Modular PLC vs Compact PLC: How to Choose, Configure, and Source the Right System

You are here for one of three reasons. Your line is being expanded, and the controller has run out of slots. You are retrofitting an old panel, and nobody stocks the cards anymore. Or a module failed this morning; the vendor says it is discontinued, and the line is down while you read this.

Most guides will tell you a modular PLC is more flexible than a compact PLC. True, and also not the question you need answered.

Quick answer: A modular PLC is usually right when the I/O count is high or expected to grow, when expansion happens in phases, when field devices are spread over a large area, and when the plant needs spare parts available in ten years. A compact PLC is usually right for a standalone machine with a fixed I/O count, limited panel space, and a tight budget.

Below: a six brand comparison, a six step BOM workflow, four paths for replacing a discontinued module, and a straight answer on lead times and authenticity. Our PLC module page covers the module families themselves; this guide goes straight to the decision.

1. What Actually Sits Inside the Rack

A modular programmable logic controller is not one device. It is a set of separate cards sharing a backplane, so you buy exactly the cards the application needs and nothing else.

Three of those mistakes come back later. They are the ones that get a rack re-quoted after it is already on the water.

Local rack vs distributed I/O. In a local rack, every card shares one backplane and all field wiring returns to the cabinet. In a distributed setup, small I/O stations sit beside the equipment and report back over industrial Ethernet or a fieldbus, as with Siemens ET 200SP, ABB S500 I/O, or Schneider TM3. Rule of thumb: past roughly 20 metres of multicore cable, price a remote I/O station before you price the cable and tray. Distributed I/O saves copper and commissioning hours. It also adds a network layer that can fail.

Where compact and modular overlap. An FX5U takes expansion adapters. A CompactLogix takes local expansion. An M241 takes TM3 cards. So the difference is not whether a controller can expand. It is the ceiling on expansion points, whether the CPU can be replaced without replacing the body and its built-in I/O, and what happens when a card dies. On a rack you swap a card. On a compact unit you swap the controller, restore the program, and re-terminate wiring. The boundary is not clean, which is why you need a decision rule rather than a definition.

2. Which One Does Your Line Actually Need?

The 7 point checklist

A rule of thumb, not a standard. If three or more answers point to modular, modular is usually correct.

• I/O count and three year growth. Fixed, under roughly 30 points, and no growth planned? A rack is overkill.
• Phased expansion. Modular lets you buy racks now and cards later.
• I/O type diversity. RTD or thermocouple inputs, high speed counters, motion axes? Compact units run out of specialty options fast.
• Field distance. Devices spread beyond one machine footprint point to modular with remote I/O.
• Downtime cost per hour. The most underrated criterion. A failed card is a ten minute swap on a rack, and a controller replacement plus program restore on a compact unit.
• Panel space and heat. Racks need width and airflow.
• Platform lifecycle. How many years of spare parts support are left?

If you cannot answer question 7 for the platform you are about to buy, ask your supplier before you order. It is the most expensive question people skip, and section 5 explains why.

Where the cost curve crosses

We are deliberately not giving a percentage. Any claim that a modular PLC saves a specific number depends on your I/O mix, and the number is usually invented. Here is the cost structure instead, which is what decides it.

A compact PLC puts the cost on the day you buy it. A modular PLC spreads it over five years. Which is cheaper depends on whether your line will change. A standalone packaging machine with 24 fixed points should be compact. A filling line built in phases should not be.

3. Brand Platform Comparison

We stock all six, so we have no commercial reason to push you toward one. Here is how they actually differ.

If you already have a dominant brand, start from that row. If your field devices speak one protocol, start from the protocol column. If neither applies, start from the fit column.

Backplane lock-in: what you are really committing to

You lock in at three levels, and only one is obvious. Hardware: backplanes, connectors, and pin assignments are proprietary, so a card from one brand will never sit in another brand's rack, and no adapter exists. Software: your program is written in one vendor's environment, and migrating means rewriting and revalidating logic, a cost frequently larger than the hardware and almost never in the budget. Supply chain: nobody prices this one, but choosing a platform means choosing whose obsolescence schedule you live with, and whose spare parts market you shop in when a card fails at 2 a.m. The CPU is a one time line item. The backplane is a subscription.

What is not locked: Modbus TCP, OPC UA, and gateway or remote I/O bridging let mixed brands exchange data. You are locked at the rack, not at the network.

