3x XM-L2 амеба

[scar]

Just finished a couple of 3x XM-L2 w/LFlex driver builds that are heading to Russia. These set-ups powered by 3S2P 11.1V packs. Just snapped a few pics this morning but don't think I will have time to do beam shots as they need to get in the mail today. Sorry about the crappy cell phone pics as my regular camera broke.














****

[tbutter]

Nice! What's the calculated runtime on high with a 6 cell pack.

[scar]

Here is what I am calculating for the 3S2P 5200mah pack-

Light head set-up in Three mode -

low (500ma's) 11 hrs
medium (1000ma's) 5.5 hrs
high (200ma's) 2.75 hrs

Please, someone let me know if I am way off on these runtimes.



***

[adrenalnjunky]

by my math, assuming 90% efficiency @ the Lflex - figuring 9.45vf (3.15v per emitter in series) on high I get the same number you do.

@ 1A - I plugged in 2.92vf per and get a little closer to 6hr runtime. Under-promise and over deliver, right?

[mattthemuppet]

nice!

Calculating runtimes for linear drivers is tricky, as the voltage they burn off as heat reduces over the discharge of the battery. I'd say a 3S linear set up is on the bleeding edge though.

Start of discharge:
LED power = 3V x 3 x 2A = ~18W
Driver heat loss = Vin - Vf x A = ((~3.8V x 3) - (3V x 3)) x 2 = 2.4V x 2 = 4.8W
Total 22.8W

Mid discharge:
same LED power, driver heat loss is ~3W as Vin has dropped to ~10.5V
Total 21W

End discharge:
same LED power, driver heat loss is effectively zero as Vin = Vf
Total 18W

To work out a total power consumption, you'd probably have to do some kind of mathmatical wizardy like calculate the area under the discharge curve or generate its slope, all of which is waaaaayyy beyond my limited/ nonexistant mathmatical ability

So if you take the mid discharge as a crude average, then your runtime should be:

Wh of battery pack / W = 11.1*5.2 / 21 = 2.75h or 2h 45m

so, pretty much bang on what you calculated

[MtbMacgyver]

The way to calculate runtime for a linear regular is to simply divide the battery capacity in milliamps hours by the led current in milliamps. Linear regulators are a special case, because by definition, input current always equals output current. So that gives

low (500ma's) = 10.4 hrs
medium (1000ma's) = 5.2 hrs
high (2000ma's) = 2.6 hrs

% efficiency doesn't really work because as mattthemuppet points out the efficiency at the beginning of discharge is dramatically different than at end of discharge. The reason using the mid discharge method give too high of an estimate is because the discharge voltage curve of a li-ion battery isn't linear.

[HEY HEY ITS HENDO]

Its a guessing game with a million variables
stop calculating and speculating
.....get a Watt meter!! Watt Meter & Power Analyzer /Watts up (Ver. 2) | eBay

[adrenalnjunky]

better yet - plug it in, set it on high, start a stopwatch and ride til it quits. Swap batteries and ride home.

[Leopold Porkstacker]

It’s good to see I’m not the only one still using Marwi mounts for my lights.

[MtbMacgyver]

.....get a Watt meter!! Watt Meter & Power Analyzer /Watts up (Ver. 2) | eBay

I have multiple of them and highly recommend them as well.

But it's not really going to help you much in this case. A switching regulator such as a buck or boost converter is a constant power device. So you can measure the wattage and it'll pretty much stay the same as the battery voltage drops.

A linear LED driver is a constant current device and the wattage changes during the discharge cycle as the battery voltage changes. You can use it to measure the current, but that's easy enough with any descent multimeter.

[epic-gamer]

how much does it cost?

[scar]

Thanks everyone for the help calculating the run times. I have been using LED Pro runtime calculator looks like it is real close.

adrenalnjunky - "underpromise and over deliver" is really a great description of how I roll. Most times I don't have the time to "test drive" some of my custom builds so I have to guess/calculate run times and cross my fingers I got close.


HEY HEY ITS HENDO - Yep, got me a Watts Up a few years ago and is one of the best tools.


Leopold Porkstacker - With the Marwi mounts, the price is right and they are really quite rugged. People really like that the base rotates a few degrees for better alignment.

epic-gamer - Shoot me an emial to the address in my signature and I will send you a bunch of info and pics on the different versions that I offer. Got to warn you I can not even come close to what you can get a light coming out of China for. I can not even buy just the raw materials for what you can get one for now a days. What you will get is a product made with "personal pride" that is made by a fellow cyclist that uses his own product on almost a daily basis.


****

[mattthemuppet]

The reason using the mid discharge method give too high of an estimate is because the discharge voltage curve of a li-ion battery isn't linear.

I was trying to figure it all out in my head, then my head started hurting so I stopped

glad to know that there's an easier way of doing it though!

[rschultz101]

linear for 3S is no good.
at full charge be 25% + for heating at 2A
at lower current, with lower Vf for XM-L2,
it be 29%+ .
the lower the Vf , the worse it is. all the advantage from newer leds,
is just burned off at linear driver.
it's no fun, still looking for a 2S 2led driver ,....
for helmet use, you'd want the most efficient driver, due weight of battery and
small formfactor on housing. just trying to help. PM if need be. cheers, Rob

[HEY HEY ITS HENDO]

Good post Rob,
why not go 2S battery 2Parallel led?

[Nitatunarabe]

I've received my parcel. Everything is OK: the lights arrived successfully, the packaging is not damaged and so on.

Here are the snapshots:
1. Low:


2. Medium:


3. High:


Some other photos:
1. On the road:


2. Classic fatbike photo:


I like these lights, they are bright and allow me to ride at night without speed or passability loss.

The build quality is great and mounting is very easy. I recommend these lights.

---
I've posted the beamshots at local forum and have a question: how many [theoretical] lumens do they produce?

[adrenalnjunky]

Scar didn't say what bin or color temp he used - but your pics look like a "cool" white tint - Cree rates the XML-2 at 728lm @ 2A - so multiply that by the # of emitters you have. 3*728=2184 theoretical lumen output on high. You'll have some loss in the optics, but I don't have any good way to guess what that loss is.

[scar]

Nitatunarabe - Glad to hear that the Amoeba's made it to Russia finally. The beamshots look great! Did whoever you ordered the other light for like their light also?


adrenalnjunky - Thanks, I really hate trying to put a lumen estimate on a light. I will take 2000 lumens on this one I used XM-L2's with a 3C tint. Kind of hard to tell since there is only white in the picture.



****