Phase Academy Interactive Power Systems Lab

Industrial Sensor & Signal Simulator

Process → Sensor → Signal → PLC
📡 Signal Chain
Physical process to operator display. Each stage can distort the value in its own way.

💡 Diagnosis

Adjust the controls to see how the signal chain behaves.

📊 Live Signal
📌 Loop Snapshot
📈 Signal Trend
Transmitted signal over the last few seconds, with NAMUR NE 43 diagnostic bands shaded.
Transmitted signal NE 43 fault threshold Saturation limit
📉 Process Variable Trend
True process value in engineering units, for comparison against what the PLC believes.
Actual PV PV as seen by the PLC
🔍 NAMUR NE 43 Bands
NE 43 defines how a 4–20 mA transmitter signals its own failure using current outside the measuring range.
BandCurrentMeaningStatus
🔧 Instrument Detail
📏 Transfer Function
Process variable to signal, with the current operating point marked.
⚡ Loop Power & Compliance
A 2-wire loop only works if the supply can push 20 mA through every series resistance and still leave the transmitter its minimum terminal voltage.
🔌 Wiring Topologies
Two-wire loop-powered versus four-wire self-powered transmitters.

2-Wire (Loop Powered)

Power and signal share one pair
24 VDC SUPPLY Class 2 XMTR 4-20 mA PLC AI 250 Ω sense +24 V LOOP RETURN / COM I

Where you see it: most pressure, level, and temperature transmitters. The transmitter regulates loop current to represent the measurement while drawing its own operating power from that same current, so its electronics must run on less than 4 mA at zero scale.

4-Wire (Self Powered)

Separate power pair and signal pair
120 VAC or 24 VDC ANALYZER own PSU 4-20 mA out PLC AI isolated PWR+ PWR- SIG+ SIG-

Where you see it: pH and conductivity analyzers, magnetic flow meters, chromatographs — anything whose electronics, display, or excitation current needs more power than a 4 mA budget allows. The signal pair is independent of the power pair, so the output can be isolated.

📋 Signal Mapping Table
Quarter-span checkpoints for the current instrument configuration. The highlighted row is nearest the live operating point.
% of SpanProcess ValueSignalSense VoltagePLC Raw
📏 Scaling Visualizer
Electrical signal on the left, engineering units on the right. The marker is the live operating point.
Signal
4 mA20 mA
Engineering
0100
🧮 Scaling Math
Every step from the wire to the operator screen, with the live numbers substituted.
🎮 Raw Count Lookup
Work the conversion backwards: pick a raw count and see the signal and engineering value it represents.
💾 Analog Input Module Resolution
Resolution sets the smallest process change the PLC can even represent. It is not the same thing as instrument accuracy.
ResolutionCountsSignal per CountEU per Count

💡 How Scaling Works

The analog input module converts the loop signal to an integer. Scaling is a straight linear map from that integer to engineering units:

EU = (Raw − RawMin) / (RawMax − RawMin) × (URV − LRV) + LRV

The most common field mistake is scaling to the module’s count range while the instrument is ranged differently — the math is right and the reading is still wrong. Always confirm the transmitter LRV/URV and the module scaling agree before you trust the number on the screen.

🎛 Controller Parameters
A first-order-plus-dead-time process with a PID controller in standard parallel form.
Repeats per minute style action. Ki = Kp / Ti in this model’s parallel form.
Transport delay before the process responds at all. Dead time is what makes aggressive tuning unstable.
Anti-windup clamping
Feed PV from the simulated loop
With this on, injected faults corrupt the controller’s feedback the way they would in the plant. Calibration and noise faults apply throughout; an open wire is injected 40% of the way into the test so you see the healthy response and the failure side by side.
📈 Step Response
Setpoint Process variable Controller output
🎯 Tuning Presets
Each preset demonstrates one characteristic response shape. Run a step test after loading.
📏 Performance Metrics

💡 Tuning Feedback

Run a step test to evaluate the current tuning.