Choose the protocol first, then the brand

Most people pick the PLC and then discover what their field devices speak. Do it the other way: inventory what your drives, sensors, and instruments already talk, then pick a controller that speaks it natively.

If your field devices speak three protocols, you are buying a gateway or a remote I/O head. Budget for it now, not during commissioning.

If your plant already runs mixed brands, and most do, standardise going forward rather than all at once: new lines share one controller brand, existing panels get bridged over Modbus TCP or a gateway, spares are pooled by brand. Ripping out a working line to unify a badge almost never pays back.

You have picked a platform. Now turn it into a parts list.

4. Building the BOM: A Six Step Workflow

Step 1. Count and classify your I/O. Count by signal type, not by device. One VFD can occupy a digital output, an analog output, and a communication connection at once, so counting devices gives the wrong number. List tag, device, signal type, voltage, quantity, and put RTD and thermocouple signals on their own row: treating a temperature input as an ordinary analog input is the most common way to order the wrong card.

Step 2. Add spare capacity. Common practice is 15 to 20 percent spare points. The part most guides miss: leave that capacity at module and slot level, not just point level. Three unused points on a 16 channel digital card will not help when the next expansion brings four thermocouples. You need a free slot and backplane current, and neither shows up on a point count.

Step 3. Select the CPU by scan time, memory, and connections, not by I/O count. Point count decides how many cards and racks you buy, not the CPU tier. Four things do: program memory, required scan time, communication connections (HMI, SCADA, MES, and every drive on the bus each consume one), and motion axes. Count comms connections before I/O points. That is usually what pushes you up a tier.

Step 4. Verify the power budget. This is the mistake from section 1, and the most common way a rack fails on site. Sum the backplane current draw of every card against what the power supply and CPU can deliver, checking the 5 V backplane current and the 24 V field supply separately, because most people only calculate the second. Get it wrong and the rack will not power up, or worse, it restarts intermittently once the cabinet warms up. You find that out after the shipment clears customs.
 

Step 5. Pick communication, and integrate the HMI and drives. Comms cards follow from the protocol decision above. Check driver compatibility with your HMI panels: same brand is normally zero configuration, mixed brand needs the driver confirmed before you order. Then check how your variable frequency drives connect. Hardwiring an analog output per drive is simple but consumes I/O; bus control saves wiring and consumes CPU connections, which loops back to Step 3.

Step 6. Define the spare parts stock. Stock by downtime cost, not failure rate. The CPU and power supply come before the I/O cards: a failed input card costs one loop, a failed CPU costs the line. After those, stock the card type you use most, usually the 16 point digital input.

This workflow assumes the modules you specify are still being manufactured. Increasingly, they are not.

5. Discontinued Modules: How to Find the Right Replacement

Every platform moves through five stages: active, mature, phase out, discontinued, and end of spare parts support. A controller family stays commercially alive for roughly ten to fifteen years. A production line runs twenty to thirty. That mismatch is the root of every obsolescence problem you will have, and it is why question 7 matters before you buy.

Four replacement paths, easiest first

Path 1. The original part number from stock. Nothing changes: no rewiring, no program changes, no revalidation. The only questions are whether it can be found and whether it is genuine. Almost always the right first move in an emergency.

Path 2. The manufacturer's official successor. Many discontinued modules have a designated form, fit, and function replacement that drops into the same slot. Not all do.

Path 3. Cross-generation migration. Move the rack to the current platform. This is real engineering: I/O remapping, program conversion, revalidation, sometimes panel modification. Anyone who calls it simple has not done it.

Path 4. Cross-brand rebuild. Last resort. New hardware, new software, new spares, new training.

Five specs to match before substituting

1. I/O type and point count, from the nameplate.
2. Signal range and electrical rating: 4 to 20 mA or 0 to 10 V, relay or transistor output, current per point, isolation. Most wrong substitutions happen here, and the answer is in the datasheet, not the listing.
3. Backplane and mechanical interface, in the selection manual.
4. Communication interface and addressing, in the configuration software.
5. Firmware and CPU version compatibility. The most hidden of the five, and the one that only shows up on site.

What to send your supplier for a cross-reference

A supplier who quotes a replacement without asking for these is guessing.

