How-to
How to compare two CLICK PLC programs for differences
Short answer
CLICK Programming Software has no built-in project compare, and the vendor has said none exists in the current release. The working method is to upload each machine's running program as a dated baseline, compare hardware, field assignments, configuration and logic in six passes, classify every mismatch, and load one merged program only after every difference has an owner.
On this page
- What the software gives you
- Quick fixes that lose information
- Capture each machine as its own baseline
- Match hardware and field assignments
- Compare the logic in six passes
- Classify every mismatch
- Worked example: two presses, fourteen differences
- Merge one program under change control
- Load one machine first and prove it
- The decision for each case
- Questions asked
Not in the software. CLICK Programming Software has no project-compare function, and AutomationDirect has said so plainly: the current release offers no way to compare two CLICK projects, and the request has been passed to the product team. That does not stop you from getting what you actually want, which is one program both machines can run. The method is concrete: connect to each PLC, use the software's read-from-controller function to upload the running program, and save each capture as a dated baseline file like Press1_2026-06-12. Then verify that hardware and field assignments agree, walk the logic in six passes, classify every difference into keep, share or unresolved, and merge only what you have reviewed. The software is a free download, so start from the current release and let discipline replace the missing tool.
What the software gives you #
CLICK Programming Software is free from AutomationDirect, so there is no license barrier to putting the current release on your engineering station. What that package does not contain, in the vendor's own answer, is any function that puts two projects side by side and reports their differences. That means three things in practice: two project files with different names can hold identical logic; two files with matching names can hold different revisions; and nothing in the tool will ever tell you which is which. The comparison is a review you perform, and the quality of the answer equals the quality of your records.
Quick fixes that lose information #
| Quick fix | Why it fails | What to do instead |
|---|---|---|
| Judge by filename | Names carry no logic; identical names can hide different revisions | Compare the captured controller programs |
| Download one file into both PLCs | Machine-specific settings may be intentional | Classify every mismatch before standardizing |
| Change several items at once | The cause of a new fault becomes untraceable | Apply one reviewed change set and test it |
A trial download is not a comparison method: it replaces the running baseline before you understand it, and if the machine then misbehaves you no longer know what you had.
Capture each machine as its own baseline #
Start from the program each controller is actually running, not from an old maintenance copy with a plausible name.
- Put each machine in the operating state your site procedure requires before a connection, and do not bypass guards, interlocks or safety functions to make the job easier.
- Connect to the first PLC and use the software's read-from-controller function; confirm the transfer direction shows the PLC as the source before you confirm.
- Save the capture with the machine tag and the date in ISO 8601 order,
Press1_2026-06-12. ISO 8601 writes the date year first, then month, then day, so the files sort chronologically in any folder listing. - Repeat on the second PLC with its own filename and its own folder.
- Set the two originally supplied files aside unchanged; they are documents under review, not baselines.
- Record which PLC each capture came from, who captured it, and whether that machine was running correctly at capture time.
If an upload cannot complete without changing something in the controller, stop and confirm the operation through the manufacturer's support channel instead of improvising.
Match hardware and field assignments #
Two mechanically identical machines can carry different electrical revisions, and logic that looks wrong in one program is sometimes correct for its wiring. Establish function, not just equal point counts.
| Item | What you compare | Failure if copied wrong |
|---|---|---|
| Inputs | Address, device function, normal state, polarity | A permissive or fault evaluates backward |
| Outputs | Address, connected load, commanded state | The wrong valve, contactor or indicator operates |
| Analog signals | Purpose, engineering range, scaling, alarm limits | Control and displayed values go wrong |
| Communications | Device, endpoint settings, data mapping, message purpose | The PLC reads or writes the wrong data |
| Stored values | Initial values, recipes, calibration data | A common program starts with the wrong machine data |
Resolve every discrepancy against the electrical drawings and the wiring you can physically see; do not edit a drawing or a program just to make the two agree. Analog and operator-panel data differences read the same whether you find them in ladder or over a network session, as in Reading a CLICK PLC float over Modbus TCP in an HMI.
Compare the logic in six passes #
Reviewing everything at once guarantees you miss the small changes that matter. Work the two saved projects in this order, one pass at a time, writing each difference into a worksheet as you go:
- Controller and communication configuration.
- Program organization and execution order, so corresponding logic lines up before you judge single instructions.
- Start, stop, mode-select, permissive and fault-reset logic.
