NESE, the no solder 18650 battery system

The reason Ive thought of cell level fusing is that when I’ve lost a single cell in a 6p, I’ve ended up losing all six cells. Correct me if I’m wrong but I understand that with cell level fusing you only lose one cell, which results in a modest voltage drop in that one parallel set.

Say you have 5 cells fuse for 100A, so 20A per cell and one blows. Now you pull 100A out of 4 cells at 25A per cell. That difference might bring those cells to thermal runaway temps and the whole board might be on fire. Its different when you have hundreds of cells in parallel as one cell gets little currents and capacity drop would be unnoticeable. You probably were playing with dodgy cells or used cells with unknown history. If one cell would act weird in small pack such as used on the boards, the fault should be noticeable strait away

1 Like

@agniusm, in your above explanation the missing information is what is each the fuse rated to blow at?

I tried to analyze it and the thermal runaway scenario didn’t come up at all. A properly designed cell level fusing would actually avoid that. Take a look:

Max amps drawn by both ESC: 100A
Cells in Parallel: 6
Max amps per cell: 16.67A
Rated Continuous Discharge per cell(25R): 25A
Fuse rated to blow at : 30A

Say one cell has gone bad then the fuse safely blows, leaving the other cells intact. The user continues to draw 100A peak(20Amps per cell) and still everything is fine.

Say another cell in that bad P group malfunctions and the fuse blows. The user continues to be able to draw 100A peak(25Amps per cell) and still everything works fine. However, he notices a significant drop in capacity and realizes that certain p-groups have lost cells. He pops in 2 new cells, replaces the fuses and it’s fixed.

Say he doesn’t notice the decreased range and doesn’t fix them. And to his bad luck another cell in that p-group goes bad and hence blowing the fuse. When he goes full throttle he starts to draws 100Aps(33.3A/cell). The fuse does it’s job and instead of letting a thermal runaway happening, they just cut off the cells without destroying them. The end effect is that he will loose braking at that point and …

But the fact is that there is no thermal runaway problem. There is a problem of loosing brakes when all fuses have failed. If we are using smart BMSs like DieBieMS, it can be programmed to realize a cell fusing has blown and warn the user in some ways.

3 Likes

This is why I don’t fuse anything but the charge port. I would rather the board burn itself to a cinder than run the risk of no brakes when you least expect it. Fuses blow all the time for no apparent reason. Charging though. Thats a different ball game and most, not all, but most fires happen when charging.

Disclaimer

This sweeping statement is based on the assumption that the battery has been built correctly. and not by me.

4 Likes

I have not seen cells that could be discharged at 20A continues. Manufacturers state that but… I was looking at 5P example, but i should use 4P, 3P, 2P examples. What then? 3P group, cell max at 30A (one of sony’s VTC’s). Set the fuse at 35A, controller is at 90-100A. One fuse blows, and then there is a fuse that does not blow (cause there are no identical fuses). Now cell can be sucked at 45A. I have seen how high power cells accelerate in heating up. I was testing with VTC5A (35A). At ~33A it quickly went to 90C, what would happen if you would not notice and fuse does not fail and you drawing 45A? It will get lit !!! Or maybe become a projectile. You can experiment if you want as you have CAD skills. See where there is unpadded tab, u extend that side by the width of the fuse, make a gap for the tab and holes for the fuses so they touch the nipple of unpadded tab. Use glass or ceramic fuse of desired rating. Smalles i have seen is 20mm so you are adding 20mm to width and that becomes 94.2 BTW, anyone seen what 30A fuse looks like? Nothing even close to tesla fuses :rofl:

image

I always recommend using brand new cells from authorized resellers. Use of additional insulation on positive end, use of at least PCM’s. If you do that, there is no point for fusing except weird nonsensical fused cell movement :smiley: I get it when people try to implement safety in powerwall builds where they use used tool, computer, medical cells ripped out of the discarded packs and they have hundreds and thousands of them. There is no benefit here. After all there is no fusing on laptops, medical devices, UPS’es, hand tools, even large backpack batteries.

2 Likes

The user continues to draw 100A peak(20Amps per cell) and still everything is fine.

It should be pretty easy to monitor each P-group and determine if one ends up with less Ah. With a BMS that exposes each group’s voltage over CAN an alert could be easily added to any open source receiver+remote. I’ll probably do this once my NESE-inspired pack is done – I plan to have a fuse wire on the + connector:

12AWG will take the place of the plated copper, with a 3d printed support to translate force from poron to the contacts. @mishrasubhransu if you are at all interested send me a PM, I would appreciate any printing advice you have as I’ve only done a bit.

Completed my NESE battery over the weekend (10S2P), now to finish the rest of the board.

image

6 Likes

Can anyone show a link to a circuit board that will fit in this: http://18650.lt/index.php/n-e-s-e-powerbank/

Just stumbled into this, pretty cool.

s-l400-1

2 Likes

Finally found the time to assemble the first NESE battery. Process was pretty straight forward and sweat less. PSX_20190512_155125 PSX_20190512_155532

8 Likes

they are great, i use them for my powerbanks, but mine are made out of wood.

Here is my complete 10S7P battery for a bike i have:

image

image

8 Likes

@agniusm

Wonderful work here! I’ve seen some of your progress between here and the endless sphere. I do have some questions. Please forgive me if they’ve been asked before.

Are the 3D models available for the BMS portion of your 10S7P battery? I see you have one for the DieBieMS but I wondering if something more generic was available.

Are there models/dimensions for the internal parallel tabs? I understand if you don’t want to share this one. Seems like this would be the money maker.

At the moment i have no other bms modules shared as i am experimenting and trying to get it as simple as i can. If you imagine there are tons of PCM’s and BMS’es out there and then i have modules ranging from 2P to 8P and to mix and match all would be impossible. I will release empty enclosures and add brackets for the boards i own so one could mate them in cad and produce desired combination. For other boards it will be up to comunity and individuals. I do not share tab dimensions and cad’s as i would like to keep some of the stuff internally as it generates some income for me to work and improve this design further. Hope You understand. Its simple part that could be easilly replicated but if you go that length, you have to put in some work :slight_smile:

5 Likes

This would be great!

I suspected this and I totally understand. No worries! I hope you continue to develop this. Solderless battery packs have so many advantages!

Here is a bit of motion picture on the thing. I think i got it sorted. I have created account on thingiverse for these, https://www.thingiverse.com/1865o/designs

3 Likes

Perfect! Those are also the common BMS boards that can be found on aliexpress.

@agniusm just to be sure, what exactly does one get when they buy a set of the tabs? Is it exactly what’s pictured, in other words enough material for 2 4P modules? I’m just confused since it looks like 2 8P tabs, 1 with the poron foam 1 without. Are you supposed to use the foam backed one on the positive side and the unbacked on the negative side?

You get 2 full tabs. You cut them in half and have 2 sets of 4P, or you can cut them to have 5p and 3p sets. Does not matter where you put the foamed tab. I usually place it on negative side

Loving these NESE modules! Making my build super clean! However the tops don’t seem to be able to fit, anybody got tips on getting them on?

IMG_20190515_181251

4 Likes