Paralleling batteries

[Distortion9936]

Hi guys I am new to this forum, and new to building electric bikes. I am working on gen 2 of my electric bike going from a hub motor to a gear reduction motor that wil be running off of the smallest front gear of my bike. My question is I have two 36v lithium ion batteries with 10ah, one of my pcb boards blew, and I was going to try to parallel the batteries together to get a larger ah. I was wondering if anyone could tell me if I need to parallel them before or after the pcb boards.

 

[SamTexas]

I assume that pcb is a battery management system (bms).

If you paralleled them after the bms, then only one of 36V battery is managed by the bms (balanced charging and LVC).
If you paralleled them before the bms, then both are managed by the bms. The question is can your bms support the now doubled capacity? In other words, is you bms capacity dependent?

In case you don't already know: "after" means that you only parallel the two main wires (positive and negative). "before" means that in additional to the main wires, you also need to parallel each individual cell in the pack.

 

[Rassy]

First issue, you need the BMS to properly charge the battery. At one point a few years ago I purchased two LiFePO4 batteries and one was delivered with a bad BMS. So while waiting to get a new BMS I put connectors on the main wires between the battery and BMS and between the BMS and controller. Then I would also unplug the balance wires plug and move the BMS between batteries. At the time I did not parallel the batteries, but only used one at a time.

Later, I did have reason to parallel these two batteries, and because of issues raised by some indicating the need for diodes to protect the BMS's, etc. I just bypassed the BMS's during discharge. Then I used the BMS's during charge. My discharges were always quite shallow.

I noticed in Ping's guidelines for the V2.5 batteries he states that paralleling the batteries is okay without diodes as long as you charge them separately and only hook them together when they are both at the same voltage (normally fully charged).

EDIT:
Look down on this page for his information on paralleling without diodes:

ttp://www.pingbattery.com/usrguide/Wiring%20Guide%20V2.5.pdf

 

[dnmun]

you can actually parallel the two packs and use the one BMS to control charge and discharge of both.

the BMS is in the negative lead from the battery, so connect both packs together through the black wire that connects to the bottom of the packs. that leaves one BMS lead to connect to the controller, to the black wire of the controller. connect both red wires from the top of the pack together and they connect to the red wire from the controller.

then you have to tie each of the cells together through the sense wires. solder the sense wire of the dependent pack to the identical sense wire of the main pack, so each cell is connected together in parallel through the sense wires. then when charging the packs will balance and when discharging the BMS will shut down when the Low Voltage Cutoff voltage is reached on any of the paired cells. or when the shunt current is exceeded.


what blew up on the BMS and do you have pictures or who manufactured the packs?

 

[SamTexas]

Make sure that the BMS is NOT capacity dependent (most are). This has to do with the discharge rate. If the BMS IS capacity dependent, then the discharge rate remains the same while the capacity has doubled. This may be ok or not depending on how much draw you want from the new paralleled battery.

 

[neptronix]

Lol, what? where are you getting that from Sam? capacity varies with discharge rate, temperature, etc. There's no way that BMSes are capacity based, they go off voltage.

( he's blocked me since i've proved him wrong too many times and he got really mad over it, but this is for the rest of the people on the forum. I don't want you falling for some misinformation. )

 

[dnmun]

he means that the shunt limiting current is that of the 10Ah pack and not what a 20Ah pack could deliver. i guess the implied deficit is the lack of the max current the two packs are capable of.

 

[neptronix]

You are overestimating Sam's intelligence

As per the original poster.. you need to parallel the batteries at the cell level for the BMS to work properly. Otherwise you risk the non-BMS pack having disbalance and thus overcharge or overdischarge cells by just using the charge/discharge leads.

That would be the proper way to do it.
Thus, there is no such thing as a before or after the BMS connection, really..

 

[Hugues]

for such an important topic, a clear and precise diagram would be welcomed i guess.

i was thinking of paralleling the other day, but i need to be very sure of what I'm doing.

At the moment, i'm not.

thanks to who ever has the knowledge and time.

 

[neptronix]

Well i hope this explains things to you somewhat...

By paralleling the charge/discharge wires, you basically link the batteries up so that the voltage of the entire pack remains the same.

