With voltage sag and added resistance from Springs/components, and vf when u are dealing with a single emitter at 10Amps there isn’t much more to give from a battery. With so many recent multi die and multi emitter lights it’s easy to forget that a single die @ 10A is still double what an XP-L HI will do.
In these tests the vF goes even higher than what a battery is fully charged. Have to put these power supply death tests into perspective.
Also, you say “NO” effect but that’s not true. Run the tests when the battery is 3/4 or 1/2 charged and you’ll see some differences I realize you only use topped off cells but they aren’t all created equal even though it may look close in the first 0-30s
Although there are Four engineers in my immediate family, I am NOT one of them[!] and do not understand the electrical components and how they are affected by Batteries/Drivers ect.
I have numerous modded[V54] single cell lights and 3 or 4 batt. torches that BENEFIT and actually require high drains for ultimate performance. I have no idea what is different[other than the driver] between those lights and this one! lol!
EDIT: Actually I do remember that I had[sold] a TK75vn Quad w/ 4 X SST-40 DE DOMED,Vinh told me it did NOT require high drains,GA or MJ1 should be used.obviously they are 10A batteries! I am getting smarter!!!
I think this speaks well of 21700’s in general to be honest. When we only dealt with FET drivers and 18650’s or crappy low drain 26650’s it probably made a bigger difference going from GA > 30Q > VTC5A
Do you think using Molicel P42A 45A 4200mAh 21700 will not only be unable to increase output, but may shorten the life of the LED and circuitry?
I was greedy and bought them on sale,never knowing or thinking that this battery is not only unnecessary[no improved output] but could be detrimental to the light itself. I have no other 21700 lights!
21700 has not been good to me! I have had Two other 21700 lights and returned them and gave away the Two 40T’s that were used with them!!! At that time they were $14 batteries,now Jon[Liionwholesale] is selling them for $11
The MJ1 and GA have a chemical mix in them to maximize their capacity at the expense of generating current. This makes their internal resistance a bit higher. Therefore they don’t generate as many amps when put under a heavy load.
BTW, Mooch considers the Molicel P42A to be a 30A continous discharge cell. So just a bit better than a Samsung 40T, competes with the 30T. For those curious.
In koef3’s (corrected) SST-40-CW-N4 test, 2000 lumens occurs at 6.8A and the maximum is ~2375lm at 8.4A. Perhaps this is where Astrolux got its “2400” figure, but it’s obviously an unrealistic number as the bond wires start failing at those currents.
nkresho’s NW light starts at ~1820lm (207,360cd at 20’ [6.1m]), which would be around 5.7A of emitter current on koef’s corrected chart, bearing in mind that the chart represents a CW-N4 emitter. Note that this is a rather substantial difference from tm07’s light, both with NW tints, though I don’t know which battery nkresho is using.
Do we have a general idea of how much is dissipated in the driver and wiring and how much actually reaches the LED? Many here are much more familiar with these circuits and emitters than I, so I thought perhaps there were some rules of thumb for different driver designs. As keof3’s corrected test was done with a C8 reflector, I assume it represents a reasonable estimate; the tint/bin would be bigger factors.
My FT03 status for order #634304XX today changed from “Processing” to “Back Order”. There are many people in line ahead of me and I don’t expect my light to ship for quite some time, but I keep mentioning changes in case others are watching their orders. I ordered via the app on 2019-03-21.
Most of those emitter tests are done with a bare LED. Once you put it in a light with a reflector and lens and driver you can typically lose 20% to 30% of the output.
Thanks. I think koef3’s “corrected” test represents an emitter in a C8 reflector, but without driver and lens. If I understand koef3’s updates and follow-up post, the reflector alone seems to cost 6. If an AR lens costs another 6, that leaves around 8-18% lost in the wiring and driver.
The koef3 numbers I quote are the “reduced” values, which are labelled “CORRECTED” in the top-most “LED TEST” chart. The other charts only show the initial bare emitter figures. I am somewhat new to the details, but I do understand what domes do and such.
I’ve been reading this and related posts as well, but I was most interested in the difference between tm07 and nkresho’s lights. tm07’s light appears to be making 2000lm or more if both cd figures are accurate.
I can tell you already you’re going to get way more loss in a reflector than 6. More like a minimum of 15 to 20% assuming it’s a really good quality reflector. Most are over 20%.
To get a 6% loss a reflector would need to use a silver coating. That’s too expensive for most flashlights. Typically flashlights use an aluminum coating which is a lot cheaper but also a lot less efficient.
Leds lose output as their temperature goes up. The blue line is at 25°C which is about room temperature. That would be like “turn-on” lumens.
Once the light has run maybe 30 seconds that’s where you see his “corrected” lumens at 85°C.
When you measure a light according to ANSI-FL1 specs you measure after the light has been on 30 seconds.
Different people on BLF like to use different times. Some care mostly about turn-on lumens, but most prefer the more realistic number at 30 seconds. I try to include both numbers in my measurements so people get an idea how fast the lumens drop. Of course, with a really powerful light that gets hot super fast, you might have to post the lumens at 25 or 20 seconds.
Note that Cree also lists output for their leds at both 25°C and at 85°C so that’s probably why Koef3 used those exact numbers. He is also probably estimating they are at 85° since you can’t actually measure the junction temperature of the die.
Koef3’s follow up post was about luminance which is different from lumens.
Lumens are to total amount of light.
Luminance is the amount of light per square millimeter of the die.