What a Transformer Actually Does

Why this matters

A control transformer is the most replaced healthy component in the trades. It gets condemned because the equipment it feeds behaves erratically, and it gets replaced with an identical unit, and the erratic behaviour comes back, because the transformer was never the fault - the load on it was, and the transformer was reporting that honestly in a language nobody read.

The reason the message gets missed is a single misunderstanding: techs treat the VA rating as a cutoff, as if the transformer refuses to deliver more. It does not refuse. It delivers whatever is asked of it and pays for the excess in two currencies, lost secondary voltage and heat. The lost voltage is what breaks the equipment, it only appears while the excess demand is happening, and a meter on the secondary with nothing calling reads perfectly normal.

What it actually does

Two windings share a magnetic core. Voltage divides by the turns ratio, current multiplies by it, and power passes through minus losses. A 120 V primary to 24 V secondary unit has a 5 to 1 turns ratio, so 1.0 A drawn on the secondary is roughly 0.2 A on the primary.

Two things ride along with that and both matter in the field.

The rating is in volt-amperes, not watts, because the transformer has no idea what the load's power factor will be and its windings are limited by current, not by work performed. A sibling article covers what power factor costs; the consequence here is that you budget a transformer in VA and you must not convert a coil's VA rating to watts and feel like you have headroom.

The secondary voltage tracks the primary proportionally. At a 5 to 1 ratio, a primary at 112 V rather than 120 V, which is 6.7 percent low, produces a secondary 6.7 percent low across the whole load range. So the first measurement on any "low control voltage" complaint is the primary, and finding it low sends you upstream instead of into the control box.

What it does not do is make power, change frequency, or make its secondary safe to handle. A 24 V secondary in a cabinet that also contains line voltage is a low-voltage circuit in a hazardous enclosure, and the meter, leads and PPE for that enclosure are sized to the highest voltage present in it, CAT III at the distribution level, under the energized-work justification in 29 CFR 1910.333(a)(1), which permits live work only where de-energizing would introduce additional or increased hazards or is infeasible due to equipment design or operational limitations.

Why the VA rating is a sag limit, not a cutoff

Every transformer has internal winding resistance and leakage reactance, which together mean its output voltage falls as its output current rises. That behaviour is quantified as voltage regulation: the no-load voltage minus the full-load voltage, divided by the full-load voltage.

Small control transformers regulate poorly. High single-digit to low double-digit percentages are ordinary. That is not a defect, it is a consequence of building a small transformer economically, and it is the reason the rating behaves as a soft limit: ask for double the rated VA and the unit will supply it briefly, sag substantially, and get hot. Ask for it continuously and it will cook itself or open its thermal protection.

So the diagnostic question for any control complaint is never "is the transformer bad." It is what is the secondary voltage at the load, at the instant of highest demand.

The worksheet, filled in

A packaged unit whose contactor sometimes fails to pull in, mostly on hot afternoons, with no failed component found on two previous visits.

Reading the worksheet: why it fails in the afternoon

The continuous load is fine. At 26 VA on a 40 VA transformer, 65 percent, this circuit would run forever, and that is exactly what a tech measuring a steady-state secondary voltage sees: a normal reading on a healthy transformer.

The failure lives entirely in the pull-in instant. Magnetic coils draw several times their holding VA while the armature is still open, because the air gap keeps the inductance low until the armature seals. Add that inrush to the loads already holding and the demand is 86 VA against a 40 VA rating, 215 percent, for a fraction of a second. The transformer supplies it and sags, and the measured voltage at the coil during pull-in is 19.8 V against a pickup requirement of 20.4 V. It is 0.6 V short of sealing, so it chatters or simply does not close.

Now the primary explains the pattern. At 112 V rather than 120, the whole secondary curve drops by the same 6.7 percent, so the 19.8 V reading would have been about 21.2 V at a healthy primary, comfortably above pickup. Read it as two effects compounding rather than one applied twice: the 215 percent peak demand buys the sag, and the low primary scales whatever is left. Strip the low primary out and the same instant reads about 21.2 V, above pickup; leave it in and the coil sees the measured 19.8 V, below it. Two independent effects, both pushing the same direction, both worst on a hot afternoon when the service is loaded and the coil is warm. Warm matters on its own: coil resistance rises with temperature, so the same applied voltage produces less current and less magnetic pull, which means a hot coil needs slightly more voltage to pick up than a cold one.

