Fish is ubiquitous and can be had many different places. Buy adhesive backed to make ur life easier. Last roll I got was from the 18650 battery store. Kapton as well. Nickel is a little harder to source but duck is great for pre-cut. 18650 club is solid as well especially for bigger sheets. If ur not in a rush, Ali will have a lot of what u need as well but caveat emptor.
Silicone balance wires have some drawbacks although I have used them with success. Specifically, when you are crimping the female dupont connector you can cut the wire if ur not careful. The stock wires are fine for 99% of applications. Just be sure to follow best practices re insulation, crossing wires etcâŚ
So Iâm building this 16s9p pack out of Samsung 50S cells, and when I measured them for voltage the standard was 3.6V, and I measured 4x cells at 3.47v, and 5x cells at 3.5v, not too big of a deviation, but I would like to mix them in together in some way that would work best, as I know car companies measure and match cells in p groups accordingly, Iâm just unsure how
Or even if there is any way to match them taking in account only voltage.
The cells have different prints on them, B03, B11, B13, B14, I assume those are batch numbers, and codes like GC55 CI, GCB5 CA.
I hope the reason for voltage deviation is just age and not something more serious
Anyone has any inputs maybe our battery genius @Battery_Mooch
Whoops, its a typo that happened 3 times in a row⌠they are 3.xx Volts
Price of listening to Joe Rogan while writing posts haha
Thanks for your input
AhhhâŚokayâŚphew.
Donât try to match cells at unknown states of charge as each one self-discharges at a slightly different rate and you donât know how long each has been sitting.
Matching internal resistances (IR) and self-discharge rates can give you the longest lasting pack. Do you have something decent to measure IR with?
You can get a rough idea of cell health and how the cells match for self-discharge by charging all of them the same way and then leaving them for at least a week. Any cells that have dropped in voltage a lot more than the others will force your pack out of balance faster and, ideally, shouldnât be used.
Well not really, I only have a solid multimeter, and at college, I know we have an RLC meter I could borrow, maybe in resistance mode it could measure the cells IR⌠I will ask my professor if something is available.
The cells are new from Nkons supplier, I donât have doubts in their quality but this is probably the last pack Iâm going to build so i would like it to be top-notch
Probably not. The R mode is most likely the same as what you get with a multimeter. Do the self-discharge test then as it only requires a multimeter and time.
Charge them all separately to 4.20V and then let them all sit for at least a week. You need to know how each will self-discharge so you need to charge them all.
AgainâŚthe voltage of the cells when you receive them from a vendor means absolutely nothing unless the voltage is very low (below 2.5V damages the cells) or near full charge (the vendor charged themâŚwhy?).
It doesnât really matter if every cell is at a different voltage, especially when the difference is small. Sure, itâs nice to have them all show up at the same voltage just for fun but itâs not worth worrying about.
Stuck a power supply in my battery instead of a cell to test the over and under voltage behavior of the bms. Worked surprisingly well, saw the output voltage collapse as soon as the voltage dropped below 2.8 and saw the balancing current kick in at around 4.2V drawing 30ish mA.
Hi, guys, am new to BMS wiring. I want make a top mounted battery connected in parallel to an enclosure battery. My first battery is a 20S8P and the top mounted one is a 20S4P. Instead of using two different BMS, I was thinking about slightly cutting each balance wire exposing the copper and soldering a second wire to connect it to the first P groupe of the top mounted battery and doing the same for all P groups effectively making a 20S 12P wired to one BMS. Is this approche to BMS wiring feasible?
Under ideal perfect world conditions it would probably be fine, but this is not a perfect ideal world. Itâs not impossible that you could end up with a great enough imbalance between the two packs under load that a significant current would flow through the balance wires, leading to unwanted and possibly destructive he4ating/melting of the balance wires.
Much safer to just have one BMS for each pack.
Even then, you need to make absolutely certain that both packs are at EXACTLY the same voltage (+/- a few millivolts) before you connect them together, otherwise a dangerous amount of current can flow between them, potentially causing damage.
There have been folks using NESE and similar no-weld pack techniques for years, and they seem to always be fraught with problems due to vibrations causing intermittent connections and micro-arcing/microabrasions between the cells and contacts.
It can work, but my general takeaway is that itâs not worth it vs a welded pack.
I agree that there is a definite risk when connecting two pack in parallel and the voltages should be close but IMO itâs not as risky as often feared.
For example, two 20S4P packs 0.5V apart, using decent cells, each pack with a total internal resistance of about 0.1 ohms. When connected you have 0.5V driving current through a total of 0.2 ohms = 2.5A flowing from one pack to the other.
Iâm not saying anyone should take two packs a volt or two apart and have no worrries about connecting them. Effort should be taken to bring their voltages as close together as possible, just as a standard âbest practiceâ to reduce arcing at the connectors and avoid risks if a mistake is made reading the voltages.
But it takes a bit of a voltage difference to cause a decent amount of current to flow.
From my limited experience and recklessness on this subject, I can confirm this to be true in my case.
I feel like the warnings are good, that care should be taken to get the voltages close, but from what Iâve seen itâd take a significant difference in voltage to produce a catastrophic result.
