Thank you for your elaborate answer i will consider doing a “normal” build but my heart still says stick with the cell level fused design. I dont want to protect the pack on the cell level against motor or VESC overcurrent. I want to protect and isolate the cells from themselves incase one has a malfunction or gets shorted. These cells have 50A peak but on a dead short the 30A SFW (Standard Fuse Wire) should definately blow.
The main fuse attached inside the case on the Pos terminal will be for simple overcurrents or battery shorts but the cell level fuses are piece of mind and last resort.
Since you mentioned the nickel strip will never even give out and act like a fuse this means there is 0 internal pack protection and one cell could runaway to the point where you have a literal toxic smokebomb in the trunk of your car. 30A Fusewire as integrated in my original design could disconnect the malfunctioning cell. Of course against internal shorts inside a cell there is literally nothing you can do and the pack will runaway with or without fuses.
Fish paper or thin plastic sheets will be used in this design only where its needed. I will wrap around the whole pack with kapton after assembly and multiple different layers of kapton will hold the balance leads in place and isolate them from the cells. On the sides and in the front where the leads run along the casing of the battery and the BMS will sit a thin sheet of plastic or fishpaper will ensure separation from the cell housings.
I mean from my experience with around 22+ battery builds my very subjective experience and with the battery failures you usually see 95% of the battery fires with Li-on cells seems to stem from shorting between series and usually with prior cell damage and high state of charge.
Li-on can be volatile but it is not nearly as volatile as Li-Po packs are, I have done all sorts of bad things when building packs from simple mishaps of shorting with nickel tabs, welding at too high current etc. and from my experience usually nothing happens if its a single cell taking damage, as long as the outer wall is intact.
Only time you see your life flashing in front of your eyes is if the series make a short circuit and you have 4-8 cells working together to make fireworks and there is a nickel to nickel short, then there is enough energy to make things go boom for, thats why good series seperation and protection is important, this is also why it takes one cell to go super sonic and the heat from it damages nearby cells and it goes boom for real.
But if there is such severe damage to the cell, the fuse probably wont prevent much, but either way I said my two cents and you can do as you like and what feels best for you of course, I also think there is a reason Tesla stopped the individual fusing of cells in that way, I also think they used to do that with a laser welder and not spot welding but I can be mistaken!.
I did some more digging down the rabbit hole about pack design and reflected on my whole project idea in the meantime.
It might have sounded like i was not really straying away from my starting design and that i was hardset on my idea.
Which is not the case.
I just want to let you know that your input and experience is really helping me out and after giving it some more thought i will most likely go for the standard method. Just sticking nickel strips on these suckers, fuse the main lead at 125-150A with a megafuse and hope for the best. I got enought engineering to do on the battery case, vesc housing, mastplate and the DIY board already. No need to fully prototype a new way of manufacturing battery packs or reinvent the wheel for every new part.
You made some really good points. I will most likely still use my copper busplate design since the materials are already here.
Instead of Fusing with the wire i will spotweld 0.2*7mm nickel strips from the busplate to the cells. Not as fuses but just as a permanent connection.
Sucks that i bought two rolls of 30A fusewire for 80€ that i probably will never use. Atleast i can still test and fiddle around with those if i want or maybe use them eventually on something else.
I also read up some more about tesla style fusing and many sources mention that they actually never saw a single broken fuse and incase of a SHTF runaway they wont matter anyway. The only time they would save you is in a series short like you mentioned. But there would need to be some substantial manufacturing flaw/error or force involved where one busplate moves far enought to short to set of a dead short to the next plate. (If we rule out human error like dropping a strip)
I wil redesign the busplates so they will accept a strip instead of the fusewire.
But before cutting anything i should test if i can actually weld the strip to the copper.
