A pressure transmitter does not send “63 PSI” through the cable.

It sends an electrical signal.

The PLC receives that signal as a raw numerical value, perhaps 6,553, 16,384 or 27,648, depending on the module and its configuration. Those numbers are meaningful to the processor, but they are not particularly useful to the operator standing in front of an HMI.

Nobody wants to see:

Tank pressure: 19,827 counts

They want:

Tank pressure: 60.5 PSI

Scaling is the calculation that converts the raw input value into an engineering unit such as pressure, temperature, flow, level, speed or weight.

The mathematics is fairly straightforward. The awkward part is knowing which numbers belong in the calculation.

Use the wrong raw range, forget the 4 mA live-zero offset or select the wrong transmitter span, and the PLC may display a perfectly stable value that is also completely wrong.

Those faults are often harder to spot than a broken wire because everything appears to be working.

What Is an Analog Signal?

A digital input normally has two basic states:

• OFF
• ON

An analog input represents a continuously changing process value.

Common analog measurements include:

• Pressure
• Temperature
• Flow
• Tank level
• Weight
• Position
• Motor speed
• Current
• Voltage
• Humidity
• pH
• Vibration

Instead of simply reporting whether pressure exists, an analog pressure transmitter can indicate how much pressure exists.

For example, a transmitter may measure from 0 to 100 PSI and represent that measurement using a 4–20 mA signal.

The signal changes continuously:

The PLC analog input module converts the electrical signal into a digital number.

The program then scales that number into PSI.

Why PLCs Use Raw Counts

The processor does not naturally think in PSI, degrees Celsius or litres per minute.

The analog input module converts the incoming electrical signal into a numerical value using an analog-to-digital converter.

That numerical value is often called:

• Raw count
• Raw input
• Digital count
• ADC count
• Input word
• Process image value

The possible range depends on the module, its resolution and its selected data format.

Examples might include:

• 0 to 4,095
• 0 to 10,000
• 0 to 27,648
• 0 to 32,767
• −27,648 to 27,648
• 4,000 to 20,000
• A floating-point value already expressed in milliamps

There is no universal raw-count range used by every PLC.

That matters a lot.

A scaling calculation copied from another machine may be completely wrong even when both systems use 4–20 mA transmitters.

Always check the manual for the exact input module and configuration.

What Does 4–20 mA Mean?

A 4–20 mA transmitter uses 4 mA to represent the bottom of its calibrated measurement range and 20 mA to represent the top.

For a 0–100 PSI transmitter:

• 4 mA = 0 PSI
• 20 mA = 100 PSI

For a −50 to 150°C temperature transmitter:

• 4 mA = −50°C
• 20 mA = 150°C

For a 0–5,000 litre tank-level transmitter:

• 4 mA = 0 litres
• 20 mA = 5,000 litres

The current signal remains 4–20 mA, but its engineering meaning depends on how the transmitter is configured.

That transmitter range must match the scaling parameters in the PLC.

Why Does the Signal Start at 4 mA Instead of 0 mA?

The 4 mA lower limit is known as a live zero.

It helps distinguish a legitimate minimum reading from certain fault conditions.

For example:

• 4 mA may mean the process is genuinely at 0 PSI.
• 0 mA may indicate a broken wire, missing transmitter power or open circuit.

This provides useful diagnostic information.

Many systems define underrange and overrange limits outside the normal 4–20 mA span.

A transmitter might output approximately:

• Below 3.6 mA for an underrange or fault
• 4–20 mA for the normal measurement range
• Above 20.5 or 21 mA for an overrange or fault

The exact values depend on the transmitter and applicable standard.

Do not assume every signal below 4 mA means exactly the same thing. Read the device documentation.

How the PLC Converts Current Into a Raw Value

The input module measures the current and converts it to a digital value.

Suppose a module is configured so that:

• 4 mA = 0 counts
• 20 mA = 32,767 counts

A 12 mA signal is halfway through the 4–20 mA span.

Therefore, the raw value should be approximately halfway through the digital range:

32,767 ÷ 2 = 16,383.5

The module may report approximately:

16,384 counts

If the transmitter represents 0–100 PSI, the corresponding pressure is 50 PSI.

