Post by reinier-geers on CODESYS crashes when adding standard OPC UA Information Model to Communication Manager ^{CODESYS Forge talk (Post)}
I only experiance a lot of problems with Version 20 and 21. Version 20 crashes at a online change Version 21 - 5 identical programms in one project. The frist runs. The thirt Crashes. NOt even a screen
Last updated: 2026-03-19

Post by reinier-geers on Resize and position Numpad/Keypad ^{CODESYS Forge talk (Post)}
HI, I download the scale dialog. But cant find were i can Set the size of the dialogs. only the position. I want the diolog bigger
Last updated: 2026-04-15

Post by skiper on Best practice to reduce CPU load ^{CODESYS Forge talk (Post)}
Hi ! I can't say exactly the best practice you can use to reduce your cpu load but you can use the codesys exemple. https://www.codesys.com/products/visualization/#c3876 Go to the midle of the page and you have a link to download an exemple.
Last updated: 2026-04-21

Post by brouwyka on JSONByteArrayWriter string result incorrect order ^{CODESYS Forge talk (Post)}
Hello everybody! We are currently working on implementing a call to an API that tracks data sent to it through a JSON body. Versions: CODESYS V3.5 SP19 Patch 7 using IIOT Libraries SL license, specifically JSON Utilities SL version 1.13.0.0. For simplicity, let's say we want to construct the following JSON body for our request (simplified to just 1 key-value pair in each of the array objects to keep things shorter, but we'd have multiple fields in each in reality): { "id": "machine_1", "readings": [ { "temp_mid": 153.8 }, { "temp_mid": 98.3 } ], "status": "RUNNING" } However, our "readings" data is added over time, not in creation, so calls to the builder would be somewhat out of order. In a simplified manner, our calls would look like this: jsonBuilder(pJsonData := JsonData, diRootObj => rootJsonIndex); // rootJsonIndex = 0. // Set "id" at root JSON object. wsValue := "machine_1"; jsonBuilder.SetKeyWithValue("id", wsValue, diParentIndex := rootJsonIndex); // JSONData KEY: diParentIndex = 0 & diIndex = 1, VALUE: diParentIndex = 1 & diIndex = 2. // ... Time passes, we get our first reading to add ... // The "readings" array does not exist yet, so we make it first at root JSON object before adding our first reading. readingsJsonArrayIndex := jsonBuilder.SetKeyWithArray("readings", diParentIndex := rootJsonIndex); // JSONData KEY: diParentIndex = 0 & diIndex = 3, VALUE: diParentIndex = 3 & diIndex = 4. // Then add the first reading, for which we first have to make an object. readingsJsonArrayObjectIndex := jsonBuilder.SetObject(diParentIndex := readingsJsonArrayIndex); // JSONData OBJECT: diParentIndex = 4 & diIndex = 5. // And then we add the field(s) in the object for the first reading. // NOTE the use of `readingsJsonArrayObjectIndex` so we set the field(s) in the first object in the array. rValue := 153.8; jsonBuilder.SetKeyWithValue("temp_mid", wsValue, diParentIndex := readingsJsonArrayObjectIndex); // JSONData KEY: diParentIndex = 5 & diIndex = 6, VALUE: diParentIndex = 6 & diIndex = 7. // ... Time passes, we want to add another root-JSON-level field ... // Set "status" at root JSON object. wsValue := "RUNNING"; jsonBuilder.SetKeyWithValue("status", wsValue, diParentIndex := rootJsonIndex); // JSONData KEY: diParentIndex = 0 & diIndex = 8, VALUE: diParentIndex = 8 & diIndex = 9. // ... Time passes, we want to add a second reading ... // "readings" array exists, so add the second reading, but first we have to make another object. // NOTE the re-use of the earlier stored `readingsJsonArrayIndex`, so we add the new object to that previously created array. readingsJsonArrayObjectIndex := jsonBuilder.SetObject(diParentIndex := readingsJsonArrayIndex); // JSONData OBJECT: diParentIndex = 4 & diIndex = 10. // And then we add the field(s) in the object for the first reading. // NOTE the use of `readingsJsonArrayObjectIndex` so we set the field(s) in the SECOND(, new) object in the array. rValue := 98.3; jsonBuilder.SetKeyWithValue("temp_mid", wsValue, diParentIndex := rootJsonIndex); // JSONData KEY: diParentIndex = 10 & diIndex = 11, VALUE: diParentIndex = 11 & diIndex = 12. // ... Finally, we want to send to the API, so we must convert the builder to data the client accepts as a body ... xTestWriteToJsonArray := TRUE; jsonArrayWriter( xExecute := xTestWriteToJsonArray, pwData := ADR(jsonDataString), udiSize := SIZEOF(jsonDataString), jsonData := JsonData^ ); IF jsonArrayWriter.xDone THEN // The writer was successful, the `jsonDataString` VAR should now contain a WString copy of the JSON object from the builder. xTestWriteToJsonArray := FALSE; END_IF We see that the JsonData STRUCT is correctly organised (in terms of the diParentIndex & diIndex set by it for each JsonElement) as we expect, as outlined in my comments in the simplified code above. However, as soon as we pass it to the jsonArrayWriter (which is a JSON.JSONByteArrayWriter), the resulting jsonDataString does not match our expectations, instead coming out like this: { "id": "machine_1", "readings": [ { "temp_mid": 153.8 } ] } { "temp_mid": 98.3 } "status": "RUNNING" The behaviour of the JSONByteArrayWriter thus seems to be the problem here. It does not seem to correctly process the children of JsonElements that are ARRAYs when additions to them are done broken up by additions to the lower level JSON OBJECT they are a part of. We have confirmed this by changing the order of the calls in our example to add both readings before adding the "status", in which case we get our expected outcome. However, in reality this is not possible for us - additions to any of the JSON's objects, arrays or array objects may happen at any time after other fields elsewhere in the JSON have been added. Does anyone know a way around this, a fix, or knows a solution we simply have not found? This same phenomenon was also noted on this forum by user @ryusoup at the end of 2023 (https://forge.codesys.com/forge/talk/Engineering/thread/c45929e2f1/#e27f) and user @mtho in early 2024 (https://forge.codesys.com/forge/talk/Engineering/thread/cd1bb450db/#1292) but both topics received no activity beyond both users' opening posts. Looking through the release notes of all the versions of the JSON Utilities SL library, I also did not see any remarks on resolutions of bugs in this vain. Thanks in advance for the assistance!
Last updated: 2026-06-10

