Time characteristics of PROFIBUS on Windows XP
Contents
1. DfP2 ProfibusSlaveExample vi Front Panel Instructions Select vathin the Slave Properties Tab the DAPAamurcatkama of the OF PROFI IT board to be used Adira A the aud rate sod a slave mackie trom module Ist Thasa parameters want correspond with the miner configs ation Run the Yi The Monitor Modify Tab shows the input and output data as wet as the Communication status of the skive To update the input data change the values of the Modify nout data control and press the Sat neut data button Detailed Inicemation about the staves communication status Is shown In the Status Info Tab Stove Properties MentorModty Ststusinto OF P2ResourceNeme zj cdmsorr 4 byte np Ka vrie oup 4 byteanp 128 byte outp Figure 11 Front panel and block diagram 15 Huang Yang 2005 National Instruments LabVIEW RT Real Time Module extends the LabVIEW development environment to deliver deterministic real time performance LabVIEW RT supports two kinds of real time platforms RTX Real Time Extension and ETS Embedded Tool Suite These platforms are also called RT targets The first one RTX adds a real time subsystem RTSS to Windows such that LabVIEW RT applications can execute on computers The deterministic real time behavior of applications running on RTSS is not affected by performance of Windows because the subsystem runs in parallel with Windows and its scheduling supersedes Wi2. 11 Huang Yang 2005 2 3 3 PROFIBUS DP Protocol Besides the first PROFIBUS protocol FMS developed 20 years ago in layer 2 there is a PROFIBUS protocol named protocol DP Decentralized Periphery which is commonly used as the main transmission technology in PROFIBUS The term Decentralized Peripherals stands for the simple fast and deterministic I O data exchange between a bus master and its assigned slave devices There are two kinds of stations within a PROFIBUS DP network DP master and DP slave According to IEC 61158 5 a PROFIBUS DP master can be either master class 1 or master class 2 A master class 1 is a controlling device that controls several DP slaves field devices It is usually hosted by a programmable controller or a process controller A master class 2 is a controlling device which manages configuration data parameter sets and diagnosis data of a DP master class 1 and that additionally can perform all communication capabilities of a DP master class 1 Also from IEC 61158 5 a PROFIBUS DP slave is a field device that is assigned to one DP master class 1 as a provider for cyclic I O data exchange In addition acyclic functions and alarms could be supported DP is now available in 3 versions namely DP VO DP V1 and DP V2 Characteristics of these versions are depicted in Figure 9 Functional Levels DP V2 m Data Exchange Broadcast Publisher Subscriber m Isochronous Mode Eguidistance plus e
3. Idle Time Figure 17 a 24 Huang Yang 2005 Time Elapsed of GET PUT DP POLL DATA Standard Deviation 4 8 16 32 64 128 240 VO bytes L2 ER D E e Average Standard Deviation s Standard Deviation of Action Standard Deviation of Idle Figure 17 b Time Elapsed of GET DP POLL DATA Mean 250 200 150 100 50 0 4 8 16 32 64 128 240 NO bytes Average Time Action Time Idle Time Figure 18 a Time Elapsed of GET DP POLL DATA Standard Deviation 4 8 16 32 64 128 240 VO bytes Time us e Average Standard Deviation s Standard Deviation of Action Standard Deviation of Idle Figure 18 b 25 Huang Yang 2005 Time Elapsed of PUT DP POLL DATA Mean 100 80 60 40 20 0 4 8 16 32 64 128 240 NO bytes Average Time Action Time Idle Time Figure 19 a Time Elapsed of PUT DP POLL DATA Standard Deviation Time ps VO bytes e Average Standard Deviation s Standard Deviation of Action Standard Deviation of Idle Figure 19 b By comparing different curves within one figure it is obvious that a function call cost very little time when no new data is available In practical application where continuous assessment to the memory is always necessary this feature makes the communication much more time efficient Both GET PUT DP POLL DATA and GET DP POLL DATA will cost more ti
4. understanding of these design options general description of LabVIEW is introduced first Then the design choices are discussed with a conclusion presented At the end of the chapter 2 Huang Yang 2005 details of designing the tests are given Chapter 4 focuses on representation of test results and analysis of the performed measurements This chapter also illustrates the implementation of a control loop and results of control loop measurements Chapter 5 gives conclusions as well as the recommendations on further research work Huang Yang 2005 Huang Yang 2005 2 PROFIBUS Background This chapter gives an overview of PROFIBUS Most of the knowledge and figures are adopted from 1 2 3 4 2 1 History The PROFIBUS Process Fieldbus project started in 1987 an association venture project supported by the public authorities in Germany Within the framework of this venture 21 companies and institutes joined forces and created a strategic fieldbus project The goal was to realize and establish a bit serial fieldbus The basic requirement was the standardization of the field device interface A first step resulted in the specification of the complex communications protocol PROFIBUS FMS Fieldbus Message Specification which is the first PROFIBUS communication protocol In 1993 a further step was the completion of the specification for the more simply configurable and faster PROFIBUS DP Decentralized Periphery protocol As for its in
5. Open channels Parameterization Normal Operation Master Function Get Set Slave Configre Activate Slave Watchdog Mast Start PROFIBUS slave Get Slave Diagnose Connect Disconnect Set Start Stop Get Put Exchange DP Poll Data Slave Function PARAM FUNCTION pua ban aa 1 J j L L L L L L L L L L L L l Exchange Process Data Data Transmission between Naster and Slave Start PROFIBUS slave Figure 13 Process of using the DF Profi2 PCI board as a PROFIBUS 18 Huang Yang 2005 3 Using LabVIEW for Windows LabVIEW for Windows does not guarantee a real time performance Also the contents of VIs supplied by COMSOFT are password protected With this option neither real time performance nor a further look inside the PROFIBUS protocol is possible 3 3 Conclusion After careful comparison among these choices finally it was decided to develop our own C C application with a Windows driver Since the Windows operating system itself is not a real time platform time characteristics obtained by measurements are not expected to be deterministic For the sake of measurement accuracy the method for timing in Windows is essential All these have been taken into account to make measurement and make tests performed on Windows XP as close as possible to a realtime operating system measurement Details of th
6. References 1 PI 2005 PROFIBUS Glossary www profibus com 2 PI 2005 PROFIBUS Overview www profibus com 3 PI 2002 PROFIBUS Technology an Application System Description www profibus com 4 ACROMAG Inc U S A 2005 Introduction to PROFIBUS DP 5 LabVIEW 2005 LabVIEW Real Time Module User Manual http www ni com labview 6 Johan Nilsson 2005 Implement a Continuously Updating High Resolution Time Provider for Windows 7 COMSOFT Germany 2005 PROFIBUS DP DPV 1 Protocol Driver 41
