View Single Post
  #29  
Old March 7th 17, 04:41 AM posted to rec.bicycles.tech
Jeff Liebermann
external usenet poster
 
Posts: 4,018
Default More About Lights

On Mon, 6 Mar 2017 18:37:14 -0800, sms
wrote:

WE passed 70 lumens per watt a long time
ago.http://www.cree.com/LED-Components-and-Modules/Products/XLamp/Arrays-Directional/XLamp-MKR.
And bicycle light manufacturers are early adopters of the latest LEDs.


I got burned badly when I specified an LED that was rated at 100
lumens/watt, and got something that was about 70 lumens/watt. If you
look at the Xlamp MK-R data sheet at:
http://www.cree.com/~/media/Files/Cree/LED-Components-and-Modules/XLamp/Data-and-Binning/XLampMKR.pdf
the footknots always says:
"Flux values @25*°C are calculated and for reference only."
I'm not sure exactly what that means, but from my limited experience,
it means that if you pulse the power to the LED with a very short duty
cycle, you might get something near the rated efficacy. If you run it
on DC, it won't even come close. Calculated also means that they
didn't bother to verify their own calculations with an integrating
sphere measurement.

Assuming 200 lumens/watt and 1000 lumens output, that would require 5
watts of input power. Looking at the same data sheet (Pg 15) the 6v
chart shows that 5w is 5.8v @ 0.85A. So far, so good. However, if we
look at the Relative Flux Vs. Current (tJ = 85*°C) graph on Pg 16, the
output is derated to 65% of 1000 lumens. Producing only 650 lumens.
Using those numbers, the efficacy is now:
650 lumens / 5 watts = 130 lumens/watt
Note that the 85C is the junction temp, not the case or heat sink
temp.

Then, there is matching LED chips. See Pg 20
"Performance Groups - Luminous Flux (Tj = 85*°C)"
where one can pick the chip output anywhere from about 515 to 1290
lumens. What is not obvious is that the efficiency also drops at the
lower output levels, resulting in even lower lumens/watt.

Then there's the loss in output due to reflectors and lenses and the
increase of input current from current regulator losses. I would
guess(tm) 15% loss of lumens to the optics, and 10% for the regulator
efficiency. So, we start with
650 lumens / 5 watts = 130 lumens/watt
and end up with:
650 * 0.85 / 5 * 1.1 = 553 / 5.5 = 100 lumens/watt

Can you see where this is going? In the end, the lumens/watt value
for the end product is the light output including all the optics,
divided by the power input as measured from the battery, and derated
by the chip junction temperature.

A dynamo can produce more than 3 watts at high speeds.


I don't ride at high speeds. I'm slowly converting one of my machines
to a "comfort" bicycle.

In the olden days
of incandescent dynamo lights there was protection circuitry to limit
the power to the bulb by cutting off the voltage over 6V. But these high
power LEDs can handle 15W so there's no need to limit the power.


Yep. So why do many dynamos still use the increased inductive
reactance at higher speeds (and higher frequencies) to provide this
voltage limiting? It would be easy enough to reduce the winding
inductance and allow the voltage to increase linearly instead of
flat-topping. The manufacturer might be afraid of blowing something
up with too much voltage at high speeds.

This light http://supernova-lights.com/en/supernova-e3-triple-2 claims
640 lumens at 4.5W. 4.5W is definitely possible from a nominal 3W
dynamo. It claims 800 lumens peak.


640 lumens / 4.5 watts = 142 lumens/watt measured "warm".
If they start out with a 200 lumens/wall LED, that might be possible.
Ok, I stand corrected, but I'm also suspicious.

I couldn't find any test reports or method used on the Supernova site.
I did find this test of the M99 model:
http://supernova-lights.com/en/blog/supernova-blog-1/post/first-test-performance-supernova-m99-pro-br-36
1100 lumens / 16 watts = 68.7 lumens/watt. Hmmm...

--
Jeff Liebermann
150 Felker St #D
http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
Ads
 

Home - Home - Home - Home - Home