Post by rossanoparis on CAN TX regression in Control for Raspberry Pi 4.21.0.0 with CAN API library (RX works, TX silently fails) ^{CODESYS Forge talk (Post)}
Environment: Target: Revolution Pi (Kunbus RevPi), armv7l 32-bit OS: Raspbian GNU/Linux 9 (Stretch), kernel 4.19.95-rt38-v7 (SMP PREEMPT RT, built 2022-03-22) CAN: via SocketCAN (can0) Runtime: CODESYS Control for Raspberry Pi SL IDE: CODESYS V3.5 SP22 + patch 3 (previously V3.5 SP22 + patch 2, where everything worked with runtime 4.18.0.0) Library: CANBusAPI from the "CAN API" package, version 1.0.0.3 Summary: The setup worked correctly with V3.5 SP22 + patch 2 and runtime 4.18.0.0. After updating the IDE to SP22 + patch 3 and the runtime to 4.21.0.0, CAN transmission via the CANBusAPI library stopped working. Reception is unaffected. There are no error indications anywhere: the application runs normally and the CAN diagnosis reports a healthy bus. Systematic testing (below) shows the problem follows the runtime version, not the IDE update. Regression isolation: The same application (compiled under SP22 + patch 3, unchanged between tests) was downloaded to the same target with different runtime versions, changing only the runtime each time: 4.18.0.0: TX and RX both work 4.19.0.0: TX and RX both work 4.20.0.0: TX and RX both work 4.21.0.0: RX works, TX does not The behavior is fully reproducible in both directions (downgrading to 4.20 restores TX, upgrading to 4.21 breaks it again). Hardware, wiring, OS, project, and application code are identical across all tests. Diagnostic data (CANBusDiagnosis FB, while the application is transmitting on 4.21): ctSendCounter: incrementing normally ctTxErrorCounter: 0 ctRxErrorCounter: 0 eState: ERROR_FREE xBusAlarm: FALSE So the library accepts and counts the outgoing messages, no TX errors are reported, and the CAN controller state stays error-free, but the messages never reach the other nodes on the bus. Expected behavior: Messages sent via CANBusAPI (CANDriver / SendMessage) are transmitted on the bus, as with runtime versions up to and including 4.20.0.0. Notes: I am aware that the CAN API package (1.0.0.3, from 2016) is old; however, the regression is tied to the runtime version, since the identical binary works on 4.18-4.20. Questions: Is there a known change in CmpSocketCanDrv / CmpCAACanL2 between 4.20.0.0 and 4.21.0.0 that could affect TX via the CAN API library? Is runtime 4.21.0.0 still supported on Debian 9 Stretch (kernel 4.19, RT-patched), or is a newer OS required? If a minimum OS/kernel requirement was introduced with 4.21, is it documented?
Last updated: 2026-08-17

Post by janderson on OPC UA Server limitations, large array crashes runtime ^{CODESYS Forge talk (Post)}
What are the limitations of the OPC UA Server? I am trying to get data off my PLC that is acquired at high rates (~50k samples/s) so I am storing them in arrays and trying to get the arrays off the PLC. When I attempt to read a ~200k element array through OPC UA the server and runtime crashes (requiring tools -> update linux arm64 -> start runtime). Is there a better way to get highspeed data off? The ACDatalog library seems a bit irritating to use so I would prefer to go through OPC UA.
Last updated: 2023-08-23

Post by rossanoparis on How to configure CODESYSControl_User.cfg in order to use a specific network card for OPCUA Server ^{CODESYS Forge talk (Post)}
Hello at all I'd like to know how to configure CODESYSControl_User.cfg in order to use a specific network card for OPCUA Server. In the past, I tried to ask in "Forge" about this topic, but I could find neither help nor document informations about it. Reading the file "codesyscontrol.log" I see these logs OPC UA Server Started: Hostname: NTLRPICTRL01, Port: 4840 URL: opc.tcp://NTLRPICTRL01:4840 Loopbackadapter activated. **All available networkadapters are used.** Multithreading activated. 3 workers used. My controller, which is a Linux PC, has got two network cards eth0 and eth1 I whant to use the card with name "eth0" as a server for OPCUA and keep the "eth1" free for other stuffs. Is it possible ? Best regards Rossano
Last updated: 2023-08-25

