1. Attachments are working again! Check out this thread for more details and to report any other bugs.

Balance at the cell low voltage limit

Discussion in 'Prius PHEV Plug-In Modifications' started by dan2l, Jan 22, 2010.

  1. dan2l

    dan2l 2014 Prius v wagon

    Joined:
    Nov 26, 2009
    654
    263
    5
    Location:
    Vancouver Wasington
    Vehicle:
    2014 Prius v wagon
    Model:
    Five
    Hello Everyone,
    On Setting up the Buddy cells, I got the 8 straps finished.
    I charged all the cells, including the new ones. I took all the cells apart and checked them all. Most were at 3.345v a couple were at 3.35v-3.36v. None were low. This is all at rest.
    CellsInParlel.jpg
    I am now balancing all the cells at the top. I have 34 cells all in parallel. I have a charger on them putting 4 amps into them. I have a second charger that can give me a boost up to 13amps but it kicks the CB out after a couple of minutes. I am getting 3.37v with 4amps and 3.38v with 13amps. I plan to leave this charging for some time so that I am sure that I have a full charge on all cells.

    More on my process later.

    Thanks,
    Dan
     
  2. Daox

    Daox Member

    Joined:
    May 19, 2009
    196
    58
    0
    Location:
    US, Wisconsin
    Vehicle:
    2004 Prius
    Model:
    N/A
    Why are you balancing at the top if you are wanting to balance at the bottom?
     
    1 person likes this.
  3. SpartanScott

    SpartanScott Michigan Member

    Joined:
    Mar 22, 2007
    60
    4
    0
    Location:
    Howell, Michigan
    Vehicle:
    2007 Prius
    Model:
    N/A
    I believe that he is putting a full, balanced, 100% charge on the cells so that he may then discharge them in series to find out the exact capacities of each cell. Buddy pairs will be chosen so that all 16 pairs have similar capacity. Then he will have to reconnect in parallel and balance at the bottom before install. Is that correct Dan?
     
    1 person likes this.
  4. Daox

    Daox Member

    Joined:
    May 19, 2009
    196
    58
    0
    Location:
    US, Wisconsin
    Vehicle:
    2004 Prius
    Model:
    N/A
    Ah yes, so he can match weak and strong cells together. I forgot about that. Thanks.
     
  5. dan2l

    dan2l 2014 Prius v wagon

    Joined:
    Nov 26, 2009
    654
    263
    5
    Location:
    Vancouver Wasington
    Vehicle:
    2014 Prius v wagon
    Model:
    Five
    Yes, this is correct.

    I have had all cells in parallel for 2 days with 8-16amps going into them for much of that time. They will sit static over night. Then tomorrow I will take them apart and assemble them in series so that I can discharge them while being sure that I have taken them down at equal power drains. Then the one that goes down to 2.8v first is the lowest capacity cell.

    Thanks,
    Dan
     
  6. dan2l

    dan2l 2014 Prius v wagon

    Joined:
    Nov 26, 2009
    654
    263
    5
    Location:
    Vancouver Wasington
    Vehicle:
    2014 Prius v wagon
    Model:
    Five
    Hello everyone,

    I left cells to sit over night in parallel. Then I broke them down and check individual voltages to be sure that they are all holding the charge equally. They were all at 3.451v at rest.



    Then I connected them in series and did a controlled discharge.
    CellsInSeries.jpg
    Attached is a picture of the setup. 34 cells=100vdc. A 100w bulb pulled 3/4amp. A hair drier pulled about 7amps. A coffee cup water heater pulled about 2amps.
    The surge protector on the power strip did not like the low DC voltage so I changed that out. As I remove the low cells the overall voltage will drop so I will need to add light bulbs.
    I drained the battery pack. All Cells were in series so that they all had the same current going through them. I drained 15 amps for 2 hours and 25 minutes. That is about 36ah. 3 cells hit the low limit.

    This is about right I think. So it confirms that I was having more of a balance problem than a weak cell problem.

    I took out the lowest capacity cells 3 or 4 at a time. I marked each cell with the number of amp hours that it took to get a full discharge on that cell. I think I have a range of 36ah to 41ah.
    When I put the pack back together I will pair the lowest cells with the strongest cells.
    EndOfDischrg.jpg

    I now have the cells back in parallel to balance them at the bottom.
    BalAtBotom.jpg
    After sitting over night all the cells were at 3.09v. I am now further discharging them with resistors at a rate of 5amps. this is to try to get them closer to a balance at the bottom.
    Thanks,
    Dan Lander
     
  7. dan2l

    dan2l 2014 Prius v wagon

    Joined:
    Nov 26, 2009
    654
    263
    5
    Location:
    Vancouver Wasington
    Vehicle:
    2014 Prius v wagon
    Model:
    Five
    Hello Everyone,
    I got the cells paired up.
    BuddyPairs.jpg
    You can see the straps that I made. Each quad pack is two buddy pairs.

