Parallel connecting high-V sub-pack modules with one (or more) BMS

john61ct

1 TW
Joined
Dec 18, 2018
Messages
8,758
I have not yet done this, but I have deduced from writings of the smartest members here that it should work.

Assume for now, you want separate 1P (or some low P-count) sub-pack modules, spot-weld serial'd, say at 6S for example

Multiple such packs to be joined in parallel at the 22V level for intra-pack power, with 2-pin connectors like XT90 or whatever in order to increase the pack's Ah capacity or decrease the weight

as needed for today's riding.

Note this thread should for clarity stick to **only** this above example, just discussing electrical design principles here

start your own thread for something else

The serial joining of these 6S packs into higher voltage units will be discussed in another thread TBA and linked, later on.

Really the above is not a technical challenge, very straightforward.

The more complex (for this noob) issue, is the desire to add cell-level* devices like

BMS
voltage monitor
cell-checker
balancer
hobby charger

*in reality could be a 6P 3P group of cells in parallel at the bottom level, but here for simplicity the example sub-packs are 1P6S, so we'll use "cell level" to mean "at 1S voltage"
 
So, the claim to be discussed here, is that, when the 1P6S sub-pack modules are only joined to each other *for power flow* at that outer 22V pair level

you can still join all the #1 cells together, all the #2 cells together, etc each in parallel directly **across** the modules

using 7-pin connectors so that a **single** hobby charger / BMS / balancer etc can be inserted and work at the cell level

as if the pack is one, at 4P or 6P or 25Por whatever.

**Not** requiring disassembly of the big pack every time to do checking / maintenance .

Note that, so long as the sub-pack modules are parallel connected say 6P the per-cell devices will not have any info about the **individual** cells

but be working at the 6P1S group level, showing the total / average information for each group.

To test just one cell at a time, then you need to disconnect the per-cell/group 7-pin connection for each pack and test the one module at a time.

This also makes these cell-level devices **easily removable** for testing / verification / replacement themselves

Choose to ride with a volt checker or BMS, sometimes go without if you don't need it.

(Please no "no BMS vs always BMS arguments here")

Your packs can be stored fully isolated when not in use, prevent the risk of a faulty or poorly designed "battery murder system" killing your expensive cells.

Decide your sub-1A balancer is too slow or not adjustable enough, buy a 2A or 5A one, or use an RC charger instead.

Depending on how you wire your intraconnects you will be able to sometimes ride light, other days need long range, beef up the pack.
 
Hey guys,

I have 98 new cells 35 of them in same production year and showing 100mAh lower the the rest of other production year. 3350mAh and 3450mAh LG MJ1

I need to have them 7s7p on top of each other to safe space and get them into a waterproof box, I thought of making two separate packs and parallel the BMS cables and also parallel the separate charge cables and charge and discharge as a one pack, would that work?

Or would I be better of having one pack 7s14p and fold the back and connect the back cells with nickel strip with one BMS.
Main reason is I need to insulate the packs as as much as possible as they will be used in for strolling moto with ip67 case

Peak drew 15A most of the time it won't even go up 10A, Max18A

Thanks Heaps
 
I guess you did not read the OP:
john61ct said:
Note this thread should for clarity stick to **only** this above example, just discussing electrical design principles here
Please post to another thread, or start your own.

 
john61ct said:
(Please no "no BMS vs always BMS arguments here")
Your packs can be stored fully isolated when not in use, prevent the risk of a faulty or poorly designed "battery murder system" killing your expensive cells.
Hold up. First you say "no discussion about BMS vs non-BMS". Then you go on to call BMSs "Battery murder systems". Doesn't matter if you preface it with "faulty or poorly designed", because you imply that BMSs are high risk (potential murderers, no less), and should be avoided/disconnected for storage.

You start the discussion, and then tell people not to discuss.

Nuh-uh...not today!
Michelle Dobyne.jpg

How many BMSs are in use (heaps) vs how many BMSs are the actual real cause of pack failure (a very very small percentage). And I'd guess that a portion of the reported cases are likely due to user error anyway (incorrect setup/programming, or just pure lazy neglect - the expectation that the BMS negates the need to periodically check on the pack).

I say BMSs are the next best thing, and if you don't have one, you're probably a complete moron :lol: (jk)

Discuss.

(You started it).
 
I am obviously not a "never BMS" guy, since the whole point of the thread is to discuss how to use one. But using a very specific limited scenario to isolate that one issue.

It will be much more valuable to the community if it can remain focused.

I am requesting that **this thread** not get sidetracked into pro's & cons, which ones, other layouts etc etc.

I would be happy to discuss those related topics elsewhere if you like.


