View Single Post
  #7  
Old March 28th 21, 11:57 PM posted to rec.bicycles.tech
Jeff Liebermann
external usenet poster
 
Posts: 4,018
Default Eyc headlight problem

On Sun, 28 Mar 2021 10:50:31 -0700 (PDT), Lou Holtman
wrote:

I have LED spotlights in the bathroom, porch with a promised
lifetime of 20000hr or something like that. That is utter BS
because they will fail almost all well within that time.


20,000 hrs would be:
20,000hrs / 8766hrs/year = 2.3 years
if you ran it 24x7. Multiply the 2.4 years by whatever fraction of a
day your bathroom and porch lights operate (duty cycle) to obtain the
estimate lifetime.

Note that lifetime for consumer LED bulbs is measured to the point
where the light output (Lumens) decreases to 70% or 80% of initial
value (L70 or L80). Here's some stuff from Australia on the topic:
http://www.liteonled.com.au/led-light-bulb-life-time/
Note that nobody is going to run an LED light test for years and
years. So, there are various schemes, such as TM-21, for
extrapolating the lifetime of the LED, again based on loss of output,
not catastrophic failu
http://www.liteonled.com.au/buying-guide/led-life-time/l70-lm-80-and-tm-21-data/

It's this extrapolation, combined with somewhat ignoring catastrophic
failures, and absurd over-control of operating temperature, that is
producing very large numbers for LED lifetime. Also, note that the
only thing that a consumer LED lighting has in common with bicycle
lighting is that both use similar LEDs. The boost or buck inverters
are quite different, the cooling problems are different, and the
operating levels are VERY different. If bicycle lighting was designed
to operate at the temperatures and output levels similar to consumer
LED lighting, we would be blowing up LED bicycle headlights at an
alarming rate. In other words, a 100 watt equivalent LED bulb (14
watts, 1350 lumens) is not the same as a B&M Lumotec Eyc running on
perhaps 3 watts. The major difference is the operating temperature
and the luminaire (housing). The typical consumer lighting fixture is
a bad joke which seems to be designed to incinerate the LED bulb.
Death from overheating is common. On a bicycle, there is moving air
to cool the housing and fewer things in the way before getting to
ambient air. Put a bicycle headlight inside a home luminaire, and
you'll probably be burning out headlights from overheating.

It
is not the LED that fails but some of the electronic component on
the (small) PCB that drives the LED. They fail in a manner you
described. I think it is a heat issue.


I agree. It's the driver board that dies first, not the LED. However,
that's not the case for all consumer LED lamps. Many of the COB (chip
on board) mounts for LEDs are lacking in heat conduction. How the LED
chip conducts heat to the COB is critical. The construction of the
COB is also critical. For example, it should be ceramic or aluminum,
not G10/FR10 PCB material. Yet, I see flashlights, light bulbs, and
cheap headlights using PCB material.

Another problem is that the consumer LED bulb drivers often use
aluminum electrolytic capacitors. These do not like high temperatures
and conveniently follow a well known temp to lifetime curve:
https://elc.kemet.com
https://www.illinoiscapacitor.com/tech-center/life-calculators.aspx
Give me the operating electrical conditions and temperatures, and I
can estimate when the capacitor will blow up. Want a 3 year lifetime?
I can provide the capacitor values needed to have most of the
components on the driver PCB blow up simultaneously.

"Selecting the right capacitor to ensure longer life of LEDs"
https://www.electronicsb2b.com/important-sectors/leds-led-lighting/right-capacitor-to-ensure-longer-life-of-leds/

"How Electrolytic Capacitors Effect the Reliability of LED Drivers"
https://www.led-drivers.com.au/blog/how-electrolytic-capacitors-effect-reliability-led-drivers
On the graph, notice how an operating temperature above 55C case
temperature has a drastic effect on capacitor life.

"Ensure long lifetimes from electrolytic capacitors: A case study in
LED light bulbs"
https://www.edn.com/ensure-long-lifetimes-from-electrolytic-capacitors-a-case-study-in-led-light-bulbs/

Note that MLCC (multi-layer ceramic capacitors) are a big improvement
over electrolytic caps, but still have (different) problems. Polymer
caps are a good compromise, but tend to be expensive. I've seen them
in computers and expect to see them in LED lighting eventually.

"Polymer Film Capacitors for LED Drivers"
https://www.powersystemsdesign.com/articles/polymer-film-capacitors-for-led-drivers/36/5690

Incidentally, a friend who specializes in lighting electronics design,
once told me that he could recognize the quality and lifetime of a
lighting driver or bulb by just looking at the capacitors. I'm not
sure I could do it today, but thinking back, his method was quite
accurate.

Enough for now.
May the light at the end of the tunnel not be a fire.

--
Jeff Liebermann
PO Box 272
http://www.LearnByDestroying.com
Ben Lomond CA 95005-0272
Skype: JeffLiebermann AE6KS 831-336-2558
Ads
 

Home - Home - Home - Home - Home