The problem with this is that when the BMS does it's thing, it's not going to balance the secondary pack. The secondary pack can get out of balance pretty bad over time..

Now when you parallel the balance wires as well, you are essentially linking the voltages of each cell to each other. By doing this, the voltages remain the same during charge/discharge/balancing.

So if you took two 10ah packs and linked them up via the charge/discharge wires AND the balance wires, it would be exactly equivalent to having a 20ah pack for all reasons...

 

[SamTexas]

he means that the shunt limiting current is that of the 10Ah pack and not what a 20Ah pack could deliver. i guess the implied deficit is the lack of the max current the two packs are capable of.

Yes, that's what I meant. To my knowledge, the only BMS that's NOT capacity limited is the "Fechter/Goodrum/Hecker bms" where the discharge does not go thru the circuit.

 

[dogman dan]

Language problem. Capacity of current what he meant. I was thinking capacity in the form of watthours, like you meant your bms was a wattmeter that shut off at a given wh.

Definetly would want a bms with enough current to supply your needs, so you might need to buy a higher current bms if you need that much current after paralelling the batteries. Follow Dnmum's plan, and then connect a bms with the ability to supply the current you need. Happy day if the remaining bms is big enough.

Bet Sam's got me blocked too.

 

[dnmun]

another posting where the initial poster evaporates.

@hugues, i gave pretty explicit directions, that will allow you to combine two lifepo4 together in parallel and have them maintain balance and protected against low voltage failures. lemme know what else is needed to make it clear.

 

[Hugues]

@Dnmun

i made a drawing below, not connected, then connected, is that what you mean ? I did not connect all bms wires not to mess up the drawing

 

[Distortion9936]

Thanks guys for all of the info. The battery is from ecobikes, and am going to try to connect them to one bms. You guys have been a lot of help thank you

 

[SamTexas]

Language problem. Capacity of current what he meant...

What makes you think it's a language problem? To protect the battery from exceeding its C rate, the BMS designer must impose a maximum draw. How does he do that? Based on the capacity of the battery, correct? Does that not make the BMS capacity dependent?

 

[neptronix]

Language problem. Capacity of current what he meant...

What makes you think it's a language problem? To protect the battery from exceeding its C rate, the BMS designer must impose a maximum draw. How does he do that? Based on the capacity of the battery, correct? Does that not make the BMS capacity dependent?

( i told you guys you were giving Sam too much credit!! )

A BMS pretty much has nothing to do with the capacity.. it looks at voltages to determine where the cells are since amp hours are a moving target.. Each BMS has a limit of how much power it can transfer and interrupt, but it's not there to limit the current coming out of the battery. If you exceed the current, it's basically like a really expensive fuse and it blows..

So the original poster could use a trillion amp hours on the same BMS. But the amount of constant current / spike current that the BMS can transfer without letting out it's magic smoke never changes.

 

[Rassy]

A BMS pretty much has nothing to do with the capacity.. it looks at voltages to determine where the cells are since amp hours are a moving target.. Each BMS has a limit of how much power it can transfer and interrupt, but it's not there to limit the current coming out of the battery. If you exceed the current, it's basically like a really expensive fuse and it blows..

So the original poster could use a trillion amp hours on the same BMS. But the amount of constant current / spike current that the BMS can transfer without letting out it's magic smoke never changes.

Neptonix, the above statements confuse me. I am sure there are different BMS's that let the magic smoke out if the controller (motor?) ask for too much current (amperage?). But the BMS's used by Ping simpley shuts down at a pre-determined amperage in order to protect the battery. This pre-determined amperage differs for the different sized (AH) batteries. After a shut down you simply have to turn everything off, turn it back on, and you are good to go again.

On an early LiFePO4 battery I tried to use it with a system that tried to pull over 40 amps, but the BMS had been set to shut down at 35 amps. After repeated shut downs I followed guidance from the forum and beefed up a resister within the BMS, after which the battery system would deliver the 41 or 42 amps requested by the controller. Of course that came at the expense of the battery, since it was no longer current protected as intended.

 

[Hugues]

that would make sense to me too, if i believe what's written on my pack

overdischarge protection 83 A