Instrumentation Calculators
Five free instrumentation engineering calculators covering 4-20 mA signal conversion, square root flow extraction, loop voltage budget, loop power supply sizing and thermowell wake frequency analysis. Each has its own page with the governing formula, a worked example, accepted input ranges and a printable report. Nothing to install and no sign-up.
This hub also keeps the thermocouple type reference table and the process-measurement design guide, both below, for the background a field engineer usually needs alongside the numbers.
IEC 60584 · IEC 60381-1 · NAMUR NE43 · ASME PTC 19.3 TW · HART
Detailed Calculator Pages
Signal and Flow
📡4-20 mA Calculator
Current to engineering units and back, with NAMUR NE43 fault-zone limits.
🔄Square Root Flow Calculator
DP transmitter linearisation with a low-flow cutoff, and where the extraction belongs.
Loop Power and Integrity
🔗Instrument Loop Calculator
Cable voltage drop and transmitter headroom across the loop at 20 mA.
⚡Loop Power Supply Calculator
AI card, isolator and IS barrier load budget against the available supply.
Mechanical Integrity
Frequently Asked Questions
What is the 4-20 mA current loop and why is it still the industry standard?
A 4-20 mA loop carries a measurement as a current rather than a voltage, so the signal is unaffected by cable resistance and voltage drop along the run. The live zero at 4 mA is what makes it valuable: a broken wire reads 0 mA, which is outside the signal range and therefore distinguishable from a genuine zero reading. The same two wires can also power the transmitter, which halves the field wiring.
How do you convert a 4-20 mA signal to engineering units?
Subtract the 4 mA live zero, divide by the 16 mA span, then multiply by the engineering span and add the lower range value. A 12 mA reading on a 0 to 250 degrees Celsius range gives 12 minus 4, over 16, times 250, which is 125 degrees. The relationship is linear except where the transmitter applies square root extraction for differential pressure flow.
When should square root extraction be applied to a DP flow transmitter?
Flow through a differential pressure element varies with the square root of the measured differential, so the square root must be taken exactly once between the element and the display. Modern smart transmitters usually do it internally, and the most common configuration error is applying it a second time in the DCS. Check where it is configured before assuming the reading is wrong.
What causes thermowell failure and how is it checked?
Fluid flowing past a thermowell sheds vortices at a frequency set by the velocity and the tip diameter. If that shedding frequency approaches the natural frequency of the well, resonance can fatigue it until it breaks off in the line. ASME PTC 19.3 TW sets the acceptance criteria, and the frequency ratio is the first check to run whenever velocity, length or process fluid changes.
How much cable resistance can a 4-20 mA loop tolerate?
The supply voltage must cover the transmitter's minimum operating voltage plus the drop across every series element at 20 mA, including the receiver resistor, any isolator or intrinsic safety barrier, and the loop cable. Work the budget at 20 mA because that is the worst case, and leave headroom for the barrier resistance and for cable resistance rising with ambient temperature.
📋
Thermocouple Type Reference
Field selection guide — IEC 60584 standard types
| Type | + Wire | − Wire | Range (°C) | Sensitivity (µV/°C) | Application | Notes |
|---|
Instrumentation Design Guide
Field reference and theory behind the five calculators — signal fundamentals, DP flow, loop budgets, thermowell vibration and thermocouples.
01 4–20 mA Current Loop
Why 4–20 mA?
The 4–20 mA standard (IEC 60381-1) has been the backbone of process instrumentation since the 1950s. Current — not voltage — is transmitted because current does not drop over long cable runs. A 500 Ω load 1 km away receives exactly the same current as one sitting next to the transmitter.
The live-zero at 4 mA is the key insight: a dead transmitter or open cable reads 0 mA — unmistakably different from the 4 mA "zero" signal. This is called fault detection by design. If you see 0 mA, you have a wiring fault, not a genuine zero process reading.
