Announcement

Collapse
No announcement yet.

Paralleling BMS protected Li-ion packs.

Collapse
This is a sticky topic.
X
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • dtawom
    replied
    Originally posted by paxtana View Post
    Not sure how you got those numbers..If you put 8 24v batteries in series it is 192v not 48v.

    The equivalent of your picture is 2 24v batteries in series and then however many parallel strings you want that are comprised of 2 24v batteries in series.

    BUT I don't recommend this.

    The bms works to prevent issues to each pack. The bms in the batteries you are referring to is not designed to function while wired up like that and it damn sure was not designed to charge like that, so you would be without protection.

    Maybe you could get it to work with cell loggers and alarms instead of bms but you would be a lot better off using a battery built for an ebike from the beginning. I see multiple critical issues in this plan and that's a big problem considering you are literally playing with fire.
    I edited my original post and added a picture of what battery configuration I want to use. It is two 24v batteries in series and then 8 banks of those in parallel. All I'm really trying to find out is if I charge with 48v from the top parallel leads from the new picture in my original post, will I damage the BMS in my lithium packs? Based on your answer I'm assuming that is highly probable. But if I use the pictured configuration for powering the bike and run individual wires from the positive and negative leads of each 24v pack to a 24 volt charger I should be okay and not damage the internal battery packs BMS correct?
    Last edited by dtawom; 05-12-2018, 05:07 PM.

    Leave a comment:


  • paxtana
    replied
    Not sure how you got those numbers..If you put 8 24v batteries in series it is 192v not 48v.

    The equivalent of your picture is 2 24v batteries in series and then however many parallel strings you want that are comprised of 2 24v batteries in series.

    BUT I don't recommend this.

    The bms works to prevent issues to each pack. The bms in the batteries you are referring to is not designed to function while wired up like that and it damn sure was not designed to charge like that, so you would be without protection.

    Maybe you could get it to work with cell loggers and alarms instead of bms but you would be a lot better off using a battery built for an ebike from the beginning. I see multiple critical issues in this plan and that's a big problem considering you are literally playing with fire.

    Leave a comment:


  • dtawom
    replied
    Click image for larger version  Name:	image_12455.jpg Views:	3 Size:	64.3 KB ID:	64393 Click image for larger version

Name:	
Views:	0
Size:	179.3 KB
ID:	64432
    I have an ebike lithium pack charging question that I haven't been able to find an answer for anywhere else so hopefully you amazing folks can help me out. After a disaster trying to use a step up controller from 36v hoover board battery packs to my 48v 1500w wheel I decided to just bump everything up to 48v. The packs I settled on were these 7-2600 mah 20650 cell series packs with their own BMS https://www.ebay.com/itm/24V-7S-BMT-...53.m2749.l2649
    and when purchased in bulk were a resonable price per cell. I intended on using 16 of these to make an @ 48v 21ah battery pack. I figured I could connect 8 sets of two batteries in series and then run those sets in parallel to get my 48v battery bank. I had originally intended to charge the batteries using a 300w 12v led power supply, connected to this solar charge controller https://www.amazon.com/gp/product/B0...?ie=UTF8&psc=1 set at 48v charging. I wanted to go with 48v because connecting a 24v +/- lead to each of the 16 batteries positive and negative terminals would be a pain. Then I started thinking about it while doing more research and began getting much more confused mainly because of the individual battery packs BMS that is involved. In a typical lead acid scenario for this you would just do like the scenario on the right in the attached image says, the only difference being in my case I'd be running 8 parallel sets of two batteries in series. The pictured scenario is for lead acid batteries, but when charging lithiums with a BMS involved aren't things a lot different? if I use a 48v charger in my scenario wouldn't the 24v BMS systems in each charging pack be getting hit with 48v and couldn't this damage the BMS/Batteries? Maybe I'm overthinking things or not visualizing how the power distribution works properly, but I could really use some help and advice before I blow up my expensive batteries. The picture on the right represents the battery configuration I'm going for.

    Any help or suggestions would be super appreciated.

    Thanks,
    Last edited by dtawom; 05-12-2018, 04:40 PM.

    Leave a comment:


  • Sebz
    commented on 's reply
    And to answer you question about the ligo, well they have a single input output, meaning the BMS can figure out if you are charging or discharging and can cut that input/ouput in case of a fault.

    on a BMS with separate charge and discharge port, the BMS does not cut the discharge port in case of an overcharged cell so if you continue charging from the discharge well I'll let you guess the outcome.

  • Sebz
    commented on 's reply
    Well I have to be honest.. I did wrote all the post and say not to do it... Mainly because lots of people don't take the necessary precautions, don't know what they are doing and so...

    Also the BMS's in yours batteries are not designed to be paralleled and they are cheap boards made in China with no detailed spec sheets... they are far from digital smart BMS used in electric cars....so who knows what they can do. But I'm a tester! I like to test stuff and don't mind breaking stuff too. I blew a BMS and wreaked 8 cells while testing paralleling 2 different batteries (capacity and chemistry) for the sake of this post.

