Servers by jDrones

Using 18650 cells in multi rotors

I have an F-450 quad with Emax MT2216-810kV motors and 1045 props. Currently, Im using a 3S, 5000mAh, 50C LiPo battery in it.

I’m considering to use LG HG2 18650 Li-Ion cells (20A CDR/3000mAh). The max current draw of my copter in Mission Planner logs is 25A during takeoff (nominal is 15-20A) and I have also measured it using a power analyzer (turnigy 150A, props inverted) and obtained similar values.

Since LG HG 2 is rated at 20A CDR, 30A burst, i think using it in 3S2P config can help my copter fly and i still have a safe margin of 10-15A.

Is there anyone who has experience of working with Li-Ion cells and can help me if my design is wrong somewhere?


I personally don’t have any previous experience with 18650 cells apart from two in my Jumper T16.

But I think your math checks out, but from other builds I’ve seen that the magnetic disturbance is far worse with 18650 cells, so keep that in mind.

I’m using Samsung 18650 cells for five years now. No issues. They last longer and have a better capacity/weight ratio then lipos.
The downside is: Higher Voltage drop. So using the voltage compensation feature of arducopter is recommanded.

@fs007 You have been using them in fixed-wing or multi-copters?
In case of fixed-wing, there is only one motor and the current draw is much less as compared to a multi-rotor. Also is this voltage compensation available in plane firmware only or copter as well?

Also among protected and unprotected 18650 batteries, which one do you recommend?

I use them in a quadcopter.
Voltage compensation is available in Copter firmware. See parameters MOT_BAT_VOLT_MAX and MOT_BAT_VOLT_MIN.
I use unprotected cells, but be aware: NEVER discharge them below their limit.

It is dangerous to use protected cells, and it increases internal resistance as well. Two cells in parallel are too few for you, you won’t be able to use of the energy in those cells before heat become a problem.
4s3p is what I use on 3DR Solo

@Andre-K What is the current draw of your 3DR Solo? And which particular cells are you using? Thanks

current is about 20A (will vary with voltage) 19A at start, 24A when voltage is close to 2.6v/cell.

Search for " 21minute flight example:" - then see the graph below.
If you had higher current load (less than three cells in parallel) the lowest acceptable cell voltage will be reached faster, with more energy left in the cells.
You need to remember that if you test a cell at and measure it’s capacity to be 3,5Ah @ 0.5A load, then you will not get 3.5Ah at 3A - at 3A , you may only get 1.5Ah out of it, due to internal heating and that the voltage drop too much under such load.
So yes, your 20A “ok” cells can handle 20A
But if you load them with 10A , deplete them to 2.5v , then they are “empty” in the sense of - “you can’t fly anymore”, but if you then load them with 0.5A , you can get out of them for example 40% more energy than what you got at 20A - before again reaching 2.5v/cell.

@Andre-K @Vabe @fs007 I’m attaching screenshots of mission planner log showing current draw herewith. Since Im using LG HG2 (20A CDR), connecting in 2P will give me 40A CDR whereas my draw is around 15A. In this way , Im having a safe margin of around 20-25A in my design. Is that not enough.

Why do you think that unprotected cells have a higher internal resistance ?

You need to understand what I tried (and failed) to explain.
It is not about about “enough” or “safe margin”
It is about the fact that if you load a Li-IOn cell with high discharge rate, you will get out less energy before hitting minimum voltage, than if you reduce discharge rate (add more cells/higher capacity)

in other words:
with 2P , each cell will work at about 7A
What I say: is that 7A load per cell will maybe make them deliver 50% of the capacity, before reaching 2.5v) while the same cell, loded with only 2A may deliver 80% of it’s energy before reaching 2.5v
hence if you add three in paralell, and reduce per-cell current, you will be able to spend more of the energy you carry anyway.

@fs007 Because
-I did test a lot of cells
-I did disassemble 18650’s - and the interconnect between protection circuit is often rather puny gauge.
-and the protection consists of two NFETs back to back with no active gate driver there(means not a perfect gate drive, a little higher IR), and there’s no such thing as perfect FET with zero internal resistance. those are usually of very small, not very great type, some I’ve seen are around 0.03ohm in on state. - then multiply for cells in series, and again… the interconnects are usually rather poor.

Finally - if a sudden current spike pulls one cell below preset not very precise voltage (often 2.5-2.7v) then that cell will switch off, the remaining in parallel with it will get all the current, and next will switch off, then next… in an instant you lose power and crash.

@Andre-K from your point what I interpret is that I will have less flight time because due to high current drain, the voltage will fall quickly leading to the cell being depleted in less time. Right?

@Shuja_Hameed - Perfectly correct :slight_smile:

So you should explore the possibility of using 3P or 4P if within MTOW.
BTW: this is not really about the cell count, but rather the capacity in parallel:
having two 2Ah cells in parallel is equal to having one 4Ah cell, or four 1Ah cells.
(except for some thermal properties that will vary with physical size)

@Andre-K the copter is already very close to the MTOW constraint. I have some payload attached to it. If i remove the payload, it will take even less current. (When i view flight logs of the flights without the payload, it shows me a lesser current draw). Now at this stage I’m less concerned about flight time and more about the copter should take off. If it can takeoff with 2P and give me flight that’s all i need at this stage.

I will use 3P but with 4S just to increase the thrust as well so that the MTOW constraint is also elevated.

@Andre-K i agree with all you wrote, but these are arguments FOR UNprotected cells.
A few posts above, you wrote:

yes, I wrote that protected cells are dangerous to fly with.
then I made arguments for using UNprotected cells

Do you see any discrepancy in that?

@Andre-K How about using a 4S2P? I would get more thurst for the same amount of current drawn with an addition of 100g in the AUW.

Yes, it your ESC can handle the voltage and be within the needed RPM, a higher voltage is better because current and heat loss will be lower.
You need to check the motor/ESC combination, higher voltage will mean higher RPM, which may require a lower pitch propeller. Or, you can get a lower KV motor and keep the propeller. can help you with the calculations.

Servers by jDrones