I'm the guy who reviews every technical deliverable before it reaches a customer. Over the last 4 years, that's meant signing off on roughly 200+ unique PLC systems and related documentation annually. In our Q1 2024 quality audit, we rejected 8% of first-time deliveries due to documentation inconsistencies—a number we're still trying to drive down. This list is built from the questions I actually get from engineers and buyers on the ground.
Short answer: Yes, but with a caveat.
The FX5U series, in particular, remains a solid workhorse. It is the natural successor to the FX3U, which you still see running thousands of lines across factories. The core strength is the same: reliability and a massive installed base. For standalone machine control or smaller distributed systems, it's a safe bet. The old 'best practice' of treating the FX series as a simple replacement for relays only doesn't apply anymore, though. The new iQ-F platform is way more network-capable than its predecessors. It can struggle if you try to force it into a high-speed, complex motion control role that's better suited for the R series.
Most people focus on raw I/O count and completely miss the architecture. That's a classic outsider blindspot. I can only speak to the selection logic we use for our clients, but it holds up:
The assumption is that it's 'just another IEC 61131-3 IDE.' The reality is the learning curve is steeper than that.
People think the drag-and-drop is intuitive. Actually, the performance of your project depends heavily on how you structure your POU (Program Organization Unit). I've seen a production engineer create a 'working' program but it ran a 50ms scan time just on the main routine. When we restructured it using a proper task configuration and separated the high-speed logic into an interrupt block, the scan time dropped to 5ms. That's the difference between a machine that synchs correctly and one that randomly faults.
A lesson learned the hard way: GX Works3 is powerful, but you must understand the task cycle. Don't just copy-paste logic from a Mitsubishi GX Developer project without refactoring it.
This worked for us, but our situation was specific. We had a batch of 30 Q68ADV modules that were out of warranty. The cost of new replacements was $600 each. A third-party repair shop quoted $180 each with a 90-day warranty.
We sent 5 test units. Two came back with faulty channels. The vendor claimed it was 'within industry spec,' but our internal test jig showed a 5% drift on an analog channel that should be 0.1%. We rejected the batch. They fixed it, but the timeline cost us a delivery delay. Now, every contract for third-party repair includes a specific calibration certificate requirement matching the original factory spec. For mission-critical or safety-related modules? Don't do it. For a simple I/O module on a non-critical conveyor? It's a cost-effective play, but you have to have a test protocol ready.
This pricing was accurate as of Q4 2024. The market changes fast, so verify current rates. An official USB-SC09-FX cable runs around $85. A generic one is $15.
I've used both. The generic cable works fine—about 95% of the time. The issue is consistency. The problem isn't that the cheap cable doesn't work; it's that it works on your laptop but not on the client's. Or it works in the office but not at the machine due to electrical noise. The official cable has better shielding and a ferrite core. It's a no-brainer for a professional service team. That $70 difference is nothing compared to the cost of a field service call to debug a connection issue. For a hobbyist bench? The generic is fine.
Yes, but not always the way you think. This is a specific question from a site engineer who was baffled.
The question everyone asks is 'what is the battery voltage?' The question they should ask is 'what is the voltage under the specific load of the PLC's memory backup circuit?'. A lithium battery for the FX3U (BR-2/3A) should be around 3.6V. If you measure it unloaded, it might show 3.6V. But if it's old, its internal resistance increases, and when the PLC tries to draw its very small backup current, the voltage can drop below the 2.8V threshold required to keep the RAM alive.
To diagnose this properly, you don't just measure the battery. You measure the voltage across the battery terminals while the PLC is powered off. If that voltage is below 2.8V, or if the battery is over 5 years old, replace it. Don't wait for a 'Low Battery' alarm on the display, because by then, your parameter settings might already be corrupted.