An Ampere Meter — almost always called an ammeter — is an instrument that measures the electric current flowing through a circuit, in amperes. Electricians and panel builders rely on it daily to verify motor loads, troubleshoot blown breakers, and commission switchgear. The meter sits in series with the load (or wraps around the conductor in clamp form) and converts current into a readable deflection, digital value, or 4-20 mA signal. Outcome: you see whether a 15 A circuit is actually drawing 4 A or 14 A, before something melts.
Ampere Meter Interactive Calculator
Vary the shunt rating, ADC range, bit depth, and test current to see shunt-meter resolution and ADC counts.
Equation Used
The shunt converts current into a small voltage by Ohm's law. The ADC step size is its input range divided by 2^N, so current resolution is one voltage count divided by the shunt resistance. Counts show how much of the ADC range is used at the selected test current, at 50% nominal load, and at full-scale peak current.
- Ideal precision shunt with no temperature drift
- Unipolar ADC range from 0 to Vadc
- Nominal discharge point is 50% of shunt rated current
- Lead resistance and ADC offset are ignored
Inside the Ampere Meter
An ammeter measures current by forcing it to do something measurable. In the classic moving coil galvanometer design, current passes through a coil suspended in a permanent magnet field. The coil twists against a hairspring in proportion to current, and the needle points to a calibrated scale. For currents above a few milliamps the meter movement itself can't take the load, so we put a precision shunt resistor in parallel — typically 50 mV or 75 mV full-scale drop across the shunt — and the movement reads the small voltage across it. That shunt is why a 200 A panel meter is still safe to wire: the meter sees 75 mV, not 200 A.
For AC, you use either a current transformer (CT) that produces a 5 A or 1 A secondary proportional to the primary, or a Hall effect sensor that reads the magnetic field around the conductor. Clamp meters use the same trick — the jaw is a split CT core. Modern digital panel meters skip the moving coil entirely and sample the burden voltage with a 16 or 24-bit ADC, then display the RMS value.
Get the tolerances wrong and the readings lie. If the shunt's burden voltage is too high, you drop real load voltage and skew the circuit you're trying to measure. If the CT is undersized — say a 100:5 CT measuring 250 A — the core saturates, the secondary clips, and you read 90 A while the actual draw is killing your motor. Open-circuiting a live CT secondary will generate kilovolts across the terminals and destroy the CT, which is why every CT install has a shorting block. Polarity matters too — reverse the CT and you'll read negative power on a perfectly healthy circuit.
Key Components
- Moving Coil Movement (D'Arsonval): A coil on jewelled pivots inside a permanent magnet gap. Full-scale deflection typically requires 50 µA to 1 mA at around 50 mV across the coil. The hairspring constant sets the scale linearity — a stretched or fatigued spring drifts zero by 2-5% and is the most common reason an analog panel meter reads low after 20 years of service.
- Shunt Resistor: A precision low-value resistor in parallel with the movement, sized so full-scale current produces 50 mV or 75 mV drop. A 100 A / 75 mV shunt is 750 µΩ. Tolerance must be ±0.25% or better for revenue-grade accuracy, and the shunt must be derated to 66% of rating for continuous duty to keep self-heating under 30 °C rise.
- Current Transformer (CT): A toroidal core wound for a fixed turns ratio — 100:5, 200:5, 1000:5 are standard. Class 0.5 CTs hold ±0.5% accuracy across 5-120% of rated primary current. Burden rating in VA must match the meter input; exceed it and the secondary signal collapses on inductive loads.
- Hall Effect Sensor: Solid-state device that outputs a voltage proportional to magnetic flux around the conductor. Used in DC clamp meters and high-bandwidth current probes. Bandwidth runs DC to 100 kHz typical, with offset drift of around 1 mA/°C — meaning you need to zero the meter every time ambient temperature shifts more than 10 °C.
- ADC and Display Front-End: On a digital ammeter, a 16-bit or 24-bit sigma-delta ADC samples burden voltage at 1-10 kHz and computes true RMS over a 100 ms window. A true-RMS reading is essential on VFD-driven motors and LED loads where current is not sinusoidal — an averaging meter under-reads these loads by 10-40%.
Who Uses the Ampere Meter
Anywhere current flows and you care how much, an ammeter shows up. Industrial motor control centres use panel ammeters as the primary diagnostic for load condition. Electricians carry clamp meters because you can't put a series ammeter on a live 200 A feeder without killing the building. Solar installers verify string current to within 0.2 A to confirm panel matching. The instrument scales from microamps in a lab bench supply up to tens of kiloamps on aluminium smelter bus bars.
- Industrial Motor Control: Panel ammeters on Allen-Bradley CENTERLINE 2500 MCC buckets, reading via 100:5 CTs to display motor running current against nameplate FLA.
