Mechatronics & Robotics

Cohen–Coon PID Tuning Calculator — Kp, Ki, Kd for FOPDT Processes

Calculate P, PI or PID controller settings from signed process gain, time constant and dead time, including Kp, Ki, Kd, Ti, Td, L/T ratio and recommended controller action.

Unit-aware inputs Deterministic calculation Engineering interpretation
Calculation workspace

Enter the known values and review the calculated result

Deterministic calculation
01
Parameters

Input parameters

Use consistent values and select the intended engineering units.

Controller configuration

FOPDT process model

02
Output

Results

Live
Ready to calculate Complete the required inputs and run the calculation.
—
Calculation tools Save, export and manage your results

Save your calculation, export available results, or store a finalized engineering artifact in your account.

Engineering reference

Method, application and limitations

Review the calculation method, intended application and engineering assumptions before using the result in a design decision.

01
Method

Formula and calculation method

Cohen–Coon P, PI and PID tuning formulas

The calculator determines initial controller settings from a first-order-plus-dead-time (FOPDT) process model defined by signed process gain K, process time constant T and process dead time L.

The normalized dead-time ratio is:

R = L / T

The implemented Cohen–Coon correlations first determine proportional gain magnitude Kp and, where applicable, integral time Ti and derivative time Td. The calculator then converts the time-form coefficients to parallel-form integral and derivative gains:

Ki = Kp / Ti

Kd = Kp · Td

The proportional gain is calculated as a magnitude using |K|. The sign of the measured process gain is retained separately to determine the recommended controller action.

P controller

Kp = [1 / (|K| · R)] · (1 + R / 3)

For P-only control:

  • Ki = 0
  • Kd = 0
  • Ti and Td are not used.

PI controller

Kp = [1 / (|K| · R)] · (0.9 + R / 12)

Ti = L · (30 + 3R) / (9 + 20R)

Ki = Kp / Ti

Kd = 0

PID controller

Kp = [1 / (|K| · R)] · (4/3 + R / 4)

Ti = L · (32 + 6R) / (13 + 8R)

Td = 4L / (11 + 2R)

Ki = Kp / Ti

Kd = Kp · Td

Symbols and outputs

  • K — signed process gain, ΔPV/ΔCO.
  • T — FOPDT process time constant.
  • L — FOPDT process dead time.
  • R — dead-time-to-time-constant ratio, L/T.
  • Kp — proportional gain magnitude, expressed as CO/PV.
  • Ki — parallel integral gain magnitude, expressed as CO/(PV·s).
  • Kd — parallel derivative gain magnitude, expressed as CO·s/PV.
  • Ti — integral time in seconds; not used for P-only control.
  • Td — derivative time in seconds; used only for PID control.

Time inputs are normalized to the calculator's canonical time unit before the formulas are evaluated, so T and L remain physically consistent when different supported time units are selected.

02
Application

When to use this calculator

When to use this Cohen–Coon PID tuning calculator

Use this calculator when a stable, self-regulating process can be represented reasonably well by a FOPDT model and you already know its signed process gain, time constant and dead time. It provides initial tuning values for P, PI or PID controllers.

Typical use starts with an open-loop process test or an identified process model from which the three FOPDT parameters are obtained:

  • process gain K — how strongly the process variable changes for a change in controller output,
  • time constant T — the characteristic response time of the first-order process,
  • dead time L — the delay between an input change and the observed process response.

The calculator is useful when you need a reproducible starting point for PID controller tuning rather than selecting Kp, Ki and Kd by trial and error. It also reports Ti, Td, the L/T ratio, a process-dynamics classification and the recommended direct or reverse controller action from the sign of K.

Use another method or a more detailed model when

  • the process cannot be represented adequately by a first-order-plus-dead-time model,
  • the process is integrating, unstable or strongly oscillatory rather than self-regulating,
  • the operating point changes enough for K, T or L to vary substantially,
  • strong multivariable interaction dominates the response,
  • actuator saturation, valve stiction or other nonlinearities dominate the loop behavior,
  • closed-loop robustness or performance must be demonstrated rather than estimated from a tuning rule.

The calculated values are controller starting settings. The calculator does not simulate the closed-loop response and does not verify overshoot, settling time, stability margins, actuator limits or disturbance rejection.

