Calculating runtime

MAD23 said:

Hello everyone!!

Im building a little flashlight with 7 xml's in series driven by a hbflex from taskled, i want to drive them at 3amps with a litium battery pack.

I was thinking in about 120 minutes runtime, but i dont really know how to calculate the mAh i need to do so.

Is here anyone that can help me?

Cheers!

Click to expand...

arek98 said:

Welcome to CPF.

Rough estimate (I like to stay on conservative side):

7 LEDS, 3.6V Vf each at 3A is 7*3.6*3=75.6W
Assuming 90% driver efficiency you need to deliver about 84W from battery.
mAh will depend on battery voltage.
hbFlex is boost, with 3A output you need to stay close to LED voltage to keep input current within limit, I would say 4 or 5 Li-Ion in series (each serie part consist of few in parallel first to get capacity you need).

Let say 5 Li-Ion pack has nominal voltage 18.5V. Voltage will sag under load and battery rating is usually fro 0.1C load, it will be less with higer load.
We are talking probably 4-5A out of battery, I would multiply battery rated capacity by 0.7 (ballpark) to get estimated capacity with high current (it may depend a lot on battery type, IMR or hi current Li-Po will be much closer to rated capacity).

Anyway, ignoring voltage sag for a while and using nominal, you will be pulling 84W/18.5V= 4.54A from battery
Using above ballpark constant of 0.7 for capacity adjustment for 2h runtime you would need about 13Ah, 5 Li-Ion pack.

If you use 18650 you will need 4p5s pack (5 blocks, 4 cells in parallel each then all 5 block connected in serie. Protection board for 5 Li-Ion pack, preferably with balancing function otherwise you would need to balance them manually at least every few cycles.

Li-Po may be better option. You could fidnf 10Ah high current cells (my 0.7 constant will become bigger because they behave better at high load).
Then you would only need 5 of them connected in series. Same protection board. Also don't forget about temperature fuze and optionally overcurrent fuze (in case protection from board does not kicks in).

Thick cables everywher, thickest you can fit/solder.

Click to expand...

jspeybro said:

HBflex has more than 94% efficiency with 5 or more XML's, so that's a good choice.
So,
7x Xml with forward voltage of 3.35V at 3A:
3A x 3.35V x 7 LEDs x 0.94 = 66,129 W

you want to sustain this for 120 minuts, so 2 hours, so your pack needs to have at least:
66,129W x 2h = 132,258Wh

So your pack needs about 130-135Wh of power.
if you use e.g. a 18650 based pack, you could do calculations with 3.7V and 2Ah= 7.4Wh
So, if you want to have about 135Wh, you'll need about 18 cells to get the power you need.

If you put all leds in series, you'll have a total forward voltage of 23,45V. Since HBFlex is a boost driver, your voltage with charged batteries should stay below this, so max 5 li-ions in series.
You could make a battery pack with 5S4P. This gives you 20 cells so more than the 18 you need.
4S5P could also work. depending on the canister, this may work or not.
Using 5S4P pack will give you an input current of about 4.5A (power divided by battery voltage, when the pack is almost empty) which should be fine for a 4P pack.
Using a 4S5P will give you 5.7A max draw from your pack. Higher current needs thicker wires as Arek already told you.

All this is actually well explained in the technical info of the HBFLEX driver ;-)

Switching high currents need considerations regarding the switch. The hall switches from taskled however can switch up to 15A, so no worries there if you want to use this.

For that many cells, I wouldn't recommand soldering them but just get a pack made.

Johan

Click to expand...