🎓 Guided Scenarios
Each scenario loads a complete instrument configuration and fault state. Read the symptom, then work the diagnosis.
📊 Scenario Trend
🔍 Troubleshooting Walkthrough

Select a scenario to load it and see the field diagnosis.

📝 Fault Signature Reference
What each failure looks like at the PLC, and what it does not look like.
Fault4–20 mA Signature0–10 V SignatureFirst Check
Open wire0 mA — hard NE 43 under-range0 V — indistinguishable from 0%Loop current with a clamp or series meter
Shorted transmitterRises above 21 mA or pins to supply limitPins to 0 V or supply railTerminal voltage at the transmitter
EMI / noiseRapid jitter about the true valueLarger jitter, worse over distanceShield continuity and cable routing
Ground loopSteady 50/60 Hz ripple on the signalSame, plus a DC offsetShield grounded at one point only
Zero shiftConstant offset at every pointConstant offset at every pointRe-zero at the LRV condition
Span errorError grows with process valueError grows with process valueTwo-point calibration, LRV and URV
Lost cold-junction comp.Reading drifts with panel temperatureSame driftTerminal block temperature and CJC sensor

⚠ Read This First

Everything below is paraphrased for teaching and follows the NEC (NFPA 70) 2023 article numbering. Article numbers move between editions — Class 1 circuits, for example, moved from Article 725 to Article 724 in the 2023 edition. Always work from the adopted code edition in your jurisdiction and the manufacturer’s listed instructions, which NEC 110.3(B) makes mandatory in its own right.

📏 Conductor Sizing & Loop Resistance
Wire resistance is what turns a healthy loop into a compliance failure. Size the run, then check compliance on the Sensor & Wiring tab.
SizeDC ResistanceLoop Ω per 100 ftDrop at 20 mA per 100 ft
Resistance values are the uncoated solid copper figures from NEC Chapter 9, Table 8 (ohms per 1000 ft at 75 °C). Loop resistance counts both conductors, so a 250 ft run is 500 ft of copper.
📙 Wiring Methods & Circuit Classification

Class 2 power-limited circuits — NEC Article 725. Most 24 VDC instrument loops are installed as Class 2 circuits. The classification comes from the source, not the wire: the circuit has to be supplied by a listed Class 2 power source (or listed equipment marked as a Class 2 source), and that inherent power limitation is what relaxes the wiring rules compared with light and power circuits.

Separation from power conductors — NEC 725.136. Class 2 conductors may not occupy the same cable, raceway, or enclosure as electric light, power, or Class 1 conductors, except under the specific conditions the section lists (barriers, listed composite cables, and similar arrangements). This is the code rule behind the practice you already know as “keep signal wire out of the power conduit” — the electrical noise consequence and the code requirement point the same direction.

Mixed systems in one enclosure — NEC 300.3(C)(1). Where conductors of different systems do share a raceway or enclosure, every conductor’s insulation rating must be at least the maximum voltage present. A 300 V instrument cable does not belong in a box with 480 V conductors.

Mechanical execution — NEC 725.24 and NEC 300.11. Instrument cable has to be supported by the building structure and installed in a neat and workmanlike manner. Cable draped over a pipe is a violation before it is ever a noise problem.

A 4–20 mA loop that fails EMI troubleshooting has usually failed a separation rule first. Check the routing against 725.136 before reaching for a filter.
⚡ Hazardous Locations & Intrinsic Safety

Intrinsically safe systems — NEC Article 504. An IS loop limits energy so that no spark or thermal effect can ignite the hazardous atmosphere, which is why 4–20 mA instrumentation dominates in classified areas. The protection lives in the whole system — barrier, field wiring, and instrument together — and the control drawing is part of the listing.

Separation — NEC 504.30. IS conductors must be separated from non-IS conductors, including by the specified spacing where they are not in separate raceways or otherwise partitioned. Running one IS pair through a non-IS wireway voids the protection concept.

Identification — NEC 504.80. IS circuits must be labeled and their raceways and cables identified. Where color coding is used for IS conductors, light blue is the identified color, permitted only where no other light blue conductors are present.

