LEDs and light persistence??

[GRHazelton]

I just bought two LED night lights, photocell controlled white light. For whatever reason I pulled one from the AC outlet in the dark while it was lit. To my surprise the glow GRADUALLY diminished, taking perhaps 5 seconds to completely disappear! Both of these night lights did the same thing. Another night light, blue, non-photocell controlled, doesn't do this. Would the photocell controller have any effect on this?

Has anyone else encountered this? I'd had the impression that LEDs pretty much were just on or off, with no "cool down." If this is common it could vitiate making a "cold light" enlarger source.

 

[DWThomas]

There could be some sort of filter capacitor that stores a bit of charge and fades slowly. I have some 60 and 75w equivalent LED lightbulbs that definitely are nowhere near instant off (which suggests maybe not too good fro enlarger use).

 

[wombat2go]

The electroluminescence of a colored LED (emission of photons by the charge carriers) should cease very quickly at switch off (nano seconds).
However white leds employ a phosphor to partially convert a discrete color band to white light. These use the principle of fluorescence, which, as I understand it, has an exponential decay after the light source is switched off. ( For example an oscilloscope's persistence or an old type fluoro light )

 

[Steve Smith]

There could be some sort of filter capacitor that stores a bit of charge and fades slowly. I have some 60 and 75w equivalent LED lightbulbs that definitely are nowhere near instant off (which suggests maybe not too good fro enlarger use).

My enlarger fas an LED light source. It will be instantly off is the switch is after the capacitor.

EDIT: I just realised that you are probably referring to an LED light which replaces a standard bulb. In which case, the power supply is integral and there's not much you can do about it.


Steve.

 

[GRHazelton]

Thanks for the quick responses! When one dies I might dissect it to see what's in there. Both the phosphorescence and filter cap theories sound reasonable. Somewhere I saw a thread about making a BW safelight with LEDs. It would seem that if the LED has a very narrow bandwidth it could make an ideal safelight, nice and bright, like the old Thomas or whatever discharge units. They were always too costly for me!

 

[Steve Smith]

Both the phosphorescence and filter cap theories sound reasonable.

Much more likely to be the capacitor. I have never known an LED to continue glowing without any current flowing through it (and I have had many years experience with LEDs).

Just for interest, here is the website of a company producing domestic and architectural LED lighting. The company I work for makes the flexible LED circuits and light guides for them: http://designledproducts.com/


Steve.

 

[wombat2go]

Much more likely to be the capacitor. I have never known an LED to continue glowing without any current flowing through it (and I have had many years experience with LEDs).

Just for interest, here is the website of a company producing domestic and architectural LED lighting. The company I work for makes the flexible LED circuits and light guides for them: http://designledproducts.com/


Steve.

I am interested in this because I just replaced all the old fluorescent lamps in my basement with
"LE 10 Watt 2 Foot T8 LED Tube, 20 Watt Fluorescent Tube Equivalent, Daylight White"

I don't open the film in this area but I do bw enlargements occasionally as the lights are in the sink area

They have quite a long afterglow, and I agree it may be partially due to the filter caps.
The best way to see the after glow is to leave the eyes closed until just after the light is switched off.

These ones have a decaying visible glow for around 2 seconds. I will try again tonight in the dark.

So on my electronics bench, I powered up a component super bright white led of 700mcd, 25mA, and aimed a fast photodiode at it.
Just with a mechanical switch, the oscilloscope indicates a fast fall of < 1 millisec of light emission to something less than 5% brightness, then a slow decay of that 5% to zero in about 800 millisec.
These are approximate, i will have to set up a switching transistor to get more accurate results.
But there is definitely an afterglow and filter capacitors are upstream of the switch in this test.

 

[Roger Cole]

I have an LED lamp house on my Omega D2. Works great and I love it. Others have talked about the reasons this might vary in a night light.


Sent from my iPhone using Tapatalk and 100% recycled electrons - because I care.

