Transformer Types (Isolation, Auto, Current)

Intuition

Three devices that look alike

Put an isolation transformer, an autotransformer and a current transformer on a bench and a casual glance finds the same thing three times: a core, some copper, two pairs of terminals. What separates them is not the ratio but how the windings are arranged, and the differences are the kind that matter rather than the kind that are merely interesting.

The isolation transformer is the one how transformers work described: two separate windings on one core, joined by flux and by nothing else. Often it has a ratio of one to one, which sounds pointless until you notice that the ratio was never the point. What you are buying is the absence of a conductive path.

The autotransformer does the same voltage-changing job with a single winding and a tap partway along. It is smaller, cheaper and more efficient than a two-winding transformer of the same rating, and it gives no isolation at all, because the output shares a connection with the input.

The current transformer inverts the usual arrangement. Its primary is the conductor whose current you want to measure — often a single pass through the core, or the cable itself — and its many-turn secondary delivers a scaled-down copy of that current into a small resistor. It is how a clamp meter reads alternating current without breaking the circuit.

The distinction that matters most is the second one, and it is a safety matter rather than a performance one: a device labelled "transformer" does not necessarily isolate anything.

Practitioner

What each one actually does

The isolating type is the simplest to specify because the ratio is usually one.

Same voltage, same current, no gain of any kind — and a complete break in the conductive path, which is the entire product.

The autotransformer replaces the second winding with a tap.

That difference current is the whole saving. Most of the load current never passes through a transformed winding at all; it flows straight from input to output through the shared conductor, and only the small remainder has to be handled magnetically.

The current transformer works from the other end, treating the measured conductor as a one-turn primary.

One volt per hundred amps, from a device that touches nothing and dissipates a tenth of a watt.

The ratio convention is worth noting before it causes trouble. A current transformer is quoted the opposite way round from a voltage transformer: 1000:1 here means a thousand secondary turns to one primary, and therefore a current division by a thousand rather than a multiplication. Commercial CTs are usually labelled by their current ratio instead, 100:0.1 or 100:5 amps, which removes the ambiguity at the cost of a longer label.

The other thing to notice is that all three of these devices obey exactly the relationships the previous two lessons established. No new physics appears in this lesson at all. What changes is which terminal the load hangs on, which winding the supply drives, and whether there is a conductive path between them, and those three choices are enough to produce three components with entirely different jobs and entirely different failure modes.

Engineer

Why the savings and the hazards both come from the construction

The autotransformer's advantage is worth quantifying, because it explains both when to use one and when not to bother.

The curve is the reason autotransformers are used for small voltage adjustments and not for large ones. Trimming 230 V to 200 V transforms thirteen per cent of the power, so the magnetics can be roughly an eighth the size. Stepping 230 V down to 12 V would transform ninety-five per cent of it, and the autotransformer would be almost the same size as a two-winding one while giving up isolation for nothing.

The current transformer's constraint comes from ampere-turns rather than from volts. In normal operation the secondary current opposes the primary's magnetomotive force almost exactly.

That near-perfect cancellation is what lets a current transformer have a small core and stay accurate. It also means the burden resistor is not optional. Open the secondary and the opposing ampere-turns vanish, leaving the full primary mmf to magnetise the core alone. The core saturates hard, the flux slams from one extreme to the other twice a cycle, and Faraday's law applied to a thousand turns and a very fast flux change gives a secondary voltage limited only by the core's saturation behaviour and the winding's insulation. The figure is genuinely dangerous on an installed CT, and it is why every current transformer terminal block carries a shorting link.

The ideal CT model breaks down in two places. Accuracy depends on the magnetising current being negligible beside the reflected primary current, so a CT loses accuracy at the bottom of its range where the two become comparable, which is why an instrument transformer's accuracy class is quoted over a stated fraction of its rating rather than all the way to zero. The burden also has to stay small: a large burden resistance means a large secondary voltage, which means real flux in the core, which means the cancellation is no longer near-perfect and the ratio drifts. That is why burden is specified in volt-amperes on instrument transformers, and why adding a long cable run between a CT and its meter is a measurement decision rather than a wiring one.

Professional

Choosing a type: the one distinction that is not negotiable

Beyond those three, selection comes down to what the transformer is for.

An isolating transformer is specified when a conductive break is required: separating a device under test from the mains supply so that a probe ground does not create a fault path, breaking a ground loop between audio equipment, or providing the isolation that a medical or a safety-extra-low-voltage supply demands. Its ratio is often one, and the specification that matters is the isolation rating rather than the turns.

An autotransformer is specified where the voltage change is modest and isolation is genuinely not required: trimming a supply voltage between regional standards, starting a large motor at reduced voltage, or providing an adjustable output as a variable autotransformer. The size and efficiency advantages are real, and the isolation loss is absolute.

A current transformer is specified by ratio, burden and accuracy class, plus a rated short-time current for fault conditions. Its cousins are worth knowing: a Rogowski coil measures the same thing with an air core and an integrator, handles very large currents without saturating, and cannot measure DC; a Hall-effect sensor can measure DC but drifts.

Construction affects all three. A toroidal core wastes less, hums less and radiates far less than a stacked-lamination one, at higher cost and with a fiercer inrush. Planar transformers replace wire with copper on a circuit board for high-frequency switching supplies. Where several separate secondary windings are provided, they are separate for a reason, and treating one as though it were another's continuation is a good way to defeat whatever isolation was designed in. The component-level treatment continues in transformers as components.