Post by arj3090 on Linux ARM64 Deploy Fail - Index out of range ^{CODESYS Forge talk (Post)}
I used Tools->Deploy Control SL in the IDE that I downloaded yesterday. So I assume it is the latest version. I did a manual install using SSH and it works.
Last updated: 2026-07-01

Post by somlioy on Beckhoff ESI import german text ^{CODESYS Forge talk (Post)}
Solved. Had to explicitly select English inside CODESYS as language instead of system default, which for my case is norwegian... Open Tools → Options → International Settings Set the development-system language to English Restart codesys
Last updated: 2026-07-28

Post by reinier-geers on Read own network device name ^{CODESYS Forge talk (Post)}
Hi. When i scan for devices on the network i get names of the devices. Is there a command to get device name in my programm ? i found : DED.GetDeviceNameString( DED.GetRoot() ); but then i get his name in the programm. not his name in the network.
Last updated: 2026-09-17

Post by patrik on Wish: CODESYSControl.cfg - again ^{CODESYS Forge talk (Post)}
Really good post. I hope that it gets seen by CoDeSys staff. I often find the documentation lacking when it comes to using CoDeSys in more complex ways. If using standard "basic PLC functionality", then it's fine and functions are documented in a sufficient way. Once you go outside of that bubble then I find the documentation not enough. CODESYSControl.cfg is a good example of this. And why isn't more of this integrated in the IDE. like changing port of the webserver. Why is it hidden in this file and not accessible through the IDE. It feels like a layer of complexity purposely left in so you don't change it if you don't know what you are doing. Like the argument is if you have the know-how to find the file you have the know-how to edit it. I find the documentation lacking when it comes to more complex libraries too. Like the element collections library. there is parameters I still don't quite know what they do. I can't find any documentation about them. There is an example project to download from CoDeSys so you can see how you are supposed to set your factories up. I leave some parameters like it is in the example. Should I? I don't know. Does it work. Yes. But I could perhaps create memory leaks if I get something wrong and I don't see that until way later. In the create method of your element factory you have a function called __vfinit and you just have to accept that it works. Why can't I get insight in what it does through documentation? Don't CoDeSys want us to know how these work so we can't use them to accidentally break something? I find the error logging lacking too. I've had the CoDeSys service crash but it doesn't say why. I've seen windows noticing the crash but no error code was sent or error message. Also in my current issue where the application fails to start after power outage. I can see in the log that it fails to load retain in the application as one log entry and application failing to start as another. But why? Why does loading the retain fail? give me more information! Is it just a skill issue with me? I don't know. If it is then I'd still want CoDeSys to step up their game when it comes to documentation. I'm sorry if a lot of frustration comes though in this post. I really like the product and what I can do with it compared to other PLC brands.
Last updated: 2024-11-21

Post by gmau on Station 'xyz': AR aborted by device (alarm.ind): AR consumer DHT expired ^{CODESYS Forge talk (Post)}
Hello, I am writing to You with the problem of Profinet communication in my Codesys V2.5 SP21 project working on Codesys Control RTE V3 x64 device. I have implemented multiple Profinet devices in the network, the Profinet itself is configured properly, based on the hardware and software parameters of the industrial PC (Windows 10 IoT). My Profinet network contains such Profinet devices as a robot, IO-Link master with some devices on it, IO modules, safety, vision sensor (the configuration is attached to this topic). Once every few hours, sometimes more often, sometimes not even once a day, the error show up at one, two or a few random modules at the same time, every one is the same: "Station 'xyz': AR aborted by device (alarm.ind): AR consumer DHT expired". The communication is then lost with that module, but it is up instantly. It is not such a problem when it comes to regular IOs, but with safety or rotation table, it causes errors and machine stops. I have tried multiple solution based on any forum suggestions, manuals, etc., but it is still happening. Mabye when I turned on the Watchdog for this communication, it is happening less times, but still, I could not turning up the Watchdog parameters at the expense of cycle time of the process. The same is with Send Clock, Reduction and DHT. I have one of the ethernet port on the PC configured with CmpEt1000Drv driver. The card itself is Intel Ethernet Connection I219-LM. These devices are connected to one EKI-2728-NI-A Profinet switch (it was changed from the standard non Profinet one, but still nothing has changed). As I said ealier, I am attaching all the configurations for each device and the communication itself - if any of You had the same problems or have any ideas what could be the cause, I would be much gratefull for any help. For now I am out of ideas, if You have any additional question, please ask. Sincerely, PF
Last updated: 2025-09-04

