VA Technology 
A turns ratio test measures the voltage ratio between two windings and compares it with the nameplate ratio for the tap under test. On one tap it is a wiring check; across the full range it is a diagnostic, because a healthy on-load tap changer produces a smooth progression of ratios and a faulty one does not.
One reading inside tolerance proves little; the sequence across the range separates a real changer fault from a measurement error.
What a turns ratio test measures
The set applies an AC voltage to one winding and measures the voltage induced in the other, with the transformer isolated and unloaded. Magnetizing current is small, so the measured voltage ratio closely approximates the turns ratio.
The phase reading is the vector group check: a Dyn11 unit shows a 30° shift, and a reading near zero degrees points at a connection error rather than a tap fault.
Connecting and exciting the transformer
Excite the high-voltage winding and measure the low-voltage winding, phase pair by phase pair. Where a neutral is brought out, include it; leaving it out changes the measured ratio by the factor the connection implies.
Automatic connection identification removes a recurring error source. The TTR-400 identifies the winding connection and confirms the vector group in the same routine, so a delta winding is not measured against wye arithmetic. It excites from a three-phase precision inverter supply up to 160 V.
Keep the excitation inside the linear part of the core magnetization curve. Too high a test voltage drags the measured ratio low, mimicking shorted turns in a sound winding.
Reading the deviation against the nameplate
Deviation is a percentage of the declared ratio: measured minus declared, divided by declared. IEC 60076-1:2011 allows ±0.5 % of the declared ratio at the principal tapping, or one tenth of the percentage impedance there, whichever is smaller. IEEE C57.12.90-2015 applies the same ±0.5 % limit to each pair of windings.
The shape of the deviation says more than its size.
| Deviation pattern | Where to look first |
|---|---|
| Every tap off by a similar amount | Nameplate misread, or the wrong reference tap |
| One tap off, neighbours correct | Selector or diverter contact at that step |
| Deviation grows across the range | Winding or tap-section fault, not one contact |
| One phase differs from the others | Shorted or open turns in that phase |
How the test finds tap-changer faults
An on-load tap changer changes the effective turns in the tapped winding in two stages: the selector picks the tap, and the diverter switch transfers the load current onto it. Either stage can fail to make cleanly, adding turns the nameplate does not expect.
That is why the ratio is plotted against tap position. A changer in good order follows the nameplate curve at every step. Where one step jumps and the next returns to normal, the transfer did not complete or the contact is disturbing the measurement. A step far from every nameplate value points at a contact not making at all.
De-energized changers fail differently: set once and left, the fault is rarely wear but a link moved during maintenance and not restored, which shifts the whole curve.
Mistakes that look like a bad tap changer
Most tap changers condemned during commissioning are measurement errors.
- Comparing against the nominal nameplate ratio instead of the ratio declared for that tap.
- Reading the tap from a remote indication rather than the changer itself.
- Running the ratio measurement straight after a DC winding resistance test, which leaves residual core flux.
- Reversing the leads, which inverts the ratio rather than reporting it.
- Exceeding the linear excitation range, which reads low on every tap.
The remedy is procedural: record the tap, compare like for like, and keep DC work away from the ratio measurement.
Frequently asked questions
Why measure the ratio at every tap position?
The tap changer is the part most likely to be wrong, and its faults are step-specific. One tap cannot distinguish a changer a step out from one that is correct, and it never reveals a contact that stopped making at a step nobody tested.
What makes a ratio reading lower than nameplate?
A turns ratio test reads low for two reasons. Shorted turns genuinely reduce the effective turns in circuit. Exciting the core beyond its linear region does the same to the reading with no winding fault present, so check the excitation level first.
Does a passing ratio test prove the changer is healthy?
No. It proves the turns in circuit are correct at each tap. Contact condition, transition timing and diverter wear lie outside what a ratio measurement sees.
A turns ratio test is a comparison, and the reference is the nameplate at the tap being tested. VA-TEK builds the transformer testers family around that work: the TTR-400 for ratio and vector group, the TDC-3100 for insulation power factor, the FRA100 for frequency response and the TWO-1100S for winding resistance. The transformer differential protection article covers the protection side. Send the nameplate, the tap schedule and the test set model through the contact form.
Sources: IEC 60076-1:2011, Power transformers — Part 1: General (Table 1 ratio tolerance; measurement of voltage ratio and check of phase displacement). IEEE C57.12.90-2015, Test Code for Liquid-Immersed Distribution, Power and Regulating Transformers. IEC 60214-1:2014, Tap-changers — Performance requirements and test methods. TTR-400 product data (vendor data).
