Parallel Batteries


Pleng

Well-Known Member
Joined
Dec 28, 2006
Messages
3,030
Ignoring for now the obvious space constraints, is it theoretically possible to wire up two Pandora batteries in Parallel? How would the charging circuit deal with this? What about the power management software?
 
yes you can!! :D

li-po needs a circuit that controls the charge. evey lipo (mobile phone, console, camera) has two componentes, the lipo cell, and the circuit.

the circuit stops charging when the voltage of the cell reaches 4,2 volts for lipo (4,1 for li-ion). the scheme is like this:

- |--- cell -
s |
+ |--- cell +

(being s as signal)

internally, the circuit ends in a + and - cables going into cells respective + / -

you can attach as many cells in parallel (NOT SERIAL!!), as you need. there are only two rules:

- the batteries should be the same type, size and age
- the charging will take more time, 2x, 3x... depending on how many cells you attach, because these circuits have a fixed current to charge
 
Hmm you know I didn't even think of 3 batteries. They're pretty slim so that could be very interesting. How would the software side of the thing work. Would it automatically 'know' that there's an additional battery connected? I mean, would it just come across as one great big battery?

Are you a sort of expert in this field? And are you interested in helping out with a little project?
 
software side is transparent, cause pandora hardware will only see one one battery circuit. there is no way to know if you connected 1 or n cells in parallel

the project is quite simple, you only need to buy the cells (same age, size, capacity...) and solder. maybe you need an external connector and an external case to keep the cells:

remove the pandora cell (unless you have a 100% simmilar cell)
keep the circuit and fin that pandora cell was connected to the circuit by only two lines ( + and - ). use those +/- to connect to the rest of the cells in parallel.

it could be something like:

lipoe.jpg


PD: there are power inputs in the board, but you cannot charge the cells with that. you could use that internal connector for example an external pack of AA batteries
 
I'm not an expert, but I imagine it could mess up the battery meter causing it to show the wrong time/percentage. It would be annoying if you had no idea when it needs to be charged. Nice idea though, OPT should sell extended batteries in the future :)
 
chame said:
software side is transparent, cause pandora hardware will only see one one battery circuit. there is no way to know if you connected 1 or n cells in parallel

the project is quite simple, you only need to buy the cells (same age, size, capacity...) and solder. maybe you need an external connector and an external case to keep the cells:

remove the pandora cell (unless you have a 100% simmilar cell)
keep the circuit and fin that pandora cell was connected to the circuit by only two lines ( + and - ). use those +/- to connect to the rest of the cells in parallel.

it could be something like:



PD: there are power inputs in the board, but you cannot charge the cells with that. you could use that internal connector for example an external pack of AA batteries

OK the project I have in mind is a bit more than just an extended battery. PM me if you're interested in working with me on something.
 
Last edited by a moderator:
This is why I asked for help with this. That comment means nothing to me :p
 
Also shouldn't there be some sort of cell balancer? If one cell reaches a full charge before the other the charger will shut off and you will have one full battery , one almost full battery.
 
so each battery need its own individual charging circuit, all linked by some master circuit?
 
DaveC said:
Also shouldn't there be some sort of cell balancer? If one cell reaches a full charge before the other the charger will shut off and you will have one full battery , one almost full battery.

yeah, that could be a problem, and batteries with many cells have their own balance methods, gives info/signal from every cell (or pair of cells)

but it is sorted somehow because the cells are the same at the begining (could be good if they are charged 100% before parallelized), and their life conditions will be the same, charge/discharge cycles, temperature, so they should age similar.

will work for a basic project with zero changes to pandora hardware

pleng I respond your PM :)
 
Last edited by a moderator:
A word of caution, do not underestimate cell balancing...
Your problem is actually not that you'll end up with a battery half full, it is that you'll end up with an exploding battery.
When you connect them in parallel and one goes out of balance, the simple charging circuits (at least the ones I've used for RC planes) will keep charging both cells until the combined voltage of both cells is above the charging threshold.
This will cause the one that was already at maximum capacity to get over charged which could cause it to catch fire.
This is not an idle warning, LiPo cells DO catch fire and sometimes explore when over charged.

example

I know they are laughing, but if this happened inside your pandora, I am guessing you wouldn't be...

On the up side, there are ways to balance while charging, you either balance while charging by giving each cell a different charge circuit (most packaged LiPo batteries with more than one cell have a balancing port which allows you to connect to an individual cell within the pack), and there was another method I saw a while back where the entire pack was being charged at the same voltage but the balancing port was connected to a small circuit which would keep drawing power from the full cells until the ones that were not full got full. It wastes more energy but works with a simple "dumb" charger.
I'm guessing the second option might be possible on the pandora with multiple cells without touching the main charging circuit.
 
Last edited by a moderator:
^ I dont fully understand. Why would anything happen to a full LiPo battery when a charger is just holding 4.2V at its terminals? While a second emptier LiPo attached in parallel would just charge until its full.

All the problems you're talking about above sound like problems associated with using LiPo cells in series, not in parallel.
 
Let's take this hypothetical situation. You have one LiPo battery at 0% charge, one at 50% charge. To charge a battery you have to apply a voltage slightly higher than its discharge voltage. LiPo batteries also explode when overcharged.

Now let's stick these two LiPos in parallel. At one point, one LiPo reaches 50%, one reached 100%, assuming equal impedance. From this point onward you cannot continue charging without exploding that 100% charged LiPo. Fortunately this can easily be fixed by letting both batteries drain entirely.

Now what if impedances weren't matched? You would *always* end up with a second battery that is below the maximum capacity because you cannot charge the system any further. This is because the batteries charge at a different rate.
 
If you attach one LiPo that is at 0% SoC and one at 50% (assuming no charging losses etc) together in parallel and let them be for some amount of time, you will have two lipos at 25% SoC. Also, lithium charging algorhytm is (for me) quite simple in deed. Hold maximum current until maximum voltage (4.2V) reached, then hold maximum voltage until charge current drops to some small value (wikipedia says below 3% of rated charge current). Done. They wont explode if you just hold that 4.2V because eventually current needed to hold that should go very close to zero. At that moment all the cells that are connected in parallel will be at 100% SoC, because if they werent at 100% SoC they would be taking in some current.

Tl;DR Everything you said applies to cells in series, not to cells in parallel.

Note: Also, letting them drain entirely could affect their life negatively, even when they have the safety cutoff circuit built-in.
 
Hey there Urjaman,

I did some researching, I take it back, you are correct, cells in parallel do not require balancing.
I was thinking RC, in 99% of RC cases we use 2, 3 or more cells in series to get a higher voltage for a higher overall wattage.

Thanks for clearing it up.
 
Back
Top