Question for battery experts: If I build a 1S12P battery pack and connect them in series to achieve the required voltage (similar to the image below), and solder 8AWG wire end-to-end to minimize current bottlenecks, would this be a good approach? My goal is to reuse batteries for different voltage projects. Is this a feasible idea, or am I overlooking anything?
I’m assuming I would use the correct BMS to match the series count.
For reference, a few months ago, I built a 13S6P battery with a 100-amp smart BMS for an efoil, branching off from 9S through 13S to get different voltages. The issue with this setup is that I have to carry all the batteries even if I’m only using 9S. Hence, I want to create something more versatile.
I’m anything BUT a battery expert but this approach sounds sort of sketchy imo
Soldering wire onto cells rather than using a spot welder would be the first issue as you might damage the cells with high temp of soldering process
If you are series connecting “independent” packs it crucial that they are in the same charge state BEFORE you connect them together
Dealing with a lot of variables in BMS would be complicated
Make your life a LOT easier and build a pack that will power your rig, don’t ignore the fact that you might face a long swim or paddle in if you have a breakdown.
IF it’s not what you want alter it or build another one so you are sure and don’t start a lithium fire
@Foilguy i was planning to solder to spot weld at every middle intersection. But yes you just reminded me the packs need to be within .2 volt tolerance of each other, which of a deal breaker to keep individual packs. Thank you for making my IQ raise by few points. I just make my own dedicated pack for the project and modify as needed. Cheers bratha.
The main problem with this will be the different aging of the cells and this will drive cell imbalance over time. It will work for sure but gives you a lot of work and possible problems while connecting and reconnecting the groups and bms wires.
What benefit do you hope to get from it? I don’t see any that would be worth the hassle.
I would look at which logical combinations you can use, like pairing multiple 6s6p bricks or 5sXp bricks that you can use in balanced numbers of cycles for each brick.
An example could be 8 6s6p bricks for your 24s12p ebike and 2 of the same bricks for your 12s6p efoil, then using and logging cycle numbers for each brick to the same total - like people in the RC community use their lipo packs
Actually the biggest issue with this is that you are using a Daly BMS which only has a balncing current of 100mA. With your 12p pack it will need more than (4ahx12 / 0,1A = 480 hours > 20 days) to rebalance it. If you charge before it is balanced either the charge will take forever or you risk a lithium fire due to overcharging of the unused bank.
All the other issues (different aging etc.) were already mentioned
Thank you all for the great feedback; it all makes sense. My goal was to test the Efoil, Tow Boogie, and several other projects without having to build specific battery packs. For prototyping, it has already helped me quickly test different voltages, but for the long term, I will be building dedicated battery packs for each project. I definitely do not want any fireworks inside the house, as it’s not worth the risk. Thank you all for your invaluable feedback.
My concern is that you are going to use thick 8AWG and yet you are using a thin strip of metal to weld 12 batteries in parallel. The bottleneck will be the thin metal strips at one end. The more batteries in parallel, the greater the current at the connections between them. The highest current will be at the very last connection.
This isn’t true. Each series string supplies part of the current and the number of parallels add up to the total. In a 10p pack each battery string supplies half of the current of a 5p pack (at the same voltage and power output). The parallel connections between the cells in the pack are not providing any high current, they are for balancing the grouped cells.
What you then do on the ends of the pack is to add multiple connections to the main wire from the series strings, this way each group shares the load equally and heating is minimised.
The series connections between cells are a different story, they carry high current but the nickel strip losses are surprisingly low due to their short length - especially compared to the losses created by the cell internal resistance. I did a calculation somewhere in the foildrive thread to show it and there’s also a great nickel loss calculator made by @ludwig_bre if you search for it here on foil.zone
You have confused parallel connection and serial connection.
In a parallel connection, the current increases while the voltage remains constant.
With a series connection, on the contrary, the current does not increase, but the voltage increases.
I’m not worried about losses, but more about heating close to the batteries. And also about the incorrect indication of the bottleneck.
If you are sure that you are writing everything correctly, then you probably do not understand my post, sorry for my English. Please see the picture in my post, it explains better.
@Pavel in my OP, the multi-volt pack shown in the photo is a 13S6P configuration, with each 8AWG wire soldered all the way across to a thin strip in the middle. For short-term testing, this setup has effectively provided multiple voltages and was well-worth the efforts. However, as I mentioned earlier, for long-term use, I plan to build a proper pack with a single voltage. Please read the OP and follow-up posts, which should clarify the initial question and subsequent discussions. Cheers.
@seantab 6P in the middle, 3P on each side is normal.
But if you have more batteries in parallel, then please check how thin strips will heat up if you make them longer, for example 20cm. Because the short strips will give up all their heat to the massive batteries, and you won’t see a rapid increase in temperature.
!