Analog Noise and Signal Oscillation in PLC Systems

Analog noise is one of the most common causes of unstable process values in industrial automation.

A pressure, temperature, flow or level measurement may fluctuate even though the physical process is stable. The PLC sees these fluctuations as real process movement and reacts accordingly.

This can create:

• Unstable HMI trends
• False alarms
• PID oscillation
• Repeated permissive changes
• Intermittent shutdowns
• Incorrect production data
• Unnecessary actuator movement

The PLC cannot automatically determine whether a changing value comes from the process or from electrical interference. It simply processes the signal it receives.

That is why grounding, shielding, cable routing and raw signal verification should be checked before modifying PLC logic or adding heavy filtering.

How Analog Noise Affects the PLC

A typical analog measurement follows this path:

1. The physical process produces a real value.
2. The transmitter converts it into a 4–20 mA or voltage signal.
3. Electrical noise is introduced into the circuit.
4. The analog input module converts the disturbed signal into raw counts.
5. The PLC scales the value into engineering units.
6. Alarms, permissives and PID loops react to the false movement.

For example, a stable pressure of 5 bar may appear in the PLC as:

4.97 bar
5.04 bar
4.91 bar
5.10 bar
4.95 bar

Small variations may be harmless. Larger or faster fluctuations can repeatedly cross alarm and interlock thresholds.

The PLC does not know that the actual pressure remained stable. It only sees the changing numerical value.

Common Sources of Analog Noise

Poor Grounding and Bonding

Different parts of a machine may sit at slightly different electrical potentials.

When instruments, cabinets and PLC systems are connected through multiple grounding paths, unwanted current can flow through signal wiring or cable shields.

Possible symptoms include:

• Constant measurement offset
• Slow value drift
• Noise that changes when motors start
• Different readings between cabinets
• Increased instability after connecting another device
• Fluctuation linked to building equipment

Good equipotential bonding helps reduce voltage differences between connected devices.

However, grounding changes should be planned carefully. Randomly disconnecting conductors or shields may remove one symptom while creating another electrical or EMC problem.

Incorrect Shielding

Shielded instrumentation cable helps protect low-level signals from electromagnetic interference.

Problems occur when the shield is:

• Not connected
• Connected at the wrong location
• Damaged
• Terminated with a long unshielded section
• Used as a signal or current-carrying conductor
• Connected inconsistently across junction boxes
• Routed beside high-current conductors

The correct shielding arrangement depends on the transmitter, analog module and site grounding design.

Follow the equipment manufacturer’s documentation and the project’s EMC strategy rather than applying one universal rule to every installation.

VFD Interference

Variable-frequency drives are a major source of industrial electrical noise.

A VFD switches its output voltage rapidly to control motor frequency. These fast switching edges can create:

• Conducted interference
• Radiated electromagnetic fields
• Common-mode currents
• Ground-potential disturbances
• Capacitive coupling into nearby cables

Analog problems associated with a VFD often appear when:

• The drive starts
• Motor frequency changes
• The drive operates at a particular speed
• Several drives run together
• The motor cable becomes longer
• The motor-cable shield is damaged or poorly terminated

A useful diagnostic test is to record the raw analog value while starting and stopping the suspected VFD.

If spikes or oscillation begin at the same moment, investigate the drive installation, shielding, grounding and cable separation.

Poor Cable Routing

Analog and communication cables should not be routed carelessly beside:

• VFD output cables
• Motor power cables
• Contactor wiring
• High-current busbars
• Welding cables
• Solenoid wiring
• Transformer conductors

Long parallel cable runs increase electromagnetic coupling.

Where signal and power cables must cross, crossing at approximately right angles generally reduces the length over which interference can couple between them.

Physical separation is often more effective than trying to remove the resulting noise later in the PLC program.

Loose or Damaged Terminals

A loose terminal may not create a complete open circuit.

Instead, it can produce:

• Random spikes
• Brief signal drops
• Vibration-related faults
• Temperature-dependent drift
• Intermittent under-range diagnostics
• Sudden changes during machine movement

The signal may appear stable while the cabinet is open and fail again after production starts.

Inspect:

• Terminal tightness
• Ferrules
• Junction boxes
• Cable damage
• Corrosion
• Moisture ingress
• Shield continuity
• Plug-in module connectors

Do not rely only on visual inspection. A conductor may appear connected while making unreliable electrical contact.

Unstable Power Supplies

Two-wire transmitters often depend on the same 24 V DC control system used by PLCs, relays and other field devices.

