Meausring incan brightness by wattage alone?

I have recently treaded into the incan area and have a question about brightness, efficiency and wattage.

With LEDs it is pretty obvious that you have bins which define the range of luminosity of the LED with a certain Vf @350mA. This gives you a pretty straightforward Lum/Watt number. (I know there are more variables involved but let's leave the LED talk there for now).

My real question is that since incans are not diodes and don't drop a specific amount of forward current controlled voltage how can one rate efficiency of incan bulbs when there seems to be no standard?

A quick search has only turned up a few threads with little information other than efficiency goes up with wattage. Now I have also read in threads that efficiency goes up with voltage as well.

Is there any way - other than searching for beamshots which are only valid if taken by the same user of the same scene on the same night with the same camera at the same settings under the exact same conditions - to compare which bulb would be brighter.

My example is the MN15, 16, 20 and 21. In specs you would think brightness would follow chronoligical order. MN16 = 225Lum and MN20 = 250Lum. But, if you look at the pics here https://www.candlepowerforums.com/threads/230857 it is clear that even on primaries the MN16 is brighter than the MN20 and by a good amount too. At a rated 2.6A the MN16 outguns the 2.45A rated MN20 by a lot. Assuming approx the same Vin for the two lamps this is only about 1 Watt difference but a very visual difference when measured. Now if I take a lamp that operates at a higher voltage but a lower current and around the same voltage where will it's brightness lay.

As you can tell I am trying to quantify a very subjective difference between lamps.

What I am really trying to do is figure out without buying every light out there what the best light will be (w/o going to bi-pin just yet). MN16, MN20, MN21, HO-M3T, EO-M3T, IMR-M3T, HO-M6R, IMR-M6.

In some simple testing so far with IMR 18500 cells I find that the MN16 and the IMR-M3T are actually VERY hard to tell apart, yet that's 2.6A vs 3.4A respectively - a pretty big difference.

Efficiency of incan lamps is measured the same way as for LED's: lumens/watt. So you get an integrating sphere and measure the lumens output and then divide by the wattage for that output.

Now, the thing is that FOR A SINGLE FILAMENT (i.e. all other things being equal), the measured efficiency of the lamp will go up as you drive it harder and harder, which is the same as saying that you apply greater and greater voltage, and let it draw whatever current it wants (which will be a monotonically increasing (but not linear) function of applied voltage. As you drive it harder, you will also see the CCT of the light go higher. It will appear "whiter" as well as brighter.

Practically speaking, for a non IRC lamp, it's hard to get much over 37 lumens/watt, although right at the melting point of tungsten you reach about 40 lumens/watt, IIRC. Typically, something like a SF lamp with 30 hours of life runs at around 30 or so lumens/watt. But that's just a rough estimate.

Finally, keep in mind that lumens output is just the total integrated light output of the lamp. If you take that output and focus it down to a very tight beam, you'll get a LOT of intensity at the hotspot--it will seem really "bright"--whereas if you spread it around spherically, more or less, like a household lamp, it will seem "dim", despite the fact that the efficiency is unchanged and the total light output and power draws are identicle. So keep in mind that lumens is a very different thing than candlepower or lux. But, efficiency is always measured in LUMENS per watt.

There are re-rating formulas that will tell you how a lamp of given specs will change with a different applied voltage than design voltage. There are also plenty of threads (or at least one) that lays all this stuff out there. Take a look in the sticky at the top of the page. IIRC, you will find links there.

However, to get you started, here is one:

Re-rated-lumens = (Va/Vd)^3.5 * design-lumens.

So if you have a lamp that puts out 100 lumens at 5 volts and you drive it a 6 volts you will get

re-rated-lumens = (6/5)^3.5 * 100
= 1.8929 * 100
=189.3 lumens at 6 volts.

https://www.candlepowerforums.com/threads/179748
https://www.candlepowerforums.com/posts/2997106#post2997106

Lux Luthors and Ictorana's incredibly well done and useful bulb data charts and graphs.

Everything you ever wanted to know about what a bulb is designed to do and what it can do when stretched. The graphs are based on the chart data and show you very visually that some bulb designs are much more flexible than others, and that some increase dramaticly in brightness when overdriven and still retain usable stability, while others have a very narrow range.

Thanks for those links - found a few others as well which had some good incan information.

Or one could simply measure it. My measurements of stock flashlight incan lamps with real out the front lumens and power consumed puts them at about 15 to 18 lumens per watt.

My new SF P61 lamp in a Solarforce Host had 349 out the front lumens and drew 20.8 watts of power. 8V and 2.6 amps. The lamp of course is brighter, typically about 30% more than what you get out the front but nobody of course runs bare exposed bulbs. I do test them from time to time to determine the difference.

349X1.42857 (the inverse of 0.7 or 70%) = 498 bulb lumens. Taking those bulb lumens divided by the same 20.8 watts of power consumed and that bumps the real lumens per watt of the bare bulb to 24 L/w.

So if you are running a 10 watts of real power consumption flashlight in a typical reflector behind some type of glass or plastic lens and you are not overdriving it to get it near the verge of fusing the filament I would use 17Lum/watt as a good rule of thumb while its running near its rated voltage with fresh batteries.