Dimming LEDs

[EFLImpudence]

No, not the usual dimming queries.

My living room has three wall lights.
When I moved in there were 60W candle bulbs so I replaced them with 9W LED and this was satisfactory.

None has failed in the six years (which is good) but recently I realised that the room was too dim to read or do anything intricate - with the change being so gradual I did not notice.

Replacing with new lamps, the difference in brightness is huge.

I do not recall seeing this discussed by anyone else. Is it normal for LED lamps?

These are the type of lamps in question:

 

[RandomGrinch]

Are there any anomalies in the individual LEDs - black spots, or some a darker colour than others?

Or are they all of uniform appearance?
...and are the new ones of the same type / lumen output as the first?
In which case - LED degradation:

 

[ericmark]

My wife said the LED's we used in the living room went dimmer, but how to show this not so easy, I suppose the exposure of pictures taken, I have told the story many times, room originally 2 BS22d bulbs, then two chandeliers fitted so 6 BA22s bulbs, and then swapped to 5 bulb so 10 x E14 bulbs.
So 200 watt went to 240 - 360 watt depending on if 40 or 60 watt bulbs, then moved to CFL so 80 watt (10 x 8 ) then 8 x 3 W + 2 x 1.8 W = 27.6 watt LED, and then 5 watt x 10 so 50 watt LED, the latter was OK, the first LED seemed OK to start with, then found too dark to read, the CFL were a failure, far too dim.

So you have to remember likely my eyes are not as good now, but seems 300 watt tungsten, = 120 watt CFL, = 50 watt LED. So LED is giving out around 6 times the amount of light got from a tungsten. However the conversion charts seem to show 7 to 10 times.

I think it was the colour temperature which made me think at first the LED was brighter than the tungsten, but to read a book seems 6 times is more of a realistic conversion. But the tungsten had just a small dark spot where the base is, but the LED shines nothing towards the base, so base up or base down also makes a huge difference.

Did experiments in University, tried over driving and flashing the light and to the eye it seemed brighter, but the lux meter showed the same, over driving did make the LED dimmer and it never recovers. So voltage spikes can make the LED dimmer.

 

[EFLImpudence]

Are there any anomalies in the individual LEDs - black spots, or some a darker colour than others?

No.

Or are they all of uniform appearance?

Yes, they all looked uniform and normal.
If there had been some dull or dead LEDs, I might have realised earlier.

...and are the new ones of the same type / lumen output as the first?
In which case - LED degradation:

Yes, the same.

Before I bought new ones, I replaced one with the original 60W incandescent and the difference was just as huge.

 

[EFLImpudence]

In which case - LED degradation:

Interesting - seems normal - surprised I have not come across it before.

Thanks.

 

[Taylortwocities]

Are there any anomalies in the individual LEDs - black spots, or some a darker colour than others?

Not an anomoly, but a 5watt daylight LED lamp will deliver more light than a 5watt warm white LED lamp.

For light output comparison you need to look at the value of LUMENS quoted.

 

[RandomGrinch]

Not an anomoly, but a 5watt daylight LED lamp will deliver more light than a 5watt warm white LED lamp.

Yes, but to be specific, these were the sort of anomalies I was thinking of

Did experiments in University, tried over driving and flashing the light and to the eye it seemed brighter, but the lux meter showed the same, over driving did make the LED dimmer and it never recovers. So voltage spikes can make the LED dimmer.

We used to use a similar experiment - running a discrete white LED through a spectrometer at varying currents - overdriving the LED red-shifts and weakens the phosphour component of the spectrum, as it is slowly degraded by the heat.

 

[JohnW2]

Interesting - seems normal - surprised I have not come across it before.

It's such a slow and insidious progress that one (certainly I) don't usually notice LEDs dimming with age (unless they get so dim as to not serve their purpose!), but I've often noticed that things suddenly get brighter if I replace an LED which has been in service for a long time with a seemingly identical unused one.

If you type "LED degradation" into Google, you'll find enough reading about this to last you a lifetime!

Kind Regards, John

 

[bernardgreen]

Many LED elements are pulsed at high frequency by the driver in the lamp. The current during the pulse is more than the element's rated current but the ON / OFF ratio of the pulsing gives a average current approximately equal to the rated current.

Some LED elements will tolerate pulse of 3 or 4 times rated current without degrading the element.
Some LED elements will be degraded by pulses of current that are as low as 2 times rated current

 

[JohnW2]

Many LED elements are pulsed at high frequency by the driver in the lamp.

Indeed.

... Some LED elements will tolerate pulse of 3 or 4 times rated current without degrading the element.

I would expect it to have some impact on life expectancy (and possibly the rate of "LED degradation"), even if the element could 'tolerate' the current in the short-term, wouldn't you?

Kind Regards, John

 

[bernardgreen]

I cannot recall what the semiconductor materials are used for high pulse current LEDs. They were developed for lamps where high brightness and low current consumption was essential. Battery operated traffic signs being one application for them

The brightness as perceived by the human eye during the pulse is significantly greater than the perceived brightness if the elements were fed with a constant current equal to their rated current

 

[JohnW2]

I cannot recall what the semiconductor materials are used for high pulse current LEDs.

I didn't realise that you were talking about 'special' LEDs developed specifically for high pulse currents, since I'm sure that many a cheap LED uses the pulsed approach, In terms of those actually intended for high current pulses, I can but presume that they are 'arted' for such use, so are not really being 'overdriven'?

They were developed for lamps where high brightness and low current consumption was essential. Battery operated traffic signs being one application for them

That makes sense.

The brightness as perceived by the human eye during the pulse is significantly greater than the perceived brightness if the elements were fed with a constant current equal to their rated current

Well, to paraphrase a notorious young lady of old, "It would, wouldn't it?" However, as above, if they are intended for pulsed use, they are presumably 'rated' accordingly, and hence are not really being 'over-driven' with the pulses.

However, going back to my comment, I would imagine that any LED (including those intended for pulsed use) would have a shorter lifespan (and maybe more rapid "LED degradation") if fed with high current pulses than if it were fed with a constant current with the same average, wouldn't it?

Kind Regards, John