LED bulb/flood as strobe lamp

Oops I missed a factor of ten in my head... Anyway I was thinking about a Joule.

George H.

A 2W light bulb? Seems pretty tiny for area illumination.

So what is your problem ?

Any "cool white" LED bulb will produce about 100 lm/W and hence a 2 W E14 bulb will produce at heat 200 lm. An 15 W incandesce bulb will produce about 90 lm, a 40 W bulb will produce 400 lm, so what is the problem ?

Not buying the numbers, is my "problem".

If you live in a cave, it's fine, I guess (100 lm/W for illumination is optimistic, I think).

** Really ?

1joule = 1 watt for 1 second or 10 watts for 0.1 seconds.

The electro inside a 10W LED bulb is not gonna be 100uF, more like 22 or 33uF consistent with a 10% ripple voltage.

.... Phil

On Sun, 23 Aug 2015 07:49:05 -0700 (PDT), Phil Allison Gave us:

Most are driven by an AC switcher 'pulser' and never 'see' 'pure' DC in their entire lifespan.

Ever take an LED bulb drive base apart?

Who said that I have only one E14 bulb in the room ?

Running a "cool white" LED well below Imax and you are close to 100 lm/W including power supply losses. For "warm white " (2700 K), getting such figures would be hard.

Under it's max current, sure, but 100W/lm is about the limit of LED efficiency. You're not going to see that in the real (retail) world. How many retail bulbs will pay for more LED (or capacitor) than is needed.

** There are many schemes, but some do have a bridge and filter cap feeding off the AC supply followed by a switching current regulator. Using electro values that make sense and will fit inside the base results in hold up times of only 50mS or so.

... Phil

Whilst I agree that nothing beats doing the experiment. The right experiment here is to close your eyes switch the light on in a dark room and then switch it off and open your eyes. The fade out of white light is visible for about 200ms even on the latest generation with diffusers. Older ones are slower at least of the bunch I have here.

It is *much* easier to see on bulbs where the individual LED dies are visible since they are clearly lit at 10uA in daylight and can be seen glowing faintly sometimes with dark adapted eyes in nominally off circuits.

It is quite possible that UK ones on 240v 50Hz ac might show more significant energy storage effects than US ones on 110v 60Hz.

I agree it doesn't have anything like a persistence sufficient to explain the delay in the LED lamps but it has to have a tiny delay absorbing and then remitting the blue photon as broadband yellow.

A video of a bulb being switched off would be definitive.

--
Regards, 
Martin Brown

I checked the labels: Osram Parathom E14 100-240 V 1.6 W made of 38 warm white 5 mm LEDs, thus each LED driven at about 15 mA. It really takes about 1 s after power switch off before any change in light output can be observed, after which it drops to invisible.

Update to this. I was wrong about the maximum length of time that an LED based mains bulb can glow for after switch off. The configuration was a

20 series surface mounted on a thick heatsinky pcb with DC ~80v applied across the chain possibly with a current limit somewhere.

Switching off resulted in the leds dimming quickly to a faint glow and the weakest one went out after 10s. The most efficient LED in total darkness stayed perceptibly lit for fully 5 minutes!!!!

The lamp was in a holiday cottage where I was staying last week. Annoyingly the fitting was out of reach. The best match I could find on Amazon was this one which is all but identical physical form:

I made one in daylight of the lamp described above you can see that all the LED dies are faintly lit for about 10s and that for the first half second it emits enough light to influence colour balance under it.

The leds each fade out one at a time as the current drops away.

--
Regards, 
Martin Brown

Interesting, So each LED has the same current through it, but some have higher voltage and still emit? Or is the voltage drop about the same.. but less non-radiative recombination for the ones that stay alight?

George H.

I don't know but I would assume that the voltage drop across the ones that stay lit must be more than that of the blue photon emitted. My instinct is that some are leakier than others and this allows the few that remain lit for longest to soldier on. I was *very* surprised by this outcome. Most I had seen before were all completely out by 10s.

It may be no coincidence that the series chain length was shorter.

I guess it represents an improvement in lattice stiffness or something in the latest generation of high efficiency blue LEDs.

--
Regards, 
Martin Brown