A transformer turns ratio tester applies an AC signal to one winding, measures the response at another, and uses the resulting voltage relationship to assess the expected transformation ratio. Depending on the instrument and configuration, it may also report phase displacement and excitation current. These observations can reveal inconsistencies in winding connections, tap selection, or transformer behavior. They do not establish the condition of every part of the transformer.

That distinction matters when a contractor must explain a handover result, a testing firm investigates an unexpected reading, or a distributor selects equipment for a customer. A satisfactory ratio is useful evidence. Calling it proof that the insulation is healthy, that all tap contacts are sound, or that no shorted turn can exist gives the measurement more authority than it has.

Is the instrument counting physical turns?

No. The physical turns ratio is the number of turns in one winding divided by those in another. A field tester normally infers that relationship from measured voltages under defined excitation conditions. In an ideal transformer, corresponding induced voltages follow the winding turns ratio. Real measurements also involve the magnetic circuit, winding connections, and the instrument’s method.

Keep three quantities separate: the physical winding turns ratio, the terminal voltage ratio measured during the test, and the ratio calculated from nameplate voltages. Megger’s technical paper on lesser-known TTR behavior distinguishes these quantities and discusses why test conditions and winding configurations matter. The practical consequence is that a comparison must use matching definitions.

For a hypothetical single-phase transformer rated 2400 V to 240 V, the nominal voltage ratio is 10. An ideal corresponding winding pair also has a turns ratio of 10. The same relationship written in the opposite direction is 0.1. Neither value is inherently wrong; the report must identify which winding is the numerator. These figures illustrate arithmetic, not an instruction to apply either rated voltage during testing.

Three-phase nameplates add another distinction because ratings commonly refer to line-to-line voltages. Individual winding voltages depend on connection. An instrument may calculate a nameplate-equivalent result automatically after the operator enters the vector group. Before applying any manual correction, determine whether the displayed value already includes that compensation. Correcting it twice can create an apparent defect entirely in the report.

Which problems can a ratio measurement help identify?

A valid deviation can indicate that the selected winding relationship differs from the expected one. Possible explanations include incorrect connections, an unexpected tap position, a winding problem, or incorrect reference information. The result narrows an investigation; it does not identify a unique cause. A wrong software setting and a real transformer problem can initially produce similarly alarming numbers.

Tap selection is a particularly useful application because selecting a different portion of a winding changes its effective ratio. A measurement can help verify that an indicated position corresponds to the expected electrical relationship. However, the expected ratio must come from the applicable tap schedule, not simply from the main nameplate voltages copied into every row.

Phase information can add evidence that a magnitude-only ratio lacks. Two winding arrangements can have similar voltage magnitudes while differing in phase relationship or polarity. If the assignment includes connection verification, request the relevant phase or vector-group evidence rather than assuming a ratio number proves the complete connection arrangement.

An unusual excitation-current result can also prompt investigation, but it needs its own comparison basis. Current measured during a ratio test should not silently replace results from a different exciting-current procedure. Preserve test conditions and compare compatible records. A numerical difference may reflect changed excitation rather than a changed transformer.

The most useful report distinguishes observation from diagnosis. “The measured relationship differs from the entered reference at this tap” is an observation. “The tap changer is defective” is a proposed explanation that needs corroboration. Keeping those statements separate makes it easier to correct a data-entry error without undermining the valid measurement or to escalate a persistent anomaly with clear evidence.

Why does testing one tap leave important questions unanswered?

A result applies to the selected winding arrangement and tap position. It cannot directly verify electrical paths that were not selected. This follows from the purpose of the measurement: the tester observes the configuration connected at that moment. The appropriate coverage depends on the approved assignment and transformer design, but the limitation should remain explicit.

Imagine a hypothetical five-position tap schedule with five different expected ratios. A satisfactory result at the nominal position verifies that measured condition. It does not provide measurements for the other four positions. If the customer expects coverage of all five, the report needs five position-specific records or a clearly documented reason for an agreed alternative scope.

A public engineering discussion about testing each tap contains questions about this coverage and anecdotal accounts of position-specific problems. The accounts are not a standard and do not establish a universal test sequence. Their useful question is whether the tested positions represent the electrical paths the customer expects to have assessed.

Even complete static ratio coverage does not prove every aspect of tap-changer performance. A ratio measured after a position settles differs from an assessment of switching transitions, contact resistance, or mechanical behavior. Define those other questions separately. Buying automatic tap sequencing can improve coverage and recordkeeping, but it does not transform ratio measurement into every possible tap-changer test.

Record the as-found and final position where the assignment requires it. Otherwise, a detailed ratio report may fail to answer the operational question that matters after testing: which position was left selected? The instrument record and the authorized restoration record serve related but different purposes and should be reconciled by the responsible team.

What can a satisfactory TTR result miss?

