TMC is an independent, primarily volunteer organization that relies on ad revenue to cover its operating costs. Please consider whitelisting TMC on your ad blocker or making a Paypal contribution here: paypal.me/SupportTMC

Info: When Balancing Occurs, and Pack Maintenance

Discussion in 'Model S: Battery & Charging' started by Todd Burch, Feb 12, 2016.

  1. Todd Burch

    Todd Burch Electron Pilot

    Joined:
    Nov 3, 2009
    Messages:
    4,734
    Location:
    Smithfield, VA
    #1 Todd Burch, Feb 12, 2016
    Last edited: Feb 12, 2016
    For years, there has been speculation about when battery balancing occurs. Some speculated it occurred at 90% or higher. Others speculated that it only occurs with the battery fully charged.

    I had my seatbelt inspection today, and while there the service tech was checking a few things in my diagnostic screens--the BMS tab happened to be up. This reminded me to ask about balancing, since everyone seems to have that question. What I learned is that the pack is *always* balancing. All the time. Which makes perfect sense and agrees with other things we've seen.

    A while ago wk057 disassembled a pack and discovered bleed resistors. These bleed resistors are what balance the pack, by very slowly consuming power from the highest voltage cells to keep everything balanced. Since the bleed resistors are so tiny, the only way they can really be effective is if they are balancing all the time, and it looks like that is indeed what's happening.

    This confirms a few things:

    1. You don't need to charge to any charge level for the pack to balance. It always happens.

    2. Running the pack down to near empty, then fully charging, appears to improve range, but of course this is only imaginary. Doing this doesn't help balance anything, it simply helps the range estimation algorithms get a better view of the voltage range of your pack. Don't do this to improve your range numbers--you're simply stressing the pack for no good reason. (This was generally accepted by many here to be the case).

    3. Charging to 100%, and letting it sit there for a few hours (which in my recent experiments seemed to be helpful for indicated range), doesn't improve balancing at all. Again, it just seems to help the algorithms *maybe* improve the *indicated* range.

    4. The long, long taper at the end of a 100% charge is not balancing. It's simply providing sips to the pack to top it off.

    5. Apparently, an update is coming soon that will improve the algorithm. (This one isn't that newsworthy, but it's nice to hear).

    So hopefully with this info owners will not go and abuse their pack to try to "initiate" a balancing session. As has been Tesla's suggestion all along, don't worry about the pack, and just let the BMS do it's thing.
     
  2. Ampster

    Ampster Member

    Joined:
    Oct 5, 2012
    Messages:
    614
    Location:
    Hermosa Beach CA USA
    My only comment is that I has to be tricky for the BMS to turn on the bleed resisters because of the flat charge curve. Perhaps this chemistry has more slope than the LIFEPO4 cells I have observed.
     
  3. Todd Burch

    Todd Burch Electron Pilot

    Joined:
    Nov 3, 2009
    Messages:
    4,734
    Location:
    Smithfield, VA
    My best guess is that they try to get a good voltage estimate at certain times. For instance, if you fully charge, they can get a better idea of the voltage differences, then they record how much needs to be bled off and slowly do that. Maybe.
     
  4. AndY1

    AndY1 Member

    Joined:
    Mar 22, 2014
    Messages:
    178
    Location:
    Slovenia, EU
    I know the system. It's similar to my RC Helicopter, when I charge the 6 cell pack. All the cells are being monitored and actively balanced when charging.

    I believe, this in one of the causes of the 'vampire drain'. So many thousands of tiny cells, being balanced 24/7 consumes power. Some of those 50 watts of vampire drain is coming from this.
     
    • Dislike x 1
  5. David29

    David29 Member

    Joined:
    Aug 1, 2015
    Messages:
    561
    Location:
    DEDHAM, MA
    This makes eminent sense. It translates to another example of letting the battery management system worry about the battery (within limits).
    Another way to say it is, don't go by folklore -- regardless of how well-intentioned the proponent or how well-versed he or she may seem to be -- if it isn't in the Tesla owner's manual, then most likely we do not need to know.
     
  6. Todd Burch

    Todd Burch Electron Pilot

    Joined:
    Nov 3, 2009
    Messages:
    4,734
    Location:
    Smithfield, VA
    Yes--certainly this contributes to vampire drain. I think wk057 listed it in his pack dismantling thread, but don't recall the value. I think it was pretty small though....maybe a few watts?
     
  7. MP3Mike

    MP3Mike Active Member

    Joined:
    Feb 1, 2016
    Messages:
    1,603
    Location:
    Oregon
    Are you 100% sure about that?

    I thought that wk057 found, during his hacking/exploration, that the balancing doesn't kick in until you hit the 93% charge level.
     