- Electrical Service & Maintenance: Fluke 376 FC clamp meters used by field electricians to verify three-phase current balance on rooftop HVAC units — phase imbalance over 10% means a failing contactor.
- Solar PV Installation: DC clamp meters like the Fluke 393 FC measuring string current on Enphase or SMA inverters during commissioning to confirm each string is within 0.5 A of its neighbours.
- EV Charging Infrastructure: Current transformers feeding revenue-grade meters inside ChargePoint CT4000 stations to bill kWh accurately at ±0.5% on 80 A circuits.
- Battery & R&D Labs: Keysight N6705C source measure units logging charge/discharge current down to 100 nA on 18650 lithium cell characterisation.
- Utility Substations: 1200:5 protection-class CTs feeding SEL-751 feeder relays for overcurrent protection on 13.8 kV distribution feeders.
The Formula Behind the Ampere Meter
The fundamental sizing formula for an ammeter with a shunt is Ohm's law applied to the shunt. What matters in practice is matching the shunt to the expected operating current range. Pick a shunt sized for your maximum current and you get good headroom but poor resolution at light loads — a 200 A shunt reading a 5 A motor will live in the bottom 2.5% of its range where ADC noise dominates. Pick too tight a shunt and the burden voltage at peak load disturbs the circuit and the resistor cooks. The sweet spot is sizing the shunt so nominal load sits at 40-70% of full-scale.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| Iload | Current flowing through the shunt and load | A | A |
| Vshunt | Voltage measured across the shunt by the meter movement or ADC | V (typically mV) | V (typically mV) |
| Rshunt | Resistance of the precision shunt | Ω (typically µΩ) | Ω (typically µΩ) |
Worked Example: Ampere Meter in a battery test bench for 18650 cells
You are setting up an ammeter on a battery test bench cycling 18650 lithium cells at up to 10 A discharge. You picked a 10 A / 75 mV shunt (R = 7.5 mΩ) feeding a 16-bit ADC with a 100 mV input range. You want to know the resolution at low charge currents (0.5 A trickle), the nominal 5 A discharge, and the 10 A peak.
Given
- Rshunt = 7.5 mΩ
- Vfull-scale = 75 mV
- ADC resolution = 16 bit
- ADC range = 100 mV
Solution
Step 1 — at the nominal 5 A discharge, compute the burden voltage the ADC sees:
Step 2 — convert to current via the formula to confirm the meter reading:
This sits at 50% of shunt full-scale — exactly where you want a precision measurement to live. The ADC code is around 24,500 out of 65,535, plenty of headroom against quantisation noise.
Step 3 — at the low end of the typical operating range, 0.5 A trickle charge:
That's only 5% of shunt full-scale. The ADC LSB at 100 mV / 65,536 ≈ 1.5 µV translates to about 0.2 mA resolution — usable, but thermal EMF on the shunt terminals (typically 1-5 µV/°C on copper-to-brass junctions) is now the same order as your signal, so readings will drift with bench temperature.
Step 4 — at the 10 A peak:
You hit shunt full-scale exactly. Power dissipation in the shunt is P = I² × R = 100 × 0.0075 = 0.75 W, which on a typical 1 W shunt means you're at 75% of thermal rating. Continuous 10 A operation will warm the shunt by 30-50 °C and shift its resistance by 0.3% (manganin temperature coefficient is around 15 ppm/°C, but solder joints and lead resistance push the effective drift higher).
Result
Nominal reading: 5. 00 A from 37.5 mV across the shunt. At 5 A you're in the meter's sweet spot — readings stable to better than ±0.05 A with a decent ADC. At 0.5 A trickle the resolution is technically 0.2 mA but thermal EMF and amplifier offset drift dominate, so trust ±10 mA at best; at 10 A the shunt is hot and self-heating shifts the reading by up to 0.3% over a long discharge cycle. If your measured current reads 4.7 A when you commanded 5.0 A, check three things in order: (1) high-side vs low-side sense lead Kelvin connection — a non-Kelvin connection adds 1-3 mΩ of lead resistance and reads 5-15% low, (2) ADC reference voltage drift if the ammeter and load share a supply rail, and (3) shunt self-heating after long runs, which you confirm by measuring shunt body temperature and recalibrating at thermal steady state.
