More project photos (2)
The problem
A single-phase energy meter calibration bench is a test rig that checks whether a household energy meter records energy correctly. It connects the meter under test and a trusted reference meter to the same load, runs both for a fixed test period, and compares their energy readings to compute percentage error.
The single-phase energy meter calibration bench answers a practical question: how do you know a meter is recording correctly? A meter that reads fast overcharges the consumer, while one that reads slow under-records consumption, yet both faults are invisible from the dial alone, with no second reading to compare against. This bench provides that second reading. It wires the meter under test and a reference meter in series on the same supply, so both meters experience the identical voltage and load current for the identical test duration. Any difference in their recorded energy is therefore a property of the meters themselves, not of the load. The microcontroller counts calibration pulses from both meters, converts the counts to energy using each meter's pulse constant, and displays the percentage error on the LCD. The bench thus turns meter accuracy into a measured quantity, while making clear its result is a comparison, not a certification.
| Method | What it compares against | What you get | Limitation |
|---|---|---|---|
| Calibration bench (this project) | Reference meter on the same load | % error at several load steps | Comparison only, not lab certification |
| Manual stopwatch and energy comparison | Hand-timed dial readings | Rough error estimate | Timing and reading errors dominate small errors |
| Sending the meter to a test lab | Laboratory reference standard | Certified test report | Costly, slow, and no hands-on learning |
230 V single-phase supply through an MCB switch.
Follow the working sequence
How it works
The single-phase energy meter calibration bench works end to end as a controlled comparison experiment. Mains supply feeds the meter under test and the reference meter in series, so the identical current flows through both meters to the same load bank. The user selects a load step by switching lamp loads on or off, then starts a test run from the push button. During the run, the microcontroller counts the calibration LED pulses from both meters and measures voltage and current, converting pulse counts to energy with each meter's pulses-per-kWh constant. When the test window ends, the firmware compares the two energy figures, computes percentage error, and displays the result on the LCD alongside the raw readings. Repeating the run at different load steps builds an error profile of the meter under test, showing whether it reads fast, slow, or within its accuracy class across its working range.
- Mains supply is connected through an MCB and isolation switch to the meter under test and the reference meter wired in series.
- The user selects a load step by switching the lamp load bank (for example 1 lamp, 2 lamps, then 3 lamps) to set the test current.
- The test run is started from the push button; the microcontroller begins counting calibration pulses from both meters simultaneously.
- Voltage and current are measured during the run, and pulse counts are converted to energy using each meter's pulses-per-kWh constant.
- At the end of the fixed test window, the firmware computes % error as (energy under test − reference energy) ÷ reference energy × 100.
- Voltage, current, both energy readings and the % error are shown on the 16×2 LCD for the run just completed.
- The run is repeated at other load steps to build an error profile, and results are recorded manually for the report.
Tech stack:
- Arduino Uno (ATmega328P)
- Single-phase energy meters (meter under test + reference meter)
- Optocoupler pulse-sensing inputs
- Voltage and current sensing (ZMPT101B + ACS712-type sensor)
- 16×2 LCD display
- Lamp load bank with selector switches
- Arduino IDE (C/C++ firmware)
- MCB + isolation switch for mains safety
| Parameter | Value |
|---|---|
| Supply | 230 V AC single-phase (nominal) |
| Load bank | 3 steps, 100 W lamps per step (design target) |
| Maximum test current | Approximately 5 A (design target) |
| Reference meter accuracy class | Class 1.0 (datasheet value of the reference meter used) |
| Meter under test class | Class 1.0 or Class 2.0 typical (datasheet value) |
| Pulse constant | 1600 impulses/kWh typical (label value, entered in firmware) |
| Test window | 60 seconds per run (configurable in firmware) |
| Error display resolution | 0.01% (design target) |
| Controller | Arduino Uno, 16 MHz clock (datasheet value) |
| Display | 16×2 character LCD |
Project features
The single-phase energy meter calibration bench gives students four capabilities in one rig: a side-by-side comparison of a meter under test against a reference meter, stepped resistive loads that test the meter at several currents instead of one, automatic pulse counting that turns both meters' calibration pulses into comparable energy figures, and a live LCD readout of voltage, current, energy and percentage error. Together these make meter error visible and repeatable: the same load step can be run again and the same error should reappear within the limits of the reference meter. The bench is built around standard single-phase metering practice, so the concepts transfer directly to utility meter testing, while its honest limits are stated plainly — it compares meters, it does not certify them, and its accuracy is bounded by the reference meter it uses.
- [Reference-meter comparison — Implemented] The meter under test and a reference meter are wired in series on the same supply, so both see the same voltage and load current, and any difference in recorded energy is attributable to the meters.
- [Automatic % error calculation — Implemented] The microcontroller counts calibration pulses from both meters over a fixed test window and computes % error as (energy under test − reference energy) ÷ reference energy × 100, shown on the LCD.
