Transformer testing and service

Confirm each winding is continuous and that there is no unexpected short between windings or to core (with proper isolation of circuits). Compare sister phases on three-phase units—large resistance imbalance suggests trouble.

Test primary-to-secondary, primary-to-core, secondary-to-core at the voltage class and procedure your shop/standards require. Discharge windings after testing. Never megger with solid-state equipment still connected.

Ratio can be checked with dedicated instruments or carefully with reduced voltage where permitted. Polarity verification prevents bank and parallel disasters covered earlier.

When forming a delta, before landing the last connection, measure voltage across the open corner. It should be essentially zero (or very small). A large reading means wrong polarity or wrong tap—closing would circulate destructive current.

Listen for abnormal buzz, check temperatures, ventilation, oil level/leaks if liquid-filled, corrosion, and loose connections. Retorque to spec when de-energized. Dust-clogged dry-types run hot and fail insulation early.

1. Visual and mechanical (leads clear, no tools left, fans free).
2. Continuity / resistance balance.
3. Insulation resistance + discharge.
4. Ratio/polarity as required.
5. Delta open-corner voltage before final close.
6. Energize and check no-load current/noise, then infrared under load.

Skipping to step 6 is how rewinds get sold.

For liquid units: level, leaks, color/smell only as preliminary clues—lab tests belong to maintenance programs. Never open sealed tanks casually. Bushings and gaskets deserve the same respect as windings.

A secondary fault that cleared may still have damaged inter-turn insulation. Megger may pass while the unit fails later under load. Trend temperature and noise after return to service.

Before you speak, write down:

1. What topology or machine you have in front of you.
2. Voltages and currents with the measurement point.
3. Frequency or rpm if they apply.
4. What the nameplate or diagram says.
5. What changes if you isolate one part of the circuit.

A diagnosis without those data is conversation, not the trade.

1. Explain the central block out loud to an imaginary helper.
2. Rewrite the field case with numbers from a real piece of equipment.
3. Complete the checklist without looking.
4. Mark which rows in the mistakes table have already happened to you.

If you cannot say the core idea in one minute, return to the first third.

Energized measurement needs PPE, a meter of the right category, and a plan if the reading does not make sense. Capacitors, inductive fields, rotating shafts, and power neutrals do not forgive haste. If the procedure says de-energize and verify absence of voltage, that rules.

Situation. Crew closes a newly wired delta bank. Instantly breakers trip and one winding smells of varnish.

How to think. The open-corner voltage was never checked—or was ignored. Polarity or labeling was wrong.

Conclusion: the gap test is cheaper than a rewind.