Candela <-> Lumen software

I've developed a simple program to convert Candela in Lumen (approximation). It's here, you need VB runtime (Micro$oft) on your computer.

If you have comments, tips or something to add, please do!

I added the program to my website on the "Other LED Links" page.
The program is hosted on my own server, so bandwidth won't be an issue for you.

THANX! What software did you use to make that program? I wanted to make my own... but I couldnt find suitable software.

-tom

Benadski: sorry to go off topic, but your signature, is it talking about the "UFO" luxeon clones in UV? If so could you make a post about them! I've bene wondering if they were any good for a long time and not many people noticed they were avalible in UV.

What algorithm do you use?

I've found 2 dirty algorithms for mcd * beam angle >> lumens - and they don't jive with what you've got. I will confess that I don't really understand them.

Both of these are in spreadsheet notation and probably just a tad dirtier than necessary.

Lambertian formula
( 1-ABS( ( ( COS( Beam angle / 2 * ( PI/180 ) ) ) ) ) ) * 2 * PI ) * ( mcd / 1000 )

point-source formula
PI * ( RADIANS( Beam angle / 2 ) ) ^2 * ( mcd / 1000 ) )

[ QUOTE ]
idleprocess said:
What algorithm do you use?

I've found 2 dirty algorithms for mcd * beam angle >> lumens - and they don't jive with what you've got. I will confess that I don't really understand them.

Both of these are in spreadsheet notation and probably just a tad dirtier than necessary.

Lambertian formula
( 1-ABS( ( ( COS( Beam angle / 2 * ( PI/180 ) ) ) ) ) ) * 2 * PI ) * ( mcd / 1000 )

point-source formula
PI * ( RADIANS( Beam angle / 2 ) ) ^2 * ( mcd / 1000 ) )

[/ QUOTE ]

ditto that.

here's what i use: 2PI * cd * [(1/(n + 1)] where n is degree in half angle in theta.

example: 550cd @ 18 degrees is 2PI * 550 * [(1/(55.95 + 1)] = 60.649 lumens.

is this around ball park?

Running it through my spreadsheet, here's what I got:
<font class="small">Code:</font><hr /><pre>
CD Ang Cd*A Lm Lam Lm pt Lm Avg
==============================================
550000 18 43.3449 42.55 42.63 42.59
</pre><hr />
Source spreadsheet is linked here.

/ubbthreads/images/graemlins/happy14.gif Goob job !

Benadski, cool program. /ubbthreads/images/graemlins/happy14.gif /ubbthreads/images/graemlins/buttrock.gif
Minor correction lm/W is "efficacy" not "efficiency" /ubbthreads/images/graemlins/grin.gif

Thanks for the positive reactions and for adding it to the LED-Museum links! I used Visual Basic 6.0 for the program. I stole the formulas from a German forum, this is what I found:

Lumen = Candela * Steradian
Steradian = 2 * PI (1 - cos ( PI / 360 * half angle))

These formula's calculate the whole thing. The efficacy (never heard of that word before, what's the difference?) is derived from that with the help of Ohm's law.

The formula's should be correct in theory, but since intensity varies over angle, I'm not sure it's really accurate, for some LEDs it will be, for some others it won't.

[ QUOTE ]
IsaacHayes said:
Benadski: sorry to go off topic, but your signature, is it talking about the "UFO" luxeon clones in UV? If so could you make a post about them! I've bene wondering if they were any good for a long time and not many people noticed they were avalible in UV.

[/ QUOTE ]

I've just got these hooked up, there bright! Only it's hard to compare them with normal UV LEDs because the wavelength is near the edge of what my eyeballs see. I think it puts out about the same as 8 to 12 normal UV LEDs.

I've noticed they are a bit hard to get. Roithner is the only supplier I found that has UV as well as IR in 1W UFO. I have the 405nm 30 deg. 11mm with M3 thread by the way.

You might have noticed Roithner has some more interesting UFO's, there is even a RGB UFO available! Too bad the light output of this sucker is VERY BAD! I hope Lumiled will make a slightly diffused RGB someday, that would kick ***!!!

Yes, I've seen the try-RGB UFO's. The only thing that intersts me about the UFO's is the UV because their is no UV luxeon. Other than that luxeons are better.

Is the 30 focused pretty good? Does it keep a tighter spot than several 5mm leds would?

Yes, the focus is actually very fine, you will see an evenly lit square spot with almost no artifacts and sharp edges. If you look well you can see a very thin hairline (from the bondwire I guess).

I have let it run all night (about 8 hrs) on a small heatsink. The heatsink temperature was about 50 degrees Celsius I think. The output has not dropped as far as I can see. /ubbthreads/images/graemlins/smile.gif

<font color="red">Should I start a new thread about these LEDs? </font>

[ QUOTE ]
Benadski said:
Lumen = Candela * Steradian

[/ QUOTE ]

This assumes that you put in the average intensity, usually you get the maximum.
And also usually the angle is defined as twice the half angle of the half intensity.
I once had a simple formula in the form: Angle in degress sqared by intensity by correction factor gives flux (was given by D.Gibson).
On the other hand, some people argue that the 'missing' flux from the center is about the same as the light has sidespill outside the 'beam'

You can never tell exactly without actually measuring it!

The intensity decrease/increase over angle is on every LED different. Some LEDs have many artifacts, some have almost none. For some LEDs the sidespill is more than the missing flux, for some it's less, so it will always be a guess!

I will update the program soon with more usefull information and calculation tools, also the spelling error will be corrected.

I think a having a box where the user can manually enter a correction factor would be a good idea. Just based on my empirical analysis so far I've found that 20° LEDs generally require a correction factor of about 1.25 to account for the light that appears outside the main beam. On the other hand, for a 60° LED your equation may overestimate the flux, and the correction factor appears to be about 0.8. No need for you to worry about what the correction factor should be. Simply have a box to enter it.

Ok, now that's a smart idea. This will surely be part of the next version! /ubbthreads/images/graemlins/smile.gif