That is the complete explanation for a fault that had defeated two visits: nothing is broken, three margins are being spent at once, and every steady-state measurement reports normal.

The measurement that found it is the one nobody had taken: secondary voltage at the coil terminals, with the meter recording, during the pull-in attempt. Use a meter with a minimum-capture or recording function, because the event is over before a display can settle.

Fixing it in the right order

First, the primary. 112 V on a nominal 120 V circuit is a supply fault, not a transformer fault, and it is the cheapest 6.7 percent of margin available. Chase it with a voltage drop measurement across the connections feeding this equipment. Correct that and the pull-in reading may clear pickup on its own.

Second, reduce the peak demand rather than the continuous load. Dropping a 2 VA board off the circuit changes 86 VA to 84 VA and fixes nothing. Reducing the coil inrush is what matters, and the standard method is an interposing relay: a small-coil relay with modest inrush, energized by the board, whose contacts carry the line-voltage load directly, so the big magnetic coil comes off the control transformer entirely.

Third, add VA. A larger transformer with the same 5 to 1 ratio sags less at the same demand because its regulation is better and its rating is further from the peak. This is a real fix, not a cheat, provided the primary protection and the conductors suit the new unit. It is third rather than first because it treats the symptom while a 6.7 percent low primary is still degrading everything else on that circuit.

Never substitute an unprotected transformer for one that was internally protected, or defeat a thermal cutout to stop nuisance dropouts. And do not judge a transformer's temperature with your hand: an overloaded unit's case can exceed the burn threshold on a surface that looks unremarkable, and a failing potted unit can vent hot compound, so use a non-contact thermometer and stand out of the line of the case.

One rule carried from the control-circuit article rather than re-derived: on a combustion appliance, do not jumper a gas valve or a safety device to test the transformer under load. If you smell gas at any point, everyone leaves the building immediately, no switches touched, no lights, no phone used inside, and you call the gas utility and the shop from outside.

The impedance side, and why a secondary short is not a primary fault

Turns ratio transforms voltage by its value and impedance by its square. At 5 to 1, impedance transforms 25 to 1.

Put a 1.0 ohm short across that 24 V secondary. Reflected to the primary it looks like 25 ohms, so the primary draws 120 / 25 = 4.8 A - a current a 15 A primary breaker will carry all day without noticing. Meanwhile the secondary is being asked for 24 / 1.0 = 24 A against a rated secondary current of 40 / 24 = 1.67 A, which is 14.4 times its rating. The transformer is being destroyed and its primary protection has no idea.

That asymmetry is why secondary-side protection is a separate requirement and why inherently limited transformers exist at all: an inherently limited unit is built so that its own impedance restricts secondary current to a safe value under a short, which is what makes a power-limited control circuit power-limited. NEC Article 450 covers transformer overcurrent protection and NEC Article 725 covers Class 1, Class 2 and Class 3 remote-control, signaling and power-limited circuits. Practically: never test a secondary by shorting it, and never replace a Class 2 unit with a general-purpose transformer of the same VA, because the rating matches and the protection does not.

Magnetizing inrush on the primary side, and why a transformer occasionally trips its primary protection on energization with nothing wrong, is a different mechanism covered in the inrush article.

Sizing the next one so this does not recur

Size on the peak, not the sum of holding values. Write down every load's holding VA and every magnetic load's inrush VA, add the continuous loads to the largest single inrush, and select so that the resulting peak sits meaningfully below the rating rather than above it. On this unit that means budgeting against 86 VA, not 26 VA.

Then leave the arithmetic behind. Write the load list and the totals inside the panel door, in VA, with the date. The next person to add a zone accessory needs the running total, and the reason this failure mode is so common is that accessories get added one at a time by people who never saw the budget.

References

• NEC Article 450 for transformer overcurrent protection, and NEC Article 725 for Class 1, Class 2 and Class 3 remote-control, signaling and power-limited circuits
• Manufacturer device data for coil holding and inrush VA, pickup voltage, and transformer regulation
• 29 CFR 1910.333(a)(1) for the energized-work justification governing live measurement inside a panel containing line voltage
• See related: How to Think in Series and Parallel at the Equipment; Why Inrush Current Trips Things That Should Hold; What Power Factor Actually Costs a Shop