 

[DWThomas]

Re: my 1st post, yes, I was e-speaking of self-contained LED bulb replacements or plug-in night lights and such. My understanding is that white LEDs are actually UV emitters with a white phosphor surround, sort of a solid state fluorescent lamp. As such, their turn-off characteristic could be a property of the phosphor more than the LED. If you used separate RGB LEDs for a variable contrast/color enlarger head, the phosphor would be out of the system.

As one who remembers the first feeble little dim and inefficient LEDs that came out, the latest generation boggles what's left of my mind.

 

[cliveh]

When we moved into our now house, the ground floor was lit with 23 X 50 watt halogen bulbs, which is 1150 watts. I replaced these with 23 LED 7 watt bulbs (50 watt equivalent). Total wattage is now 161 watts.

 

[MattKing]

Are you sure the night lights aren't the type that have a battery and are designed to light up when the power goes out

 

[bobwysiwyg]

I can't speak to the night light, but we have replaced (over time) virtually all our lights with LED's around the house. Price and performance is quite good these days. Some are standard table lamps or ceiling fixtures and some are canned ceiling floods. All are not "instant " on or off as traditional incandescent lamps, but hardly noticeable for daily living.

 

[Steve Smith]

All of our street lighting here is being replaced with LEDs. It's much better than the sodium vapour lamps they are replacing.

The light is white instead of orange but the main advantage is that it is much more directional meaning that we no longer have an orange glowing sky over built up areas. This makes the stars easier to see.

The power needed to run them is considerably less too.


Steve.

 

[ParkerSmithPhoto]

So on my electronics bench, I powered up a component super bright white led of 700mcd, 25mA, and aimed a fast photodiode at it.
Just with a mechanical switch, the oscilloscope indicates a fast fall of < 1 millisec of light emission to something less than 5% brightness, then a slow decay of that 5% to zero in about 800 millisec.
These are approximate, i will have to set up a switching transistor to get more accurate results.
But there is definitely an afterglow and filter capacitors are upstream of the switch in this test.

I'm fooling around with a Cree LED PAR 38 replacement to power my Beseler 4x5 condenser head, and it looks just like what you describe. You turn it off and most of the light dies instantly, with a fading afterglow that feels like a 1/2 second. I guess as long as it is consistent there wouldn't be any problem using it.

 

[wombat2go]

I'm fooling around with a Cree LED PAR 38 replacement to power my Beseler 4x5 condenser head, and it looks just like what you describe. You turn it off and most of the light dies instantly, with a fading afterglow that feels like a 1/2 second. I guess as long as it is consistent there wouldn't be any problem using it.

Yes, I think LEDs will be quite OK for enlarging, noting the reports of good results.
I have converted the main basement /darkroom area to tube LEDs. I do my film in a little side room under the stairs and will leave the incandescent light in there.

 

[Mainecoonmaniac]

It could be the power supply?

My brother an engineer with Intel told me that LED switch of so fast that LED dimmers work through duty cycles. Electrical pulses that turn on and off the bulb that dim the LED.

 

[ME Super]

LEDs light up when current passes through them in only 1 direction (DC). Chances are there's a bridge rectifier to convert the AC used in houses to DC for the LED. Then there's a capacitor to smooth out the rectified voltage, followed by a resistor to limit the current flowing through the bulb. The resistor also limits the rate at which the capacitor can discharge, which means that the LED will glow for some (short) time after the power is cut, unless there is a switch after the resistor. Couple this with amount of time that the phosphor may glow (on the white LEDs) and it's not instant-off. The shortest amount of time to fully off is, in theory, the amount of time it takes the phosphors to stop glowing. The longest amount of time to fully off is the amount of time it takes the phosphors to stop glowing + the amount of time it takes the capacitor to discharge to below the voltage required for the LED to conduct.

If a single diode is used as a rectifier instead of the bridge circuit (4 diodes), then the designer of the bulb will probably use a bigger filter capacitor to do a better job of smoothing out the voltage, resulting in increased amount of time for the LED to turn off.

 

[tkamiya]

There must be someone who actually took these things apart. I haven't. But based on its weight and basic characteristics of LED bulbs, I would think following is happening inside.