The live-zero at 4 mA is the key insight: a dead transmitter or open cable reads 0 mA — unmistakably different from the 4 mA "zero" signal. This is called fault detection by design. If you see 0 mA, you have a wiring fault, not a genuine zero process reading.
The Linear Equation
The conversion between current and engineering value is a simple linear interpolation:
EU = EU_min + [(I − 4) / 16] × (EU_max − EU_min)
I = 4 + [(EU − EU_min) / (EU_max − EU_min)] × 16
I = 4 + [(EU − EU_min) / (EU_max − EU_min)] × 16
where I is in mA and the span is always 16 mA (20 − 4). Key checkpoints:
4 mA = 0% span (live zero)
12 mA = 50% span (midscale)
20 mA = 100% span (full scale)
4 mA = 0% span (live zero)
12 mA = 50% span (midscale)
20 mA = 100% span (full scale)
Reversed & Split Ranges
Some transmitters use reversed ranges (e.g. 100→0 bar) where 4 mA = full scale and 20 mA = zero. The formula still works — EU_min and EU_max simply swap signs in the result.
Split ranging is common in valve positioners: two valves share one controller output, with Valve A responding to 4–12 mA and Valve B responding to 12–20 mA. Each valve uses its own 8 mA sub-span.
Split ranging is common in valve positioners: two valves share one controller output, with Valve A responding to 4–12 mA and Valve B responding to 12–20 mA. Each valve uses its own 8 mA sub-span.
HART Protocol
HART (Highway Addressable Remote Transducer) overlays a 1200 / 2200 Hz FSK digital signal on top of the 4–20 mA loop without disturbing the analog value. This gives you simultaneous analog control and digital diagnostics — access to secondary variables, device configuration, and health status — over the same two wires. HART is defined in IEC 62591.
📐 Worked Example — Pressure Transmitter
A pressure transmitter is ranged 0–100 bar. The DCS reads 14.4 mA. What is the pressure?
EU = 0 + [(14.4 − 4) / 16] × (100 − 0) = (10.4 / 16) × 100 = 65.0 bar
Now the operator notices the signal is 3.9 mA. Is this possible?
No — 3.9 mA is below the live-zero threshold. This indicates a broken wire, failed transmitter, or loss of loop power. Process reading is unavailable. Investigate the field device immediately.
EU = 0 + [(14.4 − 4) / 16] × (100 − 0) = (10.4 / 16) × 100 = 65.0 bar
Now the operator notices the signal is 3.9 mA. Is this possible?
No — 3.9 mA is below the live-zero threshold. This indicates a broken wire, failed transmitter, or loss of loop power. Process reading is unavailable. Investigate the field device immediately.
02 Square Root Extraction & DP Flow Measurement
Why Square Root?
Differential pressure (DP) flow elements — orifice plates, venturis, Pitot tubes — generate a pressure drop that is proportional to the square of flow velocity. This is derived from Bernoulli's equation:
Q = Cd × A × √(2·ΔP / ρ)
Q ∝ √ΔP (for constant ρ)
Q ∝ √ΔP (for constant ρ)
The DP transmitter signal is linear with ΔP. To get a linear flow reading you must apply the square root. Failure to do so causes severe non-linearity and low-flow reading error — at 25% DP, actual flow is 50% of span, not 25%.
Low-Flow Cutoff
At very low DP signals, noise in the transmitter is amplified by the square root function, causing the displayed flow to jitter erratically. A low-flow cutoff (typically 1–2% of span) forces the output to zero below this threshold, preventing false readings and protecting integrating totalizers from accumulating noise.
Setting it too high causes undercount at genuine low flows. Most flow computers implement this as a configurable parameter.
Setting it too high causes undercount at genuine low flows. Most flow computers implement this as a configurable parameter.
Where the Extraction Happens
Square root extraction can be implemented at three points:
1. Inside the transmitter — HART-enabled DP transmitters can be configured to output a linearised (square-root) 4–20 mA signal directly.
2. In the DCS/PLC — the raw DP signal is wired in and the square root function block is applied in software. Most common approach.
