Just sold this board, but tried my hand on pcb batteries for the GTX.
Surely no decent amount of amps will go quietly through this dinky strip on the positive side
https://esk8-news-objects.s3.dualstack.us-east-1.amazonaws.com/uploads/original/2X/c/c9d81e37beeb2d2de571460cadf2e96e7ba16e8b.jpeg
5 Likes
Isnt the "dinky strip "sposta be cell level fusing. I think lookin at it thats the idea
1 Like
Why do we fuse the positive terminal? Shouldn’t the end with both contacts have the stronger nickel?
3 Likes
I never understood cell level fusing. It seems synonymous with crappy conductivity. If the goal is to break the series connection all the four cells would have to burn through their nickel. What temp would be necessary for that and will the cells actually get it that hot? All the while will be pumping amps and from what I’ve heard a dead short will send a 30q into runaway instantly.
2 Likes
Its a lot less work/space/cost than thermal or currently triggered breakers, and it’s a decent defence from the whole pack exploding
1 Like
Will it stop the pack exploding though as it would have to short likely for awhile to burn through that nickel and by that point maybe in flames already.
Maybe the heat from the lame strip could even be a problem melting insulation or something. It’s so thin maybe the few welds would pop off. That nickel is good for maybe ten amps max even if .2mm thick.
3 Likes
Personally I’d rather see a new generation of BMS designs with adjustable shut down currents and daisy chain temperature fuses on PCBs but that will cost a lot more so it won’t happen for a while…someone has to design and make them and I have no idea where to begin
2 Likes
The temp sensors would be nice and maybe could be pretty cheap and easy but I don’t think anything will stop a fire if 30Qs burst in flames at the initial short. Tempted to dead short one just to see but I’ve heard from two people who shorted them and instant Roman candle
To me that pic above with the tiny strip looks hella dangerous.
1 Like
If you special design the pack, you can build it to eject the parallel packs if a temp sensor fires. If there is a divide between the parallel packs like on segmented enclosure then there would be minimal heat transfer to rest of the battery. Catch is you’d need electronics and a servo/solenoid/detonation bolt to open the enclosure and a plug that will release cleanly but still have a solid connection. Thats HELLA work though
2 Likes
the DieBieMS has 2 temp sensors on it and the overall unit works well for lower Amps (up to 100a) I like the setup, the flexi bms lite mini has more of what’s needed and less bulk because it’s not discharging through the bms. The size of the latest version is perfect for any pack.
Flexi lite ( still in development )
A DieBieMS with a 12s6p sanyo 20700c quad (duel) unity
6 Likes
@hyperion1 is there a reason why there is no fishpaper under ther balance wires
How you protect the balance wires in the middle of the pack to rub against the cells? There is some kind of spacer?
4 Likes
I would place one. Silicon can rub off over time. Ask @b264 about damaged silicone wires due to vibration. For sure it also depend if there is a force on it somehow beause of less space or just accidently. that fishpaper stripe is like 0.002 cent and takes not longer than 20seconds to put it there, so why not just to add it.
3 Likes
Plastic rubs off more, but silicone also can with enough vibration. Silicon is hard like glass and isn’t used for wire insulation.
4 Likes
The balance wires are all 22g silicone covered, not plastic and the coat is thicker than normal.
Most balance wires are plastic and will have more issues. The silicone balance wires are 10x better. In areas where no rough or sharp (potential) exist we dont feel it’s needed, if the balance wire gets hot enough to melt the silicone the paper won’t do much to protect anything.
We also have to think of cost, nobody wants to pay 1500$ for a battery and we can’t sell the one we put 1500$ worth of time and materials into for 800$. fictitious $ these days but not so long ago. The builds take time, materials and have a cost. But no one can sell a battery for that price these days and we never would.
Out of all the batteries we have built, 6 have had issues 3 bad bms (D140) and 3 that were pushed beyond the limits of how it was built. Quality of the materials: Cells are the base if you cheap out there you have already started with possible problems. Find a source that will deliver on every cell. Refine your technic and procedures according to the demand of the pack.
I use a rubber spacer1/4 with adequate k-tape this creates a channel to work with,
Now with the multitudes of technics and ways to do it, we have realized that a lot of it is determined on what you use. If you use balance wires that are subject to melting than yes paper should be where it’s needed and not everywhere.
When you build a pack to perform you quickly realize all the little things that can be done without like wrapping the entire group in fish paper a blanket so to speak, that with more layer trap heat, moisture.
As the demand for bigger more powerful packs increase the way they are built will need to reflect higher current and ever-shrinking space requirements.
We have also realized that nickel strip will only get you so far. With the added amps and size of cells, bigger drive systems and demand for big amps the way we build has to change.
PCBs are great for some builds, but not everything. We have gone to a hybrid of nickel strip, flat copper braid and copper bars to boost performance, lower resistance and deliver the amps safely.
If you remove the causes of the issue versus planing for them you end up with a battery pack that lasts longer and delivers.
12 Likes
Okay, I have now finished welding the other half of my battery. Which leaves me with two 5s2p packs. I am now going to hook them up in series and then connect a BMS. However I am not 100% certain on this part so I thought I would ask for help.
If I am not mistaken, the cell group I have put number 1 on should be the first cell group since the negative terminal of that group is not connected in series.
Here is the other side of the same pack:
I will then use 12 awg silicone wire to connect these two packs in series.
If I am not mistaken, putting a wire from the positive of the cell group I have labeled “5”
To the negative terminal of the cell group I have labeled “6”
It should make a series connection right? I will connect the wire like this:
2 Likes
I have one of these generic 10s 15A bms. The balance wires goes to each positive terminal of all the 10 cell groups.
So I would connect them in the order I labeled on the cells. Starting from B1+ to B10+
2 Likes
the first cell in your pack is always where the negative main lead is connected. if your bms share the balance b- with the charge wire minus you start with the first balance wire at the nickel tap on the positve of cell one or negative of cell two. it´s the same potential at this point, so it doesn´t really matter if you solder it on the positve of cell one or negative of cell two.
also, connect first the main charge leads of the bms and than solder the individual balance wires on.
2 Likes
Okay, I think I understand. So if I solder the negative main lead on the cell I have labeled “1”, then the series wire from “5” to “6” it will make the group labeled “1” the first cell in the pack? I can after that connect the balance wires like I have put the numbers?
2 Likes