Many companies start with a PLC-based system and successfully sell their products for years.

Eventually, someone looks at the bill of materials and notices that the PLC is one of the most expensive parts of the system.

The obvious idea is: “Why don't we replace the PLC with a custom PCB and reduce the cost?”

At first, this sounds reasonable.

An engineering company is hired, a microcontroller is selected, and the PLC logic is rewritten in C. The prototype works, production begins, and everything seems fine.

Then the problems start.

The new board doesn't behave quite like the original PLC. Strange failures begin appearing in the field. The system works most of the time, but reliability is not what it used to be.

Why does this happen?

One reason is the software architecture.

PLCs are based on a continuous scan cycle. Inputs are read, logic is executed, outputs are updated, and then the entire process repeats over and over again.

If a noise event temporarily changes the state of an input or output, the next scan often corrects the problem automatically.

Many embedded programmers do not write software this way. Instead, they may change an output and assume it will stay that way forever.

If noise changes the state of an I/O signal, the software may never notice. The incorrect state can remain until someone manually resets the system.

The second reason is hardware robustness.

Commercial PLCs have spent decades surviving harsh industrial environments. They contain protection circuits, filtering, surge suppression, isolation, and many other features that are easy to overlook.

When designing a custom board, it is tempting to focus only on functionality. However, simply making the system work is not enough.

The difficult part is making it survive noise, surges, static electricity, wiring mistakes, and years of real-world use.

Many engineers discover that copying the PLC's logic is easy. Matching the PLC's reliability is the hard part.

←Noise Is My Friend