• A photo of the nameplate with the complete order number (MLFB, Cat. No., part number). Families hold dozens of variants differing only in the last characters.
• Hardware and firmware version, without which nobody can guarantee compatibility with your CPU.
• A photo of the whole rack, showing the slot layout and neighbouring cards.
• Your I/O list or wiring diagram. Hand drawn is fine.
• Quantity and urgency. A line down repair and a shelf stock order are different jobs.

Line down? Send the nameplate photo and the model number.

We cross-reference a replacement across all six brands, confirm what is physically in stock, and reply within 24 hours. Contact us by email or WhatsApp.

Free consultation

6. Sourcing Reality: Lead Times, Authenticity, Compliance

Stock lead time vs factory lead time. Two supply chains sit behind every quotation. Distributor stock moves in days to weeks, depending which warehouse holds it. Factory production moves in weeks to months, and popular part numbers queue. The rule that matters: when a line is down, ask what is in stock before you ask what it costs. Downtime is billed by the hour, and a few percent saved on a part that lands in eight weeks is not a saving.

Genuine, refurbished, and pulled modules. Refurbished and pulled modules are not illegal, and for a long discontinued part they are sometimes the only option that exists. The problem is never that a part is refurbished. The problem is a seller who does not say so. A supplier who tells you a part is refurbished is more trustworthy than one who will not. Check incoming modules for intact seals, a legible batch code, a verifiable serial number, tool marks on the screws, and oxidation on the backplane connector. Treat a price far below market as a warning, not a bargain.

Export compliance. Ask before shipment, not after: CE, UL, and RoHS status for the destination; invoice, packing list, and certificate of origin; the HS code confirmed in advance, since a wrong code delays customs more often than anything else; ESD and moisture protective packaging; and the Incoterms, because FOB and DDP decide who clears customs and pays duty. Our inspection and documentation process is on our about and quality page.

Four questions worth asking any supplier. Is this new, refurbished, or pulled, in writing? Which warehouse holds it, and can you photograph the unit? Does the warranty start at shipment or arrival, and who pays return freight if the part arrives dead? Can you cross-reference a discontinued part number?

FAQ

 

 

What is the difference between a modular PLC and a compact PLC?

A compact PLC packages the CPU, power supply, and a fixed set of I/O into one body. A modular PLC separates them into cards on a shared backplane, so you configure the I/O you need and expand later by adding cards instead of replacing the controller.

When is a modular PLC worth the extra cost?

When the line will change. The premium buys cheap expansion, a ten minute card swap instead of a controller replacement, and a longer spare parts horizon. If nothing grows, it buys you nothing.

Can I mix PLC modules from different brands in the same rack?

No. Backplanes, connectors, and addressing are proprietary, so a Siemens card will never work in an Allen-Bradley chassis. Mixed brand plants integrate at network level, typically over Modbus TCP or through a gateway.

My PLC module is discontinued. What are my options?

Four, in order of difficulty: source the original part from stock, fit the manufacturer's official successor, migrate the rack to the current platform (S7-300 to S7-1500 with ET 200SP, for example), or rebuild on another brand. Send a nameplate photo and firmware version for a cross-reference before assuming migration is needed.

How do I know if a PLC module is genuine and not refurbished?

Check the seals, the label print quality, the batch code, and whether the manufacturer can verify the serial number. Look for tool marks on the screws and oxidation on the backplane connector. A price far below market is a warning, not a bargain.

What is the typical lead time for modules like Siemens ET 200SP or Allen-Bradley CompactLogix?

It depends entirely on whether the part sits in a warehouse or a production queue. Stocked parts move in days, factory scheduled parts in weeks or months. If the line is down, ask what is physically in stock before you ask for a price.

8. Choose the Architecture, Then Secure the Supply Chain

Three decisions, in this order. Architecture is decided by I/O count, expansion plans, and downtime cost, not by which technology sounds more advanced. Brand is decided by your installed base and your field protocols, and commits you to a spare parts supply chain for a decade. Sourcing decides whether any of it goes live on schedule, because the best BOM in the world is worth nothing if the parts sit on a sixteen week lead time.

Chentuo stocks modular PLC hardware across Siemens, ABB, Allen-Bradley, Mitsubishi, Omron, and Schneider, including discontinued part numbers.

Request a quote → Send the brand, model number, quantity, application, and whether the line is down or you are stocking up. Urgent enquiries are answered within 24 hours.

Free consultation