No announcement yet.

Help wanted with Li-ion charger and series/parallel pack operation.

  • Filter
  • Time
  • Show
Clear All
new posts

    Help wanted with Li-ion charger and series/parallel pack operation.

    Hi folks, first time poster, I'm building a Frankenstein's Monster. It's out of necessity, don't aim to start out arrogant but I simply can't elect to keep it simple or normal by throwing money at it.
    I am a fairly experienced and well equipped electrical engineer though haven't worked on battery charging before. I'm going to state plans and conclusions based on things I've read in my research, but I'd like to get the plan safety checked by someone that's worked with it all before.

    I have a Bafang BBS01B 48v, two used good condition 36v 10s3p "lime" brand e-scooter batteries with their BMS that has a dedicated charging input, and 18 more similar capacity (3Ah each) Panasonic ncr18650bd (salvaged from the same source of scooter batteries). I have also purchased two "40A" 3p BMS that charge through the discharge line. See it yet? I don't have a lithium charger for either, but the doing of building them is fine once I'm clear on how to do so.

    I intend to run the 36v packs in series with the 10.8v packs, and two of these sets in parallel. I do require the capacity of all of them for my fairly flat long commute, running one 48v branch then the other would be more electrically simple, but twice as aggressive on the cells, longevity is important for me, enough to warrant the complication.

    Vulnerability: BMSs in series may blow the other up when one trips due to under-voltage protection, I'm told one guards against this by placing a diode in reverse bias, parallel with each individual pack that is connected in series, I gather this ensures the motor controller will detect under-voltage and halt before frying one of the BMSs.
    Vulnerability: charging this is not trivial, to the best of my understanding one can connect BMS-managed packs in parallel if their state of charge, resting voltage level is equal, but I am uncertain how exactly to implement this connection with the BMS that has dedicated charging lines. Also voltage sources in series could go real wrong if you connected their grounds internally, I won't.

    I have discharged the slightly higher voltage of the two 36v packs so that their difference happened to be less than 6mv, then connected them in parallel via a small resistor, adjusted it sufficiently small that the voltage rebound was only just not growing. Is it now correct to connect in parallel the charging lines and charge them through it as if they were a 10s6p pack? They are now connected in parallel via the power output lines, 1ohm between their 0V, sitting at 36.8v.

    I haven't built the 3s packs yet, I do have the right materials and equipment. I am leaning towards using both 3s BMS to make it two packs instead of making one 6p pack, even though it would be simpler and electrically sufficient, because it is fewer cells in parallel wandering about over time, the two BMSs should look after them slightly better.

    I need to build chargers for these too. I understand CC-CV is the end goal, but in the immediate I'm thinking more like stepped-decaying current, up to constant voltage. I've got a 41V supply, so to charge these I can just whack it on there with some series resistance, correct? I've got a bank of sufficiently rated resistors I can short to give me a resistance from 12 ohm down to 0 over 8 steps, so I can start out with minimal current of about 2A, which across 6 cells would be about 0.1C charging! Slow I know, will probably increase it. By shorting resistors I can step the current back up as the voltage on the battery is climbing, so it's not such a decaying-current charger. Eventually as current drops the voltage across the series resistors drops, ending with the battery at 41V ~0A, so apart from wasting energy with ohmic resistance, and wasting time with inadequate current, this is a fine and dandy charge operation as far as the battery is concerned, right? Repeat for 12.3V?

    I would be building a better charger in future, this is a quick hack probably only used once or twice. Yes I've said things like that before!
    I expect the middle of the pack-bridge would be permanently connected by necessity for balanced charging of the two 3s packs, there aren't any more sneaky diodes needed anywhere in it, are there? I'd then end up with a 4-pin charging connector, possibly 3 if the ground lead of the dedicated charge lines on the scooter BMS is just connected to ground (it measures like it is), which will be connected to the V+ of my 12v charger.

    So yeah, reluctant to charge in and blow everything up, can someone give this a look-over and see if it's dumb but viable, or not?

    Cheers, fish.

    I can only suggest riding as you calculate and plan your design.