As a result of heavy snowfall yesterday night, I woke up to find out that the regular 230 volts delivered to my house had dropped to 110 volts (fixed later that day by the power company). Old fashioned light bulbs worked as if a dimmer was attached to them (only a little light), fluorescent lights did not work, one LED was blinking while the remaining LEDs worked fine!

Why is it that LED lights were unaffected by the storm, while all the other lights were affected? The LEDs were regular E27 socket LED light "bulbs".

  • 36
    \$\begingroup\$ they probably have universal buck converters for 90-240 Vac \$\endgroup\$ Nov 7, 2016 at 7:07
  • 10
    \$\begingroup\$ Just to hazzard a guess, I'd say that it wasn't too expensive to start making wide-range AC inputs (many power supplies do this, now) so they could sell the lights across different markets. \$\endgroup\$
    – jonk
    Nov 7, 2016 at 7:07
  • 9
    \$\begingroup\$ The LEDs come from IKEA, so it makes sense with the universal converters since IKEA operates in both 220v and 110v markets. \$\endgroup\$
    – sbrattla
    Nov 7, 2016 at 7:26
  • 4
    \$\begingroup\$ am i the only person horrified that voltage dropped well below brownout conditions and yet nothing tripped off? how many motors and other stuff was destroyed by this? \$\endgroup\$
    – Michael
    Nov 7, 2016 at 21:54
  • 2
    \$\begingroup\$ Internally, leds usually only require a few volts (sometimes <5V) therefore as long as that much voltage makes it through they stay on. If it has the right amount electronics it could convert almost any voltage to its needs. \$\endgroup\$
    – cybernard
    Nov 7, 2016 at 23:58

3 Answers 3


Your LED bulbs are most likely fitted with a wide-range power supply, operating from 100-240 Vac. Hence, when the voltage dropped, they were still within their operating limits and your LED bulbs were as bright as before.

Side note: You'd be surprised how many bulbs marked 220-240 Vac can actually operate down to 100 Vac or less. The nameplate numbers only tell you where it can operate, not where it can't.

  • 7
    \$\begingroup\$ +1 for the last sentence. It's important to realize this! \$\endgroup\$
    – yo'
    Nov 7, 2016 at 12:47
  • \$\begingroup\$ I assume with your side not you mean the good old incandescent light bulb? Wouldn’t those be much dimmer, redder and inefficient with 100V instead of 230V? \$\endgroup\$
    – Michael
    Nov 7, 2016 at 15:05
  • \$\begingroup\$ @Michael I was only refering to LED bulbs. \$\endgroup\$
    – winny
    Nov 7, 2016 at 15:25
  • 2
    \$\begingroup\$ I'd rather say that it tells you where the manufacturer/reseller/agent guarantees that it will work. It might work outside of the specified range, or it might not, by accident or otherwise, and you can't complain either way; but if it doesn't work inside the specified range, then that's something you can complain about. \$\endgroup\$
    – user
    Nov 7, 2016 at 16:25
  • 1
    \$\begingroup\$ It's feasible(ish) that a LED light could be in danger of failing at lower voltage as some components are forced to bear more average current to compensate; so it's not wrong for a manufacturer to give restrictive specs even if it does appear to work outside of that. OTOH this isn't usually a problem for low-power devices. \$\endgroup\$
    – sh1
    Nov 7, 2016 at 21:38

You can get quite small but sophisticated LED constant current drivers that work across a universal AC power range. Here are a couple: -

enter image description here

enter image description here

Both work down to 85V AC.

  • \$\begingroup\$ While I have some physics background I'm unable to fully grok the circuit diagrams. Can I summarize it thus: Generally, LEDs need an electronic current limiting circuit (that's why they are always attached to a little PCB); in other words, the voltage fed to the assembly is too high and would burn the LED out which has very low resistance above a certain voltage threshold, if we didn't limit the current. Since the voltage is "too high" to begin with, lowering it somewhat does no harm. As opposed to light bulbs: They are Ohm resistors. The current depends linearly on the voltage. \$\endgroup\$ Nov 8, 2016 at 10:05
  • \$\begingroup\$ @PeterA.Schneider yes that's about right, you need to control the current to an LED rather than control the voltage. If you add a series resistor and know the LED forward characteristics, voltage control becomes more feasible but, if you have a clever chip that directly controls the current and does so efficiently (by using switch mode techniques) then that's the best of all worlds. \$\endgroup\$
    – Andy aka
    Nov 8, 2016 at 10:18
  • \$\begingroup\$ @PeterA.Schneider to grok latter cct. above ( refer to any Buck Regulator) except here there are 2 Gnd references while current is sensed between them thru LED(+) while IC is charged by flyback or bootstrap diode D2 to Cvcc. So IC's GND becomes LED's V+ (Anode) \$\endgroup\$ Nov 21, 2016 at 17:03

Note that with AC voltage, it is easy to use "enough of the current as necessary" to make the circuit work.

If the circuit requires most of the energy in 230V, then your situation would have caused the circuit to fail.

If the system needed less than the energy in 110V, then whether it was fed 110V or 230V wouldn't have fazed it - it would merely have utilised more of the lower AC voltage and less of the upper. The efficiency of the system would have been less for that period - but it would still have worked.

  • 1
    \$\begingroup\$ huh? You make it sound like the light only draws current while the voltage is below the max. This is pretty unlikely; as @Andy's answer shows the usual design would be a rectifier and DC-DC converter. So current consumption would happen in spikes at the top of the voltage waveform. This works the same if the input was DC. \$\endgroup\$ Nov 7, 2016 at 13:27
  • 3
    \$\begingroup\$ @PeterCordes I think he's just saying (awkwardly) that it draws more average current at less average volts, to deliver constant average power. Which is true enough. \$\endgroup\$
    – hobbs
    Nov 7, 2016 at 18:56
  • \$\begingroup\$ This seems a bit close to saying "because it does". There's a mechanism in play to explain why this change occurs for LEDs and not incandescents. After all, mains can provide way more power than even an incandescent could use, so that must be regulating it's input too. \$\endgroup\$
    – sh1
    Nov 7, 2016 at 21:31
  • \$\begingroup\$ Nice icon ;-)... \$\endgroup\$ Nov 8, 2016 at 10:34

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.