03
Decision support

How to interpret the result

How to interpret the Cohen–Coon tuning results

Kp is the proportional gain magnitude. Higher Kp means a stronger proportional controller response for a given control error. Because this calculator deliberately reports gain magnitude, the required feedback direction is communicated separately as controller action.

Ki is the integral gain in parallel form. It is calculated as Kp/Ti. A non-zero Ki is present for PI and PID modes and accumulates control error over time. P-only mode sets Ki to zero and can therefore leave a steady-state offset.

Kd is the derivative gain in parallel form. It is calculated as Kp·Td and is present only in PID mode. The calculator does not model derivative filtering or measurement noise, so a calculated Kd value is not by itself evidence that the derivative term will be suitable in the real loop.

Ti and Td are the integral and derivative times from the classical Cohen–Coon correlations. They are also provided because many industrial PID implementations are configured using Kp, Ti and Td rather than independent Ki and Kd.

L/T describes the relative importance of delay:

  • L/T < 1: lag-dominant process, T > L.
  • L/T = 1: process time constant and dead time are equal.
  • L/T > 1: dead-time-dominant process, L > T.

When L/T ≥ 1, the calculator reports a warning and recommends explicit verification of robustness and actuator limits before commissioning.

The sign of the process gain determines the reported controller-action recommendation:

  • K > 0: reverse-acting controller recommendation.
  • K < 0: direct-acting controller recommendation.

Controller manufacturers do not all use identical gain-sign and error conventions. Verify the action definition of the actual PID block before applying the settings.

The calculator does not issue a general PASS or FAIL decision for the control loop. A numerically valid Cohen–Coon result means that the implemented tuning equations could be evaluated from the supplied FOPDT parameters; it does not demonstrate closed-loop stability, robustness or acceptable process performance.

04
Worked case

Calculation example

Cohen–Coon PID tuning example

Consider a self-regulating process represented by the following FOPDT parameters:

  • Controller type: PID
  • K = +2.0 — signed process gain
  • T = 100 s — process time constant
  • L = 20 s — process dead time

1. Calculate the dead-time ratio

R = L / T = 20 / 100 = 0.20

Because R is below 1, the calculator classifies the process as lag-dominant (T > L).

2. Calculate proportional gain

1 / (|K| · R) = 1 / (2 · 0.20) = 2.5

Kp = 2.5 · (4/3 + 0.20/4) = 2.5 · 1.383333... = 3.4583

3. Calculate integral time

Ti = 20 · (32 + 6 · 0.20) / (13 + 8 · 0.20)

Ti = 20 · 33.2 / 14.6 = 45.4795 s

4. Calculate derivative time

Td = 4 · 20 / (11 + 2 · 0.20) = 80 / 11.4 = 7.0175 s

5. Convert to parallel-form Ki and Kd

Ki = Kp / Ti = 3.4583 / 45.4795 = 0.07604

Kd = Kp · Td = 3.4583 · 7.0175 = 24.2690

Calculated PID settings

Parameter Result
Kp 3.4583 CO/PV
Ki 0.07604 CO/(PV·s)
Kd 24.2690 CO·s/PV
Ti 45.4795 s
Td 7.0175 s
L/T 0.20

Since the process gain is positive, the calculator recommends reverse-acting controller action. The numerical gains are magnitudes; the required sign or action setting must be implemented according to the conventions of the actual PLC, DCS or PID controller.

These values are initial Cohen–Coon tuning settings. Closed-loop behavior must still be checked before the tuning is accepted for operation.