Classified location wiring — NEC Article 501 (Class I) governs the wiring methods and sealing fittings themselves. Intrinsic safety is one permitted protection technique among several; explosionproof and purged enclosures are others.

Barrier resistance is a real part of your loop budget. An IS barrier adds resistance in series with the sense resistor and the wire, and it is the reason many IS loops run out of compliance voltage that an unclassified version of the same loop would have handled fine.
🌐 Grounding, Shielding & Bonding

Bonding — NEC Article 250, including 250.96(A). Metal enclosures, raceways, and equipment must be bonded together to form an effective ground-fault current path. That is a safety requirement, and it is independent of anything you do with a cable shield.

Single-point shield grounding is practice, not NEC text. The NEC does not tell you which end of an instrument cable shield to land. Grounding a shield at both ends creates a circulating path when the two ground references sit at different potentials, and that is exactly the 50/60 Hz ripple this simulator injects under “ground loop.” Ground the shield at one end — conventionally the control panel — and leave the field end insulated and terminated.

Isolation as an alternative. Where two grounded devices genuinely must connect, a signal isolator galvanically breaks the loop rather than fighting it. This is the standard fix when a 4-wire analyzer with a grounded output feeds a PLC input that is also grounded.

📝 Standards Behind the Signals
StandardCoversWhy it matters here
ANSI/ISA-50.00.014–20 mA DC transmission signalDefines the analog signal this whole tool is built around
NAMUR NE 43Failure signal levels on 4–20 mA≤3.6 mA and ≥21 mA carry diagnostic meaning, not measurement
IEC 60751Industrial platinum RTDsPt100 = 100 Ω at 0 °C; Callendar–Van Dusen curve
IEC 60584Thermocouple reference tablesType K mV-versus-temperature values and tolerance classes
ISA-5.1Instrumentation symbols and tag identificationPT, TT, FT, LT loop tagging on P&IDs
IEC 61511 / 61508Safety instrumented systemsWhere diagnostic coverage and fail-safe direction become design requirements
NFPA 79Electrical standard for industrial machineryControl panel conductor colors and machine wiring practice
Typical NFPA 79 control conductor colors: red for AC control, blue for DC control, orange for circuits energized from an external source that remain live with the disconnect open, white for the grounded AC control conductor, and green (with or without a yellow stripe) for equipment grounding. Confirm against the edition adopted for your machine.

💡 What This Tool Does Not Do

This simulator has no code-compliance engine. It does not evaluate conductor ampacity, box fill, overcurrent protection, area classification, SIL verification, or hazardous-location suitability, and its fault models are illustrative rather than instrument-specific. Use it to build intuition about the signal chain, then take real decisions to the adopted code, the listed instructions, and a qualified engineer.

🎯 Why the Signal Chain Matters

Every automated process depends on turning a physical quantity into a number a controller can act on. Pressure becomes a current, current becomes an integer, and an integer becomes a decision. Each conversion is a place the truth can change without anything announcing it.

A sensor can be telling the truth, lying, or disconnected. The control system reacts anyway.
⚡ 4–20 mA and the Live Zero

4 mA is 0% of span and 20 mA is 100%. The measurement never rides on zero current, which is what makes the scheme diagnosable: no current at all cannot be a valid reading, so it must be a broken loop.

Current also survives the trip down the cable. In a series loop the same current flows at every point, so conductor resistance does not change the measurement — it only eats supply voltage. That is why a current loop can run thousands of feet where a voltage signal cannot.

Common mistake: reading 0 mA as “the process is at zero.” Zero process is 4 mA. Zero current is a wiring fault.
🔌 Loop Compliance Voltage

A 2-wire transmitter needs a minimum terminal voltage — commonly around 12 V, though it varies by instrument — to run its electronics. Everything else in the loop takes its share of the supply as I×R at the worst-case 20 mA.

Vsupply ≥ Vxmtr,min + 0.020 × (Rsense + Rwire + Rbarrier)

Rearranged, that gives the maximum total loop resistance a supply can drive. This is the calculation behind “it worked on the bench and died in the field” — the bench had three feet of wire and no IS barrier.