7. a segment is limited to 32 However this number can be further determined by the protection type selected and bus power if any Joint operation of bus powered and externally fed devices is permissible However note that externally fed devices also consume a basic current over the bus terminator which must be taken into account accordingly when calculating the maximum available feed current Fiber optic In industrial areas with high electromagnetic noise copper based communication media may not be able to cope with induced errors This problem does not exist if fiber optic technique is used Fiber optic is also suitable for applications where particularly large distances need to be covered Line ring and star topology structures are supported with Fiber optic transmission Layer 2 Layer 2 in PROFIBUS is designated as Fieldbus Data Link FDL This layer is divided into two sub layers the Media Access Control MAC layer and the Logical Link Control LLC layer There are two main communications in PROFIBUS communication between masters and that between one master and its slave slaves As for MAC methods have to be defined to satisfy two essential requirements In the case of communication between masters all of these stations have to be given sufficient opportunity to execute their communication tasks in a specific time interval For communication between a master and its slave slaves data transfer should be performed in a
8. amount of data needs to be transferred in a short period of time 5 2 Recommendations To realize a PROFIBUS network with a faster board driver and complete the measurements To investigate the realtime performance of PROFIBUS based on a realtime OS To build a multi master and multi slave PROFIBUS network and then check its performance 35 36 Huang Yang 2005 Appendix A Numerical values of measurement results Average Time Elapsed of GET PUT DP POLL DATA 12Mbps I O bytes Mean Standard Deviation 4 75 10 8 76 10 16 79 10 32 85 11 64 94 11 128 115 11 240 151 13 Action Time Elapsed of GET PUT DP POLL DATA 12Mbps I O bytes Mean Standard Deviation 4 116 13 8 117 13 16 116 15 32 129 16 64 141 15 128 165 14 240 220 16 Idle Time Elapsed of GET PUT DP POLL DATA 12Mbps I O bytes 4 8 16 32 64 128 240 Mean Standard Deviation 24 24 25 24 25 25 25 4 a o Bh U tA N Average Time Elapsed of DFP2 PUT GET SLV DATA 12Mbps I O bytes Mean Standard Deviation 4 33 8 35 16 39 32 46 64 62 128 91 240 144 Action Time Elapsed of DFP2 PUT GET SLV DATA 12Mbps I O bytes 4 497 8 487 16 499 32 503 64 521 128 537 240 582 Idle Time Elapsed of DFP2 PUT GET SLV DATA 12Mbps I O bytes 4 8 16 32 64 128 240 Mean Standard Deviation Mean Standard Deviation 19 21 24 29 41 65 113 Huang Yang 2005 79 80 84 87 98 108 118 47 17 44 21
9. doing test in this way is the encapsulation of VIs offered by the company who provides the boards Protected by passwords these VIs are black boxes to the end users This means that it is hardly possible to measure time delay of a certain part of the whole transmission for instance time delay of those parts within Process Data before the transmission starts As shown in Figure 12 time spent in section A of the DFP2 TransferProcessData vi does not belong to the measured transmission delay 16 Huang Yang 2005 DFP2 TransferProcessData vi A Data transmission start point Figure 12 Data flow through a VI The black box approach does not give enough insight in the way things work and is therefore not especially interesting for future subprojects of the parent PhD project On the other hand it is required to purchase extra hardware the PXI devices for this approach However the cheapest PXI product made by National Instrument PXI 1031 already cost more than 1000 For our case such hardware is out of the budget of the individual project In addition for this method the PCI boards should be replaced with CPCI boards Note Configuration tool of the board currently does not support LabVIEW RT RTX version 2 Developing own application in C C Own application development is available under Windows by directly accessing the API provided by the driver of the board A general process for using the board as a master or a slave wi
10. in both factory and process automation Figure 4 shows the best known PROFIBUS versions A Process Factory Safety Motion 7 Automation Automation Control PROFIBUS PROFIBUS DP PROFIsafe PROFldrive Typical Dipa designation ET AA Profil Application p PROFIdrive ts Communication DP DP DP DP pr protocol protocol protocol protocol PMA T S P Pai i RS485 RS485 Transmission p MBP IS RS485 MBP IS R3495 Figure4 Typical PROFIBUS versions 2 3 Specification 2 3 1 Basics for PROFIBUS DP PA Some basic properties of PROFIBUS DP PA standard are summarized in Table 1 Maximum message length 244 bytes Data rate 9 6Kbit s 12Mbit s Transmission distance 100 M 12 Mbit s 1200 M 9 6Kbit s Maximum stations nodes without repeaters 32 Recommended maximum repeaters 4 Arbitration method Token Passing Error checking HD4 CRC Network topology Line star and ring Physical media Twisted pair or fiber Communication methods Master slave and peer to peer Table 1 Basics for PROFIBUS DP PA Amaximum transmission rate of 15Mbit s can be achieved when using RS485 IS Too many repeaters will cause considerable signal delay When used repeaters count as stations on the bus though they do not require an address of their own Huang Yang 2005 2 3 2 Communication Layers PROFIBUS communication is anchored in the
11. instance rapid prototyping with testing of a system prior to having the actual plant available Alternatively it can be used to test properties of the system that due to the associated safety risks cannot be tested on real plant The computer architecture of this HIL simulation setup consists of one PLC on which the control program is running and a PC where a simulation of the plant is running Connection between the two is done via a PROFIBUS connection This architecture is in first phase further simplified by using a PC instead of a PLC Thus the computer architecture that will be used in this project consists of two PCs connected via a PROFIBUS connection The framework of the system is depicted in Figure 1 PC1 DS IN PC2 Contorller PROFIBUS 4 Plant Simulation Figure 1 Framework of the experimental HIL simulation setup In this individual project focus is on the PROFIBUS communication and its applicability for this kind of control systems 2002 2006 Bojan Orlic CSP channels for field bus connected embedded control systems Huang Yang 2005 1 2 Assignment This individual project is focused on the real time characteristics of the PROFIBUS fieldbus However the time delay tests and measurements were performed only for of PROFIBUS DP boards controlled by a PC running Windows XP Still same tests are reproducible for other more real time operating systems Firstly literature study was performed in or
