A pump station needs to maintain 4 bar discharge pressure. The PLC can read the pressure transmitter, run a PID instruction, and send a speed reference to the VFD over a fieldbus — a round trip that takes two communication cycles, one PLC scan, and the fieldbus update time, typically 15–40 ms total. Or the drive can read the pressure transmitter directly through its own analog input and run the PID loop internally, updating the motor speed in under 2 ms with no fieldbus delay and no dependency on PLC scan time. Both approaches work. The decision of where to close the loop affects more than response time: it changes how the system is commissioned, how fault handling is structured, and which device owns the process setpoint. This article compares the two architectures and identifies when each makes engineering sense.
The Physical Difference: Where the Feedback Wire Lands
In a PLC-based PID architecture, the process transmitter is wired to a PLC analog input card. The PLC reads the value, executes the PID algorithm during its cyclic scan, and writes the resulting command — typically a 4–20 mA speed reference or a fieldbus speed setpoint — to the drive. The drive executes the speed command but has no knowledge of the process variable. If the pressure transmitter fails, the drive continues running at the last commanded speed because it does not know the transmitter exists.
In a drive-based PID architecture, the process transmitter is wired directly to the drive's analog input or connected via its fieldbus interface. The Siemens SINAMICS G115D distributed drive executes the PID loop in its internal firmware at the drive's control-loop rate — typically 1–2 ms — and directly controls the motor speed. The PLC may still read the process variable and drive status for monitoring and supervisory control, but it is not in the closed-loop control path. If the PLC goes offline, the drive continues regulating pressure autonomously.
When Drive-Level PID Is the Better Choice
Drive-level PID excels in single-loop applications where the process variable directly determines motor speed — pump pressure control, fan flow control, conveyor speed matching, simple level regulation. These applications share three characteristics: a single-input, single-output control structure; a process response time slow enough (>100 ms) that 1–2 ms loop update provides ample control bandwidth; and a need for the loop to continue operating during PLC maintenance or communication interruptions.
The commissioning advantage is also significant. A drive-level PID loop is configured through the drive's keypad or commissioning software — parameters for setpoint source, PID gains, ramp rates, and fault response are set in the drive parameter table. There is no PLC ladder logic to write, no PID block to configure, and no communication watchdog to program. For machine builders shipping standardized equipment — a pump skid, a cooling tower fan controller, an air compressor — this means the control functionality ships with the drive, tested at the factory, with no dependency on the customer's PLC platform or programming standard.
When PLC PID Remains the Right Architecture
PLC-based PID becomes necessary when the control strategy involves multiple inputs or coordinated outputs — cascade control, ratio control, feedforward compensation, or split-range control with two valves on a single controller output. A boiler drum level controller using three-element control (level + steam flow + feedwater flow) cannot be implemented as a drive-level PID because the loop requires three process inputs and a feedforward calculation that exceeds the drive's internal function-block capability. Any loop requiring operator-accessible setpoint entry through the HMI with ramp-soak profiling or batch recipe integration similarly benefits from the PLC's native connection to the HMI tag database.
PLC-based PID also provides superior diagnostic visibility. When a PLC PID loop oscillates, the engineer can trend the setpoint, process variable, error, P/I/D components individually, and controller output — all from the SCADA historian. When a drive-level PID loop oscillates, the engineer must connect to the drive's diagnostic parameters, which typically provide less granular trending and are not integrated into the plant historian. For processes where control loop performance is a regulatory or quality requirement — pharmaceutical freeze-drying, food extrusion, chemical reactor temperature control — the diagnostic advantage of PLC-based PID tips the scale. Browse variable frequency drives for models with integrated PID and see our process controllers for dedicated loop control alternatives.
Hybrid Architecture: PLC Setpoint, Drive Execution
A practical middle ground sends the setpoint from the PLC to the drive via fieldbus but keeps the PID execution in the drive. The PLC computes or receives the setpoint — from an HMI entry, a recipe step, or a supervisory optimization algorithm — and writes it to the drive as a parameter. The drive executes the PID loop locally, using its own analog input for the process variable. This gives the PLC ownership of the setpoint strategy while preserving the drive's fast loop execution and autonomous fault behavior. If communication is lost, the drive holds the last received setpoint and continues regulating — a defined, safe behavior that avoids the complexity of a communication-loss shutdown sequence in the PLC.
Closing the loop at the drive is not about saving a few milliseconds of response time — it is about making the machine module self-contained. When the pump skid leaves the factory with its PID loop configured, tested, and running in the drive, the system integrator connects power, a run-permit signal, and a setpoint — and walks away.