A typical field ratio test uses relatively low excitation compared with normal system insulation stress. It does not reproduce every electrical, thermal, or mechanical condition encountered in service. A weakness that does not materially change the measured relationship under those conditions can remain outside the test’s evidence. That is why a ratio result cannot be treated as a dielectric withstand result.

Similarly, a contact that establishes the expected transformation relationship at test current is not thereby proven satisfactory at service current. Contact resistance and heating are different questions. Ratio and resistance measurements can complement each other, but a good ratio does not remove the need for another method when the investigation concerns a resistive connection.

Nor should “no shorted turns detected” become “all shorted turns ruled out.” Sensitivity depends on the affected winding, fault condition, method, and measurement conditions. An unexpected ratio may support concern about a winding fault, while an expected ratio alone cannot eliminate every winding fault mechanism. Describe what the evidence supports without converting a useful screening measurement into an absolute guarantee.

Mechanical displacement, insulation aging, fluid condition, and intermittent behavior also require evidence appropriate to those concerns. The OMICRON power-transformer testing brochure presents ratio measurement alongside other diagnostic methods. That broader diagnostic context is useful: select additional evidence for the suspected mechanism rather than accumulating unrelated tests or assuming TTR answers every condition question.

A hypothetical post-event review illustrates the boundary. Suppose a transformer has an expected ratio after a protection operation. That result helps address one aspect of its electrical relationship. It does not explain the protection event or authorize re-energization by itself. The event records, inspection findings, and other justified diagnostic evidence still determine the engineering decision.

Why can a sound transformer give an unexpected result?

Start with the reference and configuration. The wrong tap, incorrect nameplate entry, reversed ratio convention, mistaken vector group, or mismatched terminal labels can create a discrepancy without a winding defect. These are inexpensive possibilities to investigate because they can often be checked against existing records before any additional measurement is attempted.

Then examine measurement validity. Lead connections, supported excitation mode, magnetic condition, and the handling of other windings can matter. Follow the instrument and transformer procedures rather than improvising changes to obtain a preferred number. A repeat measurement is useful when it tests a stated explanation and preserves what changed.

An online discussion about TTR and insulation-resistance test order shows how easily different DC tests and their effects become conflated. Treat such discussions as questions to clarify, not as authority for a sequence. Record the actual preceding tests and consult the approved procedure when magnetic history or demagnetization is relevant.

If an alternative supported excitation method changes the result, document both methods and their conditions. The changed result does not mean the physical winding acquired or lost turns. It may reflect how the transformer and measurement system behave under the different conditions. An engineering explanation must connect that difference to the method rather than choosing whichever result looks better.

Persistent disagreement after configuration and validity checks deserves review. Compare factory evidence and previous records only when their definitions and conditions are understood. A historical number without tap identity or compensation convention may be less useful than a new, well-documented result. Preserve uncertainty rather than forcing an apparently precise comparison between unlike records.

What should customers expect in a useful TTR report?

The record should identify the transformer, winding pair, tap position, expected relationship, and source of that expectation. It should also identify terminal assignments, relevant vector-group settings, excitation method, measured ratio, and any phase or current observations used in the assessment. Include instrument identity and the applicable measurement record so another reviewer can reconstruct the conclusion.

If deviation is calculated, state the convention. For example, a report may use one hundred times measured ratio minus expected ratio, divided by expected ratio. Another system may reverse the sign or present only magnitude. A small sign difference between reports is not evidence of a physical change until the formulas have been reconciled.

Do not attach a universal acceptance percentage to every transformer. Identify the applicable specification, edition, contractual requirements, and any design-specific basis used by the responsible engineer. A familiar threshold copied from an article can be misleading when the measurement definition, tap arrangement, or transformer type differs from the assumed case.

For an EPC handover, make incomplete coverage visible before approval. For a testing firm, distinguish invalid measurements from valid but concerning measurements. For a distributor, demonstrate how the supplied equipment retains the information needed for that distinction. These choices turn a table of ratios into evidence that supports a clear customer decision.

Consider a hypothetical record with the correct ratio but an uncertain winding label. The number may be accurate, yet it cannot support the intended phase comparison until identity is resolved. By contrast, a clearly identified record with an unexplained deviation is valid evidence for further investigation. Treating both as a generic failure hides the next action. The first needs record or connection reconciliation; the second needs technical assessment.

The conclusion should therefore state both the finding and its scope. A report can confirm that specified winding pairs and positions produced the documented results under the stated method, while identifying untested positions or unresolved anomalies. The asset owner then knows exactly which question has been answered and which remains open.

Conclusion: use TTR as evidence about the selected winding relationship

A turns ratio tester assesses transformation behavior under defined test conditions; it does not certify the entire transformer. Verify the ratio definition, tap basis, connections, and measurement validity before interpreting a deviation. Use satisfactory results within their documented scope, and select complementary evidence when the question concerns insulation, contacts, mechanical condition, or an unexplained service event.