Skip to content
Tech Interview Prep
Technical interview guide

Control Systems & PID

The feedback loop that keeps a robot at a target state: proportional, integral, and derivative control, and why all three matter.

Read
45 min
Practice MCQs
25
Interview Q&A
25
Edition
v4
Editorial status
Reviewed
Relevant for
Robotics Engineer

Scope: Current MathWorks, NI, Rockwell Automation, Arduino PID, python-control, SciPy, NIST SP 800-82 Rev. 3, IEC functional-safety, and OSHA guidance reviewed 2026-09-04.

Curated: · Written: · Reviewed:

25 questions

MCQ-2A PID loop's setpoint steps from 60 °C to 70 °C. At the next scan the measurement reads 62 °C, so the error is 8 °C. Which quantity does the proportional term's contribution depend on at that scan?
MCQ-3A flow loop uses a PI controller in the form u = Kp·e + Ki·∫e dt, with Kp = 2 %/(L/min) and Ki = 0.4 %/(L/min·s). A valve fault holds the flow 5 L/min below setpoint for 15 s. What is the integral term's contribution to the controller output at the end of those 15 s?
MCQ-16A 5.00 V step drives a closed position loop. The output crosses 0.50 V at 4 ms and 4.50 V at 12 ms, peaks at 5.62 V at 21 ms, and stays within ±2% of 5.00 V from 38 ms onward, coming to rest at 4.82 V. The team has to report the loop's final-value accuracy as a single number. Which metric is the 0.18 V difference between the command and the resting output?
MCQ-9A loop was tuned on a legacy controller that implements the interacting (series) form C(s) = Kc(1 + 1/(Ti s))(1 + Td s), with Kc = 2, Ti = 4 s and Td = 1 s. The team is moving it to a library that implements the parallel form u = Kp·e + Ki∫e dt + Kd·de/dt, where Ki is in 1/s and Kd in seconds. Which gain triple reproduces the tuned loop's behaviour?
MCQ-4A flow loop uses a PID controller in the form u = Kc(e + (1/Ti)∫e dt + Td de/dt) with Kc = 1 and Td = 3 s. The flow transmitter carries a 15 Hz ripple of ±0.2 % of span, and the derivative term is unfiltered. What is the peak amplitude of the derivative term's contribution to the controller output?
MCQ-5A PID temperature loop holds 60 °C with Kp = 2 and Kd = 8 (parallel form, output in %), scanning once per second. An operator steps the setpoint to 80 °C; the measured temperature is still 60 °C. The controller output pins at full scale for a single scan and trips an over-temperature alarm. Which term produced the jump?
MCQ-8A PID loop is tuned and validated at a 100 ms control tick, then moved to a 50 ms tick with the gains left unchanged. What happens to the loop?
MCQ-1A server room's cooling loop holds the cold-aisle temperature at 22 °C using a PID controller that reads a supply-air thermistor and drives a chilled-water valve. During a firmware update the thermistor reading is frozen at its last value (22.0 °C) while the valve keeps receiving controller output. The room's IT load then rises by 30 kW. Two hours later the cold-aisle temperature is 31 °C and the valve is still at its pre-update position. Which statement best explains why the loop failed to correct the temperature?
MCQ-24A team tunes a PID position loop on a robot arm joint. After raising Kp from 2.0 to 6.0 and Ki from 0.5 to 2.0, the joint reaches the setpoint faster, but the motor current now spikes to its 8 A limit on every step command and the arm audibly buzzes at roughly 40 Hz. The team wants one measurement that will show whether the new gains are demanding more actuator effort than the old ones. Which measurement should they compare?
MCQ-6A hot-water reheat coil heats a zone with a setpoint of 22.0 C. The controller output, 0-100%, drives the coil valve, and the loop has been in automatic for 20 minutes. The zone reads 19.5 C and the output has sat at 0% for the last 6 minutes. The engineer switches to manual and holds 60%: the zone climbs to 24.5 C and holds; at 0% it drops to 19.0 C. Back in automatic, the output again falls to 0% within a minute. What is the fault?
MCQ-12A flow loop commands a modulating valve with a 0–100% signal. The valve's actuator mechanically stops at 70%; commands above 70% change nothing. The controller's output limits are set to 0–100%. After a large setpoint step, the controller output sits at 100% for 30 s while flow stays below setpoint. When the setpoint drops, flow overshoots by 25% and takes 40 s to settle. Which explanation accounts for both the 30 s at 100% and the 25% overshoot?
MCQ-10A PI controller drives a furnace damper whose command is clamped to 0-100%. The loop runs at 100 ms. During a 40 s startup the measured temperature sits 60 degC below setpoint, so the integral term grows every cycle; the damper stays pinned at 100%. At t = 40 s the temperature crosses setpoint and the error changes sign. The damper does not begin closing until roughly t = 52 s, and the temperature then overshoots setpoint by 25 degC before settling. Which change removes the delayed closing and the overshoot?
MCQ-11A PI flow controller on a 6 in line runs in AUTO with Kc = 1.2 and Ti = 8 s, holding the valve at 47% to keep flow at 200 L/min. The operator switches to MANUAL, and over the next 90 s the valve is walked to 62% to chase a downstream demand; the measured flow settles at 240 L/min. The operator then switches back to AUTO with the same tuning and the same 200 L/min setpoint. On the first scan after the switch, the controller output jumps to 71% and the flow overshoots to 265 L/min. Which change removes that jump while keeping the same tuning and setpoint?
