Dave Plowman (News) wrote:
Think it's generally known that dimming a tungsten bulb reduced its
efficiency dramatically. In other words, a 100w dimmed to half brightness
used a lot more electricity than a 50w on full. Although dimming did
extend the life of the bulb.

So how about LEDs? Since they generate far less heat are they still as
efficient when dimmed? And does dimming make a difference to their life?


"The Inconsistency of Dimmers with LED Lamps"
Doug Lindsey, Electric Power Research Institute
Frank Sharp, Electric Power Research Institute
Teren Abear, Southern California Edison

https://www.aceee.org/files/proceedi...pers/1_154.pdf

Page9 Figure 4. Efficacy curves for all lamps with Dimmer 1

*******

It stands to reason, there's a reduction in bulb temperature
at low settings, which would have some impact on SMPS life.

For the LEDs, the model is less clear. The paper itself isn't worth
much, except it throws shade on using plain Arrhenius modeling
and accelerated life testing. We could probably say that Arrhenius
isn't dead, and a cooler LED last longer. But real life is
apparently not as nice as theory.

https://www.prognostics.umd.edu/calc...ity_review.pdf

The Arrhenius model... is not adequate to represent the failures of LEDs.
Light output degradation is the major failure mode of LEDs, and it
results from hygro-mechanical and electrical stresses in addition
to thermal stresses. A more realistic method of LED lifetime estimation
is required that reflects total consideration of temperature,
the level of forward current, relative humidity, mechanical
stress, and materials.

With PWM control, the current is delivered in constant amplitude pulses
(this prevents colour shift), but the temperature will drop due to the
reduction in average power.

It doesn't do much to explain how the first generation of
bulbs promised 25000 hours and now there are some 10000 hour bulbs.
Shades of past cartels ? :-) Well, everything is engineered
within an inch of death, so why should a lowly LED lightbulb
be spared such treatments ? If my car real axle only has 10%
margin, when 70 years ago there was 300% load margin, who
is the wiser ? No one will ever know. Some old timers will
look under a car and marvel at the spindly axle used. A car
executive had a larger steak dinner tonight because of that,
making the choice well worth it.

https://spectrum.ieee.org/tech-histo...ulb-conspiracy

Paul