I'm tring to choose a relay for my circuit which has a Filament LED bulb as load.

The bulb is 230VAC / 60W 6W. The nominal current should be about 0.03A which is not an issue for a regular relay. The inrush current can go from 10 to 15 times that value, from what I understand. So it can go up to 3A0.4A.

So I just need a relay that can handle a max switch current of 3A 0.4A?

How can I control the inrush current before it goes to the relay? If I add another bulb to the circuit, I'll have to get another relay that can handle twice the current. And this is not a good idea in the long-term.

UPDATE 23/03

  1. The bulb's power is 6W, not 60W as @Misunderstood pointed out.
  2. I was able to probe the bulb in an osciloscope and the inrush current max was about 2A. The average value was about 1.5A. These values may not be much accurate because the probe used was for high current values and it has low sensitivity to low current values. The real inrush current is probably below these values but since it's not much and it doesn't require a special/expensive relay, I'm not too worried.
  3. Conclusion: Since I'm not thinking of using more than 3 bulbs, I'll get a relay rated at >= 6A peak. I'll probably add a relay socket as suggested by @Harper and put everything inside a junction box. I'll add the full schematics (with the relay controller and the other switches) later, in a new post, so I can get some feedback.


  1. I have used triacs, in the past, to control the brightness of a bulb but it wasn't a LED bulb. I was told triacs dont like LED bulbs and may not be suitable to be used as dimmers. What are my options here?

Thanks everyone for your help. Feel free to comment if my conclusion is wrong of if there's something I'm missing.

EDIT: Added schematic.


simulate this circuit – Schematic created using CircuitLab

  • 4
    \$\begingroup\$ How did this get flagged for a product recommendation? It is asking about inrush current. \$\endgroup\$
    – Voltage Spike
    Mar 22, 2017 at 16:24
  • \$\begingroup\$ A NTC thermistor is commonly used for this. As it heats the resistance goes away. \$\endgroup\$ Mar 22, 2017 at 17:15
  • \$\begingroup\$ "LEDs are not subject to this limitation. LED current is controlled by a resistor, so you can use the relay's full current rating." Is this correct ? Are LED lamps not subject to inrush corrents ? \$\endgroup\$
    – nip
    Mar 22, 2017 at 17:18
  • \$\begingroup\$ When you say lamp you mean a store bought lamp that uses LEDs rather than light bulbs correct? so this would not have anything to do with LEDs if that is correct. And in this case the LED are not likely controlled by a resistor. Most commercial lighting products use multiple medium powered LEDs and a current regulator. \$\endgroup\$ Mar 22, 2017 at 17:24
  • 1
    \$\begingroup\$ @nip 60 watts of LED is a staggering amount of light, more like a 500W halogen. that would never exist in that style. More likely you have a "60W equivalent" that's about 8 watts. Conservatively assuming a terrible 34% Power Factor, that's 24VA or 0.10 amps, not considering inrush current. \$\endgroup\$ Mar 23, 2017 at 1:44

4 Answers 4


Look at the listed amp rating of your bulb

You need to look at your bulb's documentation/data sheet for its listed draw in amps. If it provides a VA rating, you can compute amps from VA/volts. If it provides actual watts and power factor, you can compute Watts/Volts/PF. LED bulbs are often marketed as "Same brightness as a 60 watt incandescent bulb", that is a meaningless number.

Look at the rating of your relay

Inrush current is a big current surge on initial startup, but no particular reactivity when interrupted. (contrast with an inductive load which has a huge kick when interrupted, and the kick can leap across contacts). The inrush current on your consumer LED-based lamp product is lesser than that on an incandescent bulb. Look at the rating on the relay. Relays are listed at different power ratings for

  • resistive (usually the highest number)
  • motor (huge inductive kick)
  • ballast (meaning magnetic ballasts, also a big inductive kick)
  • Tungsten (meaning incandescent bulb, meaning inrush current)

If it lists a tungsten amp rating, then you can use that figure straight-up; they already compensated for inrush current and UL/CSA/TUV etc. listed it as such. You don't need to.

Comply with Code - use the right relays

Since you are switching mains voltage, you must also comply with your local electrical codes. They will require that you use not random components, but assemblies listed for mains use. Contrast:

enter image description here enter image description here

The first is a bare component, you'd need to solder wires to it and then what? Wrap it in electrical tape and leave it to flop around in a junction box? Unacceptable with mains.

The second one is designed to mount on a standard junction box fitting. Note how this puts the mains wiring inside the junction box, and the low-voltage wiring outside. This satisfies separation requirements of high voltage and low voltage wiring. Once outside the high-voltage "envelope" you are subject to the much more liberal low-voltage wiring rules, same as telephone; or doorbell or thermostat in countries where that is low-voltage.

