Ebike Battery Recondition Tips?

• 

  bargaino on 15/10/2024 - 01:15

  +1

  Sometimes manual re-balancing can fix a dead pack. But not this time.

  You need to load-test the groups individually. But if it only lasts a few hundred meters after a long charge, the whole group (at least) must be dead.

  I hope you have a spot-welder?

  • 

    skid on 15/10/2024 - 09:26

    Why wouldn't manual re-balancing fix it in this case?

    • 

      bargaino on 15/10/2024 - 11:25

      Because the cells are showing at least 3.8V.
      If it was a balance problem, then when the battery cuts out, you will see a cell group run to the "flat" threshold.

      • 

        tunzafun001 on 15/10/2024 - 17:48

        I'm guessing it goes below the threshold, then instantly recovers when under no load.

        I've got a few cells in my garden solar lights. All present as 4+v when out

        But in a cree torch, some are dull, some strobe. Guessing my packs have a few of these cells. Never understood what causes a cell to strobe. It goes all night. Maybe a heat related short?

        I saw a battery charger out there that can show you the cell voltage under a load. But can't remember what it's called.

  • 

    tunzafun001 on 15/10/2024 - 10:26

    I think the BMU maybe toast, as it stops charging (light goes from red to green) when the 2 packs are still way below the 4.1v.

    I have a basic generic lithium phone charger (has 2 probes) that I might try to charge the low bank with and see what happens.

    Also the indicator light on top doesn't work either, and that seems to plug into the BMS.

    • I left the pack sitting overnight, and the total voltage hasn't changed at all. So surely a good sign of the cells being ok?
  • 

    tunzafun001 on 19/10/2024 - 23:52

    +1

    Bringing each pack up to 4.1V each manually has made a huge difference. Was stable at 41v over night.

    Took about 8 hours to bring them all up.

    But then the bike wouldn't turn on. The terminals showed only 1.3v once inserted on the bike. Completely dead…

    I inserted the charger for 1 second, and then it all fired up. For some reason the charger gets the BMS going. Bike has lots more grunt than before.

    So think it's time for a new BMS.

    • 

      bargaino on 20/10/2024 - 11:56

      For some reason the charger gets the BMS going.

      That would be a software thing, not a problem. See how it goes for a while.

      • 

        tunzafun001 on 20/10/2024 - 22:24

        Ok, cheers.

        I've got 2 battery packs. For some reason one of them needs the charger plugged in for a second to activate it, the other doesn't. But after that, they both work fine.

        So one of the batteries cut out after about 800m (medium up hill). I opened it back up and all packs were 4.1v. Connected a 12v airbed blower to each individual pack as a load to see what happens. Each pack was pretty constant at high 3.9v. No idea if that's normal, or should they not drop to 3.9v? Going to order a new BMS anyway and see (they are under $20).

        Battery number 2 went for around 4km. Then headed up a steeper hill (maybe 7:1- every 7m goes up 1m). Tough to ride up.
        In pedal assist mode it cut out after 100m.

        But I might be expecting too much from the bike? It's a pretty generic china Ebike (UniRyder). Has the controller in a specialised welded box under the crank.

        Only a 200w Bafang hub motor (so not even 250w). Pedal assist isn't adjustable. Seems to be flat out or nothing (which is annoying as it seems it's too much for the bike on steep climbs).

        Am I expecting too much? Or have I just got a batch of tired cells/ bad BMS/ cabling too thin for steep hills (I thought it would be thicker).

• 

  Bignudge on 15/10/2024 - 08:32

  I'm currently going through this myself but the differences in my pack are more extreme.

  Most group are at 3.6v, two are are 3.2v and one is at 0.5v.

  I was going to do it right and started off with the two 3.2v groups. I seperated them from the nickel strips and charged them up + capacity tested them at a 1A load. They all yielded ~2700mah over the same length of time. And once charged didnt self discharge that much over a week… so id pass these.

  Ideally I'd load test them at a higher draw but dont have any equipment to automate this.

  The 0.5v batch is on the way with testing but obviously would be busted/unsafe to use.

  You'd need a spot welder and nickel strips to put them back. Not advised but i have soldered it back with copper wire when i forst started messing with batteries and that pack from years ago is still working great.

