Last week I repaired my battery pack because one of the cells had dropped to 0 V and was draining the whole pack. I was lucky I caught it at around 2.6 V. Now all the cells are charged to 4.20 V individually.
My question is: can I add a new cell to a 2‑year‑old battery pack? My current setup is 12s7p. If I wanted to convert it to 12s8p, is that even safe or recommended?
I checked the internal resistance of my cells — they’re between 23 and 40 mΩ each. I’m not sure if mixing in a brand‑new cell is okay. I see Tesla removing modules from their packs, so I’m wondering if I can treat an electric skateboard battery the same way.
I build 3 battery pack so far if you want know my skill level.
I have 84 cells in total(12s7p). One cell was 0v(garbage). 83 cells was 2.6v. I charged the 83 cells back to 4.20v. I just need to replace one cell for this setup but I want to upgrade to 12s8p. Can I mix old cells with new cells? 12s8p?
Mixing a new cell is generally a bad idea and can’t be recommended but if you test the internal resistance of the new cell and it closely matches all your existing cells, I personally might do it. If it’s off by a statistically significant amount, I wouldn’t personally. There are a lot of variables here and how comfortable you are with risk, where your board charges, how easily you can monitor balance levels (on your phone?), and other things can affect these decisions greatly.
agree; ideally cells are meant to age together. Matching resistance is key.
But another alternative which i would do is changing a whole 8P group with new cells and monitor closelly. They might behave a bit different (meaning charging/discharging) from the rest of the pack but it would not be by a lot.
It’s key you monitor your cells with a smart bms from now on, every charge, after every ride to see how the new P group is doing compared to others
Things to look for:
We dont want the new P group discharging too fater than other groups. This might create a condition of stress. If that happens, you can observe and raise voltage cut on your vesc so you dont stress the new group too much. Here i am guessing you are by-passing bms on discharge.
Same as above, we dont want it charging too fast as well, otherwise BMS will have to fry it a lot to balance with other cells and charging will take a LOT of time
This. If there was one bad cell in the P group then cell should read zero and all the other cells in that P group would be low, hopefully over 2.5, but the other P groups would likely all be in the 3.2v range with some variability. All of the cells in the other P groups being 2.6v seems unlikely.
Have you disassembled all the other Parallel groups?
Ideally you do not mix cells. The reality is that sometimes people try to repair packs and end up doing it anyway, but it comes with risks.
If someone really wants to replace a cell, the minimum requirement is that the replacement cell must be the exact same model and the internal resistance should be very close to the rest of the cells in that group.
Your pack is already about two years old, so the existing cells already have some degradation in both internal resistance and capacity. A brand new cell will behave differently.
Usually the cell with lower resistance ends up working harder during discharge. If the difference in IR or capacity is large enough, it can create imbalance in that parallel group.
It will also tend to charge faster because it has lower resistance. If the pack does not have a good BMS with active balancing and enough balancing current, the BMS may struggle to keep that group in balance and that can become another problem over time.
Because of that, I would not recommend adding a new cell to that P group. Personally I would leave the pack as it is rather than introduce another variable that could make the pack less stable.
This is one of those “yes or no” topics, lol. You got me thinking because my cells are between 23 and 40 mΩ. I even asked ChatGPT about it, and it said to keep the brand‑new cells together and keep the old cells together, like Tesla does. Supposedly the BMS will handle the rest and prevent overheating or one series group getting stressed.
But honestly, the internet is a scary place. Lots of true and false info mixed together. lol. When it comes to you, though. You sound comfortable about it, like you’ve actually done this before. If you have, tell me! What was your result?
Right now I’m using the LLT app to monitor my battery from my phone. Last time I checked, it was sitting at 4.1 V(3 months ago). No charging because it was winter time. I also forgot to tell you. My GPS tracker was running for those three months.I forgot to turn it off . I speed up the drain process. I put 1 min refresh rate for my board location.lol
I smell gas, the CID button popped, and there’s discharged material around the positive side. My shrink‑wrap is clear, so I can see the cells without opening anything—that’s how I know. Since it’s a 12s pack, Series Group 6 is the one that got messed up. I didn’t know about this at first, but after doing some research I found out the CID actually disconnected inside the cell. It wouldn’t charge with my charger—I even tried forcing it, and nothing happened. That’s how I knew the cell was done. I’m starting to wonder if the gas had been building up from charging at 9 amps before I ride.
Alright, I’ll keep the config the same. I guess I’ll just wait until the battery starts dying or the performance drops. I only wanted to change it because I’m trying to get close to a 60‑mile range. Right now I get 54 miles, which isn’t bad. I just want the best setup. I also wanted to keep up with my friend’s bike since he’s got crazy range. Besides that, is this cell still good even with the dent?
I had to remove everything because my BMS wouldn’t let me charge from 2.6 V. On top of that, the BMS was still linked to my old phone, so I couldn’t change any of the settings. I was told to disconnect it for a while so it would reset and let me pair it with my new phone. I haven’t tried it yet. I should’ve set that up properly on the BMS from the start.
Update: I think my Bluetooth module is mess up too.lol Maybe that’s why I couldn’t connect it. Oh well.
I would say don’t use dented cells, look up some cell xray pics and you’ll see why - there is a lot going on inside a cell and there isn’t any way to dent a cell without doing some kind of damage. Especially in a high draw pack.
I would be very careful using ai to get more than a good place to start looking into finding a solution - ai has a tendency to answer a question in a way that gives you the answer you want and not necessarily the best or correctly nuanced answer. Like that friend with big charisma but not a lot of actual knowledge - it’s selling its service and not selling expertise. And even if the actual risks are not high - the consequences are big enough to exercise caution.
As far as replacing a single cell or adding one cell to each p group - it’s my opinion that the larger the pack the lower the risk of issues - the cell differences are averaged over all the cells in the p group and so a 8p pack will be less sensitive to differences in cell ir and performance than a 4p or 2p group. If you can get a new cell with similar ir and capacity to your existing cells at minimum and take the earlier advice to monitor every charge and discharge cycle.
Take into consideration the lifespan of the pack too, your pack is aging and one cell has died already - the other cells in this pack are similarly nearing their end of life too - how much work do you want to invest in this pack knowing it is already starting to develop issues? I m not saying scrap the pack but be aware that this pack may accumulate more problems and the worst case problem could be have serious consequences - do you trust yourself to always monitor and maintain the pack condition? Mistakes happen and i try to account for my personal habits in what packs i will repair or modify and keep. Ymmv. Good luck
I haven’t mixed them. The safe thing to do is to not mix them. I also don’t live on the ground floor of a single family residence and don’t have them charging in a shed or anything.
