How deep? 6P? 8P? 10P? 16P? At some point, you need to start doing it a different way. Do you have a link to that endless sphere discussion? That place is hard to navigate 
The web of 39 wires would allow every cell to be balanced as well⌠and if they were 30 gauge wires, theyâd be as small as hairs⌠and also act as fuses if there was a short⌠just sayinâ
B264, 8P is as large as i will go. 8P already is on the limit i would like it to be. There are other solutions if you need something above that. What you are saying regarding series then parallel has no technical background and will not get into that, if you want to research why, would be good start if you are or considering building batteries. To begin with, think what happens to those hair wires when soc of each series battery is not equal and power goes through that thin wire at full current?
ALnewB, there are couple couple good cells out there from sony samsung and lg. All depends what you need, capacity, power and how much you want to spend.
I think i found a BMS that i like. Its 10S 30A which would fit into NESE 6P case (might not be suitable for all as ot will add 123x74x22mm enclosure) It has blutooth module and android app so you can set cell parameters and view state of the battery wireless.
What is soc? Series operating current is a guess. I was talking about using a thin wire for balance charging only, not discharging. Hence the thinness, any use of it by accident for discharging would burn it up before something worse happened.
10s4p 30q is only 1kg Also if you have a bms for discharge its not that fast
Soc - state of charge. Parallel cells equalize capacity. That means when you connect different s groups with different state of charge, cells in parallel will want to equalize and they will do that through that thin wire and it will go vaporized:)
@agniusm Are you talking about one of these bmsâs ? They look awesome. There are even 12s 60A options and even more. Are you going to sell these? In two months Iâll be building 12s 4p setup I definitely want to go with your setup. One of the biggest advantages that I like about your packs is that I can freely take them on any flight and noone is going to say anything.
Oh no, thatâs not how I meant to connect them at all. I mean each 1P group is connected to the bms via 30 gauge wire. So, 1 cell total. For example a 10S3P pack would have 27 30gauge wires and two 8gauge wires going to the BMS. Not all connected to the same pad on the bms. Itâd have to be a bms that could balance charge all of them. Most likely 3 separate BMSs
@agniusm Also one more question would you be able to adopt your design to 21700 in the future ?? When SAMSUNG will release Samsung 30T - with 40a discharge people will jump onto 21700 format. This will allow to make smaller lighter packs like 12s2p
@b264.
In that case i could not name worse solution than yours. Its as bad as it can get and beyond. Sorry, but i will not continue to push some sense into you 
You just made your system that meny times more unreliable; a lot bulkier and hevier; a lot more expensive;
If you think its a cleaver and beneficial approach, its not, its the most stupid one. Sorry, but i had to say it
@NAF Thats not the one i was thinking about. 10s 30a is all i can fit and henze i will adopt it so batteries build with my kits have seamless design. They cost arround 60usd and i have yet not decided if i am going to sell them. Perhaps, just to save people from ordering stuff from multiple places. I have the 21700 worked out but at the moment i have no intensions of making them. There is no point on small scale applications. It works out as less dence solution when you need only few of them and the price will still be higher. It is wise to use them In Tesla cause of so many of them tips the density scales towadrs it
@NAF Just looked at that 30T cell and its worse than I thought. Its 3000mah cell at 35A. 2P volume comes out at 0.125l at 21wh and 70A discharge. Now letâs take Sony vtc5 3p which comes out at same 0.120l but at 33wh and 90A discharge. It will not replace 18650, far from it
@agniusm Your point is valid âŚbut if you look at Mooch tests - he rates these cells at 40a and 3100mah âŚsimilar thing to his rating of 30Q 18650 cells, - samsung rates them at 15A but in real life they just as good as 20A LG HG2.
@NAF Even if so, 6200mah vs 9300mah and power 80 vs 90A by the numbers is just a step back and there are cells at 2000mah with even more power. Sorry, i am not convinced, i see it as not suitable. Maybe some time later they will improove. 18650 is manufactured for so long and in such huge numbers that most bugs are ironed out from manufacturing process. Just dont see the point to switch. OK, if you are spotwelding or soldering makes some point as you do less work, but whats the point when you just shuffle cells in in my design and forget about it?
Started designing BMS enclosure to fit the NESE design. BMS will be available up to 15S and 30A charge/discharge or unlimited discharge/30A charge with external disconnect (controller logic or contactor hard disconnect). First i was gonna implement 15pin JST XH 2.5mm connector at the top of the modules (Need 16pin but that does not fit), but it introduced a second wiring loom failure point and i could not utilize full 15S capability, only 14S max. Also this meant a lot more work so more expensive. I started to look for grommets online just realizing that i can print them the way i need. here is what i came up with. The grommet will exit 16 wires to 15 cells or less depending on system voltage. Here are couple of teaser pics:
Minimum enclosure size is 6P or 123mm in length.
As seen on the Jehu Garcia vlog right?
I think he mentioned about doing review but still havenât or perhaps i missed it but yes 
Just checked and there is another mention here: https://youtu.be/Kvlh1mflOCc?t=3120
Probably he is just too busy with so many projects going on. I know those things takes a lot of time, throw in video editing and you are running out of time.
Cell level fusing would be interesting
Cell fusing is hyped. If nissan leaf cells can operate unfused, i think we are alright with couple tiny cells. Its overrated in small applications and most packs would ignite before the fuse goes off. Note how tesla has air gap arround the fuse. Most diy fised packs dont. You have to sum up failure rate and you will see that properly built 18650 pack does not need complication. Failure rate of 18650 is one in 10000000, do the math