The relationship is linear:

• Half the current span
• Half the raw-count span
• Half the engineering span

Unfortunately, not every module represents 4 mA as zero.

Another module may use:

• 4 mA = 6,553 counts
• 20 mA = 32,767 counts

Another could use:

• 4 mA = 4,000
• 20 mA = 20,000

Yet another may directly provide:

• 12.000 mA

The scaling must reflect the actual module format.

The Standard Scaling Formula

A general linear scaling formula is:

Scaled value = ((Raw input − Raw minimum) ÷ (Raw maximum − Raw minimum)) × (Engineering maximum − Engineering minimum) + Engineering minimum

Where:

• Raw input is the current value from the analog module.
• Raw minimum is the raw count representing the low end of the signal.
• Raw maximum is the raw count representing the high end.
• Engineering minimum is the transmitter’s minimum process value.
• Engineering maximum is the transmitter’s maximum process value.

It looks more complicated than it really is.

The formula does three things:

1. Removes the raw-value offset.
2. Determines what percentage of the raw range is being used.
3. Applies the same percentage to the engineering range.

PLC Scaling Example: 4–20 mA to 0–100 PSI

Assume:

• Transmitter range: 0–100 PSI
• Input signal: 4–20 mA
• Raw minimum: 0
• Raw maximum: 32,767
• Current raw input: 16,384

Insert those values into the formula:

Pressure = ((16,384 − 0) ÷ (32,767 − 0)) × (100 − 0) + 0

Pressure ≈ 50 PSI

That one is easy because both ranges begin at zero.

Now consider a module whose raw count at 4 mA is 6,553.

Assume:

• Raw minimum: 6,553
• Raw maximum: 32,767
• Raw input: 19,660
• Engineering minimum: 0 PSI
• Engineering maximum: 100 PSI

The calculation becomes:

Pressure = ((19,660 − 6,553) ÷ (32,767 − 6,553)) × 100

Pressure ≈ 50 PSI

The process value is still 50 PSI, but the correct raw midpoint is different.

This is why guessing the module range creates bad readings.

Scaling a Range That Does Not Start at Zero

Not every transmitter represents a range beginning at zero.

Consider a temperature transmitter configured for:

• 4 mA = −50°C
• 20 mA = 150°C

Suppose the analog input is exactly halfway through its raw range.

The result is not 75°C.

The complete engineering span is:

150 − (−50) = 200°C

Half of that span is:

200 ÷ 2 = 100°C

Add the engineering minimum:

−50 + 100 = 50°C

Therefore, 12 mA represents:

50°C

Forgetting the negative lower limit is a common scaling mistake.

Scaling a 0–10 V Signal

The same formula can be used for voltage inputs.

Suppose:

• Sensor signal: 0–10 V
• Measurement range: 0–2,000 RPM
• Raw range: 0–27,648
• Current raw input: 13,824

The raw input is halfway through the range.

Therefore:

Speed = ((13,824 − 0) ÷ (27,648 − 0)) × 2,000

Speed = 1,000 RPM

Voltage signals are easy to understand, but they tend to be more sensitive to voltage drop and electrical noise over long distances than current loops.

Why 4–20 mA Is Popular in Industrial Systems

Current loops are widely used because the same current flows through the complete series circuit.

The voltage may vary along the cable, but the current remains the process signal as long as the transmitter has enough supply voltage to drive the loop.

Advantages include:

• Good performance over long cable runs
• Strong resistance to electrical noise
• Live-zero fault detection
• Simple two-wire transmitter options
• Easy loop testing with a calibrator
• Wide support across industrial equipment

This does not make 4–20 mA immune to wiring faults, grounding problems or interference.

It is robust, not magical.

Two-Wire, Three-Wire and Four-Wire Transmitters

The wiring arrangement depends on the transmitter.

Two-Wire Transmitter

A two-wire transmitter receives power and sends its signal through the same current loop.

The transmitter controls the loop current between 4 and 20 mA.

A simplified circuit includes:

1. 24 V DC supply
2. Transmitter
3. PLC analog input
4. Return path to the supply

Two-wire transmitters are common because they require less cabling.

However, the loop must have enough available voltage for:

• Transmitter voltage requirement
• PLC input resistance
• Cable voltage drop
• Barriers or isolators
• Other series devices

This is sometimes called the loop-voltage budget.