12. international standards IEC 61158 and IEC61784 In the ISO OSI reference model in which 7 layers have been defined in the protocol PROFIBUS uses layers 1 2 and 7 Figure 5 Receiver Designation and function of the layers ae Interface to application program with application oriented Application layer commands read write ER Representation coding of data for analysis and Presentation layer interpretation in the next layer lenin enon etree ni 5 Session layer synchronization of communicating processes Controlling data transmission for layer 5 transport errors break down into packets PEE la ieee ie ad Network layer congestion UEM 0 0 Data link layer Control MAC including data securit Pn Definition of the medium hardware coding and speed of Physical layer the data transmission Figure 5 PROFIBUS employs 3 layers from the OSI reference model 4 3 Layer 1 Layer 1 called the physical layer defines the medium coding and speed of the data transmission PROFIBUS provides four different transmission technologies RS485 RS485 IS MBP and Fiber Optics all based on international standards and associated with PROFIBUS in both IEC 61158 and IEC 61784 standards RS485 RS485 is a simple cost effective medium It is primarily used for tasks that require high transmission rates in a range between 9 6 Kbit s and 12 Mbit s for RS485 IS the range is 9 6 Kbit s to 15 Mbit s The maximum permissible line length depends on the
13. sends data to the slave and the first polling cycle finishes the next polling starts and data stored in the slave memory is fetched to master immediately Within this period of time the slave has not yet finished data processing including copying data from its input buffer to the output buffer Another factor for the delay is that for slave a minimum interval has to be inserted between two successive memory operations Then the master can only receive the slave s updated response after the third polling finishes 30 Huang Yang 2005 Master Slave Get Data Get Data Data Flow Time Measured 3T Figure 26 Sequence diagram of a roundtrip 4 4 Control Loop Implementation In order to test the performance of PROFIBUS DP from the control engineering point of view a distributed control loop is deployed The control loop implementation is an application of the previous measurement results as well Based on the PROFIBUS DP connection a distributed simulation of the plant and associated controller are implemented Both plant and controller are modeled in 20SIM tool The plant is a simple DC motor known as Linix The 20SIM model is shown in Figure 27 Controller takes in the sampled signal from a set point generator and feedback from the Linix Plant as well After zero order hold the output steering signal of Controller is then sent to the Linix Plant Based on this simple feedback loop the output of the Plant will follow the
14. slave and receive response from the slave to the master The elapsed interval is the time latency of a single PROFIBUS roundtrip Only the master of PROFIBUS can access the bus and manage the whole process of communication Then we do timing in the master application The services of PROFIBUS DP GET PUT DP POLL DATA in master board and DFP2 PUT GET SLV DATA in slave are again used in this measurement Note the time at the start and end of the service and the roundtrip time is the difference between these two measurements The whole procedure is depicted in Figure 24 29 Huang Yang 2005 Start Timing k Ry Send Data Receive Data Ng Roundtrip Time PROFIBUS Copy Data Recei eceive Data b Send Data Er StopTiming sA Figure 24 Procedure of Measurement III N Master Slave The Measurement results are shown in the Figure 25 It can be found out that when the user data length is shorter than 128 the roundtrip time is a fixed value about 6 ms In the case of longer data length needs to be transmitted the roundtrip time increases noticeably which is coincided with the conclusion of Measurement II Roundtrip Time m e gt pu E fes 32 VO bytes mean Standard Deviation Figure 25 Although a single polling cycle time is T it takes 3T to complete a roundtrip This is illustrated in Figure 26 As soon as the master
15. transmission rate as defined in the system description of PROFIBUS RS485 supports the line topology All devices are linked in a bus structure which allows addition or removal of stations or the step by step commissioning of the system without influencing other stations Up to 32 stations can be connected in a single segment Wiring and bus termination of RS485 is shown in Figure 6 The beginning and end of each segment is fitted with an active bus terminator which has a permanent power supply to ensure error free operation If more than 32 stations are required one possible way to expand the network is utilizing repeaters to link the individual bus segments The other possibly simpler is to increase the input impedance of the transmitters for instance increasing the impedance from 12k2 to48k enlarges the number of stations within a segment from 32 to 128 Huang Yang 2005 Different cable types type designation A D for different applications are available on the market for connecting devices either to each other or to network elements segment couplers links and repeaters When using RS485 transmission technology PI PROFIBUS International recommends the use of cable type A Specification of cable type A is listed in Table 2 Impedance 135 to 165 Q Capacity lt 30 pf m Loop resistance S1102 km Wire diameter gt 0 64mm Core cross section gt 0 34mm2 Table 2 Specification of cable type A RS485 work
16. 3 4 2 Measurement Ina SAN ENG SNN EN NN NN 27 4 3 Measurement naa ann cerei it kaka NADA in an ai cr hoe UR tali 29 4 4 Control Loop Implementation ee aaa 31 4 5 DISCUSSIOTI 4 ere ee o PR EN Nan PEE EE nanah 33 5 Conclusions and Recommendations oooooooooooo ooooooooooooooooooooooi 35 PAE COfCclustOfis ceo ies te AA NN 35 5 2 Recommendations in niece tea tice ie obese deese ceu tee NN 35 Eug ECT DEDE 37 Numerical values of measurement results esses 37 Referenges AA AA AA AA AA 41 III IV Huang Yang 2005 Huang Yang 2005 1 Introduction 1 1 Context This project has two contexts The first one is local PhD research on CSP Communicating Sequential Processes based compositional programming techniques in the context of distributed real time systems connected via fieldbuses The research deals with hard real time control using several co operating processors in networked environments The network itself is embodied by an industrial field bus which are investigated with respect to real time performance The second context origins from the collaboration of our research group Control Engineering at University of Twente in the MIC Mechatronic Innovation Centre Partners of MIC are dedicated to building experimental HIL hardware in the loop simulation of a plant controlled via a PLC Programmable Logic Controller controller The purpose of such a HIL simulation can be for
17. 56 38 44 tA O0 tA l2 NE 37 Huang Yang 2005 Average Time Elapsed of Average Time Elapsed of GET DP POLL DATA 12Mbps PUT DP POLL DATA 12Mbps YO bytes Mean Standard Deviation YO bytes Mean Standard Deviation 4 76 11 4 20 14 8 TI 11 8 21 9 16 80 11 16 21 10 32 85 12 32 20 9 64 95 11 64 20 10 128 116 12 128 21 11 240 152 15 240 21 13 Action Time Elapsed of Action Time Elapsed of GET DP POLL DATA 12Mbps PUT DP POLL DATA 12Mbps I Obytes Mean Standard Deviation T O bytes Mean Standard Deviation 4 118 10 4 63 8 8 115 12 8 72 16 119 12 16 71 7 32 128 14 32 70 8 64 144 14 64 74 9 128 169 17 128 83 11 240 208 16 240 89 10 Idle Time Elapsed of Idle Time Elapsed of GET_DP_POLL_DATA 12Mbps PUT_DP_POLL_DATA 12Mbps I O bytes Mean Standard Deviation TO bytes Mean Standard Deviation 4 25 5 4 12 1 8 25 6 8 12 2 16 25 3 16 12 1 32 24 4 32 12 2 64 24 4 64 12 1 128 24 6 128 12 3 240 24 2 240 12 2 38 Cycle Time I 12Mbps I O bytes Cycle Time 4 8 16 32 64 128 240 1900 1900 1900 1900 1900 2400 2500 Roundtrip Time 12Mbps I O bytes 4 8 16 32 64 128 240 Mean 5922 5916 5926 5941 6105 7222 8146 Standard Deviation 163 207 170 201 298 301 221 Cycle Time II 8 input bytes 8 output bytes Huang Yang 2005 Baudrate Cycle Time 9 6k 93 75k 1 5M 3M 6M 12M 90000 7000 2300 2100 1900 1900 Time Unit 4 39 40 Huang Yang 2005 Huang Yang 2005