Post by vformanek on How to access files correctly in Linux ^{CODESYS Forge talk (Post)}
Hello, its still not quite clear to me how to deal with the PLCLogic directory in linux. I have the directory right here: I need to correctly access the files in this folder through CODESYSControl.cfg. Here is my CODESYSControl.cfg where on the first lines there are a few lines: [SysFile] FilePath.1=/etc/, 3S.dat PlcLogicPrefix=1 That means that if I am going to use relative paths its going to go to from the PLCLogic directory right ? For example here: Whitch paths should I use. I cant seem to successfully run any of these aproaches ...
Last updated: 2023-09-08

Post by alimans on Hex string ^{CODESYS Forge talk (Post)}
Hi kdkwhite, for Word you still can use suggested code by using a union structure and crack down your Word to two byte as bellow: TYPE CrackWordToByte : UNION InWord : WORD; OutBytes : ARRAY [0..1] OF BYTE; END_UNION END_TYPE then define your variable as this type: udInput : CrackWordToByte; now assign your Word variable input to InWord and send OutBytes[x] to the mentioned method: udInput.InWord := WordVariableInput; Input := udInput.OutBytes[x]; Regarding your question about the code: actually 48 is ascii code of "0" and while 65 is the ascii code of "A" so in above code 55 + 10 would be 65.
Last updated: 2023-09-20

Post by matthew on New Ladder Editor Issues ^{CODESYS Forge talk (Post)}
Hi There, When using the new ladder editor there is no option for function blocks to Remove uncalled FB call parameters. I also see no update function block. When you double click the function block it no longer opens the function block. I also now get warnings if I insert FB call without a contact in front. Warning is: (Impl)): C0373: Expression or part of it has no effect I take it the new editor must work differently? The only reason for using new ladder editor is the ST execute/block editing is far better, If you have alot of text in the old ST execute the editor window is very small and when online it's even worse. The new ST block seems a lot better Thanks!
Last updated: 2023-10-10

Post by matthew on Persistence Manager Backup ^{CODESYS Forge talk (Post)}
Hi, Is there a process to restore Persistence Manager files? I have backups of persistence variables used in a machine that is currently operating and tuned and all set points adjusted. There are about 500 persistence variables, they are building another identical machine. I would like to use Persistence Manager files from the first machine in the second also. I have tried copying the files and replacing them on the new RPI but then the Persistence Manager gets errors and then they are all reset to 0 etc I have tried adjusting the times to match. Instead of replacing whole file, I tried just changing the contents inside the file, but still does not work. I may be missing something here, but surely there must be a way to load a backup. Thanks!
Last updated: 2023-10-18

Post by benediktenger on Access project information of references library by script ^{CODESYS Forge talk (Post)}
Hi, I need to accesss to project information of a referenced library in the project. At least I need the "Released" flag from these information to check that no project gets shipped with unreleased libraries. Unfortunally, I do not get this information by script. I archieved to get the library reference object and the managed library object but both won't give me this flag but only basic information like company, name, version, etc. Here is my code: # get the library refernces proj = projects.primary objects = proj.get_children(recursive=True) for object in objects: if object.is_libman: for libref in iter(object): if libref.is_placeholder and isinstance(libref.effective_resolution,str): print(libref) print(libref.name) # get the managed libraries libs = librarymanager.get_all_libraries() for lib in libs: print(lib.displayname) print(lib.version) Does somebody know how do get these information?
Last updated: 2023-10-23

Post by simotion on Pointer to Softmotion axis ^{CODESYS Forge talk (Post)}
In my projet I use Softmotion. There are several Softmotion axis (type SM_Drive_Virtual). In the Global Variable List I have an array of pointers to the adress of the Axis. Axis : ARRAY[1..iMaxCntAxes] OF POINTER TO AXIS_REF_SM3 := ADR(AX01_Axis), // [1 ADR(AX02_Axis), // [2] This works, but when compiling I get the message : . [WARNING] E3D: GVL Device: PLC Logic: Application: C0564: A reference to uninitialized variable AX01_Axis is used for initialization of Axis. Accessing the uninitialized variable may result in unexpected behavior. [WARNING] E3D: GVL Device: PLC Logic: Application: C0564: A reference to uninitialized variable AX02_Axis is used for initialization of Axis. Accessing the uninitialized variable may result in unexpected behavior. Is there a way when initializing to force the hardware (softmotion axis) to first initialize and then the Global Variable List?
Last updated: 2023-10-25