3. Dedicated flow computer — used when multivariable compensation (T, P correction) or custody transfer accuracy is required (ISO 5167, AGA-3).
1. Inside the transmitter — HART-enabled DP transmitters can be configured to output a linearised (square-root) 4–20 mA signal directly.
2. In the DCS/PLC — the raw DP signal is wired in and the square root function block is applied in software. Most common approach.
3. Dedicated flow computer — used when multivariable compensation (T, P correction) or custody transfer accuracy is required (ISO 5167, AGA-3).
📐 Worked Example — Orifice Plate
An orifice plate is ranged 0–500 m³/h (at 100% DP). The DP transmitter reads 36% of span. What is the actual flow?
Q = 500 × √(36 / 100) = 500 × 0.6 = 300 m³/h
If you forgot the square root and read 36% linearly: Q = 500 × 0.36 = 180 m³/h — a 40% undercount. This is the single most common error in commissioning DP flow loops.
Q = 500 × √(36 / 100) = 500 × 0.6 = 300 m³/h
If you forgot the square root and read 36% linearly: Q = 500 × 0.36 = 180 m³/h — a 40% undercount. This is the single most common error in commissioning DP flow loops.
03 Loop Integrity & Voltage Budget
Ohm's Law Applied to 4–20 mA
A 4–20 mA loop is a series circuit. The supply voltage (typically 24 Vdc) must overcome every voltage drop in the loop and still leave sufficient voltage at the transmitter terminals to keep it powered.
V_supply = V_transmitter + I × R_total
R_total = R_cable + R_AI + R_barrier + R_iso
V_headroom = V_at_tx − V_tx_min
R_total = R_cable + R_AI + R_barrier + R_iso
V_headroom = V_at_tx − V_tx_min
Headroom must be ≥ 0 V at all times including the maximum current of 20 mA.
Cable Resistance & Temperature
Copper resistivity changes with temperature. For a 2-wire loop, cable resistance is:
ρ_Cu = 0.0168 Ω·mm²/m @ 20°C
R_20 = ρ × 2L / A_conductor
R_T = R_20 × [1 + 0.00393(T − 20)]
R_20 = ρ × 2L / A_conductor
R_T = R_20 × [1 + 0.00393(T − 20)]
A 500 m run of 1.5 mm² cable at 60°C adds about 14.4 Ω — consuming 0.29 V at 20 mA. Small, but it matters for long runs with intrinsically safe barriers.
IS Barriers & Safety Margins
Intrinsically safe (IS) Zener barriers add 80–300 Ω to the loop and clamp supply voltage. HART communicators need at least 250 Ω in the loop to function. The safety factor of 1.3× on transmitter minimum voltage accounts for:
— Supply voltage tolerance (±10%)
— Fuse resistance and connection resistances
— Future additions to the loop
— End-of-cable-life resistance increase
— Supply voltage tolerance (±10%)
— Fuse resistance and connection resistances
— Future additions to the loop
— End-of-cable-life resistance increase
Maximum Loop Load Formula
The maximum total resistance that a 24 V supply can drive while keeping 12 V at the transmitter:
R_max = (V_supply − V_tx_min) / 0.020
R_max = (24 − 12) / 0.020 = 600 Ω
R_max = (24 − 12) / 0.020 = 600 Ω
A standard DCS AI card (250 Ω) + HART resistor (0 Ω extra) + 500 m of 1.5 mm² cable (11 Ω) = 261 Ω. Well within 600 Ω. Add a Zener barrier (200 Ω) = 461 Ω. Still OK.
📐 Worked Example — Long-Run IS Loop
Supply = 24 V, Tx min = 12 V, AI = 250 Ω, IS barrier = 90 Ω, cable = 800 m × 1.0 mm² @ 50°C. Is the loop viable?