  8. Todd Burch

    Todd Burch Electron Pilot

    Joined:
    Nov 3, 2009
    Messages:
    4,734
    Location:
    Smithfield, VA
    Well nobody's information is 100% definitive I guess, but I don't see why they would impose a limit of 93% in order to initiate balancing. What would be the purpose of that, especially since there are people that almost never charge to that level? It makes more sense to just always be bleeding off higher cells. Those resistors are so small that it's a time-consuming process, and one in which you really wouldn't notice the consumption from day-to-day.
     
  9. CHG-ON

    CHG-ON Still in love after all these miles

    Joined:
    Jun 24, 2014
    Messages:
    2,766
    Location:
    Santa Cruz Mountains, USA
    I just do what Tesla tell me and I live in ignorant bliss:tongue:. It is Saint Elon, after all!

    Seriously though, I simply do not worry about it. I have never had a problem with range. Except once, when it was completely my own fault. But I made it home. I learned now much headroom I need, based on my driving habits and terrain and just charge to that. Seems to be just fine. Even on trips. I used to have my battery display miles and worry about it. Now I use percent, just like the way I used to look at my gas tank.
     
  10. Todd Burch

    Todd Burch Electron Pilot

    Joined:
    Nov 3, 2009
    Messages:
    4,734
    Location:
    Smithfield, VA
    I should add that this tech has been with Tesla I believe since 2012, has 2 EVs of his own, and seemed to be very knowledgeable about everything I asked him--his knowledge was certainly above a typical tech who might be blowing some smoke.
     
  11. wk057

    wk057 Senior Tinkerer

    Joined:
    Feb 23, 2014
    Messages:
    4,722
    Location:
    Hickory, NC, USA
    Well, a few notes based on my experience with the battery packs and the BMS.

    I think the person you spoke with may be confused a little. Yes the bleed resistors on the BMS module boards can be active at any time at any charge level. However, unless they recently changed something, they never appear to get enabled until the pack reaches the constant-voltage phase of charging, at which point the BMS knows for sure which cells are higher than others and which ones to bleed. The resistors then stay enabled for a calculated period of time (based on the delta between the highest and lowest cells) and I guess you could then consider that "always" balancing... but it only starts once the pack hits CV, which is roughly 93% SoC (can be a little higher).

    Correct, with the exception of the above about getting the bleeders enabled if it's been a while. Otherwise the rest is basically an illusion.

    Same as above. The sitting at 100% is the bad part.

    Correct, although the balancing is enabled at this time if required. The taper is the constant-voltage phase of the charge where the cells have reached peak voltage, but not peak capacity.

    This would definitely be welcomed and hope it's true.


    In any case, 99% of the time the BMS does a great job with the pack. Almost no one's pack is *actually* out of balance enough to lose any range. That's just a myth pretty much. Almost all range loss is due to the algorithm used to calculate capacity being miscalculated due to charging habits. If you never charge to 90%+ and/or never discharge to <10%, then eventually the error in the algorithm will start to add up in the form of false capacity loss. Basically Tesla just needs a way to let the BMS reset this and re-calibrate when possible, potentially by monitoring the potential sum of the calculation error and encouraging whatever type of charge or discharge is needed to help to the user when it's needed to re-calibration. Probably too complicated, so my guess is most likely they're going to refine the algorithm based on real world data from 10s of thousands of vehicles in the field in order to refine it a bit, although it still won't be perfect.

    Long story short... just don't worry about it. If you've lost a substantial amount of range, just call Tesla and let them deal with it. There isn't much that you can do to "fix" it.
     
    • Helpful x 1
  12. JenniferQ

    JenniferQ Member

    Joined:
    Sep 13, 2015
    Messages:
    974
    Location:
    San Diego, CA
    I just switched to percent from miles on battery a couple of weeks ago and it felt like I'd finally "arrived" to full EV mindset at that moment. A very nice feeling. And after 6000 miles and 4 months with the car, I really don't worry about the battery at all anymore either.
     
  13. wk057

    wk057 Senior Tinkerer

    Joined:
    Feb 23, 2014
    Messages:
    4,722
    Location:
    Hickory, NC, USA
    Well, I used percent for a while myself. Unfortunately, with only 1% resolution it's almost 3x less precise than rated miles. So, I switched back to rated miles so I had more granularity... not that the "miles" mean anything to me other than a capacity available measure. If Tesla exposed 0.1% resolution (it's there, they just round down to the nearest whole number it looks like on the UI when using percent) then I'd never look at a "miles" rating ever again.
     
  14. Todd Burch

    Todd Burch Electron Pilot

    Joined:
    Nov 3, 2009
    Messages:
    4,734
    Location:
    Smithfield, VA
    I don't think he was confused as much as just not going into details. I'll admit he didn't necessarily say what triggers balancing to start. I suppose I could interpret what he said to mean that the pack could be balancing while at any SOC--you don't have to be at 100%, for instance.

