Various Battery Questions

Sonic01

Member
Joined
Sep 19, 2017
Messages
134
Hi Guys! I have a few Q's I'm hoping you could answer.

1. Are there any issues with mixing Li-Ion and LiPo? From what I can see they both have a min of 3v and a max for 4.2v so is there any harm in mixing these together within a pack?

2. Do parallel cells balance among themselves? I read somewhere (I think on this forum) that this happens, i.e. if you got 2x 2000mah batteries, one full and one empty and soldered them in parallel then after time they would both equalize from 2000/0 to 1000/1000. Is this true?

3. How should I configure my packs? for example if i have 40x 1000mah batteries and 40x 2000mah batteries should I keep all the 2000's in one pack and all the 1000's in another, and just use a balancer to stop the 1000mah from draining to 3v before the other pack?
 
1. No, this is fine. You can do it if you want. But unless you have a good reason I wouldn't do it because of their formfactors. Obviously they won't make a neat pack due to different sizes and shapes. Also check for matching capacities. You don't want to combine a 2Ah 18650 with a 10Ah LiPo.

2. They do and they do it with unlimited current. If they have a big voltage difference then there will be a huge current which isn't good. You always balance cells (or even whole batteries) before putting them in parallel. After that they will always have the same voltage, no matter what kind of charging or discharging you do. If they have the same capacity they will share all currents 50/50, if they don't they will share the currents based on their capacity ratio.

3. That completely depends on what you want to achieve, there is no easy answer. In your example you shouldn't use them together in one pack as their difference it pretty big, but it all also depends on what you have to work with. If you don't have enough cells or other cells then you could even put them together in a pack. If you separate them and make two packs then one has half the capacity of the other. If you put them in parallel, then they will share the currents roughly 66/33. If you put them in series the resulting battery will be limited by the capacity of the smaller pack and you need to cutoff when that one is empty, just as you said. The other one is still 50% charged by that time and you can never use its capacity.
 
1. You can mix. I prefer not to mix in same string just to make it simple working with them

2. They will balance since LiIon have straight curve base on SOC compare to voltage. Dont paralell cells that are empty with full though. Make sure that they are at same voltage before you do it.

3. I would mix so that you get same amount of capacity of cells in each packs and same amount of cells in totals. Capacity in each pack then need to be same total.

This is how I do it:

 
3. I would add that you probable don't want a huge difference between highest and lowest capacity cells. So, you wouldn't want to put 2600mAh cells with 1800mAh cells. This would over stress the lower capacity cells during charging until the pack balances out. If you're using reclaimed cells, you want to be as gentle to them as you can to get as many cycles as possible from them.
 
thanks for your responses :)

so far with my capacity testing I've found most my cells are between 2300 and 2800, though I have about 12 which are 1000 to 1500. thats quite a difference, if i mix these in with my other packs is it going to cause issues? i imagine those cells will get stressed quite badly then?
 
They won't be stressed, at least not more than usual. The better cells will carry their share of the load which means they will be stressed more than you think. You will get pretty uneven current flow across the pack. The cells will share the whole load based on the ratio of their capacites (or rather their internal resistances, which is sort the same since this goes up with the capacity going down).
 
so really it doesnt seem like it matters then if ive got a couple of these low mah cells in each pack?
 
Well, yes and no.

Yes, because it will still work.

No, because it has a truckload of implications.

First of all, unless you do precise calculations you don't know how much current you are drawing from your cells. You can simplify it and make a rough estimation but this is the bare minimum you have to do. If you do what everyone does then your cells are tested at 1A. But maybe you are drawing more than 1A from your better cells. Laptop cells have a rather low maximum discharge current and when they are used it is even worse. A cell that works fine at 1A is maybe going to collapse under 1.5A, so your capacity measurements aren't valid anymore.

I would advise against such a big spread amongst the cells. It may become viable when you know for sure that you are only going to draw really small amounts of current from each cell. But even then it isn't ideal.
 
the draw from each cell in my packs shouldnt be high, im going with a 7s40p setup with a 2400w inverter, but my output will never really go above 1600w.

amperage is pulled equally across cells isnt it? so even if im pulling 40 amps, across 7 packs thats 5.7 amps per pack, or 5.7amps per 40 cells, so thats only 140 milliamps per cell?
 
No, your calculation is wrong. Current isn't distrubuted across the packs in series. Each pack in series has to supply the full current.

For 1600W output you will have about 1850W input. At nominal 25.9V that is 71.43A and 1.786A per cell. You have to re-think your setup. Even with identical cells this is a lot for used laptop cells. And if you then put a wide spread between the cells in to the mix then this is just an accident waiting to happen, so to say. It's not going to work very well and possibly not going to work at all.
 
aahhh yes, ok, now it makes sense :)

yes a 1.8A draw is too high for knackered old 1000mah cells... most of mine are good condition 2300-2800 so they should manage that load ok but again, as you said its still high.