  Remember to wear eye protection as one accidental short/spark can blind you

  • 

    bargaino on 15/10/2024 - 11:29

    +1

    You don't need to physically separate the 3.2V group, just charge it with a pair of alligator clips until it matches the 3.6V.

    Of course the 0.5V group needs to be replaced. One bad cell has killed the others.

    eye protection

    worth repeating!!

• 

  tunzafun001 on 15/10/2024 - 10:21

  Any insight into the best bang for buck charger / tester / balancer ?

  • 

    bargaino on 15/10/2024 - 11:31

    just needs a multimeter, and CC/CV power module.

    • 

      tunzafun001 on 15/10/2024 - 23:36

      I was looking at the Liito Kala devices.

      Chuck 4 cells in and push a button. Charges, does a load/ capacity test.

      Looks like it also has recovery hardware for both NiCd and Lithium cells.

• 

  tunzafun001 on 19/10/2024 - 23:54

  Any tips for a BMS repair?

  I see you can buy them on Ali Express for like $9, but wondering what makes them fail?

  • 

    skid on 20/10/2024 - 07:37

    IIRC some BMS have a suicide switch. But if the battery works after putting it on the charger for a second, it could just be updating it's solid state memory. It should work fine now. Otherwise BMS contain some logic running off a MCU.

    What did you use to charge/balance the cell in the end? A Liito Kala charger?

    • 

      tunzafun001 on 20/10/2024 - 10:36

      +1

      Just brute force with a variable power supply. Set it to 5V - 600mA per cell until they went to just over 4.2v each pack. Might of tripped the BMS.

• 

  tunzafun001 on 22/10/2024 - 01:39

  @bargaino

  I got the multimeter out, and pulled the controller out to get some numbers to analyse while riding.

  Controller is a small LSW-587.
  36v - 31.5v cutout
  6A nominal 12A max
  Throttle adjustment 1.2 - 4.4v

  Rode the bike until it cut out.

  Voltage showing as 35v when it died
  Amps 8 - 11.6A (using my DC clamp meter on the battery positive cable).

  Then did the same, but took voltages from each output of the BMS.
  Under no load, they were all 4v.

  Under load (measured as 8 amps at the battery positive)..
  Found pack 1 got down to 3V. The rest were 3.5v or higher.

  So it's either cutting out from pack 1. Will the BMS shut down if 1 pack is out?

  Amps getting close to the controller max when going up the hills. Seems this 200w bike is at it's max ability.

  Thinking of re doing the pack at 12s4P. Will be 2 less batteries (I think at least 2 are duds anyway). Guessing I will need a new controller as well. Would get a basic 36/48v controller. But would need to keep an eye on the cutout with the 31.5v cutout on a 44v pack.

  Not sure if this Bafang 200w hub has a Hall sensor?
  It isn't cutting out from total voltage, as the pack total minimum never went below 35v.

  • 

    bargaino on 22/10/2024 - 01:47

    +1

    Will the BMS shut down if 1 pack is out?

    Yes, it should, in order to protect the group. There might just be one bad cell draining it.

    Completely rearranging the cells sound like a lot of work. Can't you just replace the bad ones?
    Be careful not to overheat if you raise the voltage. you could try adding a few cells in series to the main battery, see how it goes.

    • 

      tunzafun001 on 22/10/2024 - 18:24

      Definitely one, maybe two bad cells in pack 1…

      Put a 15w load on to pack 1 and it went to 2.8v in a couple of minutes. Did the same for the other 9. They all seem stable. Came back to pack 1 and it was back at 3.9v again. So need to separate that pack.

      I was only going to rearrange the entire pack to 44v if I needed to pull the whole thing apart.

      I just rode it a couple kms up our steep hills. It was whining all the way up. But the hub was stone cold at the end. So the hub should definitely be up to it.

      There is only room for 2 more cells in this pack. Hence I was thinking of reducing it by 2 cells to make it a 48 cell - 44v 12s4p, instead of the current 50cells at 36v.

      That being said, they are early 26650's. Can now get 18650's with the same capacity. So could probably do the full 10Ah 44v inside this pack with smaller cells.

      The side benefit being the controller will have a bit more head room. Ie. It's currently pulling just over 11amps (with pedalling) to get up our hills, and it's only a 6V normal, 12A max controller. Might replace the controller for a newer version. But don't think this hub has a Hall sensor..

      • 

        bargaino on 22/10/2024 - 22:14

        Sounds like fun :) Good luck.