LED flicker from utility frequency

[A.T.]

Why do some LED bulbs show clear utility frequency flicker, while others don't? Is it there, but not visible for the weaker lamps because of less amplitude? Is the diffuse bulb coating fluorescent and smooths the flicker away?

Here a comparison of some IKEA LEDARE models, slowed down by factor 40:

Two different 200lm bulbs models (NO FLICKER):
http://www.ikea.com/gb/en/catalog/products/10255289/
http://www.ikea.com/gb/en/catalog/products/90255290/



Same type of lamp as above, but with three 400lm bulbs of this type (CLEAR FLICKER):
http://www.ikea.com/gb/en/catalog/products/80248993/



One 1000lm bulb (CLEAR FLICKER):
http://www.ikea.com/gb/en/catalog/products/20249226/

 

[Doug Huffman]

They are light emitting diodes and function for a half-cycle only unless there are two opposed in the same enclosure. I HATE buzzing LEDs

I just finished lighting our Christmas Tree and we resolved to dispose of these LED tree lights and any other odds and ends of cheap incandescents that have accumulated over the years. NEXT year we will spend what ever is required to have incandescents that will last.

At a Christmas party Saturday night, the hosts had a tree full of antique bubble-lights and incandescent globe lights.

Our home Island is heavily wooded and live Tannenbaum are traditional. They will stay up and dressed until Easter for relief for the loong nights. Outdoors lighting is too hard to take down until the thaw.

 

[A.T.]

They are light emitting diodes and function for a half-cycle only unless there are two opposed in the same enclosure

There are usually more than two diodes in those enclosures. But the newer ones use the chip-on-board technology, which looks like one huge diode.

So, it the reason for the difference, that the non-flickering 200lm use multiple opposed diodes, while the flickering 400lm & 1000lm use COB and no full-wave-rectifier??

 

[zoki85]

Prime suspect is different rectifier topology (if they use different rectifiers)

 

[A.T.]

Prime suspect is different rectifier topology (if they use different rectifiers)

That might explain why the 200lm bulbs cost 50% more than the 400lm one. Usually it should be the other way around. But maybe the 200lm models have better rectifiers, which would be too expensive for the higher power levels of the 400lm & 1000lm models.

 

[NascentOxygen]

I haven't looked closely at LED globes, but they drive the individual emitters at a few 10s of kHz, don't they? Assuming the LEDs are placed in opposing pairs, then one of the pair will be on while the other is off when driven by a square wave. But if the drive is softer, something like a trapozoid or sine wave, then there'll be a duration each half-cycle when neither of the pair is on. So flicker will be double the drive oscillator frequency.

I'd expect RFI hash on nearby radio receivers to be potentially worse with non-flickering LED globes and their sharp-edged current drive. (But I'm just speculating.)

 

[NascentOxygen]

Maybe I jumped to the wrong conclusion. Are you saying you have some that show 60Hz flicker? If that is the case, I guess the manufacturer is just saving on power supply filter capacitors.

 

[A.T.]

Maybe I jumped to the wrong conclusion. Are you saying you have some that show 60Hz flicker?

No, it looks like more 100Hz, which is double of the utility frequency of 50Hz.

 

[zoki85]

As concerns RF interference, most CFL bulbs are noticeably worse than LED bulbs.

 

[NascentOxygen]

Almost certainly just economizing on filtering, then, I'd say. The higher the power, the less effective will be modest capacitor filtering. I find the easiest way to demonstrate flicker is to wave a school ruler vigorously back and forth in a sweeping arc near the globe, at night. It's a good stroboscope.