R_cable = 0.0168 × 2 × 800 / 1.0 × [1 + 0.00393 × (50 − 20)] = 29.5 Ω
R_total = 250 + 90 + 29.5 = 369.5 Ω
V_drop_@20mA = 0.020 × 369.5 = 7.39 V
V_at_tx = 24 − 7.39 = 16.6 V > 12 V → PASS ✓
Headroom = 16.6 − 12 = 4.6 V. With 1.3× safety: V_required = 1.3 × 12 + 7.39 = 23.0 V < 24 V → SAFE
R_cable = 0.0168 × 2 × 800 / 1.0 × [1 + 0.00393 × (50 − 20)] = 29.5 Ω
R_total = 250 + 90 + 29.5 = 369.5 Ω
V_drop_@20mA = 0.020 × 369.5 = 7.39 V
V_at_tx = 24 − 7.39 = 16.6 V > 12 V → PASS ✓
Headroom = 16.6 − 12 = 4.6 V. With 1.3× safety: V_required = 1.3 × 12 + 7.39 = 23.0 V < 24 V → SAFE
04 Thermowell Vibration Analysis — ASME PTC 19.3 TW
Vortex Shedding — The Physics
When fluid flows past a cylindrical thermowell, it separates into alternating vortices — the Kármán vortex street. These vortices shed alternately from each side, creating an oscillating lateral force on the well at a frequency governed by the Strouhal number:
f_s = St × V / d
St ≈ 0.22 (cylinder, Re 300–2×10⁵)
St ≈ 0.22 (cylinder, Re 300–2×10⁵)
If f_s approaches the thermowell's natural frequency, resonance occurs and the well can fail by fatigue in minutes to hours.
Natural Frequency — Euler Beam Theory
A thermowell behaves as a cantilevered beam fixed at the process connection. The natural frequency of a uniform cantilever is:
f_n = (1.875²/2π) × √(E·I / ρ_m·A) / L²
= (β²/2π) × √(E × d_avg² / 16·ρ_m) / L²
= (β²/2π) × √(E × d_avg² / 16·ρ_m) / L²
316 SS: E = 193 GPa, ρ_m = 7 950 kg/m³. Long, thin thermowells in high-velocity gas have the lowest natural frequency — most susceptible to resonance.
ASME PTC 19.3 TW Acceptance Criteria
ASME PTC 19.3 TW-2016 requires the wake frequency ratio (Scruton criterion) to be below 0.8:
f_s / f_n < 0.8 → SAFE
f_s / f_n 0.8–1.0 → CHECK Scruton No.
f_s / f_n > 1.0 → FAIL — redesign required
f_s / f_n 0.8–1.0 → CHECK Scruton No.
f_s / f_n > 1.0 → FAIL — redesign required
This calculator implements the simplified first-pass check. Full ASME PTC 19.3 TW also requires: Scruton number (mass damping), in-line resonance check, fatigue stress evaluation, and tapered/stepped geometry correction.
Design Remedies When f_s/f_n ≥ 0.8
In order of preference:
1. Reduce insertion length — f_n ∝ 1/L². Halving L quadruples f_n.
2. Increase tip OD — raises f_n (thicker wall = stiffer beam) but also raises f_s slightly.
3. Use tapered/stepped well — larger OD at root, smaller at tip. Reduces mass and raises f_n without penalising strength at the root.
4. Reduce process velocity — route via a larger bore bypass or reduce pump speed.
5. Use helical strakes — disrupts coherent vortex shedding; effective but imposes a manufacturing cost.
1. Reduce insertion length — f_n ∝ 1/L². Halving L quadruples f_n.
2. Increase tip OD — raises f_n (thicker wall = stiffer beam) but also raises f_s slightly.
3. Use tapered/stepped well — larger OD at root, smaller at tip. Reduces mass and raises f_n without penalising strength at the root.
4. Reduce process velocity — route via a larger bore bypass or reduce pump speed.
5. Use helical strakes — disrupts coherent vortex shedding; effective but imposes a manufacturing cost.
📐 Worked Example — Steam Line Thermowell
6″ steam header, V = 28 m/s, ρ_steam ≈ 10 kg/m³. Thermowell: U = 300 mm, tip OD = 11 mm (d = 0.011 m), 316 SS.