05
Model boundaries

Assumptions and limitations

Assumptions and limitations

  • The process is assumed to be adequately represented by a stable, self-regulating first-order-plus-dead-time model over the operating region of interest.
  • Signed process gain K must be finite and non-zero. Negative K is valid and is used to determine the controller-action recommendation.
  • Process time constant T must be greater than zero.
  • Process dead time L must be greater than zero.
  • T and L must describe the same physical process response and are converted to a common internal time unit before calculation.
  • The calculator assumes the supplied K, T and L values are representative of the actual process. It does not identify these parameters from raw step-response data.
  • Cohen–Coon tuning is used as an initial open-loop tuning rule. The calculated values should be checked and, where necessary, adjusted using the real closed-loop response.
  • The calculator converts Ti and Td to parallel-form Ki and Kd. A target controller that uses another PID structure or parameter convention requires an appropriate conversion.
  • The calculation does not model actuator saturation, valve stiction, deadband, anti-windup, derivative filtering, signal noise or output-rate limitations.
  • The calculation does not include controller sample time, discretization effects, setpoint weighting or implementation-specific PID features.
  • Processes with substantial nonlinear behavior, operating-point dependence, interactions or higher-order dynamics may require system identification and a tuning method more appropriate to the actual process model.
  • A dead-time-dominant result with L/T ≥ 1 is specifically flagged for additional engineering review; the calculator does not convert that warning into an automatic rejection of the tuning.
  • P-only control can retain steady-state offset because no integral action is present.
06
Questions

Frequently asked questions

What is the Cohen–Coon PID tuning method?
Cohen–Coon is a classical open-loop controller tuning method based on a first-order-plus-dead-time process model. It uses process gain K, time constant T and dead time L to calculate starting settings for P, PI or PID control.
How does this calculator determine Kp, Ki and Kd?
The selected Cohen–Coon correlation first calculates proportional gain magnitude Kp and, for PI or PID control, the integral time Ti and derivative time Td. The calculator then converts the time-form parameters to parallel gains using Ki = Kp/Ti and Kd = Kp·Td. For P control Ki and Kd are zero; for PI control Kd is zero.
What are K, T and L in the FOPDT model?
K is the signed process gain ΔPV/ΔCO, T is the process time constant and L is the process dead time. Together they describe the steady-state sensitivity, characteristic response speed and delay of the FOPDT approximation used by the Cohen–Coon tuning equations.
How is process gain K calculated?
For a suitable steady-state process test, process gain is the change in process variable divided by the change in controller output, K = ΔPV/ΔCO. The sign must be retained: a positive controller-output change can produce either a positive or negative process response depending on the process.
Can the Cohen–Coon calculator use a negative process gain?
Yes. K must be finite and non-zero, but it may be positive or negative. The calculator uses |K| to calculate positive Kp, Ki and Kd magnitudes and uses the sign of K separately to recommend controller action. A positive process gain produces a reverse-action recommendation, while a negative process gain produces a direct-action recommendation.
What does the L/T dead-time ratio mean?
The ratio R = L/T compares process delay with the first-order time constant. The calculator classifies R < 1 as lag-dominant, R > 1 as dead-time-dominant and R = 1 as equal lag and dead time. The ratio also appears directly in every implemented Cohen–Coon gain correlation.
What happens when the process dead time is greater than the time constant?
The calculator still evaluates the Cohen–Coon formulas when L/T is at least 1, but it returns an engineering warning. Dead-time-dominant processes require additional attention to robustness, actuator limits and actual closed-loop response before the calculated settings are commissioned.
Does the calculator support P, PI and PID controller tuning?
Yes. P, PI and PID use separate Cohen–Coon correlations. P-only tuning calculates Kp with Ki = Kd = 0. PI calculates Kp, Ti and Ki with Kd = 0. PID calculates Kp, Ti, Td, Ki and Kd.
Can I enter the calculated Kp, Ki and Kd directly into a PLC PID block?
Not without first checking the controller definition. This calculator reports parallel-form gain magnitudes, while industrial controllers may use Kp/Ki/Kd, Kp/Ti/Td, proportional band, reset rate, minutes per repeat or another PID structure. Controller action and error-sign conventions can also differ between vendors. Convert the parameters to the exact form required by the target PID implementation.
What is the difference between Cohen–Coon and Ziegler–Nichols FOPDT tuning?
Both can use an open-loop FOPDT description based on process gain, time constant and dead time, but they use different tuning correlations. Cohen–Coon explicitly includes the L/T ratio inside its P, PI and PID equations. The methods should therefore be treated as alternative starting rules rather than interchangeable formulas.
Does this PID tuning calculator simulate the closed-loop response?
No. It calculates Cohen–Coon starting parameters and engineering interpretation only. It does not simulate rise time, overshoot, settling time, stability margins, disturbance rejection, actuator saturation or measurement-noise sensitivity.
Add an engineering note