PIDFF (PID with Feedforward)

The PIDFF function block in Schneider Electric EcoStruxure Control Expert (formerly Unity Pro) is the most feature-complete PID controller within the CONT_CTL library, available across:

  • Modicon Quantum
  • Premium
  • M340
  • M580

The “FF” stands for Feedforward, enabling the controller to:

→ compensate for measurable disturbances before they impact the process

Unlike simpler blocks (PID, PI_B, PID1), PIDFF supports:

  • Feedforward compensation
  • Cascade control
  • Bumpless transfer
  • Advanced mode management

It is widely used in complex industrial process control applications.


PIDFF Architecture & Control Strategy

Control Output

Output = PID(PV, SP) + FF(Disturbance)

Key Signals

  • PV → Process Variable
  • SP → Setpoint
  • FF → Measured disturbance input
  • OUT → Controller output
  • OUTD → Output tracking / display
  • MAN_AUTO → Manual / Auto selection
  • TR_S / TRI / TRS → Tracking & cascade inputs
  • RCPY → Integral preloading (startup conditioning)
  • PARA → Full configuration structure (Para_PIDFF)
  • INFO / STATUS → Diagnostics and error handling

When Should You Use PIDFF?

Use PIDFF When:

  • Disturbance is measurable and reliable
  • Fast compensation is required
  • Process interaction exists (e.g. cascade control)
  • High performance and stability are critical

Avoid PIDFF When:

  • Disturbance is not measurable
  • Signal is noisy or delayed
  • Process is simple and stable
  • Dead time dominates the system

In such cases, a well-tuned standard PID is often more robust.


Feedforward Engineering — What Actually Matters

How It Works

  • Disturbance signal is scaled
  • Converted into output contribution
  • Added directly to controller output

Key Parameters (Para_PIDFF)

  • ff_inf / ff_sup → disturbance input scaling range
  • otff_inf / otff_sup → output contribution mapping

Critical Field Gotcha

If:

ff_inf = ff_sup

Feedforward is completely disabled.


Engineering Reality

Feedforward must be:

  • Gain-matched to process
  • Time-aligned with process dynamics
  • Noise-filtered

Otherwise it causes:

  • Overcompensation
  • Oscillations
  • Output saturation

Why PIDFF Loops Fail in Real Plants

Common Failure Modes

  • Incorrect feedforward scaling
  • PID tuned without feedforward active
  • Output saturation (PID + FF exceeds limits)
  • Anti-windup not configured properly
  • Incorrect control direction (direct vs reverse)
  • Noisy or delayed feedforward signal

Result:

  • Oscillations
  • Slow response
  • Unstable operation

The Para_PIDFF Structure (Critical Configuration)

The Para_PIDFF DDT defines all behaviour of the controller.

Scaling Parameters

  • pv_inf / pv_sup → PV range (must not be equal)
  • out_inf / out_sup → hardware limits
  • out_min / out_max → operational limits
  • ff_inf / ff_sup → FF input scaling
  • otff_inf / otff_sup → FF output mapping

Tuning Parameters

  • Kp → proportional gain
  • Ti → integral time (TIME format required)
  • Td → derivative time
  • Kd ≥ 1 → derivative filter
  • dband → deadband
  • outrate → output rate limiting

Critical Pitfall (Very Common)

Ti = 10

This is interpreted as 10 ms, not 10 seconds.

Correct usage:

Ti = T#10s

This mistake alone causes severe instability in real systems.


Operating Modes & Bumpless Transfer

PIDFF supports:

  • Manual mode → operator controls output
  • Auto mode → closed-loop operation
  • Tracking mode → cascade synchronization

Key Capability

Bumpless transfer:

  • No output jump when switching modes
  • Integral term is automatically aligned

Cascade Control with PIDFF

Architecture

  • Primary loop → sets SP of secondary
  • Secondary loop → drives actuator

Critical Requirements

  • Secondary loop must be 3–5× faster
  • Primary must track secondary during transitions
  • OUTD + TRI provide back-calculation

Poor configuration results in large output disturbances (“bumps”).


AUTOTUNE — Reality vs Expectation

PIDFF integrates with the AUTOTUNE block, but:

Practical Reality

  • Provides initial estimates only
  • Struggles with:
    • large dead time
    • nonlinear systems
    • actuator saturation

Always refine tuning manually after autotune.


Known Field Issues & Practical Workarounds

Observed in Real Systems

  • Mode-switch anomalies (rare firmware behaviour)
  • Parameter reset after download
  • AUTOTUNE wiring confusion
  • Documentation gaps

Practical Workarounds

  • Explicitly initialise PARA structure
  • Use RCPY for startup stabilisation
  • Validate scaling after download
  • Avoid overly aggressive enable toggling

These are real-world issues rarely covered in manuals.


PIDFF Implementation Checklist

Before commissioning:

  • Feedforward signal measurable and stable
  • Feedforward scaling configured correctly
  • PID tuned with feedforward active
  • Output limits properly defined
  • Anti-windup mechanisms verified
  • Time alignment validated
  • Cascade loops tuned correctly

Real Industrial Applications

PIDFF is widely used in:

  • Flow control with pressure compensation
  • Temperature control with load disturbances
  • Level control with inflow variation
  • Energy and HVAC systems

It is most effective where disturbances are measurable.


Business Impact

Poor PIDFF implementation leads to:

  • Production losses
  • Energy inefficiency
  • Equipment wear
  • Unstable operations
  • Increased operator intervention

The cost of poor tuning is often hidden but significant.