I am asking what happens, from the power/current consumption point of view, when I energize a relay.

In theory as relays are coils and like to hold current there shouldn't be a current spike when I energize it. But I am unsure what happens after that when it starts to attract the contact.

Example: If I need to size a transformer for some devices including 24VAC relays I just need to consider nominal power consumption/current of relays or should I consider some increase as for motor devices?

  • \$\begingroup\$ You need to consider the voltage spike generated when you de-energize the coil. Search the site for questions about "freewheel diode" for more information. \$\endgroup\$ – The Photon Feb 27 '17 at 17:48
  • \$\begingroup\$ No problem for the diode and the voltage spike... my focus is more on power consumption and if I need to size a bigger transformer considering the nominal power consumption of a relay \$\endgroup\$ – piertoni Feb 27 '17 at 17:51

DC Relays do not take high inrush currents when first switched on. To a good approximation, the relay coil looks like a inductor in series with a resistor to the driving circuit.

The small movement of the switch does change the inductance slightly, and makes a bit of energy in the inductor appear to go away from the driving circuit point of view. However, these effects are both quite small relative to the inductance and resistance of the coil. In practice, you simply ignore this effect.

The inductor exhibits the opposite of inrush current, more like a soft start. When first turned on, the current starts at zero and rises linearly. The actual current profile over time is a exponential asymptotically approaching the applied voltage divided by the resistance.

The thing to watch out for is switching the relay off. The inductor current will not go to zero instantly. In the short term, the same inductor current will flow immediately after switch-off as immediately before switch-off. The inductor will make whatever voltage is required to keep the current flowing, which will be big enough to fry the switching transistor unless you give the current a path to flow.

A diode in reverse across the coil accomplishes this. When switched off, the inductor voltage only needs to be high enough to turn on the diode. After that, the inductor current will decay due to the resistance of the windings effectively in series with it, and the forward drop of the diode.

When particularly fast off-action is required of a relay, you add a resistor in series with the diode. You know the relay coil current when on. You size the resistor to get the largest voltage drop across it you can tolerate without damage.

AC versus DC driven

What I said above applies to driving a relay with DC. As Tony Stewart pointed out in a comment, things are a little more complicated when AC is applied to drive the coil.

The problem is that the material in the coil core can exhibit magnetic hysteresis. This means the property of the coil at turn-on depends on what state it was in when last turned off. If the core is left heavily magnetized in one direction, then the inductor that is the coil will hit its saturation current limit much sooner in that direction than normally. If you happen to switch on the coil at the right point in the AC cycle, the inductor can saturate. This means the inductance gets very low.

However, the resistance of the coil is still there. The worst case current is the peak voltage of the AC waveform divided by this resistance. The result is more current at startup, but not drastically more. Even AC relays still generally rely on the coil resistance to set the current. The inductance will reduce the average current somewhat, but usually not too much at low frequencies of the power line. This is why AC relays are specified to work up to some frequency. Above that frequency the inductance of the coil reduces the average current too much for the relay to reliably operate.

  • \$\begingroup\$ 1st statement is unclear for AC relays in question. Yes at t=0 but false, they do have ~3x inrush. so question remains .. does one size transformer dedicated to many AC Relays for inrush currents or ignore them. Pls correct \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Feb 28 '17 at 15:24
  • \$\begingroup\$ @Tony: Good point about AC relays. I had answered for DC. Still, 3x seems rather extreme for most AC relays since the coil current is still largely set by the coil resistance. The inductance is usually the not the dominant part of the coil impedance at 50 or 60 Hz power line frequency. \$\endgroup\$ – Olin Lathrop Feb 28 '17 at 16:39
  • \$\begingroup\$ yes but specs I showed in my answer indicate 2-3x for Omron and another source 3.2 for 50Hz and 3.8 for 60Hz , so the AC shaded pole motor effect must be occurring in linear mode. \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Feb 28 '17 at 17:36

Unlike DC Relays which have a current limit based on DCR, AC Relays use more reactive power when moving then reduce the VA rating when activated (sealed).


Nominal Inrush @ 50 Hz 22.5 VA 
Nominal Sealed @ 50 Hz 7  VA 
Nominal Inrush @ 60 Hz 20  VA 
Nominal Sealed @ 60 Hz 5.25 VA 

Depending on how many activate at once will determine the voltage sag on your rated transformer or the speed of the relay switch during activation.

Motors often have an inrush of 3 to 8x the max rated VA load rating. It seems the example of the 24Vac coils above (switching 20~40A) have a Surge/Sealed VA ratio of 3~4x.

I am not sure what transformer rating to choose as it depends if all your relays can be switched simultaneously or not or how critical the sag performance affects relay transition time but the power dissipation in the transformer would be a short minimal effect.

Switching time does not include contact bounce time.

Diodes are not used for AC coils. enter image description here

Ref OMRON ( best quality Relay supplier) Extending the life of relay tips by reducing the amount of arcing generated as they open is achieved by connecting a Resistor-Capacitor network called an RC Snubber Network electrically in parallel with an electrical relay contact tips. The voltage peak, which occurs at the instant the contacts open, will be safely short circuited by the RC network, thus suppressing any arc generated at the contact tips. For example.enter image description here


It makes a difference if you are talking about an AC coil or a DC coil. DC coils do not have an "inrush" current in the same way an AC coil does. They have a "pick-up" power rating that is usually no different from the "holding" power rating, except on very large ones, but only because the actual weight of the armature (thereby load) is higher.

In an AC coil however, the first instant you energize it there is no impedance yet, because there is no mutual inductance from the magnetic fields of the core. So for that instant, all you have is the resistance of the coil wire itself and that is so low that it is essentially a "short circuit", so all available fault current at the terminal is drawn. A cycle later, the core fields begin to interact with mutual inductance and create impedance, bringing the coil current down.

So is that significant in terms of power consumption to a source transformer feeding that circuit? Only in terms of the transformer's ability to react to it, or the "reactance" capacity. Specifically, the sub-transient reactance is what supplies power for that first cycle. This is what is different about a "control power transformer" and a simpler transformer, a CPT is designed to provide a high level of transient reactance specifically to avoid a severe voltage dip when energizing an AC coil on a relay or contactor. Still, you have to consider it if there is more than one coil in a circuit, that's why the vendors will provide you with a formula for determining the size of a transformer based upon the largest one plus the number of other coils that will already be energized when it is.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.