Post by smeitink on Timeout Error in Modbus Communication with WAGO PFC200 and iEM2050 Meter using 750-652 Module ^{CODESYS Forge talk (Post)}
Hi all, I'm looking for help with an issue I've come across while trying to facilitate Modbus communication between a WAGO PFC200 PLC using a 750-652 communication module and an iEM2050 Series Single Phase Energy Meter. I believe to have everything wired and setup correcty, but I keep running into a "Error time out" message, and by now I don't really know what else to try. My setup is as follows: A PFC200 Wago PLC, which has 2 750-652 Serial Interfaces extension modules connected to its field bus. I'm using one of these to talk to a Schneider iEM2050 - kWh-meter over modbus. I have connected terminal 23 (A) of the iEM2050 to connector 6 (A) on the 750-652. I have connected terminal 24 (B) of the iEM2050 to connector 2 (B) of the 750-652. I'm using 200mm of twisted together wires to connected them both, and I have placed a 120 ohm resistor between A and B at both ends. I've attached relevant pinout images to this post. I then wrote a simple program that configures the Mobus port, as per the datasheet of the iEM2050. You can find an image of the relavent page attached to this post too. This is my program: PROGRAM PLC_PRG VAR Master: FbMbMasterSerial; xIsConnected: BOOL; xError: BOOL; iIndex: INT := 1; xTrigger: BOOL; utQuery : typMbQuery := ( bUnitId := 1, // The Modbus unit or slave address bFunctionCode := 4, // Function code for reading input registers uiReadAddress := 1829, // adress for the Power on off counter uiReadQuantity := 1 // Quantity of registers to read ); iStep: INT; oStatusModbus: WagoSysErrorBase.FbResult; utResponseModbus: typMbResponse; xConnect: BOOL := FALSE; delayTimer: TON; END_VAR Master( xConnect:= xConnect, I_Port:= _750_652_24_1, udiBaudrate:= 9600, usiDataBits:= 8, eParity:= WagoTypesCom.eTTYParity.Even, eStopBits:= WagoTypesCom.eTTYStopBits.One, eHandshake:= WagoTypesCom.eTTYHandshake.None, ePhysical:= WagoTypesCom.eTTYPhysicalLayer.RS485_HalfDuplex, xIsConnected=> xIsConnected, xError=> xError, oStatus=> oStatusModbus, eFrameType:= WagoAppPlcModbus.eMbFrameType.RTU, tTimeOut:= T#5S, utQuery:= utQuery, xTrigger:= xTrigger, utResponse:= utResponseModbus); delayTimer(IN := TRUE, PT := T#3S); // Use the Q output of the timer to set xConnect after the delay IF delayTimer.Q THEN xConnect := TRUE; END_IF CASE iStep OF 0: //Wacht totdat de master de poort geopend heeft IF xIsConnected THEN iStep := 1; END_IF 1: //Stuur request naar de slave xTrigger := TRUE; iStep := 2; 2: //Wacht totdat de master klaar is met het afhandelen van de request IF NOT xTrigger THEN iStep := 3; END_IF END_CASE The TON delay before opening the port is due to a an error I encountered when opening it straight away. This seems to be a bug, as described here. However, the TON solved that particular issue. I tried reading multiple registers, but like I said, I still always end up with the "Error time out". What else can I test or try at this point?
Last updated: 2024-02-24