Choosing the Ampere Meter: Pros and Cons
Ammeters come in three families that practitioners actually choose between: shunt-based (series), CT-based, and Hall effect / clamp-based. Each one wins on different axes — accuracy, intrusiveness, AC vs DC capability, and cost.
| Property | Shunt Ammeter | CT-Based Ammeter | Hall Effect Clamp Meter |
|---|---|---|---|
| Accuracy (typical) | ±0.1 to ±0.25% | ±0.5% (Class 0.5) | ±1 to ±2% |
| AC / DC capability | Both | AC only (DC saturates core) | Both |
| Insertion required | Yes — break the circuit | No — wraps the conductor | No — clamps over conductor |
| Maximum practical current | ~2000 A (heat limited) | 10,000 A+ (transmission CTs) | 1500 A typical |
| Bandwidth | DC to ~10 kHz | 50/60 Hz to ~5 kHz | DC to 100 kHz |
| Cost (panel installation) | $30-150 shunt + meter | $80-400 CT + meter | $300-1500 handheld |
| Burden on circuit | 50-75 mV drop | Negligible voltage drop | Zero (non-contact) |
| Best application fit | DC battery & PV systems | Permanent AC switchgear | Field troubleshooting |
Frequently Asked Questions About Ampere Meter
Three things cause this on a real install. First, clamp position — if the conductor isn't centred in the jaw, the magnetic field coupling shifts and you can see 2-5% error per phase just from how you're holding the meter. Centre the conductor and re-zero between phases.
Second, adjacent conductor influence. If you're clamping one phase while the other two run within 50 mm of the jaw, their fields leak in. Pull the phase you're measuring out of the bundle by 100 mm or more.
Third, real imbalance. Above 10% phase-to-phase, suspect a failing contactor pole, a loose lug at the breaker, or a single-phasing motor. Check voltage at the motor terminals — if voltage is balanced but current isn't, the motor windings have a fault.
No, and this trips up people moving from AC panels to battery systems. A standard CT relies on a changing magnetic flux to induce secondary current. DC produces a static flux that saturates the core within milliseconds — the secondary output collapses to near zero and the CT is now useless until you demagnetise it.
For DC you need either a shunt-based ammeter, a Hall effect sensor, or a fluxgate / zero-flux CT (LEM IT-series and similar). Hall sensors handle DC to high frequency but drift with temperature; fluxgate transducers hit ±0.001% accuracy but cost 10x a regular CT.
Your analog meter is averaging-responding, calibrated to display RMS for a sine wave. A VFD output is a chopped PWM waveform with significant harmonic content — the crest factor is well above 1.414. The averaging meter mathematically under-reads non-sinusoidal current by 10-40%.
The true-RMS reading is the correct one. It's what determines actual heating in the motor windings and the cable. Use the true-RMS value for thermal sizing and protection settings. If the analog meter has been used to set overload trip points, those settings are likely too high and the motor isn't protected.
You're trapped between two requirements — inrush headroom and idle resolution — and one shunt rarely satisfies both. The standard approach is to size the shunt for steady-state plus a 25% margin, not for inrush. A 30 A shunt sees 50 A inrush as a 167% overload for a few cycles, which a properly rated shunt tolerates (most are rated for 10x for 1 second).
For idle resolution at 0.2 A on a 30 A shunt, you'll get 0.5 mV across the shunt — too low for most ADCs without front-end amplification. Add a 50x or 100x instrumentation amp ahead of the ADC, or run two parallel measurement chains: one shunt-based for steady-state precision and one Hall sensor for inrush capture. Battery management systems on EVs do exactly this.
Five percent is too much for any single component to be in spec, so it's usually a stack of small errors. Start with the CT — a Class 1 CT is allowed ±1% at rated current but degrades to ±3% at 10% of rated. If you have a 200:5 CT measuring 30 A (15% of rated), you're in the bad part of the accuracy curve.
Next, check the burden. If you've added a digital meter, energy meter, and a transducer all on the same CT secondary, total VA may exceed the CT's burden rating and the secondary signal compresses. Pull every load off except the panel meter and re-read.
Finally, the meter movement itself drifts. Analog panel ammeters lose 2-5% over 15-20 years from spring fatigue and magnet ageing. If the meter is older than the building's last renovation, replace it rather than trying to recalibrate.
You open-circuited an energised CT, which is one of the most dangerous mistakes in electrical work. With load current still flowing in the primary, the CT tries to drive its secondary current into infinite impedance. The core saturates hard and the secondary voltage spikes into the kilovolts — easily enough to arc across an air gap, destroy the CT insulation, and injure anyone nearby.
Always short the CT secondary with a shorting block before disconnecting any meter or relay downstream. Every properly engineered CT install includes one. If you find an installation without shorting blocks, that's a safety defect to flag before any maintenance work proceeds.
For permanent monitoring inside a combiner box, a shunt is the better choice — typically 0.1% accuracy, no temperature drift to speak of, and it survives lightning-induced surges better than a semiconductor sensor. The 75 mV burden is a rounding error on a 600 V string.
For commissioning and field troubleshooting, a DC clamp meter (Hall-based) wins because you don't have to break the circuit to measure. Accept the ±1-2% accuracy and zero the meter immediately before each reading — Hall sensors drift roughly 1 mA/°C, which on a hot rooftop translates to several hundred mA of zero offset if you forget.
References & Further Reading
- Wikipedia contributors. Ammeter. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