I should look to increase my cell count, if i get enough to lower the amperage a signification amount across the cells i might add those low capacity ones in.
 
Sonic01 said:
aahhh yes, ok, now it makes sense :)

yes a 1.8A draw is too high for knackered old 1000mah cells... most of mine are good condition 2300-2800 so they should manage that load ok but again, as you said its still high.

I should look to increase my cell count, if i get enough to lower the amperage a signification amount across the cells i might add those low capacity ones in.

Well even then, at 1.8a your batteries aren't going to run for very long. Maybe an hour or so. Alos they will not be able to deliver that unless they are fully charged. As they discharge the current they are capable of delivering will decrease.
 
If they are tested at 1A then you should re-test them at 2A. There is no other way to find out whether they will manage 2A or not. Also then they aren't 2300-2800mAh cells anymore.

For this amount of current you should either look into using more cells or using different cells in the first place, i.e. new ones instead of used ones or ones with another chemistry.
 
Best practice for the low draw cells (ie. Laptop cells) is to reduce overall current draw to less than an amp, and preferably as close to .5amp or less. This will drastically reduce the load on all cells.
You can calculate the average amperage across the cells by the max draw divided by the cells in parallel. But then you have to consider the variations in the high to low capacity cells. As DarkRaven mentioned, the higher capacity cells will take the heavier load on discharge. So if you have 2 cells (for simplicity) of 1000mAh and one at 2000mAh, it's quite possible that the 2000mAh cell will end up discharging almost twice the current as the 1000mAh cell. So, if they were average of 500mA, then it's possible the 1000mAh is only discharging at 250mA and the 2000mAh cell is discharging at 750mA. So, if the average was closer to 1Amp, you can see how much greater the one will be over the other. And this is only on the discharge cycle.

On the charging cycle, a cell will only take so much current to fill capacity. However, it can be overcharged voltage wise. So, the two in the example above, the 1000mAh cell will be full long before the 2000mAh cell. However, the charger won't stop charging the 1000mAh just because it's full. It'll read the two together and say, "Oh, I still have capacity to put in" and keep charging until the average between the two cells is full charged. So that 1000mAh cell is possible to be overcharged to 4.25 - 4.3 before the 2000mAh cell reaches 4.2V. Once the charging stops, then they will balance voltage out between the two.

So the lesser capacity cell gets stressed under charge, whereas the higher capacity gets stressed under load.

I'd actually like to see a rig that monitors cell-by-cell in parallel to get a real visual of this process to make it even clearer, and a graph to go by.
 
thank you all for your answers.

I'll test those weak cells again at 2A and see what happens. they all gave different mah's as well by as much as 600mah so im going to do the memory restore function on the opus to see if they level out at all.

just a side note the 1600w i mention is going to be my peak draw, and will probably only happen once a month for 1-4 hours.

typical daily load is going to be between 50-300w.
 
Doesn't matter though, you have to size your battery not only for continuous load but for peak load as well. Otherwise it just wouldn't work properly or maybe even not work at all!
Especially peak loads can be annoying because they can cause a significant voltage drop, much bigger than you have calculated for normal operation. It's possible to build a system that always shuts down on peaks because it triggers its low voltage protection and runs fine the rest of the time. That would just be super annoying.

And by the way: If you are using an Opus as almost everybody does then you can't test cells at 2A. 1A per cell is max on the Opus, you would need another device.
 
DarkRaven said:
Doesn't matter though, you have to size your battery not only for continuous load but for peak load as well. Otherwise it just wouldn't work properly or maybe even not work at all!
Especially peak loads can be annoying because they can cause a significant voltage drop, much bigger than you have calculated for normal operation. It's possible to build a system that always shuts down on peaks because it triggers its low voltage protection and runs fine the rest of the time. That would just be super annoying.

And by the way: If you are using an Opus as almost everybody does then you can't test cells at 2A. 1A per cell is max on the Opus, you would need another device.

thanks I'll bear this in mind.

btw you're wrong about the opus, it has a max 4 amp capacity, if you only use 2 slots you can up them to 2 amps each :)
 
Nope, I'm not wrong :D If I'm wrong then the specs are wrong ;)

What you said is true, but only for charging. Discharging is limited to 1A though. This is what it says in the specs. And it makes sense, because discharging generates much more heat than charging.
 
@ Korishan
If cells are charged in parallel with a decent sized bus bar they will be at the same voltage regardless of the cell capacity.
The lower capacity cell will not be at a higher voltage than the lower capacity cell.
 
Back
Top