Compliance failures hide at the top of the range. The loop reads perfectly at 4 mA and goes flat somewhere above 80% of span.
🌡 RTDs and Thermocouples

RTD. A platinum element whose resistance rises predictably with temperature. Pt100 means 100 Ω at 0 °C, and about 138.5 Ω at 100 °C. Accurate and stable, but slower, and it needs excitation current from the transmitter.

Lead wire matters. Because an RTD measures small resistance changes, lead resistance shows up directly as error. A 3-wire connection cancels it when the leads are reasonably balanced; a 4-wire Kelvin connection removes it outright and is the choice for laboratory-grade work.

Thermocouple. Two dissimilar metals produce a voltage that depends on the temperature difference between the measuring junction and the terminal block. Type K gives roughly 41 µV per °C near ambient — a tiny signal in an electrically hostile place.

Cold-junction compensation is not optional. Without it, the reading tracks the panel temperature, and the error looks like a slow process drift instead of an instrument problem. Toggle it off in the sidebar with a thermocouple selected to see the signature.
🔢 Scaling and Resolution

The analog input module reports an integer. A 15-bit module gives 0–32767 counts, a 12-bit module only 0–4095. Scaling maps that integer linearly onto the instrument’s calibrated range.

Resolution is not accuracy. A 16-bit module reading a transmitter with 0.5% error still has 0.5% error — now reported to five decimal places. Resolution bounds what the module can represent; accuracy is a property of the instrument and its calibration.

Raw counts are not a personality. They still need scaling.
📌 Dry Contacts and Discrete Signals

A dry contact carries no source of its own — it just opens or closes, and the input card supplies the sensing voltage. Limit switches, float switches, and relay contacts all work this way.

A wet contact brings its own voltage. Feeding one into an input expecting a dry contact can read incorrectly or damage the card. Match the contact type to the input type before you land the wire.

Supervision buys diagnosis. An ordinary discrete input cannot tell an open contact from a cut wire — both are simply open. Add an end-of-line resistor and a supervised input, and the card can distinguish normal open, normal closed, open circuit, and short circuit by the resistance it measures.

Sourcing outputs supply current to the load; sinking outputs accept current from it. Getting this backwards is the most common reason a correctly programmed discrete point never turns on.
🔧 Fault Signatures

Open wire: current collapses to zero. Below the NE 43 threshold, so a 4–20 mA input flags it immediately; a 0–10 V input cannot tell it from a genuine zero.

Short: current climbs past the top of the range or pins at whatever the supply can deliver. Above 21 mA is failure information, not measurement.

Noise: fast random jitter around the true value. Sources are VFDs, contactors, and unshielded runs beside power conductors.

Ground loop: a steady 50/60 Hz ripple, sometimes with an offset. Comes from two ground references at different potentials, not from the instrument.

Zero shift versus span error: a zero shift offsets every reading equally; a span error grows with the process value. Check the instrument at both ends of its range and the two are unmistakable.

If the error is the same at 0% and 100%, it is a zero problem. If it grows as you go up, it is a span problem. If it moves on its own, look at the wiring.
🎛 PID in One Card

Proportional acts on present error — fast, but on its own it settles with a permanent offset. Integral acts on accumulated error and removes that offset, at the cost of overshoot and windup. Derivative acts on the rate of change and damps the result, while amplifying any noise you feed it.

Dead time is the real constraint. During transport delay the controller is acting on stale information. The longer the dead time relative to the time constant, the more conservative the tuning has to be — no amount of gain fixes a loop that cannot see what it just did.

Windup happens when the output saturates and the integral keeps accumulating anyway. When the process finally responds, all that stored error has to be unwound, and the result is a large overshoot. Clamping the integral is the standard defense.

Never tune a PID against a noisy signal with derivative action turned up. Derivative multiplies the noise, the output chatters, and the valve pays for it.
🎓 Knowledge Check
Question drawn at random from the bank. 0 / 0