    So yeah...I do charge in parallel... all the time ;) BUT I remove the discharge parallel connection before starting the charger!!! The real danger is there: If for some reason one of the batteries BMS detected a fault and stops the charge process well the other pack in parallel will still continue to take a charge but if the parallel connection on the discharge is still present well the current will flow from the discharge of the good battery to the bad one forcing current on a possible dangerous situation .

    You should never charge from the discharge port. The BMS cant stop current if there is an issue..

    But so far haven't found anything that would be of concern if done properly... well the only thing I noticed is that charging to 100% results in a lower total voltage, like instead of 58.8V I get only to 58.3V.. but I rarely charge to 100% anyways.. and when I do (for balancing purpose) I charge them independently.

    So once they are paralleled the voltage remains the same for either charging or discharging it's the current that changes The battery that is lagging will get a little more current then the other till they are even...
    Last edited by Sebz; 03-21-2018, 07:16 AM.

  • AZguy
    commented on 's reply
    I do think the only question would be the BMS and since they are as well documented as any other chinese import it's a crap shoot the details of what they do but they are pretty simple really and so I'm guessing that the suggestion about charging them separately comes from a place of being "better safe......"

  • calfee20
    commented on 's reply
    That is what I was thinking. If one battery started taking more current the voltage would go up until it was even with other battery and then they would charge in step with each other. If they are paralleled though the voltage would have to be the same.

    Maybe you could balance them once in a while separately. I will have to look this up on endless sphere.

  • AZguy
    commented on 's reply
    I'd think you could but I recall someone expressing concern about the BMS's and balancing but I don't think this would really be a concern...

  • calfee20
    replied
    Grin has a 36 volt battery that is made to be stacked either parallel or series or both. There are 2 identical connectors coming out of each battery. Either can be a charge or discharge wire.



    So if you can parallel 2 identical Luna batteries and leave them that way can you parallel the 2 XT60s and charge them both at once? If no then why not?

    Leave a comment:


  • AZguy
    commented on 's reply
    BTW - these parts are *not* for high currents.... just an example of the circuit...

  • AZguy
    replied
    You will be far better served using a N-channel on the low-side.

    There's more to it than that configuration with the high battery voltages we work with - the gate needs to be protected against exceeding the Vgs(max). This can be done with a voltage divider but it's far better to do this with with two resistors and a zener.

    The following demonstrates this - it's a well proven design from a product intended to standoff 200V. Left side is the source and the right side is the load. The resistor values should be chosen for the application and the MOSFET - happy to help...

    Click image for larger version

Name:	i-hGgRScf.png
Views:	3264
Size:	9.6 KB
ID:	56886

    Leave a comment:


  • Sebz
    commented on 's reply
    Never tried it but looks like a very clever way to do it... Now the only thing is to find a 60V and 50A P Channel FET. and you will need a circuit t to be able to drive the gate with the pack voltage (nothing a diode and resistor can't fix)

  • gustavoya
    replied
    Hello everybody
    Did anyone try with a P channel Mosfet?
    Would it work?

    http://digitaldiy.io/articles/electr...a-voltage-drop

    Leave a comment:


  • Sebz
    replied
    Originally posted by commuter ebikes View Post


    When my BMS trips on my 72V system (two 36V batteries in series, each with a 100A BMS), the bike has no power because the LVC in the Cycle Analyst is set to 65V and the LVS in the Lyen Mark 2 controller is set to 56V.

    Under these circumstances, my Cycle Analyst reads "Low V". In order to fix this, I power cycle the system by unplugging a battery from the Y-connector, waiting a few seconds, and then plugging it back in.

    At that point, I need to be aware that I was drawing too much current so I toggle the 3-speed switch to Low and continue riding at a much lower power level.

    In order to solve this problem, I incrementally turned down the maximum current (in the CA) until the BMS never tripped on my 12 mile round trip commute. It ended up at a 30A maximum.

    In my case, this happens because my battery pack can only deliver a continuous 30A for a 12 mile ride. I can set the max to 200A and draw up to 105A, but only for a very short trip. My battery (specifically the cell capacity) is the weak link in my system.

    My next battery will be a higher capacity (Ah) battery with higher capacity (e.g. 3000mA) cells, but I will still use two 36V batteries in series.


    I made a little drawing for you to explain why the diodes are there in series.

    I replaced the BMS with a switch (it is basically that, a mosfet switch) and if let say the BMS trips then the mosfet opens and breaks a 72V circuit ....NOT a 36V circuit like the BMS was design to do. If the FETs in the BMS can handle 72V it's fine... but could be just 60V too who know....! And if let say the mosfet dies and can't be reseted then you can be stranded (depending if you can redo your connections and if the motor can take 36V)

    But add 2 Diodes and you will get 0V at the mosfet and are basically switching a 36V circuit. And yes you will get 36V (minus the V loss from the diode) at the motor.

    Click image for larger version

Name:	Screen Shot 2017-09-28 at 4.17.24 PM.png
Views:	3673
Size:	1.17 MB
ID:	46608

    Leave a comment:


  • Sebz
    commented on 's reply
    Not a paralleling unit... it's a auto switch... and probably will not hold 30a. You will not get the advantages of being in parallel.. It's basically the same as unplug and switching pack manually.

    Would not trust this at all.
Working...
X