18. University of Twente EEMCS Electrical Engineering Control Engineering Time characteristics of PROFIBUS on Windows XP Yang Huang Individual Project Supervisors prof dr ir J van Amerongen dr ir J F Broenink dipl ing B Orlic dr H Vos June 2005 Report nr 019CE2005 Control Engineering EE Math CS University of Twente P O Box 217 7500 AE Enschede The Netherlands Huang Yang 2005 Abstract This project is aiming to investigate the use of the PROFIBUS as a fieldbus for control purposes and to determine the time characteristics of PROFIBUS data transmission between two PCs The actual experiments are performed with PC s running Windows XP which is a non realtime OS The experiments are based on a one master one slave PROFIBUS network with DF Profi2 PCI boards Measurement results show that under Windows XP the transmission time for a single PROFIBUS DP polling cycle is ranging from 1 9 ms to 90 ms considering all extreme conditions This time is mainly dependent on the transmitted data length and the transmission baud rate The conclusion is that PROFIBUS used in combination with a non realtime operation system is still a reasonable deterministic cost efficient and simple fieldbus suitable for managing I O peripherals in the field level However it might not be fast enough for high speed real time motion control purposes Further investigation of real time performance with PROFIBUS is recommended
19. d data length is 8 bits output and 8 bits input Then the minimum of interval between which a steering signal or a feedback signal is sent should be 1900 ps according to the result of Measurement II When delay time of the Controller the interval is set to 1000 Hs We obtain an output of the Linix Plant shown in Figure 29 It can be seen that the output is not stable and there is obvious difference between the simulation and that in 20SIM simulator That is because some of the control and feedback signal is missing due to the too small interval value 32 Huang Yang 2005 0 200 400 600 800 1000 1200 1400 1600 1800 2000 Figure 29 Simulation based on PROFIBUS connection with 1 ms interval When the delay time is set to 1900 HS a perfect following is achieved which is shown in Figure 30 1 4 1 2 0 5 0 6 0 4 0 2 0 200 400 600 800 1000 1200 1400 1600 1500 2000 Figure 30 Simulation based on PROFIBUS connection with 1 9ms interval 4 5 Discussion With the Windows driver a single polling cycle time in PROFIBUS with a DF Profi2 PCI master and a DF Profi2 PCI slave is around 1 9 ms when transmitted user data length is equal or less than 64 bytes This value will increase noticeably with more user data for the extreme case a data transfer of 240 bytes output and 240 bytes input data will require a 33 Huang Yang 2005 cycle time of 2 5 ms Measurement II The time cost in accessing the memory on the board shou
20. data in its input buffer at a very high frequency The frequency is sufficiently high that data transmitted from the master in every polling cycle can be stored in the slave Then by changing the value of master sending interval a minimum can be found such that the slave can still catch every data transmitted 27 Huang Yang 2005 from the master This minimum is the polling cycle time A block below illustrates the whole measurement procedure PROHBUS Delay R Figure 21 Procedure of Measurement II Two factors data length and baud rate are taken into account here First the baud rate is fixed to 12Mbps and data length is changed This measurement result is shown in Figure 22 The second measurement is done with changing baud rate and fixed transmitted data length the result of which is depicted in Figure 23 Cycle Time Figure 22 Cycle Time Il 100000 80000 60000 40000 20000 0 3 6k 93 75k 1 5M 3M Baud rate Figure 23 28 Huang Yang 2005 Measurement result of Cycle Time I reveals that time elapsed within a single polling is less than 2 ms to be specific 1900 H 3 unless a data length of more than 64 is chosen That is because at a certain transmission rate 12Mbit s in this case a limited length of data can be sent within a single polling cycle When more user data is to be transmitted extra transmission time is required The theory applies in Cycle Time II as well in which the cycle time d
21. der to answer the questions what PROFIBUS is and how communication via PROFIBUS is realized Then a real PROFIBUS setup was built utilizing DF Profi2 PCI boards made by COMSOFT GmbH Germany Several options existed for constructing the PROFIBUS network After comparison and analysis and taking the limitations of this project into account the decision was made to develop an own application in C C that will directly communicate to the Windows XP board driver Figure 2 shows the setup built in this project PCI PC2 Contorller Plant Simulation DF Prof2 PCI Figure2 Framework of the setup in this project no gia Slave Several tests were carried out in order to obtain time characteristics of PROFIBUS in Windows After the transmission latency measurement a simple control loop was built Instead of real plant a simulation was used and the code for both simulation and controller was generated from a 20SIM model http www 20sim com For the highest communication speed simulation results based on PROFIBUS transmission experiment are almost identical to the results obtained in 20SIM simulator 1 3 Outline of the report Chapter 2 provides basic information about the PROFIBUS including a short history the explanation of its modular structure and some interesting points from the specification of PROFIBUS Chapter 3 describes the design choices in this project For the reader s better
22. e measurement and test will be shown in the next chapter 3 4 Test Design 3 4 1 Measurement setup structure As mentioned before DP utilizes polling to realize master slave communication In our case the Windows driver of the board handles the whole polling procedure The polling starts as soon as the board is set to be a master and activated A user application can write and read the board s memory via the provided API functions In every polling cycle the driver of the master will send command to the slave A command can be either sending data to a slave requiring data from a slave or both In such a way user data in both memory of master and slave will be updated after a single polling User Application Slave User Application Master Board Memory Board Memory FDL Board Driver BUS Physical Layer Communication Communication handled by API functions handled by Board driver Figure 14 PROFIBUS DP communication data flow 19 Huang Yang 2005 As illustrated in Figure 14 Layer 2 FDL is handled by the board driver Above Layer 2 some API functions are available in the User Interface for user to access the memory of boards 3 4 2 Priority and timing issues Considering the operating system we choose in this project is not a realtime operating system the process and thread of timing and communication may be interfered by other processes or threads In order to obtain the measuremen
