The 50 questions below are the ones that keep coming up in UK PLC, controls and automation interviews, each with an answer short enough to say out loud. Treat the answers as a skeleton: interviewers want your own project, your own fault and your own fix wrapped around the textbook point.
Which platform to revise first? According to ITJobsWatch (permanent UK adverts for "PLC Engineer", six months to 21 October 2026), Siemens is named in 11.8% of the 144 adverts and Rockwell in 5.6%. ITJobsWatch is an IT-focused index with small samples, so read that as a direction rather than a market share. It does explain why the Siemens section is the longest here.
Section 1: PLC fundamentals (questions 1 to 12)
1. What is a PLC and how does it differ from a PC? A programmable logic controller is a ruggedised industrial computer built for deterministic, real-time control of machines and processes. It runs a cyclic scan, tolerates heat, vibration and electrical noise, has field I/O built in, and is designed to be fault-found by maintenance staff on site.
2. Describe the scan cycle. Inputs are copied into an input image, the user program executes, then the output image is written to the physical outputs, with communications and housekeeping around it. Because the program sees a frozen snapshot of inputs, a signal shorter than the scan time can be missed unless you use an interrupt or a fast input.
3. Which languages does IEC 61131-3 define? Ladder Diagram, Function Block Diagram, Structured Text, Sequential Function Chart and Instruction List. Instruction List is deprecated and rare in new work. According to the IEC webstore, the third edition (2013) has been superseded by a 2025 edition.
4. When would you choose Structured Text over ladder? Use Structured Text for arithmetic, arrays, loops, string handling and data manipulation. Use ladder for interlocks and Boolean logic that electricians must read at 3am. Many good programmes mix both, with each function block in the language that suits it.
5. What is the difference between a latched output and a seal-in? A seal-in uses the output's own contact in parallel with the start condition, so it drops out if power or logic is lost. A latch uses Set/Reset instructions and holds the state until reset, and may survive a power cycle if the memory is retentive. Say which one you would use and why.
6. Why are stop buttons wired normally closed? It is fail-safe: a broken wire or lost supply looks the same as a pressed stop, so the machine stops. The PLC logic then reads a healthy input as TRUE, which surprises beginners.
7. What is retentive memory? Memory that keeps its value through power-down or a STOP-to-RUN transition, such as recipe values, counters and machine state. Non-retentive memory resets. Overusing retentive bits can restart a machine into an unsafe state.
8. Explain sinking and sourcing inputs. A PNP sensor sources current to the input, so it needs an input module that sinks current to 0 V. An NPN sensor sinks current, so it needs a sourcing input module. Vendors describe the module side differently, so always check the manufacturer's wiring diagram.
9. How do you scale a 4 to 20 mA signal? Map the raw value linearly onto engineering units: engineering value = (raw - raw_min) / (raw_max - raw_min) x span + low limit. Siemens analogue modules typically present a nominal 0 to 27648 range. The 4 mA live zero lets you detect a broken wire as a reading below 4 mA.
10. What is edge detection and why use it? A rising or falling edge instruction fires for one scan when a signal changes, for example to count products or trigger a one-shot action from a held button. Without it, a held signal would repeat the action every scan.
11. Name the standard timers and counters. On-delay (TON), off-delay (TOF) and pulse (TP) timers, plus count-up and count-down counters. Mention the accumulated and preset values, and that timers are retentive or not depending on the platform and type.
12. How do a PLC and a DCS differ? A PLC is typically programmed per machine or line and favours fast discrete logic. A DCS is engineered as one integrated plant system, with a common database, operator environment and strong process-control and redundancy features. The line has blurred, so give the process-versus-discrete context. See our DCS versus PLC guide.
Section 2: Siemens and TIA Portal (questions 13 to 22)
13. What are OBs, FBs, FCs and DBs? An Organisation Block (OB) is called by the operating system, for example OB1 for the main cycle. A Function Block (FB) has memory held in an instance data block. A Function (FC) has no memory of its own. A Data Block (DB) stores data. Use FBs for anything with state, such as a motor or valve.
| Block | Called by | Keeps its own data? | Typical use |
|---|---|---|---|
| OB | Operating system | No (temp only) | Main cycle, startup, interrupts |
| FB | Your code | Yes, in an instance DB | Motor, valve, sequence |
| FC | Your code | No | Scaling, calculations |
| DB | Accessed by blocks | Yes | Recipes, shared data |
14. What is an instance data block, and what is a multi-instance? An instance DB holds the static data of one FB call. A multi-instance lets a parent FB contain the data of the child FBs it calls, which keeps the project tidy and avoids a separate DB per call.