If the supply voltage changes when contactors, valves or other loads operate, the transmitter output may also become unstable.

Possible causes include:

• Undersized power supply
• Excessive voltage ripple
• Loose 0 V connection
• Overloaded distribution branch
• Poor DC selectivity
• Large inductive loads without suppression
• Failing power supply
• Excessive cable voltage drop

Measure the voltage directly at the transmitter or loop terminals during normal machine operation.

A standard multimeter may miss brief disturbances. A meter with minimum and maximum recording, oscilloscope or power-quality recorder may be required for intermittent problems.

Ground Loops

A ground loop occurs when a signal circuit has more than one ground-reference path.

Small voltage differences can introduce unwanted current into the measurement loop.

Ground-loop symptoms may include:

• Stable but incorrect offset
• Slow measurement drift
• Value changes when another device is connected
• Noise that disappears when test equipment is removed
• Different readings at separate control cabinets
• Oscillation associated with motor operation

Possible solutions may involve:

• Galvanic isolation
• Isolated analog modules
• Signal isolators
• Improved equipotential bonding
• Correct shield termination
• Removal of unintended grounding paths

The correct solution depends on the electrical design. Grounding should not be changed through trial and error on a live industrial system.

Current Signals Versus Voltage Signals

A 4–20 mA loop is generally more resistant to noise than a 0–10 V signal.

A current transmitter regulates loop current, making the measurement less sensitive to ordinary cable voltage drop.

A voltage signal is more directly affected by:

• Conductor resistance
• Ground-potential differences
• Input impedance
• Electrical interference
• Reference-voltage instability

This does not mean that current loops are immune to noise.

Poor wiring, grounding or shielding can still disturb a 4–20 mA circuit, especially over long cable runs or near high-frequency drive equipment.

Common Noise Symptoms

Rapid process-value fluctuation

The raw input moves continuously even though the real process is stable.

Possible causes include electrical interference, unstable transmitter power or an unsuitable input range.

Random spikes

The value suddenly jumps for one or two scans.

Possible causes include poor shielding, loose terminals, switching interference or transmitter faults.

False alarms

Noise briefly crosses an alarm threshold.

The HMI may never display the change because the spike exists for less time than the screen’s polling interval.

PID instability

The controller reacts to measurement noise by repeatedly changing its output.

A valve may hunt, a pump may change speed continuously or a heater may cycle unnecessarily.

Intermittent analog faults

Module underflow, overflow or wire-break diagnostics may appear briefly and disappear.

This often points toward loose wiring, unstable loop power or failing field equipment.

Noise Can Trigger False Permissives

Analog measurements are frequently converted into Boolean conditions.

For example:

Pressure_Healthy := Pressure_PV > 3.0 bar;

If the value oscillates around 3 bar, the permissive may switch repeatedly:

2.98 bar → FALSE
3.03 bar → TRUE
2.96 bar → FALSE
3.05 bar → TRUE

This can produce:

• Repeated startup attempts
• Output chatter
• Sequence instability
• Nuisance machine stops
• Unstable recovery logic

Adding hysteresis creates separate switching points:

Pressure_Healthy becomes TRUE above 3.2 bar.
Pressure_Healthy becomes FALSE below 2.8 bar.

This prevents small variations around one threshold from repeatedly changing the permissive.

Hysteresis does not repair electrical noise, but it can make process logic more stable after the signal integrity has been verified.

Noise and PID Control

A PID controller calculates its response from the difference between the setpoint and process value.

When the process value contains noise, the controller may interpret every small fluctuation as a real process disturbance.

This is particularly noticeable with derivative action because derivative control responds strongly to rapid changes.

Possible symptoms include:

• Oscillating controller output
• Valve hunting
• Unstable pump speed
• Excessive actuator wear
• Poor setpoint stability
• Frequent output reversals

Before retuning the PID controller, compare the raw process value with an independent measurement.

Reducing controller gain may hide the symptoms while making real process control unnecessarily slow.

Fix signal integrity first, then apply the smallest suitable amount of filtering and retune the controller if necessary.

Filtering Is Not the First Repair

A common response to analog noise is adding a moving average or low-pass filter.

The trend becomes smoother, and the fault appears solved.

However, filtering does not remove the electrical problem. It only changes how quickly the PLC responds to it.