23. ecreases sharply with the transmission baud rate increases It is measured that the cycle time comes to a constant value after the baud rate is sufficiently high Further increasing the baud rate will not improve the cycle time because the bottleneck is for high baud rates situated in a driver performance Again the minimum cycle time in the measurement is 1900 us This result has revealed that PROFIBUS is only deterministic at a fixed maximum data length If the number of slaves or the data length transmitted with a slave increases and the total transmitted data length exceed the maximum the PROFIBUS transmission time will increase incredible The term deterministic of PROFIBUS also indicates that once the data length is known the transmission time is predictable Note that the time measured during a single cycle mainly includes time elapsed for the following activities happened during transmission Time for the profibus chip to send the DataExchangeRequest to the slave The time for the slave to answer and send the data The time for the profibus chip copies the data to its memory The time the Arbitrating processors offers the new data to the windows driver that is to copy data from memory of the board The time for the slave to copy its received data to the output buffer 4 3 Measurement Ill Another method to examine the time characteristics is to utilize a roundtrip That is to send a piece of data from the master to a
24. eline 2 041 14 Huang Yang 2005 3 Design Choices In this chapter design choices in this project are discussed Since LabVIEW and its real time module is one of the options being discussed first a brief description on LabVIEW is given 5 3 1 Introduction to LabVIEW LabVIEW is a graphical development environment developed by National Instruments http www ni com labview Applications created by LabVIEW are widely used in the fields of testing measurement and control engineering In LabVIEW the block called VI Virtual Instrument takes the place of function compared to general purpose programming languages Functions can be called by other functions and similarly VIs can be encapsulated and reused by other VIs Data flow determines the execution of applications in LabVIEW environment A VI will not function until all its input data is ready There are generally two parts for a VI the front panel and the block diagram Figure 11 The front panel offers a friendly graphical user interface for instance buttons waveforms and LED lights Behind the front panel the block diagram containing LabVIEW source code defines the functionality of the VI A VI functionality can be exported to an executable file and be executed on any PC with the LabVIEW environment installed Further more the block diagrams of a VI can be protected by a password to keep the contents secret while this VI can still be used via the front panel by the end user
25. especially in the combination with some real time Operating Systems Extending the simple network to a larger scale network can help to get further insight into this fieldbus technology II Huang Yang 2005 Huang Yang 2005 Table of Contents T Introd ctioti E E E EE E N EA 1 T E Context iaa eure ain NN BNN 1 1 2 ASStPniment verte eec tte De E GINANG NEN NN E h 2 1 3 Outline of the report eere ree e e ee e bar TER e tert ee GALA 2 2 PROFIBUS Background o ee eese ee oath NAGIN pF eterne ae en Pe eI nana eee ese noe eu eT eina oso NBA 5 2 History suede Den RR teu eh REGE aa 5 2 2 PROFIBUS modular structure oooooo oo nana 5 2 3 Specificatii pce las oon ee hotte t E Hn e pa ce ehe ee b Hp Eo geb MeL iE e bot iet rope eater bas 6 2 3 1 Basics for PROFIBUS DP PA a anna ana 6 2 3 2 Communication Layers niae Hee tee LS ER Pe t es 7 2 3 9 PROBIBUS DP Protocols ertet het m eee dite 12 ANA GS DIOS ote me rete tete met ete et ene d etd nets 13 SANIT 15 3 1 Introduction to Lab VIEW A naaa nana 15 3 2 BA Choi aa AAA ha Naas aa 16 3 32 CONCLUSION iti SA easet Re ASAN IBANAG 19 34 Test NA 19 3 4 1 Measurement setup StruCtuTe oo oom Wo 19 3 4 2 Priority and timing ISSUES ANA NAKA naas 20 3 4 3 API functions of the driver seen enne 21 4 Measurement and Testing ooooooocoooo ooooooooooooooooooooooooooooooooooooooooooooooo 23 4 1 Measurement 5 3 cona BAN aa 2
26. fast and efficient way with real time performance guaranteed Consequently MAC in PROFIBUS includes the token passing method which ensures the assignment of the bus access right within a precisely defined time interval with a token Within a pre defined period of time the token passing cycle time a master in the ring will receive the token and then be able to perform data communication The master has to pass its token to the next master as soon as the time allocated to it has expired As for communication between one master and slave slaves a master slave method is provided by MAC PROFIBUS slaves are passive stations in the network They can only acknowledge received messages or transmit messages at the master s request PROFIBUS Huang Yang 2005 master is termed active station and only the master can access the bus and handle the polling procedure Within a multi master PROFIBUS network the masters without token are all passive stations and therefore can be accessed as slaves When a master takes the token it is granted the right for data transmission on the bus and is able to communicate with the slaves A master can read data from any slave on line but it can only write to those slaves assigned to it This combined method is called the hybrid medium access Figure 7 illustrates the master master and master slave communication mentioned above Passive stations slave devices are polled Figure 7 A PROFIBUS DP network The LLC la
27. his project They returns the number of ms or even seconds that have elapsed since the system was started For those processors with a high resolution performance counter two API functions are available in Windows namely QueryPerformanceFrequency and QueryPerformanceCounter 6 The declaration signatures of these two functions are listed in Code 2 20 Huang Yang 2005 BOOL QueryPerformanceFrequency LARGE INTEGER IpFrequency BOOL QueryPerformanceCounter LARGE INTEGER IpPerformanceCount Code2 Syntaxes of Timing Functions With QueryPerformanceFrequency the frequency of the high performance counter is obtained Further QueryPerformanceCounter gives the current performance counter value With the counter frequency and two counter values at the start and the end point of time the time elapsed in a certain period of time can be obtained in ms by the following calculation nElapsed int CountEnd QuadPart CountStart QuadPart 1 000000 Fre QuadPart where CountEnd and CountStart are of type LARGE INTEGER and contain the counter values Fre is also of type LARGE INTEGER and provides the frequency of the counter nElapsed stores the time elapsed in ms With this method depending on the chosen platform the accuracy of measurement up to 15 can be achieved Figure 15 depicts the general procedure of measurements in this project Note that the Read Write block represents the process to be timed in different measureme
