Best to ask Juan how he gets his numbers. Looks pretty close to me but I don't know if that's t lm or b lm.
Welch Allyn has a re-rater on its site that some use to get an idea of lumens. I pointed out a while back that because they don't use current (A) in their calculations they aren't calculating with Watts which makes sense because they are looking for lumens except for the fact that the full power should be accounted for. The full power is radiometric and the visible power is photometric. Radiometric would include IR. A Mag85 is putting out a good quantity of that. When you overdrive a lamp the calculation attributes all the voltage overdrive to lumens. Not a big deal if you are running say 30W or less but some of the crazy...erm...I mean adventurous Incan guys do 100W + or 250 or 500 or heck I don't know what the top is anymore. So using that WA system will give increasingly incorrect results in relationship to the amount of power being used. You'll see many references to lumens per Watt on CPF. lm/W is the ratio of luminous flux to radiant flux.
Worse is AWR's hotrater which uses goal seek which is muddy math. The concept of putting your best known calculations in a spreadsheet is a good one. I'm pretty sure that's what WA is doing as is evidenced when looking at the link below where V2 equals a quantity. V2 appears to be the name of a cell, not a variable.
Anyway, if you'd like to check out the WA re-rater just pop this link:
http://www.walamp.com/lpd/webstore/...&L2=3.31952948576512&adj=5&partnumber=01185-U
Within the address line of your browser you can change the V or the part number to see what they come up with for b lm. It's interesting but IMHO flawed.
Here is the calculation I think they are using plus the 65% a lot of us use to get t lm instead of b lm:
spec lm X (overdrive V ÷ spec V) ^3.4 = overdrive lm X .65 = torch lm
I haven't checked my facts scientifically on any of the above statements so treat this post as just one opinion voiced by one person...not fact.
The best way to figure what's going on with your light would be to measure it with an Integrating Sphere. I quit trying to figure exact lumens long ago and just try to approximate now. I know what several Incans are supposed to look like so for comparisons I use DM51's method of using an Integrating Tree.
?
) So in my case @ 10.8v its about 1200lumen. No wonder it its more impressive then those cheap 6-12v 55w halogen lights.:twothumbs
