Anyone have good links for the Headway 38120 on alibaba or similar? Designing an EV conversion and have almost settled on this as the battery, but wanted to test a couple cells before buying a load of them
1 Like
Each nickel strip attached to a cell in a P group will do about 15A, but the Series connections will do 15A*5.
So .15mm strips for each cell in a P group then copper series connections should work.
If you want to weld thicker you can stack two strips on each other and double it up. Might or might not need thicker than .15mm depending of your current draw requirements. I don’t think welding thiccer than .3mm is reasonable with the KWeld.
2 Likes
How many amps is that bad boy capable of discharging?
Its only 200W
1 Like
Welding above 0.2 is tricky and not every one can do it reliably as the weld current wants to travel down the strip not in to the battery welding the strip and battery together. Iv never heard of any one welding 0.4mm nickle
3 Likes
This is the idea, then connecting series in the middle.
Ideally you would solder the bus bar somewhere away from the cells like on top of a PCB so as to not overheat the cells and connect bus bars in series with flexible silicone wire. Copper takes a ton of heat to solder, so the heat quickly travels across the whole bus bar to all the cells if you are soldering directly to the nickel you spot welded to the cells.
Either way, I did this with my pack and it works great, just don’t dribble solder down the crevice between the positive terminal and inside the cell. Be as quick as possible, and be ready to cool the cells down somehow with cold forced air or something?
The only reason I did a copper bus bar like this is because my 5P pack with VTC5A is capable of 175A, and without the 8AWG bus bar there is no way a few layers of nickel would carry a fraction of that current the short distance across the P group over to the series connection without heating up.
If I were to do the PCB, and nickel bridging the terminals to the PCB, I’d personally layer up the nickel from each cell 2 or 3 times, then solder the copper bus bar to the PCB ontop of the nickel, that would mostly avoid heating the cells like I had to.
5 Likes
I know you are in Australia but batteryhookup has those cells for $11US and they are a legit site I have used.
1 Like
We all talk about these big currents but on a street board I only draw 8-15amps when I’m up to speed unless I’m acelarating or going up a good hill and wish I found that out by reading this forum but no one says what normally pull from a battry. I pull 50-80 amps when acelarating hard but for no more than 3 seconds. Any thing els is me been stupid or thrashing it up some stupid hill. If the cyclist next to you is crawling along panting and your flying past then you know you pulling a high ampage
Few basic rules I now use
Using AWG 8 or larger is hard to solder and risk cold solder joints and a bitch to bend tight
XT90 tack cable AWG10 (55A) or smaller
VESC use AWG 12-14 AWG (32-41A) as a supply cable with motor cables 14-16AWG (22-32A) (unless some one has beefed them up)
Motors use 12-14 AWG (32-42A)
30Q rated 15-20A most commonly in a 4P making it 60-80A
Hopefully you can see a common total ampage hear 60-80A in a street phane set up
Befor any one says all my A are conservative after cable factors have been applied.
3 Likes
@sedmii did some tests on professional lab equipment. He used 12x0.2mm pure nickel strip. After 7 minutes on 30A current, nickel had temperature of 47 degrees celsius. To be fair, nickel wasn’t isolated under kapton tape like in battery pack, but still this is much much higher current than in any data sheet on internet. what I want to say, we are probably using nickel way too conservatively.
for example, look at @longhairedboy video making battery for evolve. he uses only one layer of nickel (0.15mm if I am not mistaken) and his wires for series connections are not even in the middle of group, thay are between 3rd and 4th cell. That means that only one nickel strip has to carry current of 3 cells. If I remember right, evolve esc draws up to 40A, and that’s still 30A (out of 3 cells) on only one strip. and I never heard that he had melted nickel strips or burn marks.
It is better to put layer too much than layer too less, but I still think that people worry too much about nickel ampacity.
4 Likes
what was the length of the nickel strip that you tested?
1 Like
few cm, similar length like we use in batteries
2 Likes
The length of the nickel strip is not relevant to its current capacity.
It does affect voltage drop and total heat emitted but not ampacity.
2 Likes
Batteryhookup have recently run out of the stock, unless they have received more in the last 24 hours. Glad to hear a good report about them, as I wasn’t quite sure on them
Edit: they have just received more stock! Thanks heaps I’m gonna pull the trigger in a bunch of them
1 Like
You are one lucky merderficer
yup. i use 0.15mm on everything. 0.2 doesn’t reliably weld the way 0.15 does. When i was using 0.2mm i was getting packs back with broken welds after a couple months on occasion. Not everytime, but some times. That doesn’t happen anymore and there have been no negative side effects of using 0.15mm that i have seen.
3 Likes
Yes i just got my order of 500 samsung cells from them at $1 each all brand new. They are for a project I’m working on. They have great deals on randon cells all the time for cheap but you have to be fast because they sell out super quick.
1 Like
Maybe someone has an idea what is happening here but one of my groups loses voltages faster then others from time to time (when the board is just stored away). In the beginning this didn’t happen. Opened the battery but can’t see any mechanical defect.
Screenshot of my cell lvl’s:
As you can see it started balanced out on the 30/10 but in the mean time (without riding) group 2 and 6 dropped significantly more.
Could it be the bluetooth BMS?
1 Like
Have you tried leaving it with nothing connected?
Bluetooth BMS seems suspect.
unfortunate having just a subtle issue with a battery and having to repair it.
2 Likes
Yeah, maybe I should test it with the BMS disconnected!
It’s indeed annoying but luckily I can leave on passive balancing to drain the rest of the groups equally. But you still wanna know in the end 
1 Like