28. irst PC are Pentium IV 2530MHz 512 MB Memory Windows XP Settings of the second PC are Pentium II 400MHz 128 MB Memory Windows XP The most significant aspect affecting the time characteristic is the length of transmitted data and baud rate For the DF Profi2 PCI length of either input or output data can be set to 4 bits 8 bits 16 bits 32 bits 64 bits 128 bits 192 bits or 240 bits A DF Profi2 PCI board supports an input output pattern with any combination of the above data length Baud rates applicable for RS 485 cabling supported by the board are in the range from 9 6kbps to 12Mbps Measurement were performed by changing the transmitted data length and baud rate Besides Measurement II results in other measurements are provided in the form of mean and variance of 1000 samples If not specified the baud rate set in the measurements is 12Mbps 4 1 Measurement The first measurement was done in order to find out how much time is spent on the User Interface in data transmission That is the time elapsed in reading and writing the board memory In the measurement one of the following functions will be timed namely GET DP POLL DATA PUT DP POLL DATA and GET PUT DP POLL DATA Note that the API functions do not access the bus Instead they read or write to memory of the board The functions check if there is new data available in the memory or not If there is new data they copy the data Otherwise they return immediately with
29. itted data length ranging from 4 bytes input 4 bytes output to 240 bytes input 240 bytes output This indicates that with the new driver the transmission speed is about 2 times faster However more measurement is necessary to obtain a complete conclusion which is also recommended in the last chapter of this report 34 Huang Yang 2005 5 Conclusions and Recommendations The conclusion and recommendation below only applies to study and application of PROFIBUS DP with Windows driver 5 1 Conclusions PROFIBUS DP is suitable in management of field devices like sensors and I O devices etc Its open standard makes it convenient and safe to add or delete a device to from the network The simple fast and deterministic way of data exchange in PROFIBUS fits the requirement of communication in field level of a system PROFIBUS supports a wide range of transmission technology including RS485 MBP and optical fiber This feature reduces the cost PROFIBUS network and increases its compatibility as well PROFIBUS DP cannot satisfy the requirement of high speed real time motion control Which needs data transmission delay within scores of MS Since in the case of PROFIBUS communication it is necessary to have a transmission interval of around 1 9ms PROFIBUS exchanges a maximum data of 480 bytes in a single cycle that is 240 bytes output and 240 bytes input for a master It is not recommended to utilize PROFIBUS in the situation where huge
30. ld also be taken into account when considering the time factor For the master one read and write to the memory will take a period of time ranging from 115 HS to 220 4 depending on the length of user data exchanged For the slave this value is from about 500 HS to 580 5 Measurement I To do roundtrip within a single polling cycle is hardly possible The reason is that the slave has its own minimum operation interval and it needs some extra time to process receive data From the performed experiments conclusion can be drawn that 3 times of polling cycle time are required to complete a roundtrip Measurement III Control based on PROFIBUS DP is possible A transmitted data of no more than 64 bytes with a baud rate of 12Mbps can achieve a fastest transmission rate For a digital control system the maximum sample rate can achieve 526 Hz based on PROFIBUS DP transmission Considering the fact that Windows XP is not a real time operating system measurement could be interrupted by other processes Thus in the measurements timing cannot achieve to the accuracy of 1 5 and there is inevitable deviations in the measurements standard deviations are less than 100 5 in Measurement I and less than 300 US in Measurement ID At the end of the project beta version of the board s Windows driver is available and some tests are done based on this new driver Measurement shows that the roundtrip time is around 32004 at the baud rate of 12Mbps with transm
31. ll in the time gaps of every polling cycle As an introduction Figure 10 illustrates how DP manages cyclic and acyclic communication TIME LINE Polling Polling Poling Acyclic Polling Slavel Slave2 Slavel Communication Slavel GAS used by used by DP class 2 DP class 1 masters masters one polling cycle Figure 10 Sequential diagram for DP polling Due to the limitation of hardware in the project only DP VO will be further discussed in the following chapters 2 3 4 GSD files GSD an abbreviation for the German term Geritestammdaten files are device specific database files for PROFIBUS devices All PROFIBUS class 1 masters and slaves have their own GSD files A GSD file is an ASCII text file containing device specific data for instance vendor identification information supported baud rates number of input output data meaning of diagnostic messages data formats etc Many PROFIBUS configuration tools are available to edit check and use the GSD files The master of a PROFIBUS network utilizes GSD files to build its parameter record When initializing a PROFIBUS network the master imports necessary parameters of each station of the network from the GSD files before configuration starts In this way plug amp play interoperability among different devices from different manufacturers is realized 13 Huang Yang 2005 Details and format of a standard GSD file are specified in EN50170 standard and PROFIBUS Guid
32. ly Data is sent by a master to one of its slaves after which the slave returns the required data in its response if applicable in a pre defined period of time This period is also called one telegram cycle In regular PROFIBUS operation cyclic data exchange takes place between the master and its slaves by SRD CSRD Cyclic Send and Request Data with reply Communicate in SRD way but cyclically MSRD Send and Request Data with Multicast Reply CS Clock time Synchronization The time information at the master is broadcasted to all remote stations Note Not all the services are available for every PROFIBUS protocol for instance PROFIBUS DP does not support the SDA For more detail please refer to corresponding manuals In addition on layer 2 PROFIBUS also defines management related services for initialization configuration event and error handling The main communication protocol applied in layer 2 is protocol DP which realizes functionalities of this layer DP will be further introduced in section PROFIBUS DP Protocol Layer 7 The application layer provides the communication functions to the user Services offered by layer 2 are executed at the interface to the LLI Lower Layer Interface through SAPs Service Access Points Layer 7 is only defined in PROFIBUS FMS PROFIBUS DP and PROFIBUS PA define their own profiles as the user interfaces and do not use this layer In the report details on this is not presented
33. me when more user data is transmitted while time elapsed with PUT DP POLL DATA remains a more or less stable value even when user data length increases It can also be seen that GET PUT DP POLL DATA takes about the same time as GET DP POLL DATA does A conclusion can be made that reading is the most time consuming operation when accessing the board memory In the situation that both reading and writing with memory of the board is required it is recommended to use GET PUT DP POLL DATA instead of the combination of 26 Huang Yang 2005 GET DP POLL DATA and PUT DP POLL DATA In the slave board only one function is provided for reading and writing the board memory DFP2 PUT GET SLV DATA Time elapsed with this function call is depicted below Similar with GET PUT DP POLL DATA more user data indicates longer delay with DFP2 PUT GET SLV DATA Time Elapsed of DFP2 PUT GET SLV DATA Mean 32 NO bytes Average Time Action Time Idle Time Figure 20 a Time Elapsed of DFP2 PUT GET SLV DATA Standard Deviation Time us 32 VO bytes e Average Standard Deviation a Standard Deviation of Action Standard Deviation of Idle Figure 20 b 4 2 Measurement II Since PROFIBUS DP communication is based on polling the time of a single polling cycle can be measured by one way transmission That is the master keeps sending changing data to its slave at regular intervals and the slave store the