    I will Get my next two Kits tomorrow so finishing these will wait till after I see what is in the new kits. I will be gone for a week so After that I will finish the permanent install on the Gen2 and then start on the Gen3 and Gen1.

    Thanks,
    Dan Lander
     
  8. Daox

    Daox Member

    Joined:
    May 19, 2009
    196
    58
    0
    Location:
    US, Wisconsin
    Vehicle:
    2004 Prius
    Model:
    N/A
    Looking good Dan. I'm quite interested to see how well this works out.
     
  9. pbui

    pbui Member

    Joined:
    Dec 9, 2009
    190
    13
    1
    Location:
    Los Gatos, northern CA
    Vehicle:
    2004 Prius
    Model:
    N/A
    Yes, it does look quite promising; thanks for sharing your finding Dan.

    In response to this message from the other thread:
    /*
    You put 32 cells in parallel to be sure that you drain them all together to exactly the same level. So 2a/32cells is only 63ma. This is way too low. You will get a much different balance point at low currents than you will get with the current draw that the system has. This is because of the normal cell to cell variation in internal resistance. For a 4kwhr system the system draw is 22-23a.

    It might be good to move this discussion to the Balance at the bottom thread.

    Thanks,
    Dan Lander*/

    I haven't thought of that but it make sense. But with all 32 cells in parallel, I believe the optimum charge/discharge rate is 0.3 C; at least for the Thundersky batteries. If so; 0.3 for 39ah cell is around 12 A. I need a low ohm resistor; how are u getting your 5A discharge ?

    Paul
     
  10. dan2l

    dan2l 2014 Prius v wagon

    Joined:
    Nov 26, 2009
    654
    263
    5
    Location:
    Vancouver Wasington
    Vehicle:
    2014 Prius v wagon
    Model:
    Five
    Hi Paul,
    You can get 1ohm 1watt resistors at Radio Shack. But I also got a 2.3ohm coil resistor at an auto parts store that handles about 50-100 watts.

    I put 2 of the radio shack resistors in series and put that in Parallel with the coil resistor. I measured the current with a clamp-on DC ammeter. But remember that this was after I had drawn down the pack with the hair drier and light bulbs.

    What I did was not perfect. But it was the best that I felt that I could do.

    Thanks,
    Dan
     
  11. mrbigh

    mrbigh Prius Absolutum Dominium

    Joined:
    Sep 6, 2005
    3,686
    699
    2
    Location:
    Long Island, NY
    Vehicle:
    Other Electric Vehicle
    For discharging purposes, you can buy or obtain( from a discarded equipment ) a heater element from an electric clothes dryer or, the other version would be an electric water heater's heater element for about $25.00 at any hardware superstore. This one should rest in a bucket with water to help dissipate the heat.
     
  12. pbui

    pbui Member

    Joined:
    Dec 9, 2009
    190
    13
    1
    Location:
    Los Gatos, northern CA
    Vehicle:
    2004 Prius
    Model:
    N/A
    i thought of that; but my problem is my laziness. I should re-wire like Dan to raise the voltage for faster draining. Even with a cloth dryer element; which draws around 3000w per 100vac, which means ~30A current or 3+ ohm resistor. With a 3+ ohm on a 3v circuit, that's just shy of 1A.

    I think i have to overcome my laziness.
     
  13. dan2l

    dan2l 2014 Prius v wagon

    Joined:
    Nov 26, 2009
    654
    263
    5
    Location:
    Vancouver Wasington
    Vehicle:
    2014 Prius v wagon
    Model:
    Five
    Hello Everyone,
    I am finally overcoming my laziness. I was out of town for a week. When I got back I charged my New Gen 3 kit. It charged evenly with no balance issues. The new kits are much more trouble free that what we have had in the past.

    Then I worked on bolting my new box into my 2005. That was a slow process.

    Finally today I started putting my capacity paired and balanced at the bottom cells into the box. I did this with 2 sets of balancer cables. I got this all done and got my CellLogs working including the alarms. Then I started charging.

    WOW, Chan is right. The cells are charging with unbelievable balance. I am 1/3 of the way through the charge and I see less than 15mv of imbalance. I have never seen balance this good, even at rest. And this good balance is with the charger running.

    It looks like things are going well. Maybe some time tomorrow I will get it all completed.

    Thanks,
    Dan
     
  14. linuxpenguin

    linuxpenguin Active Member

    Joined:
    Mar 6, 2007
    352
    182
    4
    Vehicle:
    2005 Prius
    Hi Dan,

    Just remember that the LiFEPO4 cells have an extremely flat curve in the middle of the battery (meaning just because they are roughly the same voltage now doesn't mean they are balanced). You'll want to reserve judgment until the batteries are fully charged to determine over-all balance :). (I assume that bit about balancing to the bottom cell meant you are now using bottom balancing).

    Bottom balancing won't keep the cells in better balance necessarily, but it may end up being more practical for you given the way the Enginer balancing system..operates. I assume battery longevity isn't really a major concern?

