Comparison
How ControlLogix and S7-400 PLC architectures differ
Short answer
A ControlLogix application maps onto an S7-400 only at the requirements level. ControlLogix runs tasks that contain programs and routines on symbolic tags; the S7-400 runs organization blocks that call FBs and FCs against data blocks, mixing symbolic and absolute addressing. Match execution behavior, data ownership and I/O timing — never translate instruction names.
On this page
A ControlLogix application does not translate instruction-for-instruction into an S7-400 project, because the two platforms organize execution before any instruction runs. ControlLogix schedules work in tasks that hold programs and routines and keeps every value in a named tag; the S7-400 hands control to organization blocks that call function blocks and functions against data blocks, with symbolic and absolute addressing mixed in the same project. Compare the two architectures on execution, code units, data and I/O timing and the mapping is concrete enough to commit numbers to; compare them instruction by instruction and it collapses.
Both are cyclic controllers; the scheduling layer is different #
Both controllers run logic in a cycle and both step out of it for timed or event work. The difference is where that scheduling is declared and what it is called.
A Logix 5000 project is a set of tasks. Each task holds programs, each program holds routines in ladder, structured text, function block or SFC, and each program enters through its main routine. That program-and-routine hierarchy rewards deliberate structure, and the best practices for organizing Logix 5000 programs and routines are worth reading before your layout sets. One continuous task runs at the lowest priority and yields to every periodic and event task, which use sixteen priority levels with 0 the highest. A periodic task rate can be set anywhere from 1 ms to 10 s, and every task carries a watchdog that faults it if a pass overruns.
The S7-400 declares the same ideas as organization blocks, the blocks through which the CPU operating system calls your program. OB1 is the main cycle. Startup OBs such as OB100 run once at power-up. OB35 is a cyclic-interrupt OB the operating system calls at an interval you set in HW Config, and OB40 handles hardware interrupts from the modules. Around these sit the error OBs: OB80 for time faults, OB82 for diagnostic interrupts, OB86 for station failures, OB121 and OB122 for programming and I/O access errors. If such a fault occurs and its OB is not loaded, the CPU goes to STOP.
The mapping, line by line #
Work down a machine's requirements, not the instruction set:
| What the machine needs | ControlLogix | S7-400 | What you do in a port |
|---|---|---|---|
| Main cycle | One continuous task | OB1 | Split the old programs into FBs and FCs called from OB1 |
| Fixed-rate loop | Periodic task, 1 ms to 10 s | Cyclic-interrupt OB such as OB35, interval in HW Config | Re-enter the interval in HW Config and load OB80 |
| Event reaction | Event task | Hardware-interrupt OB40 | Re-map each trigger to a module or process event |
| Startup behavior | Prescan plus first-scan logic | Startup OBs such as OB100 | Move first-pass initialization into the startup OB |
| Code unit | Program with routines | FBs and FCs called from OBs | Give each FB an instance DB |
| Data | Tags, controller- or program-scoped, UDTs | Global DBs plus instance DBs | Assign an owner to every tag |
| I/O sampling | RPI connection per module | Process image around OB1; partitions available for timed OBs | Assign each input to an image and set its allowed staleness |
| Fault handling | Task watchdog; controller major faults | CPU STOP when a required error OB is missing | Load OB80, OB82, OB86, OB121, OB122 even if they only log |
Worked example: one 10 ms loop on both platforms #
A feed drive needs a positioning loop at a fixed 10 ms, sequence logic at cycle speed, and HMI data.
ControlLogix: create a periodic task at that interval, priority 10, and put the loop routine there; the sequence stays in the continuous task. Assume the loop body measures 2.5 ms worst case. Each pass then uses 2.5 / 10 = 25% of its interval and leaves a 10 − 2.5 = 7.5 ms margin. The task runs 1000 ÷ 10 = 100 times per second and consumes 100 × 2.5 ms = 250 ms of every second, so the continuous task can absorb at most 1000 − 250 = 750 ms of work per second before its scan balloons. A 2.5 ms pass never approaches the default 500 ms task watchdog.
S7-400: call a positioning FB, with its instance DB, from OB35, and set OB35's interval to the same value in HW Config. OB35 leaves the factory with a common default of 100 ms that varies by CPU model, and this edit is the step people miss; leave it and the loop runs ten times too slow with no error raised. The arithmetic is identical: that interval gives 100 calls per second, and a 2.5 ms body is 25% duty with 7.5 ms of margin. The failure mode differs: if a pass ever exceeds the interval, the next call arrives while the last is still running and the CPU calls OB80, the time-error OB. Load OB80 even if it only stamps a diagnostic, because without it the same overrun stops the CPU.