34. ndows scheduling The latter ETS needs NI RT series hardware to work on It provides for the hardware a real time operating system on which the LabVIEW RT applications can be executed Once downloaded to the hardware the embedded program can run without disrupting even when the host computer is entirely shut down 3 2 Design choice In the project three options are available to build a PROFIBUS network 1 Using LabVIEW RT ETS version and a PXI hardware In this approach two DF Profi2 CPCI Compact PCI boards should be first mounted into a PXI PXI 1031 for instance device as a PROFIBUS module and each PXI device can be connected to a separate PC After configuring one board into master and the other slave the LabVIEW RT application and a PROFIBUS network configuration file are downloaded onto the master CPCI board The PXI devices can be accessed by their unique IP address After rebooting the hardware the application downloaded to the PXI device will start initialize the network do transmission and measurement and finally send back required data to the host computer Note PXI PCI eXtensions for Instrumentation is a rugged PC based platform for measurement and automation systems Pros and cons It is common way utilizing LabVIEW RT to build a PROFIBUS network The PC based architecture and advanced synchronization features of the PXI modular instrument promise a high performance in real time tasks An obvious problem of
35. nts Figure 15 Procedure of Measurement 3 4 3 API functions of the driver The Windows driver of the board provides API functions for initialization parameterization diagnose and communication purposes For a better understanding of the measurements presented in the next chapter the functions for master slave communication will be introduced in this report For more detail of development with the board driver please refer to the user manual 7 GET DP POLL DATA This function reads the actual process image from the memory on the board The process image contains the slaves output data status and time stamp as well Data in the process image is updated after polling every time PUT DP POLL DATA This function puts the output commands for slaves to the memory on the board The commands will be sent to slaves in the next polling cycle 21 Huang Yang 2005 GET PUT DP POLL DATA This is the combination of the first two functions but is the fastest and most efficient way to exchange data with DP slaves The advantage of using this function will be shown by later measurement DFP2 PUT GET SLV DATA This is the only function available for a slave to prepare response data and read commands of the master from the memory on the board In addition the slave status and timestamp can be obtained with this function 22 Huang Yang 2005 4 Measurement and Testing In the performed experiments 2 PCs were used Settings of the f
36. out data operation This can be shown in Figure 16 In the figure peaks appear cyclically indicating data in the memory is updated and the function copies data which will take time The rest samples with a correspondingly smaller value mean that at that moment there is no new data available then no data copy is necessary The function does nothing but return directly Hence less time costs in this case Due to some unpreventable interrupts from other processes peak values vary from one to another 23 Huang Yang 2005 Time Elapsed 100 200 300 400 500 600 700 800 900 1000 Samples Figure 16 Time elapsed in 1000 measurement samples The following graphs show the measurement results on the time elapsed of these functions Note that in the graphs the number on I O bytes axis indicates both the input and output data length for instance 4 means the length of input data is 4 bytes and that of output data is also 4 bytes Average time and Average Standard Deviation is the mean value and standard deviation of time elapsed in all the samples Action Time and Standard Deviation of Action has the same meaning but is obtained from those samples in which the function really copies data Idle Time and Standard Deviation of Idle give the mean and standard deviation of those samples in which the function does not copy data but return Time Elapsed of GET PUT DP POLL DATA Mean VO bytes Average Time Action Time
37. s in half duplex operation only a single channel is available for communication In order to achieve synchronization PROFIBUS adopts NRZ Non Return to Zero VP 6 Station 1 Station 2 390 n Data line RxD TxD P 3 O 3 RxD TxD P RxD TxD P 3 DGND 5 C 5 DGND 2209 VP 6 O 6 vp RxD TxD N 8 D Q 8 RxD TxD N Dala lina RxD TxD N 8 Shielding Protective Protective 390n ground ground DGND 5 Wiring Bus termination Figure 6 Wiring and bus termination in RS485 transmission technology RS485 IS RS485 IS Intrinsic Safety is a 2 twisted pair version developed for strict demand of intrinsically safe application It will not be discussed further in this report MBP MBP is synchronous transmission with a defined transmission rate of 31 25 Kbit s The term MBP stands for the transmission technology with attributes Manchester Coding and Bus Powered This technology is frequently used in process automation as it satisfies demands for intrinsic safety and bus power using two wire technology Huang Yang 2005 As for topology tree and line structures as well as the combination of the two are supported by PROFIBUS with MBP transmission In practical use MBP is usually limited to a specific segment of a plant for instance field devices in hazardous areas This plant is then linked to the RS485 segment e g a control system through a segment coupler or links The number of stations that can be connected to
38. set point signal Sample Figure 27 20SIM model Figure 28 displays the simulation result in 20SIM simulator with a sample frequency of 1 KHz 31 Huang Yang 2005 model 1 5 setpoint X 20 sim rad logging ASHA OHA DIRE oo 0 5 D 0 5 1 1 5 2 time s Figure 28 20SIM Simulation Utilizing the C Code Generation tool of 20SIM Controller and Linix plant in the control loop are implemented into the application of PROFIBUS master and slave separately In realization the Controller is deployed on the PROFIBUS master application while Linix Plant is deployed on the slave application Since the steering and feedback signal is sent between intervals it is not necessary to implement the ZOH and Samplers Within both the Controller and Linix plant there is a while loop which repeats the calculation of input and output signals Euler integration method is utilized in the Linix plant to simulate the continuous performance The two models namely Controller and Linix plant are running in two different PC separately In case models are recalculated continuously without inserted time delays speed difference between these two PCs will affect the simulation seriously To prevent that a delay time is inserted into the loop of both models Necessary intervals for PROFIBUS communication limit the minimal value of the delay time From the PROFIBUS transmission point of view the baud rate is set to 12Mbps and transmitte