    Andrew
     
  15. dan2l

    dan2l 2014 Prius v wagon

    Joined:
    Nov 26, 2009
    654
    263
    5
    Location:
    Vancouver Wasington
    Vehicle:
    2014 Prius v wagon
    Model:
    Five
    Hi Andrew,
    The key to safely cycling the cells is to monitor all the cells and stop charging when the first cell gets high, and to stop discharging when the first cell gets low. The CellLogs with the obnoxiously loud alarms do that for me. I am balancing at the bottom but I am not relying on balancing at the bottom to be sure that my system is remaining safe to operate.

    Clearly what I am seeing is better than anything I have seen in the past. The data at the end of charge will tell us how good it really is. We will see if the balance that I am now seeing holds up.

    Thanks,
    Dan
     
  16. banshee08

    banshee08 Member

    Joined:
    Aug 7, 2009
    284
    72
    0
    Location:
    El Monte, CA USA
    Vehicle:
    2010 Prius
    Model:
    II
    Bottom balancing does not benefit the charge state since all the cells have a little different capacity. The important thing is that the cells all discharge to a balance state. It does not really matter what the charged state is as long as you don't overcharge any of the cells. I would be more curious on what the discharge state of balance is after your drive.
     
  17. linuxpenguin

    linuxpenguin Active Member

    Joined:
    Mar 6, 2007
    352
    182
    4
    Vehicle:
    2005 Prius
    Hi Dan,

    Very true, but my point is more that with most LiFEPO4 batteries it doesn't make much difference if you balance on the top or bottom unless the balancer in question isn't working properly. Ideally you want to balance on the top end of the battery to reduce the number of deep cycles the battery is exposed to to extend battery longevity (eg: the weaker one cell gets, the deeper you cycle all the batteries).

    The reason I suggested bottom balancing may be more practical is that it seems this system is better suited to cutoff at the low voltage during driving than it is to cutoff at the high voltage during charging--this of course makes it more practical to bottom balance but I wouldn't say it makes it more ideal.

    I assume the system is designed to cutoff around 2.8v per cell which gives you a little more of a buffer zone at the bottom (versus the smaller buffer zone while top balancing--I'm sure you've seen how fast the voltages shoot up at the top) before damaging the batteries.

    How long exactly are you all expecting the batteries to last?

    Andrew

     
  18. dan2l

    dan2l 2014 Prius v wagon

    Joined:
    Nov 26, 2009
    654
    263
    5
    Location:
    Vancouver Wasington
    Vehicle:
    2014 Prius v wagon
    Model:
    Five
    Balancing at the bottom does not mean that you discharge that far on every cycle. It does mean that if you do ever discharge that far the pack will be balanced at that point. I will run my pack towards the top. But I need to be safe if I ever get to the bottom.

    Both both botom and top were origionally based on total pack voltage, but the new kits now have additional covercharge protection that is based on the voltage of each cell. Note that this protection will undo balance at the bottom if a cell voltage goes up to 3.7v.

    The low voltage cutoff is 46v for the 16 cell pack. so that would be 2.875v per cell. So 2.875v under load is the best place to balance.

    Thanks,
    Dan Lander
     
  19. pbui

    pbui Member

    Joined:
    Dec 9, 2009
    190
    13
    1
    Location:
    Los Gatos, northern CA
    Vehicle:
    2004 Prius
    Model:
    N/A
    Dan - do you have 2 independent sirens, each connected to its wn cellog 8 ? I may have zapped the open collector on one of my cellog. It dawned on me that each cellog is referenced to its own 8-cell pack, thus the upper cellog is referenced to the 24v cell. I should have known this, what an amateur.

    what are your pack alarm voltage settings ? High/Low/Deta: 29v/23v/7v ?

    The pack does seem to behave a lot better after bottom balancing. From the bottom, it took about 5hr to recharge, which is about 4kwhr, until some cells hit 3.75v.

    Paul
     
  20. dan2l

    dan2l 2014 Prius v wagon

    Joined:
    Nov 26, 2009
    654
    263
    5
    Location:
    Vancouver Wasington
    Vehicle:
    2014 Prius v wagon
    Model:
    Five
    Hi Paul,

    Yes, all the CellLog does for the alarm is to switch the red wire to connect the red wire to the most negative cell wire. I do not use the black wire. I tap off of the cell 4 wire at the CellLog to get a 12v difference from the red wire and use that 12v to drive the siren. Also I have been removing the CellLog completely from the kit to do computer downloads because the USB power can get confounded with the pack voltage. The power distribution in the CellLog is not failsafe. Be careful with them.

    I am Running 28.5/23 I do not know what the delta is so I left it high. I turned off the Time alarm. But most important are the cell alarms.

    I had a high cell of 3.6v as the charger was lowering the current. Then it dropped back down and the charger turned off normally.
    My time was about the same.

    I am looking to get some extension wires and to move my CellLogs and sirens out from under the deck so that I will be able to monitor things an seldom never need to open the deck again.

    Thanks,
    Dan Lander