Excessive filtering introduces dead time and may cause:

• Delayed alarms
• Slow shutdowns
• Startup overshoot
• Sluggish PID response
• Hidden process instability
• False confidence in the HMI trend

Filtering should be used after checking:

• Grounding
• Shielding
• Cable routing
• Transmitter condition
• Loop power
• Terminal integrity
• Module configuration

A wiring fault should be repaired rather than averaged.

Compare Raw and Filtered Values

A good PLC project should preserve several versions of an important signal:

Pressure_Raw
Pressure_Scaled
Pressure_Filtered
Pressure_Display
Pressure_Valid

Possible uses include:

• Raw value for electrical diagnostics
• Scaled value for range verification
• Lightly filtered value for control
• More heavily filtered value for display
• Validity bit for alarms and permissives

During troubleshooting, record the raw and filtered values together.

If the raw signal oscillates while the filtered value appears stable, the filter is hiding the disturbance rather than proving the process is healthy.

Use the Correct Diagnostic Tools

A slowly refreshing HMI trend may miss short spikes.

More useful tools include:

• PLC trace recording
• Oscilloscope
• Loop calibrator
• Process meter
• Low-current DC clamp meter
• Minimum/maximum voltage recorder
• Analog module diagnostic buffer
• Managed-switch diagnostics for networked measurements

Capture related events on the same timeline:

• Raw analog value
• Filtered value
• Motor or VFD command
• Contactor operation
• Module diagnostic status
• 24 V supply
• Alarm and permissive states

This helps determine whether the disturbance is linked to a particular electrical event.

A Practical Troubleshooting Workflow

1. Verify the real process

Use an independent instrument to confirm that the process is actually stable.

2. Monitor the raw PLC value

Do not begin with the filtered HMI display.

Observe or trace the raw count from the analog input module.

3. Measure the field signal

Measure the 4–20 mA current or voltage signal using suitable equipment and safe procedures.

4. Check transmitter power

Measure the supply at the transmitter while motors, valves and other loads operate.

5. Inspect wiring

Check terminals, polarity, cable condition, shield continuity and junction boxes.

6. Review cable routing

Look for long parallel runs beside VFD output or motor cables.

7. Check grounding and bonding

Confirm the installation matches the approved electrical and EMC design.

8. Review module configuration

Verify signal type, integration time, diagnostic settings and channel range.

9. Identify event correlation

Determine whether the noise begins when a particular drive, contactor or load operates.

10. Apply minimal filtering

Only after correcting the physical causes should software filtering or hysteresis be adjusted.

Signal Isolation

A signal isolator can be useful where grounding differences or shared references create instability.

An isolator can provide galvanic separation between:

• Field transmitter
• PLC analog input
• Separate power systems
• Different buildings or cabinets

Isolation can also help protect the PLC from certain fault currents and ground-potential differences.

However, it adds another component to the signal chain with its own:

• Accuracy specification
• Power requirement
• Response time
• Calibration
• Failure modes

Isolation should be applied for a clear electrical reason rather than as a universal cure for every noisy signal.

Module Filtering and Integration Time

Analog input modules often provide configurable integration or interference-suppression settings.

Longer integration can reduce mains-frequency noise, but it also slows the channel update.

Before adding a software filter, check whether the module already applies:

• 50 Hz suppression
• 60 Hz suppression
• Averaging
• Channel smoothing
• Extended conversion time

Several filters in series can create far more delay than expected.

Document the total signal response from the physical process to the PLC logic.

Preserve Signal Quality Information

An analog value should include diagnostic context.

Useful tags include:

Analog_Raw
Analog_Engineering
Analog_Valid
Analog_Underflow
Analog_Overflow
Analog_Wire_Break
Analog_Noise_Alarm
Analog_Last_Update

The PLC can also monitor whether the signal’s rate of change is physically plausible.

For example, a tank level cannot normally move from 20% to 80% in one scan. Such a jump may indicate a spike, scaling fault or communication error.

Plausibility checks should mark the value as questionable and generate diagnostics rather than silently replacing it.

Final Thoughts

Analog noise creates false process movement.

The PLC cannot distinguish electrical interference from a genuine change unless the control system includes filtering, diagnostics and plausibility checks. Even then, software cannot repair every field installation problem.

Before changing alarms, PID settings or ladder logic, verify:

• Grounding
• Shielding
• Cable routing
• Terminal condition
• Transmitter supply
• Raw signal stability
• Analog module configuration

A smooth HMI trend is not proof of a healthy measurement.

The most reliable control begins with a clean, correctly installed and verifiable field signal.

Fix the source of the noise first. Filter only what remains.