Author Options:

lithium 18650 series first or parallel? Answered

Hey all
Looking at doing a lithium battery build but trying to cut down on soldering or spot welding.
My idea is to use a stack of 3 18650 batteries in series pushed inside a piece of PVC pipe, thus making a 12V tube and then having multiples joined together to make the final battery

Is there any reason why people mostly use a set of batteries is paralel then series?



8 months ago

I suspect what you are asking is why are cells arranged in a way such that, in a given multi cell pack, sub packs are wired in parallel initially, and then these sub packs or modules wired in series.



I have wondered the same thing, and I believe the reason is that when cells are connected in parallel, they act as a single, larger cell. There is not much concern that a particular cell in the parallel arrangement is going to get overcharged or over-discharged since they all have exactly the same voltage across them. Because parallel forces the voltages to match, all the cells must be of the same chemistry and operate over the same voltage range. It's not necessary for each cell in the parallel pack to be matched in capacity or current capability.


However, for series cell's, capacity and ESR are concerns. Generally every cell you have will have slight variations in the capacity and ESR, and other electrical characteristics that may, over time, cause the pack to become "unbalanced" where each cells has a different State of Charge (SoC). If significant, this can lead to underperforming cells (ones with lower capacities) to be over discharged when a load is connected to the pack, while the best ones may still have considerable life left in them. This will further damage the lowest capacity cells. When charged, the cells with the smallest SoC may not get to receive a full charge, or the cells that have not fully discharged may become overcharged, damaging them as well. If this gets extreme enough, then the result could be catastrophic! Generally catastrophic failures happen after a cell has been discharged so far, that it has been charged with the opposite polarity! (about -1.8 volts across a cell)


Generally the solution is to "balance" the pack with a charger that can monitor and adjust the charge current to each cell individually, or with larger banks that are more critical, have a dedicated Battery Management IC to ensure the cells always being balanced when charging and discharging, and potentially enable a safety cutoff if any cell or module voltage falls below some limit (generally 2.7V to 3V). With the parallel-then-series arrangement, this means that you will only need as many tapps as there are modules of cells in series. Otherwise you need a cell balancing solution for all the series strings that are put in parallel, which isn't cost effective.

Something that I have only implied, which is important, is that when creating modules of parallel cells to later connect in series, it's critical the total capacity all the series packs are the same in order to maximize the total energy that can be delivered from it.

For example, suppose most packs are 10AH, but one is half (!!!) of that, or 5WH. Since a series combination of cells means that the discharge current is equal for all the cells, this very weak pack will be completely drained before the other packs have delivered half their capacity! If there is no protection IC to stop the discharge, this weak pack will be damaged as it not only becomes completely discharged, but also charged in the reverse direction! While the other cells are delivering power, this weak module is robbing the total power and dissipating heat (acting like a resistor, more-or-less) and when the pack gets charged again, this pack will almost certainly fail spectacularly!

Parallel operation provides more load current. Series provides greater voltage.

Combining the two methods provides greater source voltage and greater load current.

I guess the biggest issue is finding electrical and physically identical batteries.
And for me there is always the problem of price.
If you consider the costs for (single) batteries and a suitable charger it is often much cheaper now to get a lithium battery plus charger from the discounter.
For well under 50 bucks you can get a 20V/4AH battery including charger.
At least that way you don't have to worry about the many variables.

Guessing your from australia too judging by your name. Where are you getting this stuff from for 50bucks?
Also i should say im hoping to go much bigger than 4ah, something like 12v40AH in the future.

But after doing homework last night on it i think a simple spot welder build would be the best way first.
As for the cells im planning to buy a heap from ebay of all the *fake/Dodgy* ones and test them as MAH vs cost. I know that the 9000 they claim is fake but if the value comes out about the same is the real key to this.

Spot welder on battery power? Good luck...
As fr the batteries from Ebay: Buy at least 3 times as much as you need as they will fail sooner than an icecube in your fireplace....

There is a desire for balance in a battery pack like this, and what I mean by, "balance", is that every cell in the pack has the same voltage.

When you wire a group of cells in parallel, these cells are guaranteed to have the same voltage.

Actually, any group of circuit elements wired in parallel, share the same voltage.

And any group of circuit elements wired in series, share the same current.

Those two statements are just like, corollaries to Kirchhoff's circuit laws,


That is circuit theory, and you might well wonder how this actually works in practice.

I mean supposing you have two Li-ion cells, with slightly different voltages, say 3.5 V, and 3.3V, and then you marry them together, by wiring them in parallel with some short jumper wires or tabs. What actually happens?

Well the circuit theory suggests there is going to be an enormous current, at least initially, equal to 0.2 volts divided by very small, almost zero, resistance of the short jumper wires.

So I imagine there is big initial current surge while one cell pours current into the other, and this current slows, and stops, when both cells have reached the same voltage; i.e the stronger cell gets discharged a little , from 3.5 to 3.4, and the weaker cell gets charged a little, from 3.3 to 3.4.

That's kind of diversion, I guess. But I think it useful to think about these things, in particular the possibility that large currents could flow in these jumpers connecting cells in parallel.

Like in the rare case that one of your cells gets hit by a bullet, and shorts internally. Big currents would certainly flow in that case, and the short in one cell is going to take down that whole group of parallel cells.

Mainly what I was trying to tell you here, is that I believe the reason why battery packs that have parallel cells, have them wired as parallel then series, is to force the cells in each parallel group to have the same cell voltage.

It would also be nice if the series groups had the same cell voltage too, and I think there are battery management (BM) circuits that can make that happen, by watching the voltage on each series group, while the battery is charging, and providing some small adjustment currents, to make sure each series group charges to the same voltage.


8 months ago

If each stack of three has a low resistance in series then paralleling may work and prevent the high voltage gang of three from damaging by over_current flow the low voltage gang of three.. The resistors reduce the current that will flow..

Even if you select each gang to match the other gang they will change (UN-EQUALY) with use and time !

Dont do it, you need to have a battery managment system, and it wont work if you wire the battery in series like that. Thats why they are wired in Parallel blocks, then in series.