No with 4Ah as I had calculated in the post above. The tests of were done with 2.6Ah battery running 2hours30 at 467 lumens so that is not very dissimilar indeed.
- Right about stopping distance. It would also vary with the incline and the state of my breaks. A speed of 28km/h recalculates seven meters per seconde so that an effective throw of 100 meter should give me a decent 15 seconds of anticipation time.
- That mount you link to was the one I had decided on. They go for cheap on aliexpress.
PWM, I would like to avoid too. Mostly an issue on direct drivers, no (e.g. FET +1 should have no PWM for the lumens that use the linear driver but those lower lumens that go beyond that driver would have PWM?)
I though I would just have to live with the nuisance of lights from oncoming trafic and can’t imagine how a reflector could catch it (though I could imagine a light on high or turbo being on the same level and hence reducing the hindrance though depleting the battery). Could you explain this a little further?
If you want a LED that is efficient while offering some throw, something domeless with large die will be better than a domed LED.
See this 2-years old compilation to learn more:
BTW, I thought that it would be nice to update this chart with newer data….and I can’t.
Koef3 is not here nowadays.
djozz has mostly ceased testing.
Texas Ace has ceased testing.
maukka? No LED review from him in almost a year (though I think he’s done some quick tests since then).
There are some new people filling the gap but not many of them and doing mostly quick tests rather than full fledged reviews.
While more light and throw is always better, back when me and my buddies were hitting singletrack at night, it was amazing how effectively you could see with diminishing output as the eyes adjusted. For example, a 100 lumen helmet mounted light was quite adequate in the dark. Now, if you are competing against street lights, shadows, cars, etc it will feel different. But in the dark on the singletrack it was surprising. Now, I usually rode with more than that (a triple SSC P4 on the handlebars as well) which was amazing.
Don’t focus solely on throw, either. A pencil beam is going to give you an uncomfortable, tunnel vision experience. I really liked the triple SSC P4 because it offered wonderful spill. Worked amazing with a semi-throwy helmet light. Whatever you do, don’t stick a lone thrower on your bars and expect a good experience. Something like my Convoy S2+ SST-20 with OP reflector gives and excellent helmet light beam. If you did the 7135x3 driver, 35% mode gives and estimated 126lm for 8+hrs.
Here is a chart I built with the help of BLF member tests. The lumens are generally estimated and the runtimes are calculated. And altough I forget what mAh I used for the calc, simple math can tell you that
I would use a head lamp and a single cell light handlebar mounted, that’s what I do on my Xc bike. I do a KR4 on the handlebar mount with the Olight handlebar mount (nice piece for about $12) and a sofirn sp40 headlamp. Both can maintain around 400 lumens (guestimate) for a while (over a couple hours) if you don’t put them on turbo/blast mode for more then a couple seconds here and there.
Thanks for that link! If I understand and ’gestimate” the graph correctly, the non-dedomed SST40 trups all others in the graph up to ca. 350 lm. And from there up to 530 lm it trumps the dedomed SST40 (but not the XHP35).
If that is correct then the non-dedomed SST40 may actually be one of the better choices among existing leds for my purpose, a usage btw. say 100 to 400 lm? (P.S. I would be to worried to try dedoming a led)
That quote was also useful for calculating efficiency of a linear driver. If I apply the formula for user case the efficiency is less great. If I usually ran a SST40 @ 250lm it would itself use ca. 2.75V. (see Luminus SST-40, a N4 BA bin tested)
Using the formula, I get an efficiency of 1/2 * (2,75/4,2 + 2,75/2,9 [if I use the battery slightly below 3v]) = 80%.
I read somewhere that the efficiency of switching drivers runs between 75-85% but also that Simon Mao claimed 95% for his 8A buck driver. Perhaps I should hold off until an equivalent 4A driver for C8+ shhould trickle down. Could that happen?
What I really would like to know now, is how Simon’s C8+ SST40 driver —also with an 8A mode (and 1.5A mode; said to be a regulated FET driver)— compares with that of say a regular M21A or a C8G?
Like someone mentioned above, Not so sure I would go with a thrower on the bars. I liked a flood light when I was riding. I was also a slow rider at night, so my needs were different.
I would lean to any quad light and 26650/21700battery to run it as it is going on the bars and not on your helm to add to possible neck strain.
Again my needs most likely do not match yours. As I was riding in a large combination of night environments from on the side of the well lit roads, to mile long stretches of single tracks through parks and bike paths that were unlit.
I am thinking rather than the FT03 You may like the FT02S as a better choice, More spill and 200-300 yards of useable light. This has been my go to this year as I am out in the dark more than the last few years here up north.
S2+ sst40 and w1, S21A sst40, EC01 sst40 and xhp50.2, FW21 Pro and X9L, M21A sst40. This are the ones I have used,
I find more than EC01 too big for me
What did you experience, is W1 not too much a pencil beam and XPH50.2 not too wide for throwing over a distance of a gravel road? Most of yours were SST40 then. Which of all of those torches had the longest runtime when used with a same battery?
Not quite this when talking about average efficiencies over the discharge course of the battery. Please see my comment right above yours: ca. 80% efficiency on average at 250lm, with indeed a Vf 2.75.
As for flooder. Again, as mentioned a few times before, I have a flooder on the way (headlamp). I could always swich postion the thrower on head and the flooder on bar of course.
At X luments, whatever that X is, domed LED will always be more efficient than a dedomed one. But at the same time it will be less throwy.
At the same throw, dedomed LED will be typically more efficient but produce more lumens (except for low intensities where a domed LED always prevails).
In the case of SST-40 the cross-over point when the dedomed LED starts producing more throw is at about 50 cd/mm².
Domed LED needs 3.4A and 1370 lm to reach it, dedomed 1.6A and 612 lm.
Thanks for clarifying! That makes sense. Since I would use the lamp at lumens much lower than 600 lm, the SST40, as is, would be a good choice.
A pity that the reviews are less frequent now. Do you have data of other relevant leds that might do better in throw/lm/W? Otherwise I’ll stick with the regular SST40 which has just about enough spill as well.
Thanks everybody for your hulp. Before I bite the bullet I have only two remaining questions:
Has anybody info on the relative efficiency of the driver of C8+ ramping (regulated FET) versus C8+ 12 group/(statements, runtimes or tailcap measurements) [or FT03 Fet driver]?
Does anybody know if TIR lenses exist for these optics?
That “ramping” driver from Convoy should rather be called a driver with two adjustable modes + 1 special mode, it’s very different from the ramping firmware from the M21C-U.
I find the timing to get into the mode adjustment a bit hard, but maybe there is something wrong in the switch from the host that I used, I can’t tell.
If you know which two output levels you need then the ramping driver is suitable, but I wouldn’t use it if you need to adjust it often in the field.
Some members like that driver, so maybe they had a better experience with it.
Thanks. I am specifically looking for info on the respective efficiencies: C8+ with regulated fet (the “ramping” firmware) versus C8+ with 7135s (the “12 groups” firmware).
(I am aware of the fitmware’s modes. I hope that the ramping is not too quick to find out —with some patience— and then select a specific efficient current level in the 1,5A mode 2 -if that mode is more efficient with 100-400 lm (and I’d just set mode 1 on very low) and once I know leave these levels in memory most of the time.)