Post by thejesh on WebVisu: switch-frame destroys and rebuilds referenced visualization - is retention configurable? ^{CODESYS Forge talk (Post)}
Hi, HTML5 Conrtrol elements on CODESYS V3.5 SP22 Patch 2, WebVisu only (no TargetVisu). The client is Chromium 147 in kiosk mode on a Raspberry Pi CM5 panel; the runtime is CODESYS Control Win V3 x64/CODESYS Control RTE V3 x64 on a separate IPC. I have measured that changing a Frame element's switch-frame variable destroys the currently referenced visualization and fully rebuilds it on return, including all embedded HTML5 controls. We would like to know whether this lifecycle is configurable. Measurement 1: Curently, our production page contains a good number of custom HTML5 controls and will only increase , each of which the WebVisu client instantiates as a separate about:srcdoc iframe. I counted live targets via the Chromium remote debugging endpoint (curl http://localhost:9222/json/list) before and after switching. Production displayed (index 0) 108 iframes Switched to a light page (index 1) 46 iframes Switched back to Production 108 iframes 62 controls are destroyed on leaving the page and re-instantiated on return. Observed latency was approximately 2.5 -5s away from the heavy page and approximately 4.0 -5 s back to it. This was measured with the panel otherwise idle: 2.8% system CPU, 97.1% idle, 2 h 56 min uptime, no remote-desktop session running, hardware acceleration confirmed active in chrome://gpu. I had previously measured the same behaviour under a much higher CPU load and have since resolved that load separately (VISU_TASK cycle period), so I am confident the rebuild is the frame lifecycle itself and not CPU starvation. Measurement 2 — frame offset instead of switch index I then displayed the same overlay page by a different mechanism. Instead of using a switch-frame index, I placed the page in a fixed Frame parked off-screen below the visible canvas, and drive its Y offset from an INT in Structured Text to move it into view. With this approach the overlay appears immediately, the production page underneath is never torn down and keeps updating, and no visualization is destroyed or rebuilt in either direction. The iframe count does not change at any point. This is the same visualization, containing the same HTML5 controls, as in Measurement 1 — the only difference is how it is brought on screen. So the rebuild appears to be a property of the switch-frame mechanism specifically, not of the page or its control count. Questions: Is the destroy-and-rebuild behavior on switch-frame the intended and documented lifecycle for WebVisu? Is there any setting, property or client option that causes a referenced visualization to remain instantiated when it is not the active index — the equivalent of keeping it mounted but not displayed? Does the runtime or client offer any form of caching or pre-instantiation for referenced visualizations, so that a return switch does not pay the full construction cost? Is the cost specific to HTML5 controls, or would the same rebuild occur with an equivalent number of native visualization elements? We ask because it would inform whether reducing our HTML5 control count is a viable solution. Is there a recommended pattern for showing a full-screen page over a live production screen without tearing the production screen down? My current workaround is to keep every page permanently instantiated and move it on and off screen by driving a frame offset variable, which avoids the rebuild entirely but means every page stays live and consumes client resources for the life of the session. This works, but I would prefer to use the supported mechanism if one exists. Any help on this is greatly appreciated. Thank you!
Last updated: 2026-08-07

Post by mubeta on Some 'pathetic' errors in SoftMotion program ^{CODESYS Forge talk (Post)}
Yes, this is the error the sometimes show up. What make me crazy is the fact that it happens randomly and not each times. I know very well where the problem is, in wich one program row it's located. For each actions of the state machine I have all events recorded with log on text file. it is not problematic for me to find the application point of the fault, but I need to understand why occasionally and for no apparent reason, switching the state machine and thus changing the motion FB, sends the axis into failure (but only occasionally). For example, one case that is often problematic is the execution of the Axis Halt instruction. When, after a MoveAbosulte instruction this returns the event as 'done' and indeed the axis is in standstill, the state machine first sets the move instruction to FALSE, and the next cycle sets the Halt request to TRUE. Some of the time everything works out fine. Occasionally, however, in this exchange, the axis goes into fault, also losing the OPERATIONAL state. Meanwhile, I would like to understand why the motion FB instances must still be called even after the Execute is set to FALSE, especially in view of the fact that the next instruction is programmed to abort the previous one, with BufferMode set to 'Aborting'. All these unnecessary FB calls are an unnecessary overhead on the CPU anyway. Is there any precise rule about when to cease calling the various instances? (It should precisely be the 'done' status that says this one has finished its work).
Last updated: 2024-07-18

Post by otbeka on CAA File Handling: "Read only" error ^{CODESYS Forge talk (Post)}
Hi, On CODESYS V3.5.19.50 I am using the latest 3S File Access library (CAA File & CAA Types) to save and retrieve configuration data from the User Data section of an SD card, via file called config.txt. The application reads the file to memory once upon initialisation, and then the only other operation is overwriting the file with the updated configuration whenever the user changes the settings. An issue is ocurring, seemingly randomly, where the config.txt file, and all surrounding directories, cannot be written to by the unit, returning a READ_ONLY_CAA error upon file open. This sometimes does not clear even after power-cycling the PLC several times, yet checking the file permissions with ls -l shows normal read, write, and execute access. I am wondering if this is caused by the PLC itself or by an issue in my application. I have included my CAA File handler function block (SD_Card_Bin(FB).txt) and any associated types/calls below: TYPE ConfigData : STRUCT Config : ARRAY[0..256] OF STRING(64); NumberOfStrings : UDINT; END_STRUCT END_TYPE TYPE Buffer : STRUCT Data : ARRAY[0..SIZEOF(ConfigData)] OF BYTE; DataLength : UDINT; END_STRUCT END_TYPE A case within my MAIN PRG: mode.WriteConfigDataToFile: VisuElems.CURRENTVISU := 'Screen_Write'; IF xModeHasChanged THEN iErrorCount := 0; iRoute := 0; END_IF CASE iRoute OF 0: inBuffer := ConfigData_To_Buffer(GVL.UnitData); xConfigError := FALSE; xDataConfigured := FALSE; iRoute := 1; 1 : //Read Data from Card SD_Card_Bin.strFileName :=FileName; SD_Card_Bin.iFileOperation := 2; SD_Card_Bin(xExecute :=TRUE, WriteBuffer := inBuffer, xDone =>xDataConfigured, xError=>xConfigError, strErrorCodes=>strError); IF xDataConfigured THEN SD_Card_Bin(xExecute :=FALSE); IF xConfigError THEN iRoute := 32767; ELSE iRoute :=100; END_IF END_IF 100: // The end bCurMode := mode.DisplayProcessVariables; 32767: // Config error iErrorCount := iErrorCount + 1; IF iErrorCount > 2 THEN // Repeats 3 times, if no progress abandons and returns error bCurMode := mode.WriteError; ELSE iRoute := 0; // Try again END_IF END_CASE I'd like to rule out my application as the cause of this bug - help via comments or criticism would be greatly appreciated.
Last updated: 2025-03-19