MCQ-25A joint runs a PID position loop (Kp = 8, Ki = 40, Kd = 0.1, derivative on measurement) at 1 kHz with a ±10 V output limit. In simulation a 30° step settles in 120 ms with 4% overshoot. On the bench, commanding a 30° step while the joint rests against its hard stop drives the output to +10 V; when the command reverses to −30°, the joint stays pinned against the stop for about 2 s before moving. What is the cause?
MCQ-20A gas pressure loop runs PI control on a 100 ms scan with Kp = 5 %/bar and Ki = 1.5 %/(bar·s), setpoint 40.0 bar, output range 0–100%. The transmitter publishes on a fieldbus every 100 ms. The last sample before a 6.0 s fieldbus outage reads 38.0 bar, and the integral term is 30.0% when the outage begins; no stale-sample handling is configured. The bus recovers with the line at 39.0 bar. What is the controller output at the first scan after the bus recovers?
MCQ-22A temperature loop runs a PID at 100 ms with gains Kp=2.0, Ki=0.5, Kd=0.1, output clamped to 0-100%. Operators report the heater "sometimes slams to 100% for a minute, then sits at 40% with the temperature 3 °C below setpoint." The controller logs only the final output value. Which single change to the logged telemetry would let you tell whether the 100% period is integral windup during saturation or a genuinely large proportional term?
MCQ-7A servo velocity loop must cross over at 40 Hz to meet the machine's spec, and the analog design already meets the 45-degree phase-margin target there. The team budget for the digital path is at most 10 degrees of extra phase lag at 40 Hz. They plan a 2 ms sample period, with negligible jitter and computation delay, and model the sample-and-hold as an average delay of T/2. What should they do?
MCQ-17A position loop's open-loop gain crosses 0 dB at 20 rad/s, where the measured phase is −158°. The team plans to add a digital filter that inserts pure transport delay in the loop. What is the largest additional delay the loop can tolerate before it reaches the stability boundary?
MCQ-15A flow loop has a 6 s transport delay between the control valve and the downstream flowmeter. A step test on the loop gives an ultimate gain Ku = 4.0 and an ultimate period Pu = 24 s. The loop runs with a PI controller at Kc = 1.6 and Ti = 20 s, and a setpoint change produces a 24 s oscillation that decays over a few cycles. A colleague proposes doubling Kc to 3.2 to make the loop 'respond faster to the flowmeter.' What is the most likely outcome?
MCQ-14A distillation column has a tray-temperature loop (outer) trimming the setpoint of a reflux-flow loop (inner). The reflux valve sticks, so a 5% step in reflux flow takes about 40 s to settle; the tray temperature responds to a reflux change with a time constant near 8 min. The team is deciding whether to keep the cascade or run tray temperature directly on the reflux valve. What does the timing say about the cascade?
MCQ-13A furnace's fuel valve is positioned by a feedforward path u_ff = 2·F (u in %, F in kg/s) plus a PI trim on outlet temperature. Over the operating range the plant's true steady-state fuel requirement is linear in F: at F = 20 kg/s the outlet holds setpoint with the valve at 45%, and at F = 30 kg/s it holds setpoint with the valve at 75%. Feed flow steps from 20 to 25 kg/s and the outlet settles back to setpoint. What is the final valve position?
MCQ-18A level loop on a 4-20 mA DP transmitter is being auto-tuned with a relay-feedback (Åström–Hägglund) test. The vessel must stay between 20% and 80%; the inlet valve is air-to-open and fails closed on loss of air. The engineer runs the test with the controller in MANUAL, driving the valve output as a square wave of ±5% around 50% with a 30 s period, and the level swings between 42% and 58%. The test completes and returns Kp = 3.1, Ti = 45 s. What should the engineer do next?
MCQ-19A gas turbine's fuel-control loop runs one PID tune. At 100% power it settles in 0.5 s with 4% overshoot; at 30% power the same gains give a sustained 6 Hz limit cycle and 25% overshoot. The spec is settling under 0.7 s at both points, and power ramps continuously between them. Available signals: exhaust gas temperature (EGT), sampled at 20 Hz, tracking power with a 0.1 s lag; and compressor inlet temperature, sampled at 1 Hz, flat across the power range. A tune that damps the 30% case stretches 100% settling to 1.3 s. What should the engineer do?
MCQ-23An exothermic batch reactor runs one temperature loop: TT-101 feeds a PID in the BPCS PLC that throttles cooling-water valve FCV-101, and the same PLC drives FCV-101 fully open at 130 C as an over-temperature trip. A stuck-open feed valve can start a runaway, and the LOPA for that scenario sets a target PFDavg of 1e-2 for the trip. The team's reliability calculation, covering TT-101, the PLC and FCV-101 over a one-year proof-test interval, gives 6e-3, and they ask to credit the loop as the reactor's over-temperature safety function. What should the reviewer conclude?
MCQ-21A pen tester on a plant network captures one legitimate setpoint write: target register 40001, value 72.5: sent from the SCADA master to a compressor RTU, and replays the identical packet 25 minutes later from the same subnet. The RTU accepts it and the compressor setpoint changes. The RTU already authenticates the operator's credentials and checks that the source IP is the SCADA master's. Which command-format change closes this replay path?