Three-Wire Transmitter

A three-wire transmitter commonly uses:

• Positive supply
• Supply common
• Separate signal output

The output may be current or voltage.

The signal common must be connected correctly. Poor common wiring can create offset or unstable readings.

Four-Wire Transmitter

A four-wire transmitter uses a separate power supply pair and a separate signal pair.

This is common for devices that need more power than a two-wire loop can provide.

Examples may include:

• Flowmeters
• Analysers
• Complex weighing instruments
• Devices with displays or heaters

Never assume the wiring type from the number of terminals visible on the device. Check its wiring diagram.

Active and Passive Analog Inputs

Current-loop terminology can be confusing because manufacturers do not always describe it consistently.

An analog input may be:

• Active
• Passive
• Sourcing
• Sinking
• Internally powered
• Externally powered

An active input may provide loop power.

A passive input only measures the current and requires an external loop supply.

Connecting two active devices together can cause problems. Connecting two passive devices together produces no loop current at all.

Before wiring a transmitter, determine:

• Which device provides loop power
• Expected polarity
• Input resistance
• Required external supply
• Whether channels share a common terminal
• Whether channels are isolated

The PLC module wiring diagram is not optional reading here.

Using Built-In PLC Scaling Instructions

Many PLC platforms provide instructions or function blocks for analog scaling.

Examples may include:

• SCP, or Scale with Parameters
• SCL or SCALE instructions
• NORM_X and SCALE_X
• Linear conversion blocks
• Function blocks created by the machine builder

These instructions reduce the need to build the calculation manually.

A typical scaling block asks for:

• Input value
• Input minimum
• Input maximum
• Output minimum
• Output maximum

It then calculates the engineering value.

Built-in functions can reduce errors, but only when the parameters are correct.

A scaling block with the wrong raw minimum is simply a very efficient way to calculate the wrong result.

Scaling in Siemens PLCs

A common Siemens approach uses two operations:

1. Normalize the raw input into a 0.0–1.0 value.
2. Scale the normalized value into engineering units.

For example:

• Raw input: 0–27,648
• Normalized value: 0.0–1.0
• Pressure output: 0–100 PSI

The normalization step determines what fraction of the raw range is present.

The scaling step applies that fraction to the process range.

For a 50% raw input:

• Normalized value = 0.5
• Scaled pressure = 50 PSI

The concept is the same as the standard formula; it is simply divided into two clear steps.

Scaling in Rockwell PLCs

Older Rockwell platforms commonly use the SCP instruction.

The instruction typically includes:

• Input
• Input minimum
• Input maximum
• Scaled minimum
• Scaled maximum
• Output

In Logix 5000 systems, scaling is often performed using:

• Arithmetic instructions
• Function blocks
• Add-On Instructions
• Raw/proportional data formats
• Engineering-unit data provided directly by the module

Always check the selected module data format.

The same input card may produce different values depending on whether it is configured for raw counts, engineering units or percentage of range.

Use Floating-Point Arithmetic

Scaling calculations often involve division.

Using integer-only arithmetic can remove decimal information through rounding or truncation.

Suppose:

Raw input ÷ Raw maximum

If both values are integers, some PLCs may perform integer division.

A raw input of 16,000 divided by 32,767 could become zero rather than approximately 0.488 if the result is stored as an integer.

The final scaled result would then also be zero.

Use floating-point or real data types where needed.

A safe approach is to convert the raw input into a real value before division.

For example:

REAL(Raw input)

The exact conversion instruction depends on the PLC platform.

Clamping the Scaled Value

Real signals can move slightly outside the normal range.

A 4–20 mA transmitter may occasionally output:

• 3.98 mA
• 20.02 mA

Noise, calibration tolerance and intentional overrange behaviour can all produce small excursions.

Without clamping, the scaled value may show:

• −0.1 PSI
• 100.2 PSI

That may be acceptable for diagnostics, but it can look strange on an operator display.

Clamping limits the displayed or controlled value to a defined range.

For example:

• Values below 0 PSI are displayed as 0 PSI.
• Values above 100 PSI are displayed as 100 PSI.

Be careful not to hide faults.