Relay the relays

There are many relays like the second one. Some may be 5V coil. But other voltages will be much more readily available, e.g. 24V is popular in North America. This may not be a problem: some have a built-in transformer that matches their coil. These devices supply you 2 wires - if you short them, the relay picks up. Simple as that. You can do that with the 5V relay of your choice.

Even better, 24V travels well. That means you can switch the mains voltage in the location that makes sense, i.e. down at the service panel/mains supply or somewhere it's easy to access an electrical box to mount it... keep the 5V near your PC... and let the intermediate coil voltage do the traveling.

  • \$\begingroup\$ Thanks for the write-up. Honestly, I havent thought of how I'd assemble the circuit yet. I was thinking of testing it on a breadboard and then print a pcb. Its the first time I see those types of relays, and I've searched in 2/3 electronics stores in my city. But the thing is, you don't really answer my question. I know there are relays capable of dealing with 2/3 bulbs but my issue is the long-term scability. If I'm dealing with only 1 bulb, I can have a 4A relay. If by some reason I decide to use 3 , I need to get a 12A relay. I'd rather have a solution that wasn't dependent on the relay. \$\endgroup\$
    – nip
    Mar 22, 2017 at 19:10
  • \$\begingroup\$ Having this is mind, what do you think about the solution suggested by @Misunderstood of using a NTC thermistor ? \$\endgroup\$
    – nip
    Mar 22, 2017 at 19:17
  • \$\begingroup\$ See edits. If the relay has a tungsten rating, that accounts for inrush current and you don't need to factor for it. That's a solid gold number some listing agency (UL/CSA/TUV) stands behind. So a 10A tungsten relay can drive dozens of 0.20A lamps. Don't need that thermistor, the system is built to handle inrush current. You won't find these parts in any electronics stores because mains electrical is a totally different discipline with totally different supply chain. Check it out, more arrows in your quiver! \$\endgroup\$ Mar 23, 2017 at 1:36
  • \$\begingroup\$ I'll have to look more into that. So far I've seen dozens of relay datasheets and I've never seen a Tungsten rating. \$\endgroup\$
    – nip
    Mar 23, 2017 at 10:16
  • \$\begingroup\$ The bulbs dont have any documentation, they are a cheap "replica". That's why I wanted to check with the osciloscope, to make sure it wasn't making weird things. \$\endgroup\$
    – nip
    Mar 23, 2017 at 13:08

These bulbs are made to be switched on with regular household switches.

You need not do anything other than adding the relay. Current will be about 30mA. Nearly any relay will do the job.

All the other "60W" filament LED bulbs are in the same class 7-10 Watts.

There is a table on this site: The Next Generation of LED Filament Bulbs

5 Watt, 375 Lumens, Filament LED Bulb

enter image description here

This is a 60W Equivalent Soft White by Westinghouse, it draws 7.5 Watts.

enter image description here

6.5 Watt, EcoSmart 60W Equivalent Soft White A19 Dimmable Filament LED Light Bulb

 EcoSmart 60W Equivalent Soft White A19 Dimmable Filament LED Light Bulb


You may find something like this in the base of an LED light bulb.

enter image description here

What I found at the patent office was this image. The base (10) "contains a control board". This base is smaller than the above circuit board.

enter image description here

The LED light bulb 8 includes, in its base 10 hermetically sealing the opening of a glass cover 12, a control board (not shown) that converts commercial electric power into electric power for driving the LEDs.


enter image description here

As shown in figures, the LED light bulb provided by the present invention at least includes: a bulb base 10, an insulation part 20, a power module 30, a support post 40, a light source module 50 and a lampshade 60.

The bulb base 10 is e.g. but not limited to an E26/E27/B22 connector. If being applied in a small LED light bulb, an E12 connector used in a small nightlight can be adopted. If being applied in a large illumination lamp, an E40 connector can be adopted. According to one embodiment of the present invention, an E27 connector is adopted for illustration and shall not be a limitation to the scope of the present invention.

The insulation part 20 is disposed on the bulb base 10, formed as e.g. but not limit to a hollow barrel-shaped structure, and formed with an accommodation space 21, the bottom thereof is formed with a thread 22 for being screw-fitted in the bulb base 10, two sides thereof are respectively formed with a fasten post 23, the top end of the fasten post 23 is formed with a fasten hole 24. Wherein, the insulation part 20 is made of a plastic material. In addition, a slide groove 25 is respectively formed between the two fasten posts 23 of the insulation part 20. Moreover, the insulation part 20 can be served to insulate the power module 30 thereby complying with relevant safety regulations.

The Filament

enter image description here

The substrate 10 is set to be of an elongated bar-shaped construction to constitute a main body of the LED filament. In present embodiment, the length of the substrate ranges from 5.00 mm to 200.00 mm, the width thereof ranges from 0.50 to 10.00 mm, and height thereof ranges from 0.10 mm to 5.00 mm. the light emitting unit 20 is fastened onto at least one side surface of the substrate 10, and includes plural regularly distributed blue light chips 21 and red light chips 22. The blue light chips 21 and red light chips 22 are sequentially connected to one another in series by a metal conductive cable 40. Two ends of the substrates 10 are provided with electrode pins 50 connected respectively to the two ends of the metal conductive cable 40.


