Lithium battery passive balancing during charge cycle

Discussion in 'The Projects Forum' started by inwo, Jul 16, 2016.

  1. inwo

    Thread Starter Well-Known Member

    Nov 7, 2013
    2,433
    315
    What exactly is meant by passive balancing?
    Have not been able to find exactly how it works.
    I notice that the bms boards I've been looking at only use a few 10's or 100's of ma.

    I'm charging a 18s 15ah battery with 10 ohms across each cell. 2.3v nominal per cell.
    Unable to do the math, so I'll monitor and see if it helps balance.

    Seems the high cells will dump more and low cells will dump less. Correct?

    Not efficient, I know, but on the right track? MVC-008F.JPG MVC-009F.JPG
     
  2. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,282
    1,236
    2.3 volts.... Lead acid??
    Usually there is another component (like a zener diode) that allows the resistor to dissipate power only when the cell is fully charged, not all the time.
     
  3. Externet

    AAC Fanatic!

    Nov 29, 2005
    758
    57
    What is 18S ? bms ? What are cells supposed to dump ? Are you missing the charger for the cells/battery on your picture ?
     
  4. Dr.killjoy

    Well-Known Member

    Apr 28, 2013
    1,190
    156
    With lithium the S indicates the amount of cells in the battery pack... Have you checked out a cheap hobby charger??? You might be able to mod the wiring to get it to charge
    .
    BMS
    Battery Management System
     
  5. inwo

    Thread Starter Well-Known Member

    Nov 7, 2013
    2,433
    315
    Thanks for replies.
    I am able to charge fine. Not worth the trouble to build or buy a real battery management system.
    The hobby chargers don't do enough cells.
    Yes, 18 cells in series.
    I have several pallets of lithium LTO batteries. They have a range of 1.5 to 2.8v per cell.
    The resistors do seem to have a small effect. Not enough to bring gross imbalance in line.
    May try the zener idea or lower shunt value.
    MVC-001F.JPG
    12 cell charger for another project.
     
  6. Dr.killjoy

    Well-Known Member

    Apr 28, 2013
    1,190
    156
    What are you charging them up too because I though 2.8v would damage the cells ?? Sorry I can't help out more but I was hoping by explaining more someone might be able to help ..
     
  7. inwo

    Thread Starter Well-Known Member

    Nov 7, 2013
    2,433
    315
    As of now I'm using the original 36v lead acid charger. Which charges to 44v or 2.5v per cell.
    There is no real problem with balance, as I am no where near the high or low limits of any one cell.
    As time permits, I will try to raise the charger voltage a few volts, to get more out of my pack.

    Lto is a great chemistry. Tolerates running batteries dead. Overcharging swells the cells but doesn't overheat them.
    Can't find high current zeners. May try 18 bipolar transistors with small zeners C to B across each cell.
    Anything more complex than that isn't worth the effort. IMO Would then just put together a small 18s charger.
     
  8. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,282
    1,236
    First time I have heard of them. Interesting!
    Here is a basic circuit base on a TL431. I think you would need the low voltage one (1.25 volts) and adjust the resistor divider for your batteries.
    [​IMG]
     
    inwo likes this.
  9. inwo

    Thread Starter Well-Known Member

    Nov 7, 2013
    2,433
    315
    Thank you.
    That is the circuit that I was just going to post about.
    Didn't know about the low voltage version 431.
    Or the led portion. I'll do some more testing.
    Here is the circuit I tested and a board for 12s.
    My thought is bulk charger with this circuit plugged into my bms plug.
    As of now I'm using Li-ion at 4.2v limit as that's what I have in my nevs.
    Had not yet decided on current limit resistor, or if needed.
    Thought about "off-board".
    Now with the led, may incorporate "on-board".
    shunt regulator board.gif shunt regulator.gif
     
  10. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,282
    1,236
    I didn't trace out your circuit, but I think you want the current shunt resistor where it is shown in the schematic I posted. Otherwise all the charge current needs to go thru the resistor shown in your circuit.
     
  11. inwo

    Thread Starter Well-Known Member

    Nov 7, 2013
    2,433
    315
    Yes, I'm re-doing the board with the 5w collector resistor and led across it.
    Big pads next to to-220. On end for more flexibility.
    Still checking for errors.
    shunt regulator board.gif
     
  12. inwo

    Thread Starter Well-Known Member

    Nov 7, 2013
    2,433
    315
    Surprised with the low value of R103 and high value of 101 & 102. I'm using 20t pot.
     
  13. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,282
    1,236
    Think of it as a comparator that can sink 100ma. Made it easier for me.
    upload_2016-7-25_7-58-37.png
     
  14. inwo

    Thread Starter Well-Known Member

    Nov 7, 2013
    2,433
    315
    shunt feedback.png
    Q101 didn't immediately saturate with voltage rise. Darlington no better.
    Adding positive feed back makes it sensitive to 50mv change, and snaps on.
    Any thoughts?
     
    Last edited: Jul 29, 2016
  15. inwo

    Thread Starter Well-Known Member

    Nov 7, 2013
    2,433
    315
    If it isn't too much trouble, would someone simulate this circuit to find optimum values?
    Works ok, on bench, but my equipment isn't up for changing a few mv at a time at high current.
    Set to turn on at 4.15 volts
    I have these parts:
    2sa1725
    3ohm/12watt
    10k, 50k, and 200k 20t pots.
    Will this circuit have a region where it overheats?
    I'm hoping not to need heatsink, although board layout makes it easy.
     
  16. inwo

    Thread Starter Well-Known Member

    Nov 7, 2013
    2,433
    315
    Built as shown there is no limit to base current.
    Current thru tl431 is over 200ma.
    That can't be good.
    100 ohms between 431 and base,still drives pnp and base current is down to <50ma.
    I can deal with the heat from pnp better than tl431.
    It also switches sharper.
     
    Last edited: Jul 29, 2016
  17. inwo

    Thread Starter Well-Known Member

    Nov 7, 2013
    2,433
    315
    Everything is working again. As per original diagram.
    Not sure what is going on.
    431 running cool with no base resistor.
    Clothes pin holds temp. probe.
    Red to black clips to measures base current.


    I'm moving the one rat nest circuit cell to cell.
    Set to bleed down to 4v.
    Starting point is about 4.05v.

    MVC-004F.JPG MVC-005F.JPG
     
  18. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,282
    1,236
    Sorry @inwo I have been on my way to vacation. Is all well?
     
  19. inwo

    Thread Starter Well-Known Member

    Nov 7, 2013
    2,433
    315
    All was working well. After 12hrs or so tl431 overheated again.
    The first time, seemed a fluke, as it worked well without a base resistor.
    Tonite 431 was too hot to touch. >300ma.

    I can go as high as 22 ohms and still keep npn saturated. .19 VCE
    Even 47 ohm is ok. .23 VCE

    Any thoughts?
     
  20. inwo

    Thread Starter Well-Known Member

    Nov 7, 2013
    2,433
    315
    On the plus side. I'm half way thru balancing 12 X 47ah cells.
    Taking them from ~4.06v to set-point of 4.00v. About 6hrs for each cell.
    That's fine.
    When I have 12 running, it will balance overnight.
     
Loading...