15. What does "optimised block access" mean? The compiler decides the memory layout and you address by symbol, not by byte offset. It is faster and more robust, but you lose fixed absolute addresses, which matters when a third party reads the data over an old protocol.
16. How is a PROFINET device identified? By its device name, which the controller uses to assign the IP address at start-up, plus a GSD file that describes the device in the hardware configuration. A common fault is a replaced device with the wrong or blank name, so it never joins the network. See the PROFINET guide.
17. How does PROFINET compare with PROFIBUS DP? PROFIBUS DP is a serial fieldbus on RS-485 with speeds up to 12 Mbit/s. PROFINET runs on Ethernet and supports standard IT traffic alongside real-time data. According to PI (PROFIBUS and PROFINET International), PROFINET mechanisms can enable cycle times down to 31.25 microseconds.
18. How do you find out why a Siemens CPU faulted? Read the diagnostic buffer first, in the order the events happened, then check the module and device diagnostics. Organisation blocks such as OB82 (diagnostic error interrupt), OB83 (pull or plug of modules) and OB86 (rack or station failure) let you react in code instead of letting the CPU stop.
19. When do you use a cyclic interrupt OB? When code must run at a fixed interval independent of the main scan, for example a PID loop every 100 ms. The cyclic interrupt OBs start at OB30 and you set the interval in the OB properties.
20. What is OB100? The start-up OB, run once when the CPU goes from STOP to RUN. Use it to initialise values, reset sequences to a safe step and clear stale commands.
21. Compare the S7-1200 and the S7-1500. The S7-1200 is a compact controller for small machines. The S7-1500 is the higher-performance family with more integrated diagnostics, more memory and technology functions. Both are programmed in TIA Portal. Our S7-1200 versus S7-1500 guide goes deeper.
22. What can PLCSIM not prove? It simulates the CPU logic, so it validates sequences, interlocks and HMI links. It cannot prove real wiring, sensor behaviour, drive response or true timing. Say that you test in simulation first and still run a staged site test.
Section 3: Rockwell and Allen-Bradley (questions 23 to 27)
23. How do CompactLogix and ControlLogix differ? Both are programmed in Studio 5000 Logix Designer. ControlLogix is the larger chassis-based platform for bigger systems with redundancy options, and CompactLogix is the smaller, cost-optimised platform. MicroLogix is legacy and Micro800 is the small-machine family.
24. What are controller tags and program tags? Controller-scoped tags are visible to all programs, while program-scoped tags are local to one program. Prefer program scope for encapsulation, and use aliases to give I/O meaningful names.
25. Explain tasks: continuous, periodic and event. The continuous task runs whenever nothing of higher priority is running. A periodic task runs at a fixed interval, and an event task is triggered by an event. Priorities run 1 to 15, and a lower number means higher priority.
26. What is an Add-On Instruction? A user-defined, reusable instruction with its own parameters and local tags, comparable to a Siemens FB. It lets you version and reuse tested logic, such as a motor starter.
27. What is a UDT and why use one? A user-defined data type groups related tags into one structure, such as a valve with command, status, fault and timer members. It keeps tag naming consistent and makes HMI faceplates reusable. Our Studio 5000 guide covers the basics.
Section 4: SCADA and HMI (questions 28 to 32)
28. What is SCADA? Supervisory control and data acquisition: software that connects to many controllers, shows live graphics, logs history, manages alarms and lets operators send commands. The PLC still does the real-time control. See PLC versus SCADA.
29. What is a tag database? The list of every data point with its address, data type, scaling, engineering units, alarm limits and archive settings. Naming and structure decisions here determine how maintainable the system is.
30. How do you design alarms well? Every alarm must need an operator action, have a clear priority and be rationalised against the standard ISA-18.2 lifecycle. Avoid alarm floods by suppressing consequential alarms and removing nuisance alarms rather than silencing them.
31. What makes a good HMI screen? A clear hierarchy, muted backgrounds, colour used only to signal abnormal conditions, and trends beside numbers. ISA-101 is the HMI design standard to name. Our HMI design guide has examples.
32. How do you choose between WinCC and Ignition? Choose by installed base, licensing model, required scale and the team's skills. WinCC integrates tightly with TIA Portal projects, while Ignition is a general platform with server-based licensing. Be ready to say you would follow the site standard.