The 6 filaments are connected in series as are the 25 LEDs in each filament.
this is about 150 LEDs, red and blue mixed, so a forward voltage on average of 2.5v so about 375 volts total forward current. At 5 Watts that about 13 mA.

So there must be a rectifier and tiny boost regulator in the base.

Not much to create massive inrush.

The 1.5 to 2 Amp inrush your are seeing sounds a bit high for these circumstances.

How was the current measured. The voltage across a shunt resistor?


The video was enlightening. The rectifier and linear current regulator is what I'd expect. The light bulb in the video, the filaments were not connected in series like the patent. The current regulator is the equivalent of a dynamic resistor that adjusted its resistance to the voltage applied.

For a linear regulator to work the LED voltage must be less than the power supplied.

This means no capacitor and no inductor which equals no inrush.

The one bulb with the capacitor in the beginning of the video would create negligible inrush.

Now this is a 553 Watt LED grow light. It has 144 3 Watt and 48 2 Watt LEDs. These put out the equivalent light of a 1000 Watt HPS lamp.

An indoor farm may have dozens of these. When you flip the switch on them I'll expect some inrush. Not yours.

enter image description here

Now compare the little LED PCB above to this power supply. This power supply has some serious inductors and capacitors.

It's the capacitors (fast charging batteries) and inductors (building magnetic field on startup) that create inrush.

enter image description here

  • \$\begingroup\$ But other types of bulbs (incandescent or halogen, e.g) are also made for that purpose and they also have inrush currents. At least, thats the idea I got by researching about it. Either way, I'm going to try and take the bulb to uni and measure it on an osciloscope. \$\endgroup\$
    – nip
    Mar 23, 2017 at 10:06
  • \$\begingroup\$ Your current is only about 30mA, not much of an inrush. Where I see the inrush thermistor used is in high wattage power supplies and amplifiers. I'm not sure what you are reading that has you so concerned. You have no reason to be concerned. \$\endgroup\$ Mar 23, 2017 at 15:33
  • \$\begingroup\$ The inrush corrent is about 10-15 times the nominal current. If the bulb is 30mA, the inrush current can be 3A-4.5A. Using 3 bulbs, it goes up to 9A-13.5A. This was my concern. By testing the bulbs I have here, I was able to measure the inrush current, which is about 2A (max). In case I have 3 bulbs, I only need a 6A peek relay. If I didnt test, and guessed the value, I'd probably get a 15A peek relay. A NTC thermistor is used in similar scenarios but I'd have to make ajustments because they are not designed for fast-switching. \$\endgroup\$
    – nip
    Mar 23, 2017 at 15:47
  • 1
    \$\begingroup\$ @nip No, the inrush cannot be that large. There is nothing there to generate it. There has to be a low impedance that creates the inrush. An LED light bulb does not have that kind of capacity to absorb 3-4.5 amps like a motor or amplifier. Electrons will not go where there is no emf to compel the electron to go there. You do not need an NTC because there is not enough current to generate heat. I do not know why fast switching and adjustments were mentioned. It's just a relay, a mechanical switch, nothing going on here. \$\endgroup\$ Mar 23, 2017 at 16:15
  • \$\begingroup\$ Okay I searched for inrush light bulbs and something did come up on LED lamps. You have a 7 Watt bulb. These inrush discussions are relating to turning on thousands of watts of LED fixtures. An LED fixture can require 500 Watts and there may be many of them that need to be turned on. The inrush is caused by the power supplies providing all that wattage. Your 7 watt bulb has no huge capacitors or inductors that would suck up electrons. Especially yours, there is no place to put a power supply like what is used in other LED light bulbs. The base is too small. \$\endgroup\$ Mar 23, 2017 at 16:34

In rush can handles with a circuit to track the voltage crossing and enabling an AND logic switch at about the zero crossing.

  • 1
    \$\begingroup\$ Welcome to EE.SE. The zero-cross solution does not solve the problem of the inrush on the first cycle or two when the internal LED power supply capacitors are discharged and have to be charged from empty. There's an edit link below your answer if you wish to improve it. \$\endgroup\$
    – Transistor
    Dec 22, 2019 at 13:37

In the bulb, it seems that the electronic circuitry is simply a rectifier full-wave feeding a capacitor (filtering) then a resistor and Leds. Formely and statisticaly, the inrush current peak may be very high if turn-on occurs on the maximum voltage (need of fuse). In fact, it can be a short-circuit, thus current not limited. A NTC may be needed inside before the capacitor but is costly. That is the reason why electronical switches can not be used for dimming with some kind of bulbs.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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