As some might have noticed, I have been working on a flexy battery PCB with “wireless” technology
I actually worked with bestech to make this happen since they have loads of experience!
It’s kinda similar to the pcb evolve uses in there GTR but without BMS so you’re free to use the bms of your choice!
Some quick specs:
10s4p
50A contious, peak up to ?
0,3mm custom nickel strips
4P custom battery holders
What are the advantages of using this battery pcb kit? 1) Allow your battery to flex with your board 2) No hassle with soldering the balance wires on the nickle strip 3) Easly removing a cell group when faulty 4) Reduce difficulty and time to make a battery pack 5) No need to use glue to keep the cells togheter
It’s more expensive than I hoped it to be but it was a big investment to make the custom nickel, holders and pcb But you will have a nice piece of new esk8 tech
these were 5 bucks each! have more copper than 12awg and will have more balanced cells as each cell is connected with the same length nickel and will have the same amount of resistance.
it looks like the ones here being sold don’t have double sided traces as there’s no vias to tranfer the current and it looks like the trace is pretty thin…how much copper is on there for the series connections?
so I should let people buy stuff that looks the equivalent of maybe an 18awg wire for series connections and not bring it up? at a price that’s like 8 times what it costs to get them made yourself in something over 12awg? whats the point of the forum.
ill post the design if anyone wants it.
the traces are thin. unless they’re more than 2oz copper, which is rare, it looks way under spec. you don’t know what it is.
maybe im just pissed because I paid to get these designed right, and I offered them for no profit and people didn’t want them. I think its all marketing. and im pissy about selling motors with a lifetime warranty that have a great rep with the best stuff all hand made…but people would rather pay more for what seems less. im just pissy. but what I say is true about the board.
I think bestech is trying to make bucks and the copper is money.
ask them how wide the traces and if 2oz copper. it neeeeeds to be double sided at minimum for enough copper if 2oz and with vias connecting both sides. the ones I got were 4 bucks each if like 40 bought.
I also don’t see how you could connect these groups of four in the middle of the board if they’re already connected to the cells. I don’t see how you can connect any of the tabs to the board period actually. I can see the sides being wrapped around the board and soldered to the backside but the connections in the middle of the board how would you do it? have you built one in real life? two rows of cells is a big complication with assembly
seems so. I don’t know everything about it and don’t want to discourage you just saying what comes to mind from my experience. if you can take the rough criticism I think will be better. the copper content being the biggie.
im confused by the layout. Id have thought youd want to have the series path go up and then back down so both terminals are at one end but looking at this it seems theres a long trace going the whole way back under a solder guard and that’s how the terminals are both at the one end. what’s the long trace under the solder guard for instead of just having the two rows connected at the far end and then snake back through the cells again. I don’t see why theres the long trace coming back
continuing my criticism!! hey im interested. as I said the 4 cells wont have the current path as best as possible with that shaped nickel and the further cells with more resistance. maybe not a big deal. and youre using .3mm nickel for essentially a series connection as well as all the cells current is forced into that one path. maybe its wide enough.
if I had my way…id integrate our designs. I like your double wide version with the slots to fit the tabs. and I guess the balance leads imbedded is nice. but id do a slot for each cell and not the groups of four. then you don’t have to make those plastic things and electrically will be better and I think mechanically as well. that’s what I think and hope to hear disagreement. @EboardSolutions it would then be cheaper and easier to make I think. easier to get standard nickel. I used a 3d printed spaced to space the cells but if you had a slot for each piece of standard .2mm nickel that was 20mm or whatever it would orient cells without anything else needed. easier to do each cell individually. wanted to do two rows on a board but seemed a risk but those slots are really nice
I think @hummieee has a point (all be it not the best worded as I think he a little frustrated pleas keep it civil) about the amount of copper/nickle used to join the P Groups but id say aim for AWG10 /6mm2 give you a minimum 30A continues rating.
I think its a good compromise of exceeding it a Little on acceleration and up hills but not been total overkill the rest of the time. by the looks of it it highly sceptical its any were near that.
a 18650 in a e sk8t it expect to be 20 a cell these days 4p that’s 80A your quoting 50A aim for 80A
as for current path in the battery it can be argued both ways
Advantage
there is no voltage potential in the middle of the pack (eg between p1 and P2 so no need to insulate it as it won’t go boom)
Disadvantage long lead on the bat the Pack + and - out the same end making the pack higher resistance
Advantage
The + and - come out the same end of pack minimal resistance in battery pack as less material needed to connect P Groups making the pack lower resistance
Disadvantage
voltage differential in middle of pack eg P1 and P10 face each other risk of going boom if poorly insulated
what I don’t understand is how you spot weld the middle as there’s no holes in the PCB to pass the nickle table throw for welding
using 0.3 mm nickel is kind of needed to carry the current from 1 cell if using single layer but a lot of the cheep battery welders the community uses can barely do nickel this think resulting in potenchal bad spot welds. its why 0.15 is so commonly used in multiple layeres