Post by bernd on Version 3.5.19.40 probleme bei FUN mit _to_ im Name ^{CODESYS Forge talk (Post)}
Hallo, ich habe von Version 3.5.19.0 auf 3.5.19.40 hochgerüstet, jetzt bekomme ich bei allen FUNs die "to" im Namen haben einen Fehlermeldung. Oft in der OSCAT.LIB verwendet. Gleiche Meldung wenn ich einen FUN neu anlege. Sobald ich das to weglasse z.b. nur "t" ist der Fehler weg. Programm ist unverändert. Fehlermeldung: Typ.... wird nicht unterstützt. Hat jemand das gleiche Problem? Bzw. wo liegt das Problem? Bzw. nutzt jemand das SP4 und bei ihm kommt der Fehler nicht? Läuft bei mir auf Rasperry. Beispiel: DT_TO_SDT -> Fehlermeldung Typ DT_TO_SDT wird nicht unterstützt. DT_T_SDT -> Übersetzung Fehlerfrei
Last updated: 2023-11-09

Post by bahrt on Mapping GPIO and adding GPIO configurations ^{CODESYS Forge talk (Post)}
Hi I am about to configure ALL GPIOs on a Raspberry Pi 3 model B V1.2. It seems that I have limited understanding of why not all the IO’s become available with the corresponding mapped IO’s – the 'dwInUse' contains mapping of the inputs from GPIO0 to GPIO27. But 'dwInputs' are only showing GPIO0 to GPIO8 and GPIO15 as available inputs. These mentioned inputs work fine You will be able to watch the limited mapped IO activity here: https://drive.google.com/file/d/1zz4PBNBfRZBF2YkiJ2VQxw6L9MBVji7R/view?usp=sharing Can you please bring your best idea on why GPIO09 to GPIO14 and GPIO16 to GPIO27 state are zero in the dwInputs register? Thanks in advance. Best regards Andreas Bahrt
Last updated: 2023-11-16

Post by ewi04 on Recipe Manager - RecipeManCommands, load & write wrong values, Bug? ^{CODESYS Forge talk (Post)}
I'm getting closer to the error. The combination of STRUCT and ARRAY probably causes the strange behavior. As soon as the ARRAY is followed by a variable, the problem arises. Examples: X: ARRAY[1..2] OF INT; // okay Y: ARRAY[1..2] OF STRUCT; // problem Z: STRUCT; Z.ArrayOfInt[1]; // okay Z.ArrayOfStruct[1].Variable // problem // *update - array of array also causes the error A: ARRAY[1..2] OF ARRAY[1..4] OF BOOL; // problem I can't be the only one who has this problem. Or? Maybe there is a solution. Anyone? Or is it a bug after all?
Last updated: 2023-11-17

Post by bahrt on Raspberry Pi GPIO mapping ^{CODESYS Forge talk (Post)}
Hi I am about to configure ALL GPIOs on a Raspberry Pi 3 model B V1.2. It seems that I have limited understanding of why not all the IO’s become available with the corresponding mapped IO’s – the 'dwInUse' contains mapping of the inputs from GPIO0 to GPIO27. But 'dwInputs' are only showing GPIO0 to GPIO8 and GPIO15 as available inputs. These mentioned inputs work fine You will be able to watch the limited mapped IO activity here: https://drive.google.com/file/d/1zz4PBNBfRZBF2YkiJ2VQxw6L9MBVji7R/view?usp=sharing Can you please bring your best idea on why GPIO09 to GPIO14 and GPIO16 to GPIO27 state are not active (at a fixed zero state) in the dwInputs register? Thanks in advance. Best regards Andreas Bahrt
Last updated: 2023-11-21

Post by felipemsgarcia on Edge Gateway online, but PLC is not online ^{CODESYS Forge talk (Post)}
Hello, Up until yesterday (2023/11/20) I was able to connect to the PLC remotely with no issues through Edge Gateway. However, today I can see the Edge Gateway online but I can find the PLC even if I do a network scan. Following the same path as janber0206 @ https://forge.codesys.com/forge/talk/automation_server/thread/e78b11d7e9/, I looked into certificates and found that one of the Trusted publisher certificates has expired earlier today (2023/11/21). Please see image. Questions are: Does it affect the connection to the PLC through Edge Gateway? How do I fix it? Thank you very much in advance!
Last updated: 2023-11-21