Wake frequency: f_s = 0.22 × 28 / 0.011 = 560 Hz
Natural frequency (simplified): f_n = (1.875²/2π) × √(193×10⁹ × (0.011)² / 16 × 7950) / (0.300)²
≈ 1.401 × √(14,793,750 / 127,200) / 0.09 ≈ 1.401 × 341.2 / 0.09 ≈ 5,314 Hz (simplified; actual tapered well will differ)
f_s / f_n = 560 / 5314 = 0.105 → SAFE ✓
Now consider a longer well U = 600 mm: f_n scales by (300/600)² → f_n ≈ 1328 Hz. f_s/f_n = 560/1328 = 0.42 → still safe. But at U = 900 mm: f_n ≈ 591 Hz, f_s/f_n = 0.95 → FAIL. This shows why thermowell insertion length is critical in high-velocity gas.
Wake frequency: f_s = 0.22 × 28 / 0.011 = 560 Hz
Natural frequency (simplified): f_n = (1.875²/2π) × √(193×10⁹ × (0.011)² / 16 × 7950) / (0.300)²
≈ 1.401 × √(14,793,750 / 127,200) / 0.09 ≈ 1.401 × 341.2 / 0.09 ≈ 5,314 Hz (simplified; actual tapered well will differ)
f_s / f_n = 560 / 5314 = 0.105 → SAFE ✓
Now consider a longer well U = 600 mm: f_n scales by (300/600)² → f_n ≈ 1328 Hz. f_s/f_n = 560/1328 = 0.42 → still safe. But at U = 900 mm: f_n ≈ 591 Hz, f_s/f_n = 0.95 → FAIL. This shows why thermowell insertion length is critical in high-velocity gas.
05 Thermocouple Theory & Selection
The Seebeck Effect
When two dissimilar metals are joined at one end and the junction is heated, a small voltage (EMF) develops that is proportional to the temperature difference between the hot junction and the reference (cold) junction. This is the Seebeck effect (1821).
V_emf = S_AB × (T_hot − T_ref)
S_AB = Seebeck coefficient [µV/°C]
S_AB = Seebeck coefficient [µV/°C]
Type K: S ≈ 41 µV/°C at 500°C. IEC 60584 polynomials convert millivolt tables to °C with ±1°C accuracy.
Cold Junction Compensation (CJC)
The transmitter/head amplifier measures the voltage at its terminals, which are at ambient (reference) temperature, not 0°C. CJC corrects for this by measuring the reference junction temperature (typically with an RTD or thermistor) and adding the corresponding offset voltage to the TC output before converting to temperature.
Without CJC, a Type K TC in a 40°C ambient would read approximately 1.6 mV low, corresponding to a ~40°C undercount.
Without CJC, a Type K TC in a 40°C ambient would read approximately 1.6 mV low, corresponding to a ~40°C undercount.
Extension & Compensating Cables
TC signal cables must be the same alloy pair as the thermocouple (extension grade) or have the same Seebeck coefficient at the reference junction temperature (compensating grade). Using ordinary copper cable introduces a second, uncontrolled junction — a systematic error of 5–30°C depending on ambient excursion.
Extension cables: same metals as TC (marked with matching colour code)
Compensating cables: cheaper alloys with matched EMF up to ~120°C
Extension cables: same metals as TC (marked with matching colour code)
Compensating cables: cheaper alloys with matched EMF up to ~120°C
Type Selection Rules of Thumb
Type K — General industrial default. 0–1260°C. Use unless a specific reason demands otherwise. Avoid in vacuum/reducing atmospheres (K decalibrates above 750°C in these conditions).
Type J — Legacy equipment only. Cap at 540°C to prevent rapid oxidation of the iron leg.
Type T — Food, pharma, cold storage, cryogenics. Best corrosion resistance of base-metal types.
