14s packs instead of packs in 14s

Dear DIYers,...

I have a someskill in a lot of topics but not one deep enough that's DIYfor me.

I have been playing around a lot with the idea, possibilities and since i'm moving in a few months i'd like to start making things more concrete.

As money might be an issue for now, and i need to get familiar with the deep skills i would like to start small and expand later on. However i would like to start off with a 14s system. Not to waste exess money on upgrading later on. For a starter it would be overkill but i'm fairly certain it would pay off in the end and i believe in the idea

So i might start of simple, enough to led-light my shed and slowly expand. This withoutthe hassle of reconfiguring and so on made me think about the topic.

Please don't make the topic about my epic mspaint skills, they are not the topic but it is the fastest i could do. But it might help a little explaining my typed explaination.

my idea is about the expandibility more than rahter my mcgyver skills which i think are awfull. I see most people follow the same proven concept but i'd like to think diffferent.

So i was thinking about milling out slots for 14 rows wide enough for the 18650 to fit, insidemill a clear opening for the fuse wire to fit trough. afterwards i would be able to slide the cells in. But even with one block i would have a 14s configuration 'well if connected the right way at least).

Later on I would be able make another pack and with some rewiring the capacity would have expanded, and all let's say what now are called "packs" expanded evenly. But without mayor hardware changes, if assumed cabling used for let's say 10 packs this could be done with only losening the cabling on the expansion side and extending the bus bars made of heave copper plating. I have the luck I can get my hands on shorter lenghths reasonable copper wire for free.

I would place the "rows" of cells upward with some air gap to provide some ambient cooling. I am also able to "buy" the 18650 cells with my company so they become "less expensive" being "new" ones.

For now the idea is to learn from and use it as an offgrid solution.

Of course the would also be a batrium, any things i might overlook in this configuration might be welcome before i start investing money.

Thanks in advance and may the power flow greatly for you!
Bert

In case you wonder,I would use wood since it is the most basic element i can work with pretty handy And IF it were to catch fire I don't think it matters but freely correct me if i'm wrong, more minds make more ideas.
Also i made a mistake in the bank top view picture, the pos should be on the top right.

Not sure if I got this...
You want to put 14 cells in series and then expand by putting another 14S battery in parallel to the first one? Is this correct?

tought so my explaination was very confusing,... the way you put it, definately no... i'll see if i can make it a little more clear

In expanding on DarkRaven

If this is what you plan on doing, there is nothing wrong with that setup. It allows for per pack expansion instead of needing another 14 packs to add to the installation. This idea is actually my plan for layout.
I plan on making a pack that has 10-20 cells in parallel (haven't fully decided on final number yet), and then 14 sets of those in series. This would give me a pack that is 48V nominal with a capacity of about 10Ah per pack (assuming 10 cells / parallel). Then I would add the packs together in parallel to further increase my capacity. As I collect more cells and run them together into a pack, I add another pack to the bank.
The biggest drawback to this though, is this: You will need a separate BMS for each pack. Not saying that Batrium can't do it, but it's a lot more topology to the mix. I'm going to build BMS's based off of the bq chipset by Texas Instruments. It basically does most of what the Batrium does in a single chip, but on a much smaller scale. This would allow for monitoring of such a small setup (per pack, that is). And then the bq chips would all talk to a central MCU to collect all the data and compile it into something useful/readable.


Addendum: Just saw your reply to DarkRaven. So I guess the above doesn't apply to you, then

Darkraven (and all others)




, I hope these help a little more.

Korishan, please elaborate me, once the new "pack" is connected, would not the parallel bank become (let's say from one to two "packs") basically a doubleparallel bank?

forgive me my terminology,
but let's say 10 cells in "row1", pack1 together with the 10 from "row1" pack 2 would make up for 20 batteries in parallel instead of 10, would they not balance among themselves?
Is this not the same as adding 10 cells to your existing 14s packs?
Why would they need to be individually balanced? since IF so, this soul not be taht cost effective in the ling run

Ohhh, ok. I see what you're doing. You are doing somewhat of a combination of what Pete/Daromer/other have done and what I was suggesting.

I see. What you are doing is have a pack that is 14s1p. Then you connect another 14s1p in parallel at the cell level. This electrically gives you 14s2p. Add a 3rd pack of 14s1p and this gives you 14s3p, etc, etc, etc.
Then yes, all the cells in slot 1 of all the packs that are connected in parallel will balance with each other.

This will work too. However, you will have losses at all your pack connections unless everything is soldered together. I don't think you'd be able to make the wires thick enough to handle the load after awhile down the line. Mostly for safety reasons.
Also, I think you'll have a certain limit to how long you could make this setup. Physically and electrically.

It's an interesting idea, don't get me wrong. I would like to see an end result if you decide to go this route.

Did I get it right as to how you are thinking of doing it?

Korishan, please elaborate me, once the new "pack" is connected, would not the parallel bank become (let's say from one to two "packs") basically a double parallel bank?

Click to expand...

Not sure what you are referring here, though. Unless I already mentioned it in the above.

Yes, that is right on, the main reason is to start off with a 14s system but to be able to slide in capacity on the go, but without having to reconfiguring the existing packs. (apart from the end connections and the extention bars, on the expanding side)

I will have a look into their names tomorrow, allmost bedtime for now.

Well since i am very macgyver DIY one of the first things i was planning to get was a thermal camera of some sorts to find these weak spots.
As for the cabling, i know there is the law of pouillet. I should have to do some calculations there. For the pack connections i was thinking of copper bars 2.5 cm by 0.4 cm. (these i have) and seemed fairly ok (without doing soms calculations).
As for the packs, would off course not add packs of 14 cells would be useless work, but how big no idea yet.

nevermind the quote, bad explaination, you got it

Oh I see, so you are making individual 14S batteries and put them in parallel as I thought, but on each cell. I can't really see an advantage though. I think you are making your life harder by doing that.

It will work. More work in the end.

Try to build as big as you can with first. Then set for instance for 14s20p packs or something and expand. You need to beware of that each connection will hold some resistance and more importantly some work.
You could keep it to balance wires between each cell on the cell level and only keep the outher negative and positive as main lead. I run somewhat like that on my big bank but its 4s100p... That i have 3 in paralell with smaller wires in between for balancing. That saves me BMS money.