It is often better to keep:

• An unclamped diagnostic value
• A clamped operator-display value
• Separate under-range and over-range alarms

That gives the operator a tidy display without throwing away useful information.

Detecting Open Circuits and Signal Faults

A normal 4–20 mA loop should not operate near zero current.

Therefore, a very low reading can be used to detect a fault.

A typical program may define:

• Below 3.6 mA: underrange or open-loop fault
• 3.6–4.0 mA: below calibrated range
• 4–20 mA: normal measurement
• 20–21 mA: above calibrated range
• Above 21 mA: overrange or transmitter fault

These values are only examples.

Use the transmitter and module specifications for the actual thresholds.

Add a time delay where appropriate so a brief disturbance does not trigger a nuisance alarm.

Common PLC Scaling Mistakes

Using the Wrong Raw Maximum

A programmer assumes the module uses 0–32,767, but the actual configured range is 0–27,648.

The result will be wrong across most of the measurement span.

At the top end, a real 100% input would display only:

27,648 ÷ 32,767 ≈ 84.4%

A 100 PSI signal could appear as approximately 84.4 PSI.

The reading may look believable, which makes the error particularly sneaky.

Forgetting the 4 mA Offset

A programmer scales 0–20 mA as though 0 mA represents the transmitter minimum.

But the real signal is 4–20 mA.

At 4 mA, the process should show 0%, yet the calculation may show 20%.

This creates a substantial offset across the entire range.

Reversing Minimum and Maximum Values

If engineering minimum and maximum are entered backward, the displayed value decreases as the real process increases.

For example:

• 4 mA displays 100 PSI
• 20 mA displays 0 PSI

This can happen intentionally on reverse-acting devices, but most of the time it is a configuration error.

Using the Wrong Transmitter Range

The PLC is scaled for 0–100 PSI, but the installed transmitter is configured for 0–250 PSI.

At 20 mA, the PLC displays 100 PSI while the real pressure is 250 PSI.

The electrical signal is healthy.

The scaling is stable.

The measurement is dangerously wrong.

Always compare the transmitter nameplate or configuration with the PLC logic and HMI units.

Ignoring Non-Zero Engineering Minimums

A transmitter configured for −20 to 80°C has a 100°C span, but its minimum is −20°C.

Scaling it as 0–100°C creates an offset of 20°C.

Scaling an Already Scaled Value

Some analog modules can return engineering units or a percentage directly.

Applying another scaling calculation to that value produces nonsense.

Check the configured data format before adding custom math.

Performing Integer Division

The calculation loses fractional resolution and may produce stepped or completely incorrect values.

Use REAL or floating-point data where appropriate.

Scaling the Wrong Channel

Copied logic sometimes references the previous channel.

The value may appear reasonable because both sensors operate in similar ranges.

Cross-check:

• Rack
• Slot
• Channel
• Tag alias
• Physical terminal
• HMI tag

Wiring: Where Many Analog Problems Begin

The mathematics gets blamed frequently.

The wiring deserves at least equal suspicion.

Analog signals are more sensitive than ordinary digital inputs. A digital input may continue functioning despite moderate noise because it only needs to distinguish ON from OFF.

An analog input is expected to measure small changes accurately.

A little unwanted voltage can become a false pressure change, unstable temperature reading or wandering flow value.

Use the Correct Cable

Use cable suitable for analog instrumentation.

A good cable commonly includes:

• Twisted conductors
• Overall shield
• Appropriate conductor size
• Suitable insulation
• Environmental resistance
• Correct capacitance for the application

Twisting helps reject electromagnetic interference because noise tends to affect both conductors similarly.

Shielding provides additional protection against external electric fields.

Cheap cable may work on the workshop bench and behave terribly after installation beside a 90 kW VFD.

Keep Analog Wiring Away From Power Cables

Avoid routing analog signals alongside:

• VFD output cables
• Motor leads
• Contactor wiring
• Heater circuits
• Welding cables
• High-current busbars
• Transformer primary wiring

Maintain physical separation where possible.

If signal and power cables must cross, crossing at approximately 90 degrees generally reduces coupling compared with running them parallel.

Do not place a 4–20 mA cable in the same conduit as motor conductors unless the design and applicable standards specifically allow it.

A few saved metres of cable are not worth months of unstable readings.