Post by manuknecht on Maximum dynamic limits of MC_MoveLinearAbsolute and MC_MoveLinearRelative ^{CODESYS Forge talk (Post)}
Hello all I use an axis group with two linear axes in a kinematic system. Using MC_MoveLinearAbsolute and MC_MoveLinearRelative, I am trying to realise dynamic movements (set velocity: 1 m/s, set acceleration: 30 m/s^2) with the highest possible jerk, as I want to compare this with a movement with a trapezoidal velocity profile. However, my measurements show that the axes accelerate with a maximum jerk of approx. 800 m/s^3, even if the value in the function block was set to 10,000 m/s^3. I have already increased the dynamic limits of the two axes accordingly (Velocity: 5000 u/s, Acc-/Deceleration: 250'000 u/s^2, Jerk: 100'000'000 u/s^3, where 1 u = 1 mm), but this had no effect. I have checked the input parameters of the function blocks and the VelFactor, AccFactor and JerkFactor factors are all set to the maximum (value 1). Reducing the cycle time has resulted in a higher jerk, but as we are already working with a cycle time of 1 ms, there is nothing more that can be done here. The axes themselves are still a long way from drawing the maximum current, which is why I assume that this should not be a problem either. Are there other factors or parameters that influence the dynamic limits and therefore allow faster movements? The PLC I am using is a Raspberry Pi 4 and Codesys V3.5 SP20 with all libraries up to date. Thanks in advance Manuel
Last updated: 2024-05-08

Post by mondinmr on Patched Wine packages for Ubuntu and CODESYS 3.5 ^{CODESYS Forge talk (Post)}
I have now published Ubuntu amd64 packages of my patched Wine build for CODESYS, targeting Ubuntu 22.04, 24.04, 26.04 and 26.10. The packages are based on Wine 11.20 and include all five compatibility patches covering NCrypt, thread suspension, security tokens, service tokens and local WMI marshalling. Both 32-bit and 64-bit Windows applications are supported through WoW64. In my working environment, CODESYS 3.5 SP21 successfully compiled and went online with a very complex project containing hundreds of thousands of lines of code and extensive use of polymorphism. PPA and installation All four binary builds have completed successfully and are now published and available through APT. Wine ADV for CODESYS PPA sudo add-apt-repository ppa:mondinmr/wine-adv sudo apt update sudo apt install wine-adv The runtime installs under /opt/wine-adv and provides dedicated commands, allowing it to coexist with your existing Wine installation. For example, to launch CODESYS from an existing dedicated prefix: WINEPREFIX="$HOME/.wine.codesys" wine-adv \ 'C:\Program Files\CODESYS 3.5.21.30\CODESYS\Common\CODESYS.exe' Adjust the executable path to match your installed version. Use the same WINEPREFIX when running the CODESYS installer or installing its prerequisites. The package contains the patched Wine runtime. CODESYS and its required Microsoft .NET components must be installed separately inside the prefix. Source code, patches and packaging are available on the project page. This is an experimental community build. Feedback and test results from other users are welcome! CODESYS Installer and CodeMeter licensing are also working in my setup. For CodeMeter, I use the native Linux CodeMeter runtime, while the Windows applications run under the patched Wine build.
Last updated: 2026-10-31

Post by hwillems on Ranges, Lambdas, on Fixed arrays of structs ^{CODESYS Forge talk (Post)}
I do datastructures and algorithms in Codesys. For example a Struct of Person with thing's like IdNumber, Name, Age etc. as example. Now i do all kind of calculations, filters. So i have this pretty big Fixed Array with Structs. On this struct i want to do simple stuff you can do easily in C++/Python/Rust etc. For example i want to do this: AvererageAge := Average(Peoples.Age); Then it will return the average of all members ages. Or Sort struct on age etc. Or sort on alphabetical Name. Or use Lambda functions to filter/mutate out things like, filter out everybody above 18 years old. Or remove people who it's name start with "A". Currently i have to write my own custom function for example sorting on Age. And make a super specific function based on that particulare datastructure. Here an Example: (*Before calling this FIlter method, set the mNodeFilterSwitch to the desired filter.*) CASE mNodeFilterSelect OF (********************************[ Status Filters ]***********************************) NodeID: FOR x := ACS_OUT_BEGIN TO ACS_OUT_END BY 1 DO FOR y := ACS_IN_BEGIN TO ACS_IN_END BY 1 DO IF marrNode[y].Status.oiNodeID > marrNode[y + 1].Status.oiNodeID THEN mNodeTemp := marrNode[y + 1]; marrNode[y + 1] := marrNode[y]; marrNode[y] := mNodeTemp; END_IF; END_FOR; END_FOR; Started: FOR x := DES_OUT_BEGIN TO DES_OUT_END BY -1 DO FOR y := DES_IN_BEGIN TO DES_IN_END BY -1 DO IF marrNode[y].Status.oxStarted > marrNode[y - 1].Status.oxStarted THEN mNodeTemp := marrNode[y - 1]; marrNode[y - 1] := marrNode[y]; marrNode[y] := mNodeTemp; END_IF; END_FOR; END_FOR; Starting: FOR x := DES_OUT_BEGIN TO DES_OUT_END BY -1 DO FOR y := DES_IN_BEGIN TO DES_IN_END BY -1 DO IF marrNode[y].Status.oxStarting > marrNode[y - 1].Status.oxStarting THEN mNodeTemp := marrNode[y - 1]; marrNode[y - 1] := marrNode[y]; marrNode[y] := mNodeTemp; END_IF; END_FOR; END_FOR; END_CASE; I have like 30+ of these in the enum. Not really DRY code right? These are custom made bubble sort filters in a function. You pass in the Datastructure, and say what function you want. (This is an enum collection of sorting functions) And then the Array with Nodes of Structs gets ordered. Why can't we have Iterators and Lambda's and build in standard functions like regular languages? Also i use bubble sort because it's the easiest to implement because i can't get this to code DRY. Problem with ST (Even the new one with classes) that it's very limited for programming datastructures and algorithms. Yes you still not want dynamic memory and you need to choose the correct algorithm so you know the most extreme edge cases regarding the time it takes to execute the algorithms.(Real-time execution) How are other people dealing with this? Here for example saw some software using an adjusted ST language and having FOR EACH possibility: https://www.fernhillsoftware.com/help/iec-61131/structured-text/st-for-each.html You can then build your own custom Iterator functions. I wish the IEC 61131-3 standard would be more expressive and having more standard modern features, but still keep close to the fact of no dynamics memory and real-time systems.
Last updated: 2023-08-31