The port decision: same interval, same measured body, same duty — different failure behavior. On the Logix side a late pass faults that task while the rest of the controller keeps running; on the S7-400 side a late pass is a time error you must have provided for. Set the interval deliberately on both platforms and budget from the measured 2.5 ms, never from the name of the block.
Data: named tags versus data blocks with addresses #
On the ControlLogix side you never manage addresses: logic reads and writes tags, each tag is either controller-scoped or program-scoped, and structure comes from user-defined data types. Scope is the access rule, and renaming a tag is a compile-time operation.
The S7-400 project is different in kind. Function block data lives in instance DBs — one per call unless you multi-instance — free data lives in global DBs, and the same operand can be reached by symbol or by absolute address. A contact can use the symbol Fwd_PB or the absolute spelling behind it, for example DB10.DBX 2.0 in global DB 10; those are two views of one location until someone edits the DB layout, and then the absolute spelling can silently point somewhere else.
So the port rule is ownership. A controller-scoped tag becomes a global DB element or a static variable in an FB's instance DB; a program-scoped tag becomes data owned by that program's FB. Decide each one before writing logic: the S7-400 gives you more ways to be inconsistent than the ControlLogix does, and it will compile either way.
I/O timing: RPI connections versus the process image #
ControlLogix I/O lives on connections: each module has a requested packet interval you set per module, and input data arrives asynchronously to whatever routine is running — a task reads the latest packet, whenever it landed. Timing control is per module, not per routine.
The S7-400 samples differently. OB1 works on the process image, which the operating system updates at the start and end of each cycle; the timed OBs can be assigned their own process-image partitions on the S7-400, so the fast loop reads a snapshot taken for it rather than reaching into OB1's image; and a direct peripheral access bypasses both, reading the module at the moment the instruction executes.
Port decision, per input: a value the main sequence consumes goes in OB1's image; a value the fast loop consumes goes in that OB's partition; a value that must be sampled at instruction time, such as a fast event bit, uses direct peripheral access. Staleness is a design number on the S7-400; on the ControlLogix it is an RPI.
The decision for each job #
Maintaining an S7-400 plant as a ControlLogix person: read the project as OBs and DBs, not as tasks. Spend the first hour on OB1, the cyclic-interrupt OBs and the symbol table, because the instance DB list tells you where the data actually lives. Budget the software too: an S7-400 project opens in the STEP 7 classic line, and which STEP 7 a project needs is worth settling before you install anything, while a ControlLogix 5571 opens only in Studio 5000 — see why only Studio 5000 programs a ControlLogix 5571 before planning offline work.
Converting ControlLogix logic into an S7-400 project: write the requirement list first — every loop's rate, every event's trigger, every tag's owner, every fault's expected reaction — then apply the table above. There is no instruction-level translation, so port behavior the way you would port a PLC-5 filter routine to structured text: rebuild it, then measure the timing on the target and hold the result to the numbers in your specification.
Choosing between the two for a new machine: both architectures do the same machine work, and the axes above decide engineering effort, not capability. Those same axes decide the wider platform question, so see how PLC and PAC architectures actually differ before settling on a class, not just a brand.
Whatever the path: moving a loop onto a new controller re-creates its fault behavior, so reprove the interlocks and the emergency-stop chain on the target before the machine runs.
Questions asked
Does the S7-400 have anything like produced/consumed tags?
No direct equivalent exists. The S7-400 exchanges periodic data through shared data blocks over configured communication links, through S7 communication functions such as PUT and GET, or through PROFIBUS and PROFINET application relationships. When porting, replace each produced/consumed tag pair with a global DB the sender writes and the receiver reads, and set the update rate from the link configuration instead of the RPI you used before.
Which software opens and edits an S7-400 project?
The STEP 7 classic line (SIMATIC Manager) up to V5.x. TIA Portal does not program standard S7-400 CPUs, so the engineering station must run a STEP 7 V5.x release compatible with the project. Older projects can require a migration step inside the V5.x line before they open cleanly, so match versions before editing rather than after a failed download.
Is PCS 7 the Siemens equivalent of ControlLogix?
Not at the controller level. PCS 7 is a process control system built on S7-400-class controllers, so comparing ControlLogix with PCS 7 compares a controller platform against a system layer. Compare ControlLogix with the S7-400 first, and bring PCS 7 into the discussion only when the job really is a process-control system rather than a machine controller.
What happens on an S7-400 when a fault OB is missing?
The CPU goes to STOP for most asynchronous faults. If a programming error (OB121) or an I/O access error (OB122) occurs with the matching OB not loaded, the S7-400 stops rather than skipping the failed call. A ControlLogix task fault does not stop the other tasks the same way, so a ported application should have those OBs defined even if they only log the event.