120V AC mains is coming in from one end.
This is rectified and filtered.
Then this is modulated via PWM (pulse width modulation) and goes through a transformer at higher frequency
This is rectified again and filtered.
The current flowing to LED arrays are measured here and negatively fed back to the PWM mentioned earlier. If too much current, reduce pulse width. If not enough, increase pulse width.
LED lights at constant current as a result.

Because of size, there can't be very large capacitors in these things. But every time AC is rectified and filtered, there is a fairly large capacitor there which acts like quick charge/discharge battery. So there is a time constant. Once power is removed, it takes a bit of time to discharge completely.

That - I think - is what may be happening.

 

[tkamiya]

I actually found circuit diagram for such LED bulbs by googling "LED bulb circuit". The circuit is very VERY simple.

AC main goes into a bridge rectifier. It then goes into a filter capacitor. Then there is a current limiting resister, then into a series-parallel connected chain of LEDs. Seems most of them do not involve PWM.

The part that could cause persistence is the capacitor right after the rectifier. Common value appears to be anywhere between 10 microfarad to 220 microfarad. But because of the current involved, even at 220 microfarad, there should not be much of perceivable latency in light going out. It should almost be quick. (of the top of my head, 1 second or less) Maybe OP's particular bulb has some unusual circuitry....

I'm puzzled.

 

[Steve Smith]

I actually found circuit diagram for such LED bulbs by googling "LED bulb circuit". The circuit is very VERY simple.

AC main goes into a bridge rectifier. It then goes into a filter capacitor. Then there is a current limiting resister, then into a series-parallel connected chain of LEDs. Seems most of them do not involve PWM.

The part that could cause persistence is the capacitor right after the rectifier.

I think that's the only thing which can cause it. And that circuit is exactly how I would do it. No need for any fancy switch mode power supplies for something so simple.


Steve.

 

[alanrockwood]

Just with a mechanical switch, the oscilloscope indicates a fast fall of < 1 millisec of light emission to something less than 5% brightness, then a slow decay of that 5% to zero in about 800 millisec.
These are approximate, i will have to set up a switching transistor to get more accurate results.
But there is definitely an afterglow and filter capacitors are upstream of the switch in this test.

Two theories were mentioned, slow decay of fluorescence and slow discharge of a capacitor. Either could fit this observation.

The fluorescence theory would be as follows. (Warning: spectroscopy/physics/photochemistry ahead) It is possible for the emission to come from a combination of both fluorescence (usually fast) and phosphorescence (usually slow). There can also be something called intersystem crossing wherein a singlet state (responsible for fluorescence) converts to a triplet state (responsible for phosphorescence.) For those inclined to go deeper into these concepts, emission from a triplet state is "forbidden" in most molecules, which means it is slow, whereas emission from a singlet state is "allowed" in most molecules, which means it is fast.

The capacitor theory can also work if there is an RC network in the circuit, not of the simple RC type but a more complex RC network. This can give a more complex decay than a simple exponential.

 

[GRHazelton]

Are you sure the night lights aren't the type that have a battery and are designed to light up when the power goes out

Nope. Just the usual auto type, an upgrade from the incandescent type.

 

[tkamiya]

So.... Let's smash that LED lamp and find out! Anything for science and to settle an Internet argument, right?? hehehe....

 

[GRHazelton]

So.... Let's smash that LED lamp and find out! Anything for science and to settle an Internet argument, right?? hehehe....

Argument?? This has been a civil, informative discussion! Wanna see an argument? Visit another photo site, its name begins with U..... Say no more, say no more....

 

[DWThomas]

So.... Let's smash that LED lamp and find out! Anything for science and to settle an Internet argument, right?? hehehe....

Hehe, actually I may have a candidate, I had an "infant mortality" in a 75W equivalent Cree bulb a week or so ago. Since it's only about a month old I thought I might dredge up the receipt and see if I can get it replaced. If that doesn't succeed, I go for the razor saw!

I tend to think switching supplies might come into play with some, especially the ones that work off a standard dimmer, but one of these days I'll find out.