Post by brouwyka on JSONByteArrayWriter string result incorrect order ^{CODESYS Forge talk (Post)}
Hi @TimvH, As discussed in our email contact, your example does not actually test/reproduce the bug I am describing: Adding to the JSON builder on later cycles works fine. Your example works because you do the following: 1. You finish the first cycle (xFirst) with adding an object to the array with 2 fields; 2. When the next cycle is triggered (xAdd), you immediately start with adding another object to that array. This works fine as the last thing you did before #2 was the addition of the array and an object to it. To reproduce the bug you should be doing the following instead, as I showed with the code I shared with my first post: 1. Create the array - save the index that is returned; 2. Add an item, either a primitive or an object, to the array; 3. Add an item outside the array (anything: a primitive, a new array, a nested object, etc); 4. Add another item to the array using the index you saved at #1. You will now see, as I shared in my initial post, that the second item is placed completely outside the root JSON object. This also happens to anything else you try to add after step #4: everything after this point will be added outside the root JSON object: the JSON is completely broken. Important to note is that this not only happens with arrays, but also with nested JSON objects. Once you add something outside of a nested JSON object, you can no longer add anything to that nested JSON object, as that causes the exact same bug. This also applies to arrays of objects, so if in your test you had tried adding a new key-value pair to the first nested object in your array after you created the second nested object, you would also run into this bug. It seems that the JSONByteArrayWriter (I haven't tested the other writers in the JSON Utilities SL library, so I don't know if they suffer from the same problem) simply does not handle any JSON fields that add brackets (so arrays with "[" & "]" and nested objects with "{" & "}") well, and closes them prematurely instead of checking if any later JSONElements in the JSONData's array belong to any of these bracketed fields. After reviewing the objects & functions of the JSON Utilities SL library, my guess is that either the JSONByteArrayWriter linearly goes through the array of JSONElements in the JSONData and only checks the diParentIndex of each JSONElement in direct ascending order, OR if it does use the JSONElement.GetChildren(); method, that this method is either broken and doesn't give all children correctly to the writer. Neither explains why everything completely breaks, and you cannot even add to the root JSON object anymore, however, so there is probably more than just that going wrong in the writer. To me, after a full week of testing and attempting workarounds, this seems like a bug in the library that needs to be fixed by Codesys, as I cannot see anything wrong in the JSONData constructed by the JSONBuilder - this seems purely a problem in the writer.
Last updated: 2026-06-15