 
Relevant

https://endless-sphere.com/forums/viewtopic.php?p=1561764#p1561764
 
If the OP question is:

"Can he parallel several 16s1p subpacks at the bike (using a setup that looks like parallel charging boards but with only 2 pin power connectors)?"

The answer is, off course yes, as long as every single cell's voltage remains within the safety range (for 18650 e.g. 3.6-4.1V) and that no dangerously excessive current is permitted to flow from any one of those 16S1P to any other.

Remember, every single cell (even LiPo and Lead Acid) is composed of some serial connection of smaller sub-cells and when a car battery for example fails, it is actually one of the 2V cells inside of it that fail. The internet is full of photos and videos of LiPo packs that are dissected to show the same principle. Even electronic components (e.g. capacitors) are made like that. The only thing is that all those things are factory matched to make the probability of one series part fail so rare that it justifies selling the series combination as one thing (which is a fact that pushes us all to the OP way of thinking, i.e. cheaper repairs), while here it is an end user paralleling in addition to to the restrictions that come with Lithium batteries.

I deleted my previous long discussion here and working on starting a new thread that may lead to solving the issues with this different from mainstream method
 
> Note this thread should for clarity stick to **only** this above example, just discussing electrical design principles here

> start your own thread for something else
 
There is a problem that I did not think about in my last reply. For 1s sub-packs (1 cell) to be connected in parallel (say two cells in parallel), the worst case within the recommended range for 16850 cells of 4v2 to 3v6 (assuming there is some mechanism or electronic circuitry that keeps them there) theoretically happens when the two are at the opposite extremes and that is 4v2-3v6=ov6 and if your cells can supply 3A safely, then 0v6/3A=.3 ohm which is closer to the resistance of the nickel strips/fuses everybody is using, while if you are paralleling, say 10s1p sub-packs, you'll have a theoretical worst case of 10X0v6=6volts which will require 6v/3A=2ohms resistance of 3AX6V=18watt tolerance, which resembles carrying an electric space heater near the battery at all times, or if the connection was .3 ohms, you'll get 6/.3=20Amps worst case between the paralleled sub-packs which is more like distributing the faults that may happen
 
john61ct said:
I have not yet done this, but I have deduced from writings of the smartest members here that it should work.
Wouldn't the following reply by amberwolf in that other thread ... " Would it be possible to connect 2 Battery packs to 1 BMS?" started on 6/5/20) also be relevant to this thread started 2 days earlier on 6/3/20) ?

amberwolf said:
Basically, yes, but: Wiring them permanently into a single pack means if a cell in one dies and takes the rest of it's group with it, it kills all the cells in both packs in that group. Similar failure occurs if the single BMS has a channel that fails and allows either overcharge or overdischarge of a group, and also if the BMS channel fails shorted and kills the cells.

If left as two separate packs with their own BMS, then if one fails for the above reason, you've only lost half your pack, not the whole thing.

If paralleling a pair of packs at the mains only, with BMSes that have just one port (common charge/discharge), then there's no issues with charge or discharge.

If doing that with separate C/D ports, then it is safer to disconnect the charge ports during discharge, and vice-versa.

Being the other thread with above reply by amberwolf has been linked by you to this thread seems like aw's reply is not only relevant, but perhaps the smartest reply so far ... :thumb:
 
eMark said:
john61ct said:
I have not yet done this, but I have deduced from writings of the smartest members here that it should work.
Wouldn't the following reply by amberwolf in that other thread ... " Would it be possible to connect 2 Battery packs to 1 BMS?" started on 6/5/20) also be relevant to this thread started 2 days earlier on 6/3/20) ?

amberwolf said:
Basically, yes, but: Wiring them permanently into a single pack means if a cell in one dies and takes the rest of it's group with it, it kills all the cells in both packs in that group. Similar failure occurs if the single BMS has a channel that fails and allows either overcharge or overdischarge of a group, and also if the BMS channel fails shorted and kills the cells.

If left as two separate packs with their own BMS, then if one fails for the above reason, you've only lost half your pack, not the whole thing.

If paralleling a pair of packs at the mains only, with BMSes that have just one port (common charge/discharge), then there's no issues with charge or discharge.

If doing that with separate C/D ports, then it is safer to disconnect the charge ports during discharge, and vice-versa.

Being the other thread with above reply by amberwolf has been linked by you to this thread seems like aw's reply is not only relevant, but perhaps the smartest reply so far ... :thumb:
the paralleled packs can use 2 diodes, Shotkey diodes, or even ideal diodes (made with mosfets and sold made for less than a couple of bucks on ebay). A simple series of analog switches (like good old 4066 or 4051) can be used to time share a single bms or the single bms can be made to spend more time on the pack that "needs it more"
4051a (367x271).jpg4051b (192x229).jpg
 
Back
Top