A transformer turns ratio test is the fastest way to confirm that the unit in front of you is the unit on the nameplate. It takes about fifteen minutes, needs a variable AC source and two multimeters, and catches three failures that quietly ruin low voltage installations: swapped windings, shorted turns and a mis-set tap. This guide covers the applied-voltage method used on transformateurs basse tension from 100 VA to 10 kVA, shows the arithmetic, and explains when a dedicated ratio tester earns its price. Transformer testing does not need to be complicated to be useful.
The turns ratio is primary turns divided by secondary turns. Since voltage per turn is the same on both windings, it also equals the voltage ratio:
N1 / N2 = V1 / V2 = I2 / I1
A transformer marked 220 V / 110 V should produce a ratio of 2.000: apply 100 V to the primary and a healthy unit delivers 50 V. Winding construction sets that ratio, so a drift from the nameplate is physical evidence of a winding problem. The test answers three questions:
It does pas measure insulation health, core loss or temperature rise, so it is one entry in a test schedule.
Safety first. On a small transformer it is easy to get careless because the voltages look harmless, so treat every winding as live until proved otherwise. Isolate and lock out every source, ground the windings, disconnect downstream loads, and wear insulated gloves and glasses even at 110 V.
For the test itself you need a variable AC source, two true-RMS multimeters and short test leads.
| Objet | Objectif | Note |
|---|---|---|
| Variable AC source (variac) | Sets a safe, adjustable primary voltage | A 0-250 V, 0.5-5 kVA variable autotransformer covers most bench work |
| Two true-RMS multimeters | Read primary and secondary simultaneously | True-RMS matters; averaging meters misread on distorted waveforms |
| Test leads with fused probes | Safe connection to terminals | Keep leads short to reduce pickup |
| Nameplate data or datasheet | Supplies the expected ratio | Record rated voltages and tap schedule before you connect anything |
This is the standard bench method for small transformers: you control the primary voltage and calculate the ratio from two simultaneous voltmeter readings.
Read the nameplate for rated primary and secondary voltage and tap schedule, then divide primary by secondary to get the expected ratio – 2.000 for 220 V / 110 V, 1.727 for 380 V / 220 V. Write it down before you touch a terminal; every reading is judged against it.
Feed the variac from a protected supply and connect its output to the transformer primary. Put meter one across the variac output and meter two across the secondary. Both must be on AC volts and a fixed range, because auto-ranging mid-measurement shows as a false dip.
Wind the variac up slowly to 20 to 50 percent of rated primary voltage – for a 220 V primary, 50 to 100 V. That is enough for a stable reading without stressing insulation. Never exceed the rated primary voltage.
Most bad results come from here. Let the voltage settle, then read both meters at the same moment. If the secondary drifts, a lead is making poor contact. Repeat twice; readings within half a percent are trustworthy.
Divide primary voltage by secondary voltage. That is the measured ratio.
Measured ratio = Vprimary / Vsecondary
Record the applied voltage, secondary voltage, measured ratio, nameplate ratio and percentage error – five fields worth more than a pass/fail tick. Worked example: on a 1000 VA, 220 V / 110 V transformer, meter one reads 100.0 V and meter two reads 49.8 V, so the measured ratio is 2.008 against a nameplate ratio of 2.000.
| Étape | Action | Record |
|---|---|---|
| 1 | Read nameplate, compute expected ratio | Expected ratio, tap position |
| 2 | Connect variac to primary, meters to both windings | Winding identified as primary |
| 3 | Raise to 20-50% of rated primary voltage | Applied voltage |
| 4 | Read both meters simultaneously, twice | Vprimary, Vsecondary |
| 5 | Divide primary by secondary | Measured ratio |
| 6 | Compare with nameplate, compute error | Ratio error in percent |
Ratio error is how far the measured ratio sits from the nameplate ratio, as a percentage:
Ratio error (%) = (Measured ratio – Nameplate ratio) / Nameplate ratio x 100
In the example above, (2.008 – 2.000) / 2.000 x 100 = +0.4 percent. A positive number means the primary has relatively more turns than expected, or the secondary fewer. Acceptance limits are not universal: they depend on the transformer class, the standard specified and the manufacturer’s tolerance. The bands below suit low voltage dry-type units, but the datasheet always wins.
| Transformer class | Typical acceptable ratio deviation | Comment |
|---|---|---|
| Small dry-type LV transformer (100 VA – 10 kVA) | +/- 1% | Winding and measurement scatter both contribute |
| General purpose LV transformer, closer tolerance build | +/- 0.5% | Specified for control and instrumentation duty |
| Measurement or instrument transformer | +/- 0.1% to +/- 0.5% | Class-dependent, verified with a calibrated ratio bridge |
If a unit lands just outside the band, repeat the measurement before condemning it. Lead placement, meter accuracy and waveform distortion can each account for a few tenths of a percent.
Three-phase testing adds two complications: test every phase, and know whether the nameplate voltages are line or phase values. A delta winding has no neutral, so comparing a line-to-line reading against a phase ratio produces an apparent error of exactly the square root of three. The three-phase transformer connection guide covers the terminal arrangements.
