There's No One 'Best' PLC. Here's How to Find Yours.
I get asked this at least once a week: "What's the best Omron PLC?" My answer is always the same: "It depends." And that's not a cop-out. It's the truth. The right controller for a high-speed packaging line is a bad choice for a simple pump station. The best CPU for a project with a 6-week deadline is different from one where you have 6 months to spec and test.
After coordinating over 200 rush orders for controls projects—some with less than 48 hours of lead time—I've stopped believing in a universal recommendation. Instead, I think in scenarios. Here are the three I see most often, and what I'd actually do in each.
Scenario 1: You're Retrofitting an Existing Machine (or Panel)
This is the most common, and the most painful when done wrong. You have an existing control panel. Maybe it's a Frigidaire dishwasher test fixture—I've seen that before. Or an old CJ-series system that's been running for a decade. The goal? Swap the brain without redesigning the entire panel.
What to Look For
In this scenario, form factor and wiring compatibility are king. You're not trying to invent a new architecture. You're trying to minimize downtime. Your key specs are:
- Same I/O footprint. Can the new PLC module fit the existing wiring layout? If you're replacing a CJ2, the direct replacement is often another CJ2, or an NX-series with an adapter. Or rather, to be specific: an NX1P2 with the right terminal block adapter mimics the form factor closely.
- Pin-to-pin compatibility. This is where I've seen projects derailed. A CP2E looks similar to a CP1L on paper. The pinouts are different. You'll be rewiring. (Note to self: always verify this before promising a 4-hour swap.)
- Existing programming software. If your maintenance team is comfortable with CX-Programmer, don't force Sysmac Studio on them unless there's a clear benefit. I still kick myself for specifying an NJ on a retrofit where the techs had only used CJ. We added two days of training costs.
The Advice (Not Popular, but True)
Don't upgrade the CPU generation just because you can. A CJ2 is still a perfectly capable controller for most discrete manufacturing tasks. The new NX or NJ series offers better motion control and IIoT connectivity, but if none of that matters to the current application, you're introducing unnecessary risk. In Q1 2025, we had to rush-ship a discontinued CJ2 CPU because the replacement NX unit had a 3-week lead time and the client's line was down. Was it worth the premium? Absolutely. The alternative was a $50,000 penalty clause.
If you must upgrade, plan for a 20-30% buffer on your timeline. That covers the inevitable wiring mismatch (I've yet to see a direct swap that didn't need at least one terminal change) and the Sysmac Studio learning curve.
Scenario 2: You're Building a New System from Scratch
This is where you have freedom. No legacy constraints. But that freedom is deceptive—too many choices lead to paralysis. I've seen teams over-spec and under-spec equally often.
What to Look For
Your selection matrix should prioritize:
- CPU performance vs. I/O count. An NJ-series is overkill for 32 discrete I/O points. A CP1H handles that beautifully. Conversely, a CP1H will choke on a multi-axis motion control application. I learned this the hard way: we specified an NJ3 for a simple packaging line in mid-2024. It worked, but we paid for unused capacity. The client's alternative was a cheaper Mitsubishi setup, and honestly, they had a point.
- Network and communication requirements. EtherCAT? EtherNet/IP? Built-in or add-on module? An NJ or NX with built-in EtherCAT simplifies wiring. A CJ2 will need an additional module. That adds cost and panel space.
- Future expansion. This is the one place where I'll advocate for spending a bit more upfront. A CPU with 20% headroom now saves you from a full replacement in two years. (Based on our data from 150+ projects, the typical expansion is about 15% within 18 months.)
The Advice (That You Might Not Expect)
Start with the I/O modules, not the CPU. This sounds backward, but it works. Choose your analog input module (like an NX-AD3603) or your safety relay module first. Those determine the form factor and backplane. The CPU is then picked to match that architecture. I started doing this after a project in March 2024 where we bought an NJ3, then realized it needed a special adapter for the NX safety modules. The adapter had a 10-day lead time. Alternative: we could have chosen an NJ5 with built-in safety support. That's done.
Also: download the full datasheet before you quote. Not just the brochure. I once quoted a project based on a CP2E datasheet that listed 'Ethernet port.' Turns out it was only for programming, not for fieldbus. We had to add an Ethernet/IP module. —Actually, it was the CP1L manual that was misleading. Or maybe both. Point is: verify.
Scenario 3: You're on a Tight Budget (or Tight Deadline)
This is where my role as an 'emergency specialist' kicks in. You need a PLC, you need it fast, and you might not have the budget for the top-tier model.
What to Look For
In this scenario, availability and simplicity trump performance. You want something:
- In stock somewhere. Not just theoretically available. Actually on a shelf. We use distributor networks for this. (Note to self: call distributor before committing to a lead time.)
- Simple to configure. The CP1L or CP2E are great here. They're straightforward, well-documented, and the programming environment (CX-One) is familiar to most techs.
- Widely supported. If something goes wrong, you need replacement modules fast. Standard models (CP1L, CJ2) have better stock availability than specialized ones.
The Advice (Counter-Intuitive)
Pay for the rush fee if it buys certainty. This is where my core belief—time certainty premium—kicks in. In October 2024, we paid $400 extra for next-day air on a CP1L CPU. The alternative was a 5-day standard delivery that would have missed the client's commissioning window. The client's lost production was valued at $12,000 per day. The $400 looked like a bargain. (Disclaimer: prices as of Oct 2024; verify current rates.)
But here's the flip side: don't pay a premium for a model you don't need. I've seen engineers spec an NJ just because it's 'newer,' then pay extra for rush delivery. The CP1H or CJ2 would have shipped standard and arrived in time. The 'new model' premium is real, and often unnecessary.
One more thing: the 'cheapest' option often isn't. A discount vendor with a 2-week lead time vs. a premium vendor with a 3-day lead time? The 'cheap' one might cost you a week of downtime. I've seen the math fail on that too many times.
How to Know Which Scenario You're In
Still unsure? Here's a quick litmus test:
You're in Scenario 1 if: You have an existing panel. You are swapping the PLC. You can't touch the I/O wiring or the enclosure layout.
You're in Scenario 2 if: You are designing a new system from a blank page. You have time (3+ months). You want to optimize for performance and longevity.
You're in Scenario 3 if: You have less than 4 weeks to delivery. Or the budget is fixed. Or the project is 'mission-critical' and you can't afford a long lead time. This worked for us, but our situation was a mid-size B2B company with predictable ordering patterns. If you're a seasonal business with demand spikes, the calculus might be different.
I can only speak to domestic operations. If you're dealing with international logistics, there are probably factors I'm not aware of.
One final thought: the right PLC is the one you can get, configure, and deploy reliably—not the one with the most features on the datasheet. That's a lesson I learned the hard way.