Section 5: Industrial communications (questions 33 to 37)
33. Compare Modbus TCP with PROFINET. Modbus TCP is simple, open and polled, so it is easy to integrate and troubleshoot but has no built-in device engineering or deterministic timing. PROFINET is configured from the hardware catalogue, supports real-time cyclic exchange and device diagnostics, and suits tightly coupled I/O.
34. What is EtherNet/IP? An industrial Ethernet protocol that carries the CIP protocol, dominant on Rockwell systems. Here "IP" stands for Industrial Protocol, not Internet Protocol.
35. What is OPC UA and why does it matter? A platform-independent, service-oriented architecture that replaces COM-based OPC Classic. According to the OPC Foundation it provides encryption, message signing and certificate-based authentication, which is why it is the usual route for exchanging data between PLCs, SCADA and IT systems. See OPC UA connectivity.
36. Managed or unmanaged switch on a plant network? Managed. VLANs, port mirroring for diagnostics, loop prevention, monitoring and port security are worth the extra cost, and unmanaged switches hide faults.
37. A device will not communicate. What do you check? Physical layer and link lights, then IP address, subnet mask and duplicate addresses, then the protocol-level identity (such as the PROFINET device name), then firewall or VLAN settings, then the configuration in the controller. Change one thing at a time and record it.
Section 6: Safety (questions 38 to 43)
38. What is a Safety Integrity Level? A discrete level from SIL 1 to SIL 4 defined in IEC 61508 for the risk reduction of a safety function. For low-demand operation, SIL 1 means a risk reduction factor of 10 to 100, SIL 2 100 to 1,000, SIL 3 1,000 to 10,000 and SIL 4 10,000 to 100,000.
39. What is a Performance Level? The measure used for machinery safety functions in ISO 13849-1, from PL a (lowest) to PL e (highest). It combines architecture category, diagnostic coverage, component reliability and common-cause failure measures.
40. What separates Category 3 from Category 4? In both, a single fault must not cause loss of the safety function. Category 4 additionally requires that single faults are detected at or before the next demand and that an accumulation of faults does not cause loss of the function, which demands higher diagnostic coverage.
41. What is a safety PLC? A controller certified for functional safety, with redundant processing, self-test and safe I/O, programmed with certified safety function blocks. The safety program is normally kept separate and protected by a signature.
42. Describe stop categories 0, 1 and 2. Category 0 removes power immediately. Category 1 is a controlled stop with power available, then power removed. Category 2 is a controlled stop with power maintained. These come from IEC 60204-1.
43. What must you do before working on an electrical or pneumatic machine? Follow the site's isolation and lock-out procedure: identify the energy sources, isolate, lock off, prove dead or depressurised, and only then work. See our functional safety guide.
Section 7: Practical and behavioural (questions 44 to 50)
44. Walk me through diagnosing an intermittent machine stop. Reproduce or timestamp it, check the alarm and diagnostic history, look at what changed, isolate by splitting the problem in half, then fix the cause and add a diagnostic so it is visible next time. Use the STAR structure (situation, task, action, result). Our troubleshooting guide has a method.
45. How do you approach an unfamiliar PLC programme? Read the functional specification and I/O list, take a backup, go online in a safe state, read the alarm and diagnostic history, trace one sequence end to end, and speak to the operators about known problems before changing anything.
46. How do you make an online change safely? Agree it with production, check the safety implications, back up first, make the smallest change, test it, record it and keep a rollback. Never force outputs on a machine others may be near.
47. Tell me about a time you disagreed with a technical decision. Pick a real example where you listened, brought evidence and accepted the outcome. The interviewer is judging collaboration, not who won.
48. How would you test a new machine before it ships? Check the I/O one point at a time against the drawings, test interlocks and safety functions, run each sequence in manual then auto, test fault and recovery cases, and record results in a factory acceptance test sheet.
49. How do you keep up to date? Name specific things you do: vendor documentation and release notes, free courses, simulators and a home project. Our free lab simulators give you somewhere to practise.
50. Why do you want this job? Connect the employer's industry, technology and projects to something you have actually done or built. Specifics beat enthusiasm.
How should you practise answering?
Practise out loud against a timer, with one real example for each of fault-finding, commissioning and teamwork. The National Careers Service advises reading the job description closely and preparing examples from past experience in advance. For a walk-through of the format, this video covers common automation interview questions.
How EDWartens can help
The Professional Module (£2,140 + VAT = £2,568, five-day live online class plus self-paced Basic Engineering) includes a one-to-one resume session, 7 role-specific CV templates and interview-preparation sessions. There is no job guarantee. For free practice, use the browser simulators and see /courses/professional or /signup.