Type N — Direct replacement for K above 1000°C. More stable, especially in cyclic thermal service.
Types R/S — Precious metal. Lab furnaces and precision high-temperature process only. Cost 10–30× Type K.
Type J — Legacy equipment only. Cap at 540°C to prevent rapid oxidation of the iron leg.
Type T — Food, pharma, cold storage, cryogenics. Best corrosion resistance of base-metal types.
Type N — Direct replacement for K above 1000°C. More stable, especially in cyclic thermal service.
Types R/S — Precious metal. Lab furnaces and precision high-temperature process only. Cost 10–30× Type K.
06 Loop Load Analysis — Component Impedances
Series vs Parallel Loads
In a 4–20 mA current loop all loads are in series — the same current flows through every component. This means impedances add arithmetically. Unlike a voltage bus, adding a load to a current loop does not affect other loads — only the total voltage budget matters.
If two devices must share the same 4–20 mA signal (e.g. DCS + local indicator), they are connected in series (not parallel — that would split the current and corrupt the reading).
If two devices must share the same 4–20 mA signal (e.g. DCS + local indicator), they are connected in series (not parallel — that would split the current and corrupt the reading).
Typical Component Impedances
AI card input resistor: 100–250 Ω
HART termination resistor: 250 Ω
Zener IS barrier: 80–300 Ω
Galvanic isolator: 0–50 Ω (powered)
Local indicator: 50–200 Ω
Positioner feedback: 100–600 Ω
Cable (1.5 mm², 100 m): ~2.3 Ω
HART termination resistor: 250 Ω
Zener IS barrier: 80–300 Ω
Galvanic isolator: 0–50 Ω (powered)
Local indicator: 50–200 Ω
Positioner feedback: 100–600 Ω
Cable (1.5 mm², 100 m): ~2.3 Ω
HART Communication Requirement
For a HART communicator to work reliably, the loop must present a minimum of 230–250 Ω to the FSK signal. If the AI card has only a 100 Ω input (some older systems), you must add an external 150 Ω HART resistor in series. Most modern DCS AI cards include 250 Ω internally. Check the hardware manual before commissioning.
07 Standards, Codes & Further Reading
⚠ Calculator Scope & Engineering Limitations
All calculators in this suite are first-pass sizing and checking tools. They should be used to identify feasibility, catch gross errors, and guide early design decisions. They do not replace rigorous engineering analysis or code compliance sign-off. Specific limitations:
4–20 mA / Loop Load: Resistances are treated as purely real (no inductive/capacitive effects). In practice, cable capacitance can limit HART bandwidth on very long runs (>1 km).
Thermowell: Uniform cross-section assumed. Reynolds number correction to Strouhal number (0.19–0.22 range), Scruton number (mass damping), and tapered/stepped geometry are not modelled. ASME PTC 19.3 TW software (e.g. REOTEMP TW Pro, STS TW) should be used for final qualification.
Square root / DP flow: Ideal fluid, incompressible single-phase assumed. No compensation for gas expansion factor Y, velocity approach factor, or discharge coefficient Cd variation with Reynolds number.
All results are indicative — verify with certified engineering documents and applicable standards before ordering equipment or signing off designs.
4–20 mA / Loop Load: Resistances are treated as purely real (no inductive/capacitive effects). In practice, cable capacitance can limit HART bandwidth on very long runs (>1 km).
Thermowell: Uniform cross-section assumed. Reynolds number correction to Strouhal number (0.19–0.22 range), Scruton number (mass damping), and tapered/stepped geometry are not modelled. ASME PTC 19.3 TW software (e.g. REOTEMP TW Pro, STS TW) should be used for final qualification.
Square root / DP flow: Ideal fluid, incompressible single-phase assumed. No compensation for gas expansion factor Y, velocity approach factor, or discharge coefficient Cd variation with Reynolds number.
All results are indicative — verify with certified engineering documents and applicable standards before ordering equipment or signing off designs.