- Every output path, from its command conditions through its interlocks.
- Sequence transitions: timers, counters, constants, math, scaling, alarm thresholds and data movement.
- Operator-panel command and status mapping.
Proving a program you did not write is the same discipline whether the previous author was a contractor, a coworker or a code generator; How to verify AI-generated PLC code before you trust it lays out the same evidence-before-trust structure.
Classify every mismatch #
Record each difference in a worksheet with five columns: location, machine A value, machine B value, operational effect, disposition. Every row gets one of three dispositions:
- Intentional machine difference, traced to wiring, calibration or a documented operating requirement: keep it, per machine.
- Correction that should be shared: apply it to the common program.
- Unresolved: it blocks the merge.
A different constant is as often tuning or calibration as defect, and a different contact condition is as often wiring as coding style. The rule is hard: any unresolved row blocks the merge, because a merge with an unexplained difference is a guess wearing a filename.
Worked example: two presses, fourteen differences #
This is an illustrative example with invented numbers to show the arithmetic; it is not a measurement. Two presses with the same mechanics are captured as Press1_2026-06-12 and Press2_2026-06-12. The field pass counts 13 inputs and 9 outputs on each machine, so the point counts agree. The six logic passes find 14 differences:
- 6 corrections that exist only in Press2: one fault-reset interlock plus five timer preset changes.
- 2 machine-specific clamp setpoints: Press1 holds 4.2 bar and Press2 holds 5.0 bar, a gap of 5.0 − 4.2 = 0.8 bar, traced on the drawings to a heavier part number.
- 6 rung-comment and rung-order differences with no operational effect.
Check the total: 6 + 2 + 6 = 14 differences, of which 0 are unresolved. Decision: merge approved. Press2 becomes the common program because it already carries the six corrections; the 0.8 bar difference leaves the logic and becomes per-machine data at 4.2 bar and 5.0 bar; the comments are reconciled during the merge. One program, two data rows.
Merge one program under change control #
Choose the baseline that already carries the corrections, then bring across only reviewed changes, recording for each one what moved, from where, and why. Keep machine-specific values as data, not as buried logic. Do not mix edits from both files in one uncontrolled online session. If a machine must return to production first, restore that machine's own captured baseline and standardize later: a machine that runs is not proof that the two applications are interchangeable.
Load one machine first and prove it #
- Schedule the change under the site's PLC modification procedure.
- Clear any forces and temporary test conditions before you connect.
- Confirm the target controller and the transfer direction at the final prompt.
- Load machine 1, then check for configuration and communication faults at startup.
- Use Data View, the CLICK tool that shows live values while the project runs, to confirm each input indicates in both states and each output commands its device.
- Run one full production cycle on machine 1 before touching machine 2, then repeat the same checks there.
- Keep both dated baselines available for rollback.
The decision for each case #
- The six passes find no differences: load the reviewed common program to both machines and keep a single baseline going forward.
- Every difference is classified: merge it, share the corrections, keep machine values as data, and load one machine first.
- Any difference stays unresolved: each machine keeps running its own captured baseline, and standardization waits for the answer rather than the deadline.
- A supplied file's origin is unknown: it is reference material only, never a download.
Questions asked
Is there a paid or newer edition of CLICK software with a project compare tool?
No. CLICK Programming Software is one free package from AutomationDirect, and the company has stated that the current software offers no way to compare projects, with the feature request passed to its product team. Install the current release from the manufacturer before you start, since it costs nothing, then run the manual passes described here.
Can I diff the two project files with a text-comparison utility?
Not usefully. The project files are meant to be opened in CLICK software, and there is no documented text export of the ladder logic for an external diff tool. A byte-level comparison only confirms the files differ, which you already know. The comparison has to happen as a structured review of both projects inside the software, pass by pass, with every difference written down.
How do I prove which file belongs to which machine when the filenames disagree?
Upload. The program inside each controller is the truth, so read it from each PLC and save it under the machine tag and the capture date. Any supplied file whose origin you cannot prove stays a reference document. Never download it to a machine to find out where it belongs, because a wrong download replaces the running baseline you are trying to identify.
Where does Data View fit in a comparison?
Data View is the CLICK software window that shows live values while the controller runs, and it verifies behavior rather than code. With both machines idle, watch the same inputs and outputs in Data View on each one: matching normal states confirm the field side before you start comparing ladder. It will not find a rung difference; the six logic passes do that.