Shield Grounding

Shielding practices depend on the installation, frequency environment, equipment design and site grounding philosophy.

For many low-frequency analog circuits, the shield is bonded at one end to reduce the chance of circulating ground-loop current.

The PLC or control-cabinet end is commonly selected.

However, “always ground at one end only” is too absolute for every industrial application.

High-frequency interference may require bonding at both ends using a proper low-impedance connection. Some manufacturers specify their own shielding arrangement.

Follow:

• Transmitter instructions
• PLC module manual
• Site grounding standard
• Applicable EMC practices

Do not connect shields randomly based on whichever terminal is nearest.

Also avoid using the cable shield as the signal common or protective-earth conductor.

Ground Loops

A ground loop occurs when different points in the signal circuit sit at slightly different electrical potentials and unwanted current flows through the grounding or shielding path.

Possible symptoms include:

• Stable offset error
• Reading changes when another machine starts
• Different values with laptop connected
• Fluctuation related to building loads
• Repeated analog-module overrange faults

Solutions may involve:

• Correcting grounding
• Using isolated analog channels
• Installing a signal isolator
• Improving equipotential bonding
• Separating supply commons
• Reviewing shield termination

Do not simply disconnect protective grounding to make the reading stable.

That may create a much more serious safety problem.

Check Polarity

Current-loop polarity matters.

Reversing positive and negative connections may produce:

• No reading
• Zero current
• Negative indication
• Module fault
• Transmitter failure in extreme cases

Before energizing the loop, verify:

• Power-supply polarity
• Transmitter positive and negative
• PLC input positive and negative
• Shared common arrangement
• Terminal numbers

A quick wiring check is cheaper than an analog input card.

Long Cable Runs

Long cable runs introduce:

• Additional resistance
• Greater exposure to electrical noise
• Increased capacitance
• Larger ground-potential differences
• More junctions and failure points

Current loops are generally better suited to long runs than voltage signals.

For a two-wire transmitter, calculate whether the supply voltage can overcome all voltage drops in the loop.

A simplified loop budget is:

Available supply voltage ≥ transmitter minimum voltage + input voltage drop + cable voltage drop + barrier or isolator drops

If the transmitter runs out of voltage headroom, it may fail to reach 20 mA at the top of the process range.

The result can look like incorrect scaling even though the real problem is insufficient loop supply voltage.

Analog Input Isolation

Analog modules may provide:

• No channel isolation
• Group-to-group isolation
• Channel-to-channel isolation
• Isolation from the backplane

Non-isolated channels are cheaper but more vulnerable to ground-potential differences.

Isolated channels can help when signals originate from equipment supplied by different sources or located far apart.

Check whether channel commons are internally connected.

Accidentally tying together supposedly independent circuits can introduce measurement errors or damage equipment.

Troubleshooting PLC Analog Input Problems

Analog faults come in several flavours:

• Reading stuck at zero
• Reading stuck at maximum
• Stable but inaccurate value
• Slowly drifting value
• Randomly jumping value
• Intermittent open-loop alarm
• Correct PLC value but wrong HMI value
• One channel faulty while others are normal

The quickest approach is to start at the source and test each layer in order.

Step 1: Understand the Expected Signal

Before measuring anything, determine:

• Transmitter type
• Process range
• Electrical signal range
• Wiring type
• Power source
• Normal process value
• PLC raw-count range
• Scaling parameters
• HMI engineering units

Without those details, a measured 12 mA could mean 50 PSI, 50°C, 2,500 litres—or something else entirely.

Step 2: Inspect the Transmitter

Check:

• Device power
• Local display
• Fault indicators
• Process connection
• Sensor condition
• Configuration
• Range settings
• Output mode
• Terminal tightness

A pressure transmitter connected to a blocked impulse line may output a stable but incorrect value even though its electronics are working perfectly.

Not every analog problem is electrical.

Step 3: Measure the Signal at the Source

Use suitable test equipment to measure what the transmitter is actually producing.

For a 4–20 mA loop, a loop calibrator is ideal.

A multimeter can also measure current, but it must be connected in series. Connecting a meter configured for current directly across a power source can create a short circuit and blow the meter fuse—or worse.

Where suitable test points are provided, use them according to the device instructions.

Compare the measured current with the expected process condition.