Post by hammer on Cartesian XY Chart — 2-D array binding: "Element to use" enumerates the wrong dimension ^{CODESYS Forge talk (Post)}
Title: Cartesian XY Chart — 2-D array binding: "Element to use" enumerates the wrong dimension; documented example (Dialog: XY Chart Configuration, "Two-dimensional array of base type REAL") appears transposed Product / version: CODESYS Development System V3.5 SP21 Patch 40 (3.5.21.40). Cartesian XY Chart visualization element (System_VisuElemXYChart / VisuElemXYChart). Summary: When a curve's Data X/Data Y is bound to a two-dimensional array, the element treats the last array dimension as the per-point record (the one selected by "Element to use") and iterates the first dimension as the sequence of curve points. The online documentation's worked example has the dimensions in the opposite order, so following it produces a non-functional curve and a nonsensical "Element to use" list. Steps to reproduce (fails — matches the documented layout): 1. Declare aData : ARRAY[0..1, 0..4095] OF LREAL; (2 data sets × 4096 points, i.e. first dimension = set, second = point). 2. In XY Chart Configuration, bind a curve's Data X (and Y) Variable to this array. 3. Observe the "Element to use" combobox. Actual result: The "Element to use" combobox lists 4096 entries (the indices of the second dimension) instead of 2 (the data sets in the first dimension). Selecting a row does not produce the expected curve; at most a single/degenerate point is plotted. Steps to reproduce (works — transposed): 1. Declare aData : ARRAY[0..4095, 0..1] OF LREAL; (points first, set second). 2. Bind Data X/Y to this array; "Element to use" now correctly lists [0] and [1]. 3. Populate aData[point, set] and the curve plots as expected. Expected result: Either the documented example should select the data set from the first dimension (as its [1]/[2] "Element to use" implies), or the documentation should state that the last dimension is the record dimension (stride) and the first dimension is the point sequence — consistent with how the element actually behaves. Analysis: The behavior is self-consistent if the element always walks a single array by a fixed byte stride, reading one element per point — the same mechanism as the documented "one-dimensional array of a DUT" case, where it strides by SIZEOF(DUT) and "Element to use" picks the member offset. For a 2-D array the record is the last dimension: to plot column/set c it starts at [0, c] and strides by the last-dimension length to reach [1, c], [2, c], … (row-major layout makes this a contiguous column walk). Under that rule, ARRAY[0..N, 0..1] is correct (record length 2 → "Element to use" [0]/[1]), and the documented ARRAY[1..2, 1..50] with "Element to use" [1]/[2] is transposed. Documentation reference: Dialog: XY Chart Configuration → Data X / Data Y → "Example: Two-dimensional array of the base type REAL" (ARRAY[1..2, 1..50], Element to use [1]/[2], "100 array values are displayed"). This example does not plot as described in 3.5.21.40; transposing to ARRAY[1..50, 1..2] does. Request: Please confirm whether this is (a) a documentation error (example dimensions reversed) or (b) an element defect (combobox enumerates the wrong dimension). Either way, a corrected example plus an explicit statement of which dimension is the point sequence vs. the record/stride dimension would prevent others from hitting this. Two separate 1-D arrays for Data X and Data Y work correctly and are unaffected.
Last updated: 2026-08-14

Post by mondinmr on Frustration-Fueled Feedback on Project File Management and Git Integration ^{CODESYS Forge talk (Post)}
Good day, I’m writing this message out of frustration regarding the current way project files are saved as encrypted XML and single-file format in CODESYS. I find this approach to be quite tedious for several reasons: Limited Access to Structured Text: Not being able to access Structured Text (ST) externally makes it impossible to work with alternative editors like VSCode. Tools like VSCode are incredibly responsive and feature advanced systems such as GitHub Copilot, which would be a real game-changer when working with ST IEC. While CODESYS works well for small code snippets or debugging, when the codebase grows, switching to VSCode to boost productivity becomes essential, and copy-pasting between environments is a cumbersome workaround. Poor Integration with Git: This file format also makes it very difficult to integrate effectively with Git. I have tested the internal demo, but for advanced merges, it is unusable. Without properly formatted plain text, it’s impossible to leverage the vast ecosystem of external tools around Git that allow smooth merges in heterogeneous teams. File Corruption on Network Drives: I often work from multiple locations with shared network drives. When the development environment saves a file and something goes wrong midway (which can occasionally happen when using VPNs and network drives), the entire project becomes irrecoverable. There’s no way to cancel the save process, and the development environment freezes. This has happened to me at least four times over the past two weeks, and one of those incidents cost me an entire day of work. All of this is particularly disappointing because I truly believe that the libraries, runtime, and overall work done by CODESYS are exceptional. I find it fantastic that there is a platform allowing development of PLCs and control systems using OOP, which is a huge advantage in modern control engineering. I apologize for the rant, but this issue has been extremely frustrating. Best regards.
Last updated: 2024-10-15

Post by mondinmr on Frustration-Fueled Feedback on Project File Management and Git Integration ^{CODESYS Forge talk (Post)}
In reality, other vendors are already moving towards integration with VS Code. B&R, for example, is likely the only one that could truly challenge Codesys in the market: they’ve overhauled their once-archaic environment (previously lacking ST OOP) and are now catching up significantly with Automation Studio Code, offering modern programming capabilities and native support for tools like Git and GitHub Copilot. The other big names in the market, frankly, are still light years behind. For smaller projects, typically handled by “wire strippers” (with all due respect!), a classic environment might be just fine. But once you’re dealing with thousands or even millions of lines of code—possibly organized in an OOP paradigm, spanning large teams, integrating third-party libraries, requiring version control, and performing comprehensive tests—a lightweight, responsive editor supercharged with AI-based features becomes essential. The real strength of something like VS Code lies in its vast ecosystem of extensions and integrations: from advanced Git support to automated build/test pipelines, and even AI-assisted autocompletion that already makes software development much more agile and productive. I understand that a PLC project involves more than just ST code (task setup, fieldbus configuration, libraries, etc.). However, there’s no technical reason not to at least provide ST files in a pure text format, making version control and collaboration far easier. B&R has clearly adopted this approach; I hope Codesys will accelerate its own efforts to offer a native, text-based integration and modern development tooling. Once you go beyond a certain level of complexity—dozens of configurations, OOP hierarchies, and distributed teams—you really do need “engineer-level” software tools, not just solutions targeting “wire strippers.” For large-scale automation projects, a more open and flexible approach is the only way to avoid chronic frustration. Let me emphasize that I truly love Codesys and consider the work done so far to be brilliant—I’d be sorry to see it overshadowed by others moving faster in this direction.
Last updated: 2025-02-18

