that the transmission end time of the synchronization data frame transmitted for synchronization is synchronized with the main synchronization signal. When the number of subscribers to the network is small, some or all of the subscribers that transmit prior to frame transmission and that are subordinate stations of the synchronization are subscribed to the network. There signal synchronization scheme and transmits a null data frame. 2. A subscriber station, which is a slave station of the synchronization, detects a reception end time of the synchronization data frame,
The signal synchronization system according to claim 1, wherein the station synchronization signal of the own station is synchronized with the main synchronization signal. 3. A subscriber station which is a master station of the synchronization, when the number of subscriber stations to the network is small, instructs to send an empty data frame to a part or all of the subscriber stations which are slave stations. Claim 1 which is characterized by the above-mentioned.
The signal synchronization system according to item 2 or item 2. 4. The signal according to claim 1 or 2, wherein the subordinate station serving as the slave station determines whether or not to transmit an empty data frame based on the judgment of each subscribing station. Synchronous method. 5. The subordinate station, which is the slave station, is generated by the reception end when the reception end time of the synchronous data frame is earlier or later than a normal expected reception end time beyond a certain allowable range. The signal synchronization method according to claim 2, 3, or 4, wherein the synchronization signal is ignored. 6. The destination address of the empty data frame is
The signal synchronization system according to any one of claims 1 to 5, wherein the signal is a group address or an individual address that does not indicate a subordinate station that is a slave of any synchronization in the network. 7. The destination address of the synchronous data frame is a group address including all subscribers that are slaves of synchronization in the network, according to any one of claims 1 to 6. The signal synchronization method according to any one of items. 8. The synchronous data frame length is fixed,
The signal synchronization method according to any one of claims 1 to 7, which is known in advance by all the subscriber stations that are subordinate to the synchronization.