PLC programming: 5 mistakes to avoid

Stick with your architecture and remember to reuse your code

By Daniel Fenton January 15, 2014

Every PLC programmer will make a mistake when he or she is writing code. Whether it’s the result of pressure by the customer, lack of coffee in the morning, or simply getting distracted at the wrong time, here are the top five mistakes to avoid.

1. Not sticking to an organized architecture

When a program is first laid out, all of the code for each portion should fall into neat, orderly sections. Keeping each part of the code separate allows the person working with the program to easily visualize the overall execution of the program.

It also helps to debug the program if something goes wrong, since each piece of code is encapsulated into its own structure. Finally, it looks better, because there isn’t the mess of an entire program on the screen at one time.

Unfortunately, as the developer and post-deployment maintainers work on the code, sometimes the code begins to creep out of each section, and intertwine with the supposedly isolated functions around it. This leads to confusion as tags hop from one sheet to the next, or outputs are written twice during execution.

It also leads to disorganized code. Finally, this can be dangerous, as minor tweaks across the board cause little program issues to build up until something breaks. Maintaining code organization and encapsulation is critical to the longevity of a PLC program.

2. Not documenting the code

Documenting code as it is written and, further, as it is maintained is critical to keeping a PLC and its program operational during the long periods of time between updates and adjustments. A quick few sentences on each major portion can save a lot of time and headaches later on. It can also help the programmer put his or her thoughts on paper, which can be helpful in figuring out the next step.

Even though the code may make sense at the moment it is programmed, the five minutes spent explaining thoroughly why a particular technique was used can save hours when, months later, a user must decipher what is going on. All too often, code originally installed in a system will change due to fixes, updates, and feature additions. If the documentation is not updated with the code, this can lead to confusion and misinterpreted programming.

3. Creating redundant tags and variables

As logic statements and ladder rungs are written out, often a program can have multiple branches of logic that eventually lead to a tag that functions as a sort of “flag.” This flag will usually wind up activating other pieces of logic in the program. Sometimes, these pieces of logic are mutually exclusive, such as a Boolean variable, which requires the program to send a message to one remote station or another, depending on the status of certain inputs.

The IEC 61131 Structured Text statements IF and ELSE are purpose-built for this exact logic, so that only one tag needs to be tested. If the tag is true, one set of code is executed. If the tag is false, another set of logic is executed.

If a programmer is not paying attention, he or she can be put in the situation of creating a “flag_1” and a “flag_2,” with both of these tags performing the same duty. Now, both variables need to be tested, which leads to increased code complexity and greater memory utilization. Not only are both of these results poor programming practice, but repetitive use of redundant tags and variables will require that a program utilize a more powerful (and therefore expensive) processor than otherwise required.

4. Not reusing existing code

Aside from ease of maintenance, debugging, and readability, the reason it is so important to encapsulate and isolate code in a program is that code encapsulated inside of a function block can be reused over and over throughout the program. This removes the requirement to rewrite the code that already exists inside of a function block. Therefore, the programmer can generate a standard, all-use function block that can be debugged and modified once, and have that change reflected throughout the entire program.

Collections of these general use function blocks are called function block libraries, and many manufacturers offer application or industry-specific function block libraries for download on their websites. These function block libraries heavily cut down on the amount of development time required for a given application or project.

Sometimes, without realizing it, a programmer will find him- or herself writing code that’s already been written. Where five or six of the same pieces of equipment need to operate similarly, it is easy to create a function block that accepts the inputs of the sensors and outputs solutions for the actuators. On the other hand, it can be more difficult on seemingly one-off algorithms to find a reason to write generic code.

The effort needed to write a function block that is generic enough to be reused must be balanced against the time spent for development on the project. It is worthwhile, however, to bear in mind that a little extra time spent immediately can save a lot of time later on.

5. Not utilizing version control

Keeping your code organized is another major practice that is sometimes lost in the commotion of trying to complete a project for commissioning. Version control systems can be as simple as naming a project for a particular PLC once it has been deployed, or as powerful as a dedicated application used specifically to track what the changes are in each major revision.

The use of such versioning schemes mitigates situations where a particular fix or feature might have gone out to the field, and the PLC has a program that may or may not have those changes.

Programmers often forget to keep track of what changes are in the program on the PLC without a proper versioning strategy. Documentation, a good naming scheme, and possibly dedicated version control software can lessen these issues and increase the detail available to the programmer.

Dedicated version control software also allows a system to track the user that made every single change, documented or otherwise. This can be particularly useful if multiple programmers are working on the same program. 

Daniel Fenton is a product specialist with Phoenix Contact.