One phase several percent off while the other two are clean points at a fault in that phase winding, not the transformer as a whole – which decides repair versus scrap.
A turns ratio tester, or TTR set, reports ratio, polarity, phase angle and sometimes excitation current in one pass, and calculates the error for you. It earns its place when you need vector group confirmation, fast testing across many taps, documented commissioning results, or phase angle measurement that two voltmeters cannot provide.
| Consideration | Variac + two multimeters | Dedicated ratio tester |
|---|---|---|
| Measured quantities | Voltage ratio only | Ratio, polarity, phase angle, excitation current |
| Three-phase vector group | Not possible | Direct reading |
| Typical accuracy | 0.2-0.5% with good meters | 0.05-0.1% |
| Cost and setup | Low, uses existing bench gear | Higher, needs calibration and training |
| Best suited to | Single-phase bench checks and incoming inspection | Commissioning, tap verification, QA documentation |
If the transformer has off-circuit taps, the ratio should change by the tap step at each position. A +/- 5 percent schedule on a 2.000 base ratio reads about 2.100 at +5 percent and 1.900 at -5 percent. Test every tap; one that reads correctly at nominal but wrong at the extremes usually means poor contact, not a winding fault.
A variable autotransformer is different: its brush moves along an exposed winding, so the ratio changes continuously. Testing one means checking that the output tracks the dial smoothly with no dead spots or jumps. A variable autotransformer with a pointer scale helps here, because the dial reading can be compared against the measured output across the full sweep.
The brush contact that makes a variac adjustable also makes it a possible source of erratic readings. If the secondary jumps while the primary reading is steady, suspect the variac before the transformer under test.
| Erreur | Symptom | Fix |
|---|---|---|
| Using an averaging meter | Reading 1-3% low on a distorted waveform | Use true-RMS instruments |
| Testing with the load still connected | Secondary reads low, error looks like a winding fault | Disconnect all downstream equipment |
| Comparing a line voltage against a phase ratio | Error of about 73% on a delta winding | Confirm line versus phase reference first |
| Auto-ranging during the measurement | Brief dip, unstable reading | Lock the meter range manually |
| Reading the meters at different moments | Random error of several percent | Read simultaneously, repeat twice |
| Over-volting to get a clearer reading | Insulation stress, heating, saturation | Stay at 20-50% of rated primary voltage |
| Measuring at the source, not the winding | Good unit rejected | Verify applied voltage at the winding terminals |
The pattern across phases, taps and time is what identifies the fault.
| Observation | Likely cause | Next action |
|---|---|---|
| Measured ratio below nameplate | Shorted turns in the primary, or fewer primary turns than specified | Repeat test, then inspect for discolouration and burnt varnish |
| Measured ratio above nameplate | Shorted turns in the secondary | Compare with a DC resistance check on the same winding |
| One phase deviates, others clean | Phase winding fault or a bad internal connection | Treat as a winding failure; do not return to service |
| No secondary voltage at all | Open winding or broken connection | Continuity and resistance test to localise the break |
| Ratio changes smoothly with the tap | Normal tap behaviour | Record and trend over time |
| Ratio jumps erratically at one tap | Poor tap contact or misalignment | Clean and re-seat the tap, retest |
A one-off check tells you whether the unit is good today; a trend tells you whether it is degrading. Keep four numbers per unit – applied voltage, secondary voltage, measured ratio and test date – alongside your insulation resistance readings and your DC resistance measurements of the transformer windings.
Repeat the test at incoming inspection, after any repair, and at your maintenance interval. If the ratio has moved beyond the datasheet tolerance, investigate before the unit returns to service. Load capacity is a separate calculation, which is why transformers are rated in kVA rather than kW.
The turns ratio test is the cheapest high-value check available to anyone working with low voltage transformers. A variable source, two decent meters and fifteen minutes confirm that the unit transforms at the ratio it claims. Do it at incoming inspection so a wrong unit never reaches the panel.
If you are specifying a transformer and need the ratio, tap schedule and winding arrangement documented up front, browse the step up and step down transformer range or send your rated voltages and load details and we will match a unit to them.
The turns ratio equals primary turns divided by secondary turns, and it equals primary voltage divided by secondary voltage. A 220 V / 110 V transformer has a ratio of 2.000.
Yes, on a low voltage transformer. Feed the primary from a variac set to 20 to 50 percent of rated voltage, read both windings with true-RMS multimeters at the same time, then divide primary volts by secondary volts. A resistance setting cannot measure ratio.
For small dry-type low voltage transformers, plus or minus 1 percent is a practical working limit, and closer-tolerance builds are specified at plus or minus 0.5 percent. Confirm against the datasheet or the standard the unit was ordered to.
Test each phase individually – A-a, B-b and C-c – and compare the results against each other and against the nameplate. Confirm whether the nameplate voltages are line or phase values first, because on a delta winding a line-to-line reading shows a false error of about 73 percent.
No. A variac and two true-RMS multimeters are sufficient for single-phase bench checks in the 100 VA to 10 kVA range. A dedicated tester becomes worthwhile when you need phase angle and polarity, a three-phase vector group check, or documented commissioning results.