Example:

• Actual pressure: approximately 50 PSI
• Transmitter range: 0–100 PSI
• Expected current: approximately 12 mA

If the transmitter outputs 7 mA, the problem is before the PLC.

Step 4: Measure the Signal at the PLC

Compare the signal at the transmitter with the signal reaching the analog input.

If the transmitter sends 12 mA but only 9 mA reaches the input circuit, investigate:

• Wiring
• Loose terminals
• Parallel loads
• Incorrect series connection
• Junction-box corrosion
• Cable damage
• Isolation barrier
• Signal conditioner

In a healthy series current loop, the same current should flow throughout the loop.

Step 5: Compare Electrical Signal With Raw Counts

Go online with the PLC and inspect the raw analog value.

Use the module documentation to determine what raw count should correspond to the measured signal.

Suppose:

• 4 mA = 0
• 20 mA = 32,767
• Measured signal = 12 mA

The expected raw value is approximately 16,384.

If the meter shows 12 mA but the PLC reports 5,000 counts, check:

• Module configuration
• Selected channel type
• Raw data format
• Input module
• Terminal wiring
• Channel damage

Step 6: Compare Raw Counts With the Scaled Value

If the raw value is correct but the engineering value is wrong, focus on software.

Check:

• Raw minimum
• Raw maximum
• Engineering minimum
• Engineering maximum
• Data type
• Scaling instruction
• Register mapping
• HMI conversion
• Units
• Clamping logic

This usually indicates a configuration problem rather than a transmitter or wiring problem.

Step 7: Inject a Known Signal

Disconnect the transmitter safely and inject a known signal using a loop calibrator or signal generator.

Useful test points include:

• 4 mA
• 8 mA
• 12 mA
• 16 mA
• 20 mA

For a 0–100 PSI range, those values should correspond approximately to:

Test several points, not just zero and span.

A system may be correct at both endpoints and still have a linearity or configuration problem in the middle.

Step 8: Try Another Analog Channel

Where the hardware and configuration allow it, move the signal to a known working spare channel.

Configure the spare channel correctly first.

If the problem follows the signal, suspect:

• Transmitter
• Wiring
• Signal conditioner
• Process connection

If the problem remains with the original channel, suspect:

• Analog input channel
• Terminal block
• Module configuration
• Backplane or module fault

Change one thing at a time and document the original wiring.

Step 9: Look for Noise Patterns

Electrical noise often creates readings that jump or fluctuate.

Observe when the disturbance occurs.

Does the reading change when:

• A VFD starts?
• A contactor switches?
• A welder operates?
• A heating bank turns on?
• A cable chain moves?
• A cabinet fan stops?
• Another machine begins production?

A repeatable relationship points toward routing, shielding, grounding or power-quality problems.

Move a temporary test cable away from the noise source as a diagnostic test.

If the signal becomes stable, you have learned something useful. Do not leave the temporary cable draped across the factory floor as the permanent repair.

Step 10: Check PLC and HMI Filtering

Filtering can smooth a noisy signal.

Possible filtering methods include:

• Module hardware filtering
• Digital averaging
• Moving average
• Low-pass filter
• Time-based debounce
• Rate limiting

Filtering can make a display easier to read, but it also slows the response.

Too much filtering may hide rapid process changes or delay alarms.

Do not use software filtering as a substitute for repairing damaged shielding or poor cable routing.

Fix the noise source first where practical.

Common Analog Faults and What They Usually Mean

Reading Stuck at Zero

Possible causes include:

• Open circuit
• No transmitter power
• Reversed polarity
• Broken cable
• Incorrect channel configuration
• Failed transmitter
• Blown input fuse
• Wrong PLC address

On a 4–20 mA loop, zero current usually indicates a fault rather than a valid minimum measurement.

Reading Stuck Near Minimum

Possible causes include:

• Process genuinely at minimum
• Transmitter output fixed at 4 mA
• Sensor not responding
• Blocked process connection
• Wrong transmitter range
• Insufficient loop voltage
• Scaling minimum configured incorrectly

Reading Stuck at Maximum

Possible causes include:

• Process overrange
• Shorted signal
• Transmitter fault mode
• Channel configured for the wrong signal type
• 20 mA simulation left active
• Incorrect scaling
• Damaged analog input

Stable but Incorrect Reading

Possible causes include:

• Wrong scaling values
• Wrong transmitter range
• Offset error
• Incorrect HMI units
• Calibration drift
• Wrong register or channel
• Incorrect signed/unsigned data type

Stable does not mean accurate.

