we’re currently running a TwinCAT 3.1 project (version 4026) with Git as our version control system, and we’ve run into a workflow issue that we can’t seem to solve.

Setup:

• 3 developers working in parallel

• Multi-PLC project (separate PLC task for each developer)

• One “main” development PC where the hardware configuration is activated

• Each colleague clones the project from Git and works on their PLC

• The goal is that all of us can be logged into our own PLC at the same time and perform Online Changes

The problem:

After the hardware config is activated on the main PC and we all pull the project from Git, TwinCAT refuses Online Change on the other machines.
We get an error saying that the mapping does not match or is invalid, even though nothing has been changed in the I/O configuration.

It seems like TwinCAT thinks the project is different from what’s running on the target, although everything (in theory) is identical – same commit, same configuration, same PLC code.

What we tried:

• Full rebuild on every workstation → no effect

• Deleting the .compileinfo, .tmc, and generated build folders before building → no effect

• Re-activating hardware locally on every PC → works, but then every developer ends up committing different hardware GUIDs to Git → merge chaos

• Using “PLC only” activation → same mapping error when trying Online Change

What we think is happening:

TwinCAT stores some hardware-related GUIDs / internal hashes locally, and after cloning the project, they don't match the target runtime, even if the code is identical.
As soon as one person re-activates the solution, the others can no longer do Online Change.

Question:

Has anyone found a clean workflow for multi-developer TwinCAT projects with Git, where multiple people can:

• clone the repo

• go online with their PLC

• do Online Change

• without everyone having to re-activate the I/O tree individually?

Is there a recommended “.gitignore” setup to avoid committing machine-specific files?
Or is TwinCAT simply not made for true parallel development with shared hardware config?

Any best practices, workarounds, or real-world experiences are highly appreciated!

Thanks in advance!

Note: English is not my native language, so I used ChatGPT to help write this post. Hopefully everything is understandable – feel free to correct me if something sounds unclear.

Read through the one post I could find on the topic here, and since it was more helpful than anything else I've found in the past several weeks, I figured I'd post looking for help. If anyone has some good resources for learning, or wouldn't mind downloading their knowledge/understanding into an example and post it as a reply to this thread, I'd be incredibly grateful.

Is this possible? If so, any pointers and references would be greatly appreciated! 

I'm trying to use TwinCAT on a Virtual Machine created with Virtual Box, but when I try to put TC in Run mode I get the following errors:

The first with the Previrtualization Interface set to "Hyper-V" and the second error when it is set to "None" (in the Virtual Box Settings --> System).
I tried already many suggestion is possible to find on the Network, like: deactivation of Hyper-V, Virtual Machine Platform and Windows Hypervision Platform on the Host system, setting of shared CPUs leaving one Insulated and putting it as Default, running as administrator win8settick, etc...

From the second Error, I though that the Guest machine is not able to find an active virtualization (Intel VT-x) by nested virtualization.

Indeed in the Setting of VM (in System --> Processor) the setting of "Nested VT-x/AMD-V is not selectable, and I was not able to make the selection active.

Here my system parameters:

Host machine:

HP ZBook Power with processor Intel Ultra 9 185H

OS Win11 Pro

Virtualization Technology and Intel VT-x in the BIOS are active

Virtual Box version 7.2.0 r170228

Guest machine:

OS Win 10 Pro

TwinCAT 3.1 build 4024.67

In the following screenshot all the setting of the Virtual Machine:

I decided to use a virtual machine, thinking was a good idea to start using TwinCAT and perform testing and prototyping in a saparated part of my PC. I wanted in the end install TwinCAT directly in the win11 real machine , but I'm very worry these problem could happen also there.

I'm really close to be disparate, and I hope someone could know a solution for this issue.
I thank you in advance for every help and apologize if my post is not orthodox.

With best regards,

Simone

I have the object model for the UDT in FT Optix, and we pivoted to a Beckhoff platform for everything but the HMI (given it's a rather complex HMI project that wouldn't be realistic to redo given our project timeline for runoff). Given the size and scope of the existing UDT structure, I would love to avoid having to redo it from scratch. Every other PLC brand exports these in standard PLCOpen formats, except Rockwell. Was told from a Beckhoff AE that it might be possible to pull the structure in via OPC UA and generate the analogous Struct in TC3.

Anyone out there done this? If so, any pointers or instructions you've found that kind of spell this out?

I’m working on a project where I want to connect two simulated PLCs (TwinCAT 3 and CODESYS) to a SCADA system (Ignition) via OPC UA. I already have a connection between Ignition and CODESYS, and now I want to add a connection between TwinCAT 3 and Ignition.

In CODESYS, I’m using the default port 4840. I think using the same port in TwinCAT as in CODESYS could cause problems. Is it possible to change the default port in TwinCAT 3?

Thanks,
Klaus

I am getting a very strange and challenging issue.

I have a fully working PLC software that has been running for a long time no issues.  Working in the field all operating 100%.

In MAIN, I have maybe declared variables (as can see several hundreds of variables currently working without issues) of BOOL, INT, REAL etc.  No issues.

But

As soon as I add any new variable declaration (i.e. as highlighted in the picture) and then click the button to "Activate Configuration", the code etc uploads but then the PLC goes into configuration mode and give the errors as in pictures.

The Errors seems to point to licensing issues which makes no sense.  Nothing else changed in the working programme.  Literally just declared a new variable:

bSuffMassflowToStandardise:BOOL;

has anyone tried OSCAT function Trend?, I can't get it work. i'm using an analog PLC signal as the X input, but the problem as  i noticed is the last_x is always equal t X. I doesn't store the previous value in memory. Can anyone help me or test on own system this block.