39. t data as accurate as possible the corresponding process and thread are set to the highest priority In Windows the highest priority for a process and a thread are REAL TIME PRIORITY CLASS and THREAD PRIORITY TIME CRITICAL separately A process of REAL TIME PRIORITY CLASS has the highest possible priority The threads of the process preempt the threads of all other processes including operating system processes performing important tasks For example a real time process that executes for more than a very brief interval can cause disk caches not to flush or cause the mouse to be unresponsive In this way the measurement process is protected from being interrupted by most processes Stable and reliable measurement results can be obtained A template is shown below that can be used to set the measurement process and thread to the highest priority SetPriorityClass GetCurrentProcess REALTIME PRIORITY CLASS SetThreadPriority GetCurrentThread THREAD PRIORITY TIME CRITICAL measurement done here SetThreadPriority GetCurrentThread THREAD PRIORITY NORMAL SetPriorityClass GetCurrentProcess NORMAL PRIORITY CLASS Code 1 Template for Setting Priority PROFIBUS DP offers the time stamp profile which is used for precise time assignment However the precision of this time stamp is only one ms which is a too long time period for the measurement Many Windows functions are available for timing but most of them are not suitable in t
40. ternational standardization PROFIBUS achieved national standardization in 1991 1993 in DIN 19245 Part 1 3 and Europe wide standardization in 1996 in EN 50170 Then since 1999 it was standardized in IEC61158 In 2002 the completion of activities to update IEC61158 was established 2 2 PROFIBUS modular structure PROFIBUS has a modular design Various communication technologies application and system profiles as well as device management tools are available for specific requirements The system structure is illustrated in Figure 3 below o Application Profiles II 0 Hi lt o E 2 2 a CE D gt 5 Ng o5 8 K5 FA 2 8 lt 2 es 58 mi 2 e 5 5 wo E8 5 B5 55 Application Common Application Profiles optional Profiles PROFIsafe Time Stamp Redundancy etc 2 as 32 Ot og Communication E a 8 u Technologies 2 8 8 E go 53 SE Spe a z0 Ov E Transmission NRZ Fiber Glass Multi Mode MBP Manchester Bus Powered 5 2 s 3 Technologies intrinsic Safety Optics Glass Single Mode MBP LP Low Power 03 a t 5 PCF Plastic Fiber MBP IS intrinsic Safety or 2 0 Figure3 Technical system structure of PROFIBUS Huang Yang 2005 Notice that the Application Profiles I II are user interface layers above the application layer Which is not depicted in the graph since the popular PROFIBUS techniques do not utilize this layer Based on this structure several mature PROFIBUS modules are widely used
41. th the driver is depicted in Figure 13 Pros and cons Apparently writing own programs in the lower level might take more time to implement the desired tests More detailed knowledge about the PROFIBUS protocol is necessary in order to do application development without an advanced development environment like LabVIEW RT However demo codes supplied with the board driver will definitely shorten the development period sharply since it is possible to construct our own application by easily reusing functions of the demos instead of starting programming from scratch Doing test in C C way gets closer to the need of the associated PhD project With more specific knowledge of PROFIBUS measurement result can be obtained and explained in a more transparent and logical way On the other hand the research group is now working on Debian Linux patched with RTAI COMSOFT the producer of the board used here does not support this real time system QNX a real time operating system that is supported by the board manufacturers becomes an impractical choice in this project because too much time would be spent on studying this OS Thus Windows becomes the only choice if this option is chosen Besides since the operating system here does not satisfy the real time demands it is hardly possible to measure the real time characteristics of PROFIBUS 17 Huang Yang 2005 Initialization Search for device Hardware Initialization
42. xtensions Clock Synchronization amp Time Stamps HARTonDP Up Download Segmentation m Redundancy DP V1 Acyclic Data Exchange between PC or PLC and Slave Devices plus extensions Integration within Engineering EDD and FDT ma m Portable PLC Software Function Blocks IEC 61131 3 0 m Fail Safe Communication PROFIsafe x m Alarms 2 DP VO 9 m Cyclic Data Exchange between PLC and Slave Devices plus extensions m GSD Configuration m Diagnosis ml Time Figure 9 Functionality of the PROFIBUS DP version with key features 12 Huang Yang 2005 DP VO specifies the basic functions of PROFIBUS transmission including master master s token passing method and master slave s polling data exchange DP V1 mainly defines the acyclic data exchange functions of DP master class 2 DP V2 introduces advanced functionalities that enable communication between slaves As for acyclic data exchange it can only be done between a DP class 2 master and a DP slave In addition the acyclic communication takes place when all the slaves of DP class 1 master are polled and there is still time left until next polling cycle The DP class 2 master can take the token and is able to access the slaves acyclically after all the DP class 1 masters have used up their time As soon as the next polling starts the acyclic communication has to stop and the token has to be passed to the first DP class 1 master In other words the acyclic data exchange can only fi
43. yer mainly controls frame synchronization flow control and error checking Based on RS485 all PROFIBUS characters are comprised of 11 bits in which one is the start bit 8 data bits one parity bit and one stop bit If no data is being transmitted the idle state potential on the data line is 1 A start bit causes the data line go to 0 A PROFIBUS NRZ coded character frame is given below Start DO D1 D2 D3 D4 D5 DE D7 Parity Stop ee ee RE RE ERES HL AA AA Figure8 A PROFIBUS character frame In the case of MBP transmission no start bit parity bit and stop bit is available That is only 8 data bits exist within one PROFIBUS character 10 Huang Yang 2005 Layer 2 of PROFIBUS operates connectionless In addition to the peer to peer data transmission only in FMS and PA broadcast and multi cast communication is also possible in PROFIBUS According to IEC 61158 3 layer 2 provides the following data transmission services to its layer above SDA Send Data with Acknowledgement Data is sent to a master or slave with a short acknowledgement returned in response SDN Send Data with No acknowledgement Message is sent simultaneously to a group of slaves multi cast or all slaves broadcast including other masters connected to the bus but without token However those slaves do not respond or acknowledge this message Broadcast is realized by SDN SRD Send and Request Data with rep
Download Pdf Manuals
Related Search