Spreadsheet for easy calculation of light output

[EDIT]: Heavily revised version april 2012.

Here is a simple spreadsheet "LED-CALC " for calculating output from torches with Cree LED's XM-L, XP-G, XR/XP-E. Now also includes XB-D and XT-E. Formulas are fitted to CREE's data sheets.
Good for trying out "what if".

It is preloaded with 5 cases for each type of LED. You can change these cases freely or add more cases by copying columns.
The spreadsheet is provided "as is" with no responsibilities for my part.
I have found it pretty good for predicting output.
But most important: Have fun with it.

This is a bit of the spreadsheet for explanation:

Only 5 numbers has to be set (the upper part in green color):

LED Current [mA]:
The max current through the LED.

PWM [%]: The PWM level 0-100 %. 100 if constant current.

Heat Resistance ['C/W]: The heat transfer resistance for the flashlight body. An empirical number (normally 10-15). Guidance is provided.

OTLfactor: This is also an empirical number, how much of the light find its way out the front lens (normally 80-90%).

bin (lm@700mA or 350mA):
Put here the middle output for the LED bin e.g. 290 for a T6.

Now the results:
LED cold (lm):
is the output from the LED calculated with a formula fitted to Cree's output vs. current curve.
LED warm (lm):
is the derated output after warm-up.
LED temp ('C):
is calculated so that a cooling value of 3 will give about 150 'C at the highest allowable current.
is the output derate factor calculated with a formula fitted to Cree's output vs. temperature curve.
OTF cold (lm):
Out The Front lumens when cold.
OTF warm (lm):
Out The Front lumens when warm.
apparent light:
A number (for output) which is better to compare with because of the eyes logarithmic behavior. A difference of 1.0 is a doubling of output but is not perceived that much.
Torch efficacy (lm/W):
how many lumens do I get from 1W of battery power.
runtime (hr):
with a 2400mAh 18650 battery.

Thanks for doing this and making it available.

That must have been a lot of work.

Thanks for sharing your work!

Thanks for the appreciation guys.
This have evolved over a few months when I found that looking up in Cree's graphs for each case was a little too troublesome. The work involved? well I did'nt notice, it was fun all the way (I like optimizing second degree equations by hand ).
Now I hope that my fellow members will not just read about it but try it out. It can tell a lot of how overdriving don't pay off.

sixty545, this spreadsheet could prove to be very helpful. Would you please explain how to determine the first two numbers: max mA and average mA?

Do you mean the maximum mA that that particular LED is capable of, or something else? And you say that average mA is lower with PWM. How much lower? Would we have to look at a scope trace and calculate it based on duty cycle? Or measure it with a multimeter? For most lights' High mode, there would be no PWM, so the first two numbers would be the same, correct?


Very useful - thanks for sharing!

Could anyone please explain how to determine the first two numbers: max mA and average mA?

sixty545 just explained it to me in a PM:

"avg mA" is the average current you chose to drive the LED with. Without PWM drive "max mA" is the same as average, but with PWM the "max mA" will be higher. For instance if you drive a LED with 1A PWM and have 3 modes, 100%, 50% and 10% then "max mA" is 1A in all 3 cases but the "avg mA" is 1A, 1/2A and 1/10 A respectively.

It is necessary to insert both values as a higher "max mA" / "avg mA" lowers the efficiency (called efficacy when concerning visual light).

My apologies for resurrecting an old thread. However this topic intrigues me since I’ve often wondered how many lumens are actually getting out past the glass as apposed to what the the LED is rated at. In doing a search, I found this thread.

I have 3 questions.

1. Does this spreadsheet work with XML2 U3 (change bin to 320?)?
2. Has anyone tested the estimates produced by this calculator with an integrating sphere to see how close it is?
3. The approx run time is using simple division of the amps and rating of the battery. Is it really that simply?


Are there threads out there that I missed that address my questions?

Sorry for another resurrection of the thread, but does anyone have this spreadsheet or a similar tool? The OP’s link has expired.