HVAC damper actuators: the 24V AC sizing trap

Why this happens

The trap is not that people get the arithmetic wrong. It is that they take a number from a datasheet that describes one operating state and use it to size equipment that has to survive a different one.

The plunger position is part of the magnetic circuit, so it is part of the impedance. On AC the current is set by impedance, and the impedance is resistance in series with reactance, which depends on inductance. With the plunger out, the air gap is large, the reluctance of the magnetic path is high, so the inductance is low and the impedance is low: the current is high. Seat the plunger and the gap closes, reluctance falls, inductance rises, impedance rises, and the current falls. That is not a transient — it is the steady-state behaviour of the two positions, and the ratio between them is typically somewhere between three and ten to one.

Two consequences follow, and they pull in opposite directions. If you size the transformer and the wire from the holding figure, the inrush exceeds what the circuit can supply, the voltage sags, the plunger does not seat, and you stay stuck in the high-current state that you sized for the low-current one. If you size everything from the inrush, you carry margin you do not need for the rest of the installation. The only way to get this right is to specify both numbers.

An AC coil held with its stroke blocked will destroy itself. If the plunger cannot seat — a jammed damper blade, a binding linkage, an obstruction, or simply an actuator installed against a stop it cannot reach — the current never falls to the holding value. The winding then dissipates several times the power its rating assumes, continuously. This is the standard cause of an AC solenoid burning out on a correct supply, and it is a mechanical fault that presents as an electrical failure. The thermal side of this is covered in what actually kills a coil.

AC coils need a shading ring, and a damaged one produces a specific symptom. The force from an AC winding passes through zero twice per cycle, so without a short-circuited ring in part of the pole face the plunger chatters and buzzes at twice line frequency and never holds cleanly. A cracked or missing shading ring is therefore a classic “it hums and will not pull in” fault that looks like a weak coil and is not one. Buzzing diagnosis is set out in how to split magnetic noise from mechanical noise.

And then there is the unit itself: on AC, a volt-amp is not a watt. A solenoid is inductive, with a power factor commonly around 0.4 to 0.7. Apparent power in VA therefore exceeds real power in W by roughly a factor of two, and every device in the chain — transformer, wire, thermostat contacts, relay, triac — is limited by current, which means it is limited by VA. Sizing from watts understates the current by that factor of two, and this single mistake explains a large share of overheated transformers and welded contacts.

One honest caveat about the hardware. Most HVAC damper actuators are small motors, not solenoids — shaded-pole or synchronous motors in two-position units, and stepper or brushless motors in modulating ones. For those, the inrush is the motor starting current, which is highest at stall, and the gap-dependent impedance story above does not apply. The VA budget and the inrush logic apply to all of them, because the transformer does not care what kind of load is pulling the current. What is solenoid-specific here is the gap effect and the blocked-stroke burnout — and both apply wherever a solenoid is doing the work, gas valves in particular.

Check these in order

1. Start with the torque the damper needs, not with the voltage. The damper manufacturer’s torque table for the size and the linkage is the input. An actuator selected on supply voltage alone will be wrong at the extremes of its own range, and the failure will appear as a stall or an incomplete stroke.

2. Get inrush VA and holding VA from the datasheet, not just watts. If the datasheet gives only watts, convert with the stated power factor or ask. This is the single highest-value number in the selection.

3. Sum the VA on the transformer, and check the worst case rather than the average. Class 2 transformers are rated in VA, and every load shares it. If two actuators can start together, the transformer must supply the sum of the inrushes at that instant.

4. Check the switching device against an inductive load, in VA. Thermostat triacs and small relays switching an AC inductive load need a snubber, or the contact arcs and welds and the triac latches on. Confirm whether the controller’s rating is expressed in VA or in amps, and what load type it assumes.

5. Check voltage drop over the run, remembering that it bites at 24 V. A long run of a small gauge costs volts in proportion to current, and current is at its worst exactly during the inrush. The permissible drop is a much smaller fraction of 24 V than it would be of 120 V.

6. Where the application allows, use a modulating actuator. It takes a constant supply and separate control, so the load leaves the controller’s contacts. That is often the cheapest fix for an existing installation that is stressing its switches.

7. Verify the fail-safe behaviour and the supply that supports it. A spring-return unit has to reset the spring on power restore, which takes the full rating; a supply that sags at that moment leaves the damper in the wrong position, and that is a safety outcome rather than a comfort one.

8. Verify that the stroke can be completed. Prove that the linkage lets the actuator reach the end stop, because if it cannot, you are energising a solenoid with the gap still open and paying for it in winding temperature.

What actually to change

Finding	What to change	Why not the other thing
Sized from holding current	Size the transformer and wire from inrush VA	The sag prevents pull-in, which keeps you in the high-current state
Sized from watts	Convert to VA using the power factor before anything else	A factor of two is more than most margins
Coil failed with correct supply	Find the mechanical reason the stroke never completed	Replacing the coil restores the fault, not the function
Actuator buzzing, will not hold	Inspect and replace the shading ring	A stronger coil does not fix a zero-crossing force collapse
Contacts welding or triac latching on	Add a snubber, or switch to a modulating actuator	Replacing the controller repeats the same failure
Voltage sag over a long run	Increase the conductor size and check the drop at inrush current	A larger transformer does not recover volts lost in the run
Several actuators on one transformer	Stagger the starts, or upsize on the summed inrush	Averaging the loads understates the instantaneous demand
Fail-safe reset unreliable	Verify the supply at the reset condition, not at steady state	The spring reset is a high-demand event, not a standing load

When it IS the harder problem

Several actuators share one transformer and one controller. Then the inrush currents coincide, and the transformer sees a demand that no individual datasheet describes. Either stagger the starts deliberately in the control logic, or size for the sum and accept the cost. Measure the supply voltage at the furthest actuator during a simultaneous start — that measurement settles the argument faster than any calculation.

You are retrofitting onto a transformer of unknown or already committed VA. This is the common field situation, and the honest move is to measure rather than to estimate: record the supply voltage at the actuator while it operates, and treat a figure that sags appreciably below nominal as evidence that the budget is full. An actuator running on a sagging supply may still work while quietly sitting closer to a stall than anyone intended.

The actuator holds a damper against a spring or a binding linkage. If the return spring or a stiff blade keeps the stroke from completing, the unit never reaches its low-current state and runs at or near the pull-in current continuously. That is a thermal failure that arrives on the coil’s calendar rather than the damper’s, and it is diagnosed mechanically — measure where the stroke actually ends.

Power loss and restore on a freeze-protection damper. The unit must close on loss and re-open on restore, which means the transformer has to deliver the inrush at the worst ambient in the year, with the linkage stiff and the lubricant cold. Sizing that system on a mild-weather measurement is how a building discovers in January that the damper it relies on cannot reset.