- [Stepped load bank — Implemented] Switched lamp loads let the user test the meter at several load steps, so error can be observed at light, medium and full load rather than at a single operating point.
- [Live voltage, current and energy display — Implemented] A 16×2 LCD shows supply voltage, load current, energy counted from each meter, and the computed % error during every test run.
- [Pulse-constant entry — Implemented] Each meter's pulses-per-kWh constant (for example 1600 impulses/kWh on a typical single-phase meter label) is entered in firmware, so energy is computed from pulses for both meters on the same basis.
- [Multiple test runs and reset — Implemented] Each test run is started and reset from a push button, so a run can be repeated at the same load step to check repeatability.
- [PC data logging — Optional customization] Serial logging of test runs to a computer for tabulation can be added on request; it is not part of the base bench.
- [Three-phase extension — Future scope] Extending the bench to three-phase meters is possible future work; the base project covers single-phase meters only.
What is included
- Working calibration bench prototype (meter under test, reference meter, load bank, controller and display)
- Arduino firmware with pulse counting and % error calculation
- Circuit and wiring documentation with mains-safety notes
- Component list with meter pulse constants and accuracy classes
- Setup and test-procedure instructions (load steps, test window, recording sheet)
- Project report PDF (background, metering theory, methodology, results format)
- PPT presentation for final review
- Viva Q&A preparation document (accuracy classes, pulse constants, error formula, why series wiring matters)
Limitations & prerequisites
- The bench compares a meter against a reference meter; it does not certify meters and carries no NABL or laboratory accreditation.
- Results are only as good as the reference meter: a Class 1.0 reference meter bounds the smallest error the bench can meaningfully resolve.
- It is a single-phase bench only; three-phase meters, CT-operated meters and prepaid meters are out of scope for the base build.
- This bench calibrates meters; it is not a prepaid energy meter and does not sell, recharge or disconnect electricity supply.
- Lamp loads are resistive only, so the bench does not test meter behaviour under inductive or capacitive loads unless such loads are added.
- All work uses 230 V mains and must be done under supervision with the MCB and isolation switch; the bench is not a toy or an unsupervised household device.
- Accuracy figures for the built bench are design targets only until a specific build is tested against a known reference.
Frequently Asked Questions
What does this bench actually test?
This bench tests whether a single-phase energy meter records energy correctly by comparing it against a reference meter wired in series on the same supply and the same load. Both meters experience the identical voltage and current for the identical test duration, so any difference in recorded energy is the meter under test's error, computed as a percentage. The result is a comparison, not a certification.
How is the percentage error calculated?
The percentage error is calculated from the energy recorded by each meter over the same test window, using the formula (energy under test − reference energy) ÷ reference energy × 100. Each meter's energy comes from its calibration pulses divided by its pulses-per-kWh constant, so both figures rest on the same measurement basis. A positive result means the meter under test reads fast; a negative result means it reads slow.
Why must both meters share the same load current?
Both meters must share the same load current because the comparison is only valid if the two meters measure the same quantity. If the currents differed, an energy difference could come from the loads rather than from the meter, and the percentage error would be meaningless. Series wiring guarantees the identical current flows through both meters, which is the builder's key design decision in this bench.
Does this bench certify a meter or replace a test lab?
No, this bench does not certify a meter and does not replace an accredited test laboratory. It compares a meter under test against a reference meter and reports percentage error, which is excellent for learning and for spotting faulty meters, but a certified test report requires a laboratory reference standard and accredited procedures. The bench's own accuracy is bounded by its reference meter's class.
Is this the same as a prepaid energy meter project?
No, this bench is not a prepaid energy meter and does not sell, recharge or disconnect electricity. A prepaid meter controls supply against a paid balance, while this bench tests meters: it measures how accurately a meter records energy by comparing it with a reference meter. The two projects share metering hardware but solve different problems, and this project is the testing side.
Is this project suitable for a final-year project?
Yes, this project is suitable for a final-year project in Electrical Engineering, because it demonstrates metering principles, accuracy classes, pulse-based energy measurement, microcontroller interfacing and experimental error analysis in one working rig. Students can explain the series-wiring reasoning, the error formula and the error profile across load steps, which together make a strong viva discussion.
Components & software requirements
- Arduino Uno (ATmega328P)
- Single-phase energy meters (meter under test + reference meter)
- Optocoupler pulse-sensing inputs
- Voltage and current sensing (ZMPT101B + ACS712-type sensor)
- 16×2 LCD display
- Lamp load bank with selector switches
- Arduino IDE (C/C++ firmware)
- MCB + isolation switch for mains safety
Delivery information
Built-to-order project. Delivery timeline is shared after order confirmation based on current queue.
Support terms
Complete documentation, setup guide, and viva preparation included. Support for setup and explanation provided.
Download free synopsis (PDF)
Enter your name and WhatsApp number — the synopsis PDF downloads immediately. We may send occasional project updates on WhatsApp; no spam.