Post by kurtk on SysProcess Execute Command unable to run commands with special characters ^{CODESYS Forge talk (Post)}
Thank you for this additional information. Could you please provide the references to the official documentation where you found this information ? I got as far as: Then insert this: [SysProcess] BasePriority=Realtime Command=AllowAll devoid of references of where the CodesysContol.cfg file was located or of the need to run codesys as root I incrementally located bits of information with perplexity searches but remained mired at error 25 ERR_NO_ACCESS I am curious whether root access is MANDATORY to run SysProcessExecuteCommand (- and/or any other SysProcess functions) or if a more access-constrained User=quasiroot could suffice without the security concerns ? Apparently codesys runs fine without User=root for almost everything... That raises the question: What IS the default codesys user ? I notice in the prototypes: VAR sCommand : STRING := '/usr/bin/whoami'; sOutput : STRING(255); refCommand : REFERENCE TO STRING; refOutput : REFERENCE TO STRING; resultCmd : UDINT; END_VAR whenever I attempt to build this Codesys v3.5 SP2 whines that "cannot convert type REFERENCE TO STRING to STRING" don't remember the exact message... Apparently this happens whenever a fixed-length string is involved. if I get rid of (255) and just set the string := ' '; that is sufficient to make the compiler happy. Is this a recently enforced constraint ? Why does this compile for you and not for me ? There are references to functions / utilities which are used/needed to massage the string references - but I see no specific details... I also notice you specify: Make sure the lib SysProcessImplementation, SysTypes, and CmpErrors is on the project. I don't comprehend why all of the dependent libraries are not automatically referenced, added ? I don't recall seeing anywhere but here that SysProcessImplementation must be included... It just seems odd that the default is to exclude essential libraries, or require alchemical conjuring to assemble all the ingredients for gnat soup ;-) Thanks a lot... I'm still a codesys newby getting my wheels on - or sometimes just partially on
Last updated: 2025-09-12

Post by adamrobotic on llcallback() iec exception caught cmpid=0x00000002, eventid=0x00010008 ^{CODESYS Forge talk (Post)}
Hi All, Recently I have been experiencing a issue in CodeSys SP16 Patch 3, with a exception which comes up after I have made several online changes. callcallback() iec exception caught cmpid=0x00000002, eventid=0x00010008, after I experience this fault then I receive a error message for any subsequent online change "online change failed, no task gap within 10000ms". To recover either the controller needs a orgin reset, which will allow online edits thereafter until falling back into the loop. I am using a TURCK TX707 product which is controlling robot trajectory, We have hard coded the coordinate data rather than using persistence variables so hence the need to do online changes consistently. I have read on the forum other people have experienced similar issue with the online task gap, but I have no task which have same priority and also monitored the CPU usage and have not seen it go above 13%. Anyones help would be much appreciated, thankyou.
Last updated: 2023-08-18

Post by sunfloweeerrry on Edgelogix-rpi-1000-cm4102032 SPI port connection to PiFace I/O Driver in CodeSys. ^{CODESYS Forge talk (Post)}
Hello, I hv a problem. Couldn't solve this for 2 days now. So I'm quite new at CodeSys and I have to use the spi port x30 on the edgelogix rpi 1000. So what I have to do is just 'add device' at the SPI where I hv to pick the 'SPI master' after that I hv to change the parameters according to the edgelogix I did found in the spidev.h but nothing got updated so I 'add device' again where I pick the PiFace I/O Driver and I put my variables at the I/O Mapping, which i put the input startbit at the in part and led at the out part. I want to make the led blink according to my structured text code but somehow the hardware didn't connected with the CodeSys spi port. Nothing happens. Any idea what I have to do?
Last updated: 2023-08-18

Post by ofey on Codesys Soft PLC OPC UA server ^{CODESYS Forge talk (Post)}
Hi! I'm sorry if this has been asked before. I have tried looking up previous topics regaring this, but even though I'm following tutorials and other posts I cant seem to get this to work consistently. I actually got the connection established on a test project after some trial and error. I only got it to work with the "Anonymous login". But after i deleted the test project and tried to get the same thing to work on another project I cant get it to work anymore. What I'm trying to do: Set up and OPC UA Server using CODESYS with the free soft PLC. I then try to connect to the OPC UA Server with the program "UaExpert". What happens: I get the error Error "'BadUserAccessDenied' was returned during ActivateSession". What I have done in CODESYS: - Added the object "Symbol configuration" and checked some test variables - Checked "Support OPC UA features - In "Communication settings -> device -> Change runtime security policy" I have checked "Allow anonymous login" What I have done in UAExpert: - Found the OPC UA server under "local". (It is showing two servers though, I do not know why. It showed the same when it worked yesterday). Se attachment - Checked "Anonymous" under "Authentication settings in the OPC UA server. I'm new to CODESYS, so it could be something elementary wrong in my settings. Maybe something in the user settings in CODESYS regarding user rights or something?
Last updated: 2023-09-14

Sort by relevance or date