Hey all, I was just wondering if it was okay to mix cells of different capacities in a parallel battery bank.
This questions came up in a Korean flashlight forum and basically one party said that it was critical to use the same models from the same manufacturer with the same capacity along from the same batch. And I claimed that it was okay to use different li-ions, as long as they aren’t CrapFire branded. But obviously I agreed that using ones from the same batch would be best.
I wouldn’t use different cells. Same brand/voltage/ah with protection on it and they should be ok, but a diff batt brands/v/a in the same parallel could run down quicker causing it to do many things, it’s unknown to me what diff lithiums would do but i’ve seen nimh depolarize as the one that ran down first did and then the ones that were drained quicker because of that one did, maybe unprotected li-ions would but pcb protected should shut the power off at 2.75-2.8v, but even with pcb protected cells theres always the chance of draining them past the point of safe rundown voltage making them duds. (there is a way to bring duds back but you shouldn’t get them to that point as the chemistry becomes unstable after the 2.75-2.8v mark, that’s why they have that protection)
If you are talking of a parallel setup then it isn't possible for one cell to run down before the others as they will all be kept at the same voltage during discharge. A newer or stronger cell may contribute more to the overall energy delivered but unless one of your cells is actually faulty, it can't damage the others.
However, some cells like the modern Panasonics can be discharged safely to 2.5V whereas most other types shouldn't be taken that low. In this case you probably shouldn't mix these different types.
Having said that, I'd always recommend matched cells if you can manage it. If you are talking of the battery bank type that takes multiple cells and outputs to USB, then these impose a smaller load on the cells than a typical modern XM-L light, and that load is spread across multiple cells. They don't highly stress the cells. Most of the ones I've seen have low-voltage shutoff built-in. If you have access to a multimeter, it might be good to check what level the battery bank charges your cells to. If it's over 4.2V you may not want to let it fully charge each time for fear of damaging the cells inside. Many people load these boxes with laptop pulls and report good results. They are a great way to utilize old or tired cells that aren't up to high loads in your lights.
Another odd thing is that some of these battery banks say you should NOT USE PROTECTED cells, citing dire warnings about damaging the circuitry inside. Can't see how but.......... Oh, and some protected cells may be too long to fit in.
About the only type of "mismatch" of any real concern is the charge levels when you first put the cells into the box. If one cell is fully charged and another fully discharged, the voltage difference may cause problems (self-discharge, self-charge, current flows?) I'd check them with a multimeter first to make sure they are all at a similar level.
The OP of the CPF page came to the conclusion that it was relatively safe to put cells that are in slight difference in terms of voltage because the peak current flow only lasts a very short time.
It pretty much confirms what we thought Just need to be careful not to mix a fully discharged cell with others, and in general try to start with cells within 0.5V of each other.
parallel cells will float at whatever voltage the bank is at, either on discharge or charge, so theoretically you can put a completely discharged cell with a fully charged one and the one that can’t accept or provide current will float at the voltage of the other cell until they’re both the same. I don’t think that there’s much point doing that personally, but it’s still worth saying
Now, series cells is a whole different kettle of fish - here you want to have the cells equally matched and at the same state of charge when assembling the pack. It’s also a very good idea to either have balance taps or an appropriate PCB as even cells from the same batch may have small differences that build up in time to an unbalanced pack (which can then go kaboom as one of the cells is charged to >4.4V).
Actually, there will flow more or less substantial equalising currents to force the cells to the same voltage.
With some cell chemistries this may be problematic (NiCd), with others not so much(NiMh).
I would not mix cells with different health status or discharge characteristics.
I doesn’t need to end somewhat bad, but may result in low overall capacity and performance. (Cells capable of high discharge rates might damage the wiring though.)
hmm, I’m just repeating what Mtbmacguyver on mtbr told me and he’s forgotten more about battery tech than I’ll ever know. Certainly, if you have a weak cell in a parallel bank it will increase the per cell current draw of the healthy cells, which may end up with the healthy cell voltage getting pulled down further. Similarly, higher capacity cells may end up charging at a higher charge rate once the weak cell stops accepting current. I can’t see how any equalising currents can flow though, unless that was what you meant?
you’re right though, I should have pointed out that there’s no point mixing different cells. I was just trying to say that it’s unlikely anything bad will happen if you do.
The current sharing aspect is correct but not what I meant.
Sorry, it’s not so easy for me to express myself in English. Because of that I will use an analogy like I would for children. That is not meant to be offensive in any way, just way more easy for me!
Imagine you have three rain water barrels. One filled to 4/5, one to 3/5 and one to 2/5. If you connect them at the bottom, water from the 3/5 barrel will flow into the 2/5 one and from the 4/5 one to the other two. These water currents will equalise the water height, thus equalising currents.
With the battery pack it’s the same thing. The cells with higher voltage will discharge into the ones with lower voltage until they are equal. It is not different to attaching a charger.
If the cells have the same discharge characteristics, differences are marginal and so are the equalising currents.
Let’s take for example the very popular Sanyo and Panasonic 18650 cells. As we can read in different threads, the voltage of the Panasonic cell will decrease sooner under load. Paralleling these cells will result in the Sanyo cell discharging into the Panasonic cell.
With these Li-Ion cells, you might never notice. With bigger arrays of high discharge cells this is no fun though. I just wanted to add that because some people are experimenting with more potent cells these days and melting wire isolation is somewhat unpleasant.
Matt, did you read the link ryansoh3 posted? http://www.candlepowerforums.com/vb/showthread.php?161299
Balance currents of up to 6.5A were generated (for a short time) in an extreme test.
As ryansoh3 said, the conclusion was that "probably" nothing bad would happen but best practice would be to try to keep the initial voltage differentials low.
dave_ gave you the rainwater barrel analogy as another way of visualising this. Think of the connecting "pipe" in his example representing the resistances between the cells. The cell's own internal resistance (including any PTC if applicable) plus the resistance in the connecting wiring will limit the current flow just as varying the diameter of the "pipe" would vary the water flow.
um, ah, I didn’t :zipper_mouth_face: Thanks for the link and analogies guys, I now know something that I didn’t when I woke up this morning I’m happy to stand corrected.
you’re right too, that it’s not just theoretical - I’m in the process of rebuilding my powertool packs with salvaged li-ion/po cells and the last thing I would want is an pack short, however unlikely. Those 18.5V packs can dump a serious amount of power too.