Randomly Jumping Reading

Possible causes include:

• Electrical noise
• Loose connection
• Broken conductor
• Poor shield termination
• Ground loop
• Unstable transmitter power
• Moisture in a junction box
• Damaged moving cable
• Failing sensor

Reading Changes When a Motor Starts

Suspect:

• Voltage dip
• VFD interference
• Poor grounding
• Shared power supply
• Signal cable routed with motor leads
• Weak shield connection

Correct Raw Value but Wrong HMI Value

Suspect:

• Incorrect PLC scaling
• Additional HMI scaling
• Wrong tag
• Incorrect units
• Decimal-point error
• Communication register mismatch

One Channel Is Bad but Others Work

Possible causes include:

• Failed channel
• Damaged terminal
• Configuration difference
• Field wiring fault
• Shared common problem
• Channel-specific fuse or protection

All Channels Are Bad

Possible causes include:

• Missing module power
• Incorrect module configuration
• Backplane fault
• Shared common disconnected
• Power-supply problem
• Program addressing issue
• Entire module failure

Calibration: Keeping Measurements Accurate

Scaling and calibration are related, but they are not the same thing.

Scaling converts one numerical range into another.

Calibration compares an instrument against a known standard and adjusts or records its accuracy.

A perfectly scaled PLC cannot correct a pressure transmitter that is physically reading 5 PSI too high—unless software compensation is deliberately added, and even then, the underlying instrument should usually be investigated.

Zero and Span

Two important calibration concepts are:

• Zero
• Span

For a 0–100 PSI transmitter:

• Zero corresponds to 0 PSI or 4 mA.
• Span is the difference between maximum and minimum: 100 PSI.

For a −50 to 150°C transmitter:

• Zero or lower-range value is −50°C.
• Span is 200°C.

A zero error shifts the entire measurement.

A span error changes the measurement slope.

Zero Error Example

A transmitter outputs:

• 4.8 mA at 0 PSI
• 20.8 mA at 100 PSI

The output is shifted upward by 0.8 mA across the range.

That suggests a zero-offset problem.

Span Error Example

A transmitter outputs:

• 4 mA at 0 PSI
• 18 mA at 100 PSI

The lower point is correct, but the upper point is too low.

That suggests a span or gain problem.

Five-Point Calibration Check

A useful verification may test:

• 0%
• 25%
• 50%
• 75%
• 100%

For higher-accuracy applications, points may also be checked while decreasing the input to identify hysteresis.

Record:

• Applied process value
• Expected current
• Measured current
• PLC raw value
• PLC scaled value
• HMI displayed value

This reveals where the error enters the measurement chain.

Use Suitable Calibration Equipment

Depending on the signal, useful equipment includes:

• Loop calibrator
• Process calibrator
• Pressure calibrator
• Temperature dry-block calibrator
• Resistance decade box
• Precision voltage source
• Reference thermometer
• Deadweight tester

The equipment should have suitable accuracy and a valid calibration status for critical measurements.

A cheap handheld multimeter can be useful for fault-finding.

It may not be suitable as the reference standard for calibrating a critical process instrument.

Do Not Calibrate Around a Wiring Fault

Suppose poor wiring causes the PLC to read 11.6 mA while the transmitter outputs 12 mA.

Adjusting the transmitter until the PLC displays the expected value does not fix the system.

It hides the wiring error by deliberately miscalibrating the transmitter.

Repair the signal path first.

Then calibrate the instrument.

Document Calibration Work

Record:

• Instrument tag
• Manufacturer and model
• Serial number
• Calibrated range
• Date
• Reference equipment
• As-found values
• Adjustments made
• As-left values
• Technician
• Next calibration date

The as-found reading is especially important.

It shows how far the instrument drifted before adjustment and helps determine whether the calibration interval is appropriate.

Do Not Adjust Every Instrument Automatically

Some devices remain stable for years.