Post by captaincookie on SP19 Runtime Zugriff auf lokales Dateisystem ^{CODESYS Forge talk (Post)}
Hallo, das konkrete Problem konnte ich seither leider nicht lösen, allerdings habe ich mit SP19 Patch 4 einen Workaround gefunden, der zumindest für meine Anwendung funktioniert: Zunächst besteht das Problem nicht für alle Funktionalitäten gleichermaßen. Die Anzeige einer .bmp oder .tiff-Datei im Visu-Elementtyp Bild beispielsweise funktioniert aus dem lokalen Dateisystem. Die Anzeige von .cnc-Dateien in einer Tabelle dagegen nicht. Allerdings funktioniert das wiederum, wenn der Dateipfad in den Ordner der PLC führt, genannt '_cnc'. In der PLC wird dieser Ordner über den Pfad _cnc/ erreicht. Auch lohnt es sich einen Blick in die PLC Config-Datei zu werfen und ggf. Anpassungen vorzunehmen. Einzelne Einträge definieren Zugriffsrechte und ähnliches. Die beschriebene Lösung ist nicht sonderlich schön und es erklärt das Problem nicht. Vielleicht kann das noch jemand anders beantworten.
Last updated: 2023-11-22

Post by orbis-tertius on 3.5 P20 hangs - no response ^{CODESYS Forge talk (Post)}
Is there anybody can help me with the same problem? I have to migrate created project in ver.3.5.17.0 32BIT with runtime 4.7.0 to ver.3.5.20.0 64BIT with runtime 4.7.11.0 due to the new license. All the libraries and dependencies are updaded and there are no any errors. The first compilation and loading to the controller is being compleated well. After this being done and if I am forced to change any values or change any code in the current project, another compilation will be freezed. This problem can be solved if pressed Build/cleen or cleen all. There is another method to solve this; if project is loaded with excluded object (Alarm Configuration), repeated compilations will be done well.
Last updated: 2024-06-27

Post by hwopereis on How to use Application Trigger in Modbus TCP Server ^{CODESYS Forge talk (Post)}
Hi, I'm currently reading and writing several registers with a Modbus TCP server as a subdevice of a Modbus TCP Client. If I read and write cyclically, there are no issues. The problem is this: I see there is also the option to write by "Application", which I would like to use. However, I don't know With the Modbus RTU client, I see there is a trigger variable, xTrigger, to trigger the writing tasks. However, I can't find any such trigger for the Modbus TCP Client. Does anybody know how to implement the trigger behavior for the Modbus TCP client? In particular, I struggle to find the exact "trigger" variable. I'm running Modbus Client version 4.4.0.0 Thanks in advance!
Last updated: 2024-07-01

Post by mikek10 on HTTP Client TCP Init Error ^{CODESYS Forge talk (Post)}
Hi, I tried posting in the discussion for the HTTP Client example but perhaps that is not monitored? I have been using http client successfully for approximately 12 months, however in recent weeks I have begun to get TCP_INIT_ERROR when attempting to communicate. The only resolution seems to be power cycling and then the connection begins to work again for a time. I am not aware of any changes to code or libraries between working ok and now. Cold reset does not resolve, only power cycle. It seems the error from NBS.TCP_Client is 6002 which I think is invalid_addr However the address is the same when working and not working?
Last updated: 2024-07-04

Post by adaml on ScriptEngine - execute "Export -> Bootproject and firmware (SD Card)" from python ^{CODESYS Forge talk (Post)}
Hi everybody, recently I was checking how many time consuming manual tasks I can perform with python scripts through ScriptEngine. I have in the project many devices and currently I export bootprojects for them by right clicking on Application, then go Export -> Bootproject and firmware (SD Card) and then point folder where files should be exported. It's fine if project has few devices, but I have some of them with over 30 devices and manual clicking like that is very time consuming. How can I do it with python script? So far I found "create_boot_application" method from Application object (taken by projects.primary.find('Application', True)) but it only generates .app and .crc file instead of full structure to store on SD card (FIRMWARE, USERDATA, SDCARD.INI, Version, Version.txt.sig). Should I use some method to control user interface? Best regards, Adam
Last updated: 2024-07-05