If that wont work ill use pure strips or get a 0.2mm nickel sheet and cut single sheet busplates that can be welded directly in place.
if i adjust the holes with these flat spots shown in red it should offer a good edge where the strips could bend down nicely. Sadly i cant seem to find a supplier with 0.2x7mm or 8mm strips in stock for now. But i saw that nkon has these precut 30mm long 0.2x7mm strips for nearly the same price as buying per meter.
I adusted the spacing from the top of the cell to the bottom of the busplate to 0.5mm (2mm before). Now I will use fishpaper stickers on the positve poles of the battery for more insulation since the spacing of 2mm before was plenty. I moved the busbar closer to the cell so the nickel strips bend down more easily
I just built a pack with a copper nickel sanwich. Maximum copper thickness I was able to spotweld with kWeld was 0.15mm copper but this needs a lot of energy with kweld (around 150J). I used 0.1mm copper with 0.1mm Nickel on top. Welds with 80J with kweld. After 30 welds the electrodes got too hot and I had to make a break.
The pack works well so far.
What kind of power supply are you using with your k-Welder? I have some old tattu 4S 1550mah lipos that i could wire in parallel to make the welder more flexible and portable. I do also have a good car battery that i could use but it would be rather heavy to move around on a desk due to the wire lenght
I use 2 Hobbyking Turnigy Graphene 3S 5000mAh in parallel, they are allmost too capable, if I fully charge them I get overcurrent (>2000A). If I charge them to 3.9V I get 1900A:
Car battery should work as well. There is also a capaictor bank available from keenlabs, if you want it lighter.
I also had a 4S in the beginning but 3S produces less sparks.
Take it from me a single car battery is not enough Amps, Li-po setup is nice and simple but can get very hot, however from my testing you need two car batteries in parallel to have enough juice, otherwise a PSU and the capacitor bank works really well.
I have 3 older Optima BT DC 5,5 BlueTop 75Ah laying around from an old solar project from my dad. I will test with one and see.
Btw what kind of connectors are you using on your battery packs? I was looking at amphenol for the IP67 Model but for the love of anything thats holy i cant find the surlok female socket with the m8 thread (saves the most space and i could bolt the fuse directly to it) anywhere or if i find it its always bulk order only. I really want it to be atleast IP67
Problem with the amphenol connector is that it is absolutely Massive (I assume you mean the two pin DC connector) however the smaller Surloks are just more practical and compact and more appropriate for our builds.
In most cases the 5.7mm 120A rated are fine, if you don’t give your efoil to a 100+ kg beginner. Motor amps are higher than batt amps, so it also depends where you use them.
My mistake. I meant the Male threaded version of course. Threaded will be perfect for the box that im building since the cables will run to either side of the box and i can safe some space that way. Thanks for the Link!
Really weird that Digikey didnt show up in my google search i did look for atleast half an hour for these. Only found them on mouser without any good description. Now that i understand the coding and naming i found the ones i need there aswell. Seems like digikey has the wrong picture for the black male and orange male 120A screw type connectors tho. They show up as crimp type. I already reported it to them.
I think i will still go for the 200A Version since the female plug being a little bigger isnt a deal breaker for my install case. The screw type will be ~7mm deeper but it will fit.
If i use 35mm² cables the rating with 150A should be on the safe side and completely fine. The price difference is also not that much.
I bough the ones without key as I use them for the motor phases. I used colors as a “key”. The only thing that happens if you mix them up is the motor turning backwards. Different Story for battery connectors. For the battery you can also separate the connectors and keep the cables short.
Does anyone know of a waterproof IP67 connector model with a screw top that can handle 30A similar to this XT60 Model? I cant find a good connector for the charging port. Since the connector wont be plugged in while riding regular connectors wont work. They are all only rated while plugged together. I could make a dummy connector but thats just extra stuff i can forget to connect or loose.
Thanks for the help. I found the Waipu SP21 connectors right away. Looks good. Probably not as good as the “Amphenol Conec” i finally found 10min ago but the price is alot better and for the amphenols half the things are out of stock anyway.