Others drift more quickly because of:

• Process vibration
• Temperature cycling
• Pressure shock
• Chemical exposure
• Mechanical wear
• Sensor ageing

Calibration should follow the process risk, manufacturer guidance, regulatory requirements and historical performance.

Adjusting a stable instrument unnecessarily can make it worse.

Verify first. Adjust only when justified.

Analog Signal Problems by Symptom

Protecting Analog Inputs From Damage

Analog input channels are designed for a specific signal range.

Common mistakes include connecting:

• 24 V DC to a low-level voltage input
• Mains voltage to an analog channel
• Current signal into a channel configured for voltage
• Voltage signal into a current input
• Incorrect external power
• Two powered outputs against each other

These mistakes can damage the channel immediately.

Before connecting a wire, verify:

• Signal type
• Expected range
• Channel configuration
• Terminal number
• Polarity
• Common connection
• Maximum permitted voltage
• Need for external resistance

Do not identify a conductor by colour alone.

Measure it and trace it against the drawing.

A Practical Analog Troubleshooting Example

Imagine a 0–10 bar pressure transmitter using 4–20 mA.

The HMI shows 6 bar, but a calibrated mechanical gauge shows approximately 5 bar.

Step 1: Calculate the Expected Current

The process is at 50% of the transmitter range.

Therefore, expected current is:

4 mA + (50% × 16 mA)

4 mA + 8 mA = 12 mA

Step 2: Measure the Transmitter Output

The loop calibrator measures 12.02 mA.

The transmitter appears accurate.

Step 3: Check the PLC Raw Value

The module is configured for:

• 4 mA = 6,553
• 20 mA = 32,767

At approximately 12 mA, the expected raw value is around 19,660.

The PLC reports 19,670.

The module and wiring are likely working correctly.

Step 4: Inspect the Scaling

The PLC scaling block contains:

• Raw minimum: 0
• Raw maximum: 32,767
• Engineering minimum: 0 bar
• Engineering maximum: 10 bar

There is the problem.

The module does not produce zero counts at 4 mA. Its raw minimum should be 6,553.

After correcting the raw minimum, the displayed pressure changes to approximately 5 bar.

No transmitter replacement required.

No new cable.

Just one incorrect parameter quietly lying to everyone.

Building Better Analog Logic

Good analog programming includes more than one scaling calculation.

A robust signal block may provide:

• Raw input
• Measured current or voltage
• Scaled engineering value
• Low-low alarm
• Low alarm
• High alarm
• High-high alarm
• Underrange fault
• Overrange fault
• Open-loop detection
• Quality or validity bit
• Filtering
• Calibration offset
• Simulation status
• Alarm delay

This creates a clear and reusable structure.

The rest of the program should not blindly use a process value when its quality is bad.

For example, if a pressure signal is invalid, the system may need to:

• Stop a pump
• Hold the last valid value
• Switch to a backup transmitter
• Generate an alarm
• Prevent automatic startup

The correct response depends on process risk.

Use Clear Tag Names

A useful signal structure might include:

• PT101_Raw
• PT101_mA
• PT101_Pressure_bar
• PT101_Underrange
• PT101_Overrange
• PT101_BadQuality
• PT101_HighAlarm

That is much easier to understand than:

• N7:42
• TempReal3
• AI_12_Final_New

Good naming does not make the transmitter more accurate, but it makes future troubleshooting far less painful.

Final Thoughts

Analog scaling is the bridge between an electrical signal and a meaningful process value.

The transmitter sends current or voltage. The analog module converts that signal into a raw number. The PLC then scales the raw number into PSI, degrees, litres, kilograms or another engineering unit.

Every stage must agree:

• Transmitter range
• Signal type
• Module configuration
• Raw-count range
• PLC scaling
• HMI units

If one part is wrong, the system can display a believable number that has little connection to reality.

When an analog reading misbehaves, begin at the source. Verify the real process, measure the transmitter output, compare it with the PLC raw input and only then inspect the scaling calculation.

For unstable readings, look closely at wiring, shielding, grounding, cable routing and power. A surprising number of mysterious analog faults turn out to be loose terminals, damaged cables or electrical noise.

The calculation matters.

The wiring matters more often.

And before adjusting a transmitter to make the screen look right, make absolutely sure the screen is doing its part correctly.