Post by naks on Change IP address from IEC code ^{CODESYS Forge talk (Post)}
Hi, I'm working with wago 750-8212 controller. I want to change the IP address of the Ethernet adapter from webvisu. I have read some other threads in this forum and I have tried the following: Added the following text to the config file CODESYSControl.cfg: [SysSocket] Adapter.0.Name="br1" Adapter.0.EnableSetIpAndMask=1 Used the below code : IF xUpdateEthernet THEN Ethernet.Enable := FALSE; Reconfig ( xExecute := TRUE, itfNode := Ethernet, eError => ErrorReconfig, xBusy => xBusy, xDone => xDone, xError => xErrorReconfigure); IF Reconfig.xDone THEN ErrorCode := Ethernet.UpdateConfiguredIPSettings(IpAddress := NewIp, gateway := newGateway, subnetmask := newSubnetmask); END_IF IF Reconfig.xDone OR Reconfig.xError THEN reconfig(xExecute := FALSE); END_IF Ethernet.Enable := TRUE; Reconfig(xExecute := TRUE, itfNode := Ethernet ); END_IF I could change the IP address but when i restart my Controller then the IP address resets to the original, Is there anything I have missed, what else can i try ? Thank you in advance.
Last updated: 2024-07-08

Post by rcanfield on Code Coverage and Test Manager ^{CODESYS Forge talk (Post)}
I am exploring the possibility of measuring test coverage using the Test Manager and Profiler tools. It appears there are some significant limitations. Has anyone been successful trying to use these tools? Here are my observed limitations. Am I misunderstanding anything? * Profiler only works for one "run cycle". Even warm resets will clear the coverage calculations. Thus, all tests would have to manually reset values when complete to prevent contamination to other tests. * It is not possible to see which statements were/weren't executed. One would have to use the flow control tool which cannot be used in conjunction with the Profiler, and must be viewed in realtime. * The profiler control cannot be automated and therefore must be executed through a manual procedure. Thank you!
Last updated: 2024-07-11

Post by rcanfield on Code Coverage and Test Manager ^{CODESYS Forge talk (Post)}
I am exploring the possibility of measuring test coverage using the Test Manager and Profiler tools. It appears there are some significant limitations. Has anyone been successful trying to use these tools? Here are my observed limitations. Am I misunderstanding anything? * Profiler only works for one "run cycle". Even warm resets will clear the coverage calculations. Thus, all tests would have to manually reset values when complete to prevent contamination to other tests. * It is not possible to see which statements were/weren't executed. One would have to use the flow control tool which cannot be used in conjunction with the Profiler, and must be viewed in realtime. * The profiler control cannot be automated and therefore must be executed through a manual procedure. Thank you!
Last updated: 2024-07-11

Post by alez on CANopen safety , SRDO error behavior ^{CODESYS Forge talk (Post)}
I have a project in which there is a Safety device in CANopen, and the PLC interfaces with it as Master. I would like to understand if it is possible to set the behavior in case one of the SRDOs that I receive from the slave is no longer received. From what I have seen, if the SRDOs are not received, the CANopen safety status becomes "Not in progress" and the mapped variables freeze at the last value detected. I want to understand if it was possible to set the variables connected to the SRDOs to take a default value (e.g. zero) in the event of a CANOpen error during reception.
Last updated: 2024-07-15

Post by pazderai on CanOpen write issue ^{CODESYS Forge talk (Post)}
Thank you for an answer. Unfortunatelly, it doesn't work. The slave needs command byte as follows 0x22. It means expedited transfer where s-bit is not set, therefore data length is not specified. If I use segmented mode by CiA405.SDO_WRITE_DATA it produce command byte as follows 0x21 which is unfortunately supported by the slave and it returns with error 0x05040001 - SDO Abort Code -Client/server command specifier not valid or unknown. I can use CAN API low level to overcome this problem where it is possible to set 0x22 as a command byte, but than I can use user interface to set the communication probably, because I thing that machine expert uses library CiA405.SDO_WRITE for setting the slave. Or am I wrong?
Last updated: 2024-07-16

Post by ucconversions on displaying all incoming CAN bus messages ^{CODESYS Forge talk (Post)}
Hello Im using CAN API Im receiving CAN messages with no problem. below is my code. METHOD ProcessMessage VAR_IN_OUT ( Incoming message ) Message : CAN.RxMESSAGE; END_VAR VAR i: USINT; END_VAR recvCANid := Message.udiCanID; IF recvCANid = 16#18FF8247 THEN UserVarGlobal.g_countMsg_RPMset := UserVarGlobal.g_countMsg_RPMset + 1; FOR i := 0 TO 7 DO myDataFF82[i] := Message.abyData[i]; END_FOR END_IF but in this way I can only display CAN messages which I know its ID. I want to display all incoming CAN messages for example first incoming CAN message - I assign it into a variable Message1. second incoming CAN message - I assign it into a variable Message2. third incoming CAN message - I assign it into a variable Message3. Can you give me guidance? Thank you BR
Last updated: 2024-07-18

Sort by relevance or date