    Regardless, I think what you state and what he stated can be pretty much in agreement.

    Agree, and that makes sense too--it makes sense that the pack might have to be closer to full to perhaps get a better view of the voltage state of the cells and identify which should be bled off, and by how much.

    Makes perfect sense.

    I'm hoping it's a more substantial improvement in the algorithm, versus the tweaks we usually see. (Disclaimer: I understand that getting a true picture of the charge state of a pack is really hard, so I'm not dinging Tesla on this one).

    I'm in complete agreement with you.

    I think that based on your findings from disassembling the pack, and other corroborating evidence such as this, that we have a pretty good idea of what's going on now. And now we understand why Tesla never put any information in the manual about pack maintenance. You just don't have to worry about it.
     
  15. apacheguy

    apacheguy Sig 255, VIN 320

    Joined:
    Oct 21, 2012
    Messages:
    3,711
    Location:
    So Cal
    Consensus amongst the technical crowd of this forum says that Tesla top balances. However, once the off balanced bricks light up during CV charging, the bleed resistors will continue to do their thing regardless of charge state. Perhaps this is what the tech meant when he said the pack is always balancing. Here's why:

     
  16. Kipernicus

    Kipernicus Model S Res#P1440

    Joined:
    Dec 2, 2009
    Messages:
    1,171
    Location:
    Belmont, CA
    Thanks for reporting your findings, Todd, and to wk057 for weighing in as well.

    How do those bleeders work? Are they actively and individually switched on and off for those cells that need it?
    Or is it through clever circuit design that they are bleeding off cells that are at higher voltage than other cells?
    Otherwise it seems like they would be bleeding everyone at the same rate and won't help with balancing at all.

    If the bleeders are only doing their thing when the cells are at peak voltage, then they shouldn't contribute to vampire loss when below 93% right?

    Is this 93% peak voltage why when we first got the cars, before the battery slider was implemented, that the "standard" charge was at 93%?

    The capacity miscalculation: is this why some people seem to be able to keep driving even after the range shows zero miles, while other cars shut down right at zero?
     
  17. apacheguy

    apacheguy Sig 255, VIN 320

    Joined:
    Oct 21, 2012
    Messages:
    3,711
    Location:
    So Cal
    The bleed circuits each control a separate brick so only the bricks with excess Ah will trigger the bleed. There are 16 modules and 96 bricks in an 85 kWh pack.
     
  18. Todd Burch

    Todd Burch Electron Pilot

    Joined:
    Nov 3, 2009
    Messages:
    4,734
    Location:
    Smithfield, VA
    Actively switched on and off, yes. Someone else can correct me if I'm wrong, but this bleeding occurs at the module or brick level (not the individual cell level).

    Well yes again, kinda. Again wk057 or someone else can correct me if I'm wrong (and wk057's battery disassembly thread has details on this), but I believe each module or brick has a circuit board that controls the initiation of bleeding for that module or brick.

    They don't only do their thing at peak voltage. Wk057 is saying that bleeding is merely triggered at higher SOC. The bleeding can continue because the BMS can determine how much energy to bleed off of which modules or bricks--it can do that at virtually any SOC. And keep in mind these are very very small power resistors. We're talking maybe a couple watts here and there, so it won't be a significant contributor to vampire loss.

    My guess is yes. My speculation is that the BMS calculates where it thinks zero ought to be, but if the algorithm is off and the actual voltage ends up being higher than expected, the BMS allows the car to continue to drive.
     
  19. Ampster

    Ampster Member

    Joined:
    Oct 5, 2012
    Messages:
    614
    Location:
    Hermosa Beach CA USA
    Yes, the BMS is guessing where zero is in the aggregate and is letting the car use the pack until the pack Low Voltage Cutoff is reached or one of the modules is approaching its minimum voltage, whichever occurs first.
     
  20. garygid

    garygid Member

    Joined:
    Aug 11, 2014
    Messages:
    612
    Location:
    Laguna Hills, Orange County, CA
    If the equilization is only activated when above 93% SoC, my car, and many others,
    have never begun equalization. Also, the measurement of the 6 "cell" voltages
    in each module is best done when neither charging nor driving (and not when heating
    or recharging the 12v battery). Best to measure when there is essentially no current
    going into, or coming out of the traction battery.

    So, if making the equalization decision only while charging, and when only when over
    93% SoC ... that would seem to be a substantially sub-optimal design decision.

    IMO, it is more likely that the start and stop decisions are made on a "cell" basis,
    in real time, and perhaps only when stopped or parked.
     

Share This Page