Charging parallel battery packs

I'm working on a project that uses a 24V 25A controller. I've tentatively decided that the best option for a battery, in terms of packaging, cost, and ease of use, is two identical BMS protected bottle batteries attached in parallel. Together they'll have as much energy and power as the bike requires.

My question is in regard to charging these batteries. Clearly the safe and sound way to charge them would be to dismount both of them and charge them separately on separate chargers. But I'm curious as to what is practicable without dismounting and separating the batteries.

If the batteries are left in parallel, and each one is plugged into its own charger, will that work or will the chargers interfere with each other? Does it make more sense to use a single higher current charger with two sets of parallel leads for the two batteries?

Is there some fundamental problem I'm not seeing with charging two BMS equipped batteries that are wired in parallel?

Probably a good idea to keep them isolated using a Schottky or ideal diode. I use two lithium packs, both with BMS, through a Schottkey on one trike (originally from ebikes.ca), and one using an ideal diode (from Sphere member Tiberius in England)..Nary a problem in 18000 miles and no need to worry about one battery charging the other...makes parallel operations foolproof, even with a better fool like me, as occasionally one of the packs is forgotten to be charged. I use two chargers, charging them separately, at the same time as they are connected to the "Y" type 2 for 1 diode.

I run seven of those 36v "hoverboard" packs. They are 10s2p packs and each has its own BMS. My understanding is that the BMS on these packs is not a balancing BMS, but is there for protection. I keep them all paralleled together when they power the bike and when I charge them. I spot check individual packs from time to time and after about 1500 miles of use, they have all stayed in nearly perfect balance with each other.

Chalo said:

If the batteries are left in parallel, and each one is plugged into its own charger, will that work or will the chargers interfere with each other?

Click to expand...

If they have separate charge and discharge ports, then if you don't disconnect the discharge ports, then when one BMS tries to shut off charging current to it's pack because of a full cell, it can't--it's still getting charged thru the discharge port from the other pack and charger.

Does it make more sense to use a single higher current charger with two sets of parallel leads for the two batteries?

Click to expand...

This is functionally the same as the other, if you have separate charge and discharge ports, if you don't disconnect the discharge ports from each other.

You could just as well wire the charge ports in parallel on the vehicle itself...except that in the case of separate charge and discharge ports, if one battery needs to shut off it's output due to LVC on a cell, it can't really, because power can still flow out the charge port to the other pack that's still active.



If you don't want to disconnect anything (via switches or connectors), you can use diodes on the discharge ports as noted above, as long as you don't use regen on the system. If you have to use regen, you can't use diodes. You're then stuck with disconnecting the discharge ports from each other in some fashion.

The easiest way for that is to just put a plug on one pack that's easy to connect/disconnect. But you want to make sure it *has* to be disconnected in order to charge the pack.

So you can use a charging port connector that is actually enough pins to also run a jumper wire from just one wire of the discharge port of one pack to the controller (two pins). So, four pins total if you're using one charger for both packs, or six pins if independent chargers.

Then you unplug the jumper connector used for discharge, and connect the charge connector for charge (which does not have the jumper wire in it).



If the charge and discharge ports are the same, then it doesn't matter whether the packs are paralleled or not, or if you use separate chargers or not, the BMS in each will disconnect each pack as needed from charge or discharge. (assuming they're designed correctly)

amberwolf said:

If they have separate charge and discharge ports, then if you don't disconnect the discharge ports, then when one BMS tries to shut off charging current to it's pack because of a full cell, it can't--it's still getting charged thru the discharge port from the other pack and charger.

Click to expand...

I'm guessing this would be more of an issue with packs that are either badly out of balance, different in capacity or chemistry, or at different points in their cycle life.

I would have no problem remembering to unplug the controller cable before attaching the charger cable, but for my friend I would like the thing to be as fault tolerant as possible-- and part of that is reducing the number of things to plug in and unplug.

A major misgiving I have with parallel batteries that must be unplugged to charge every time is that when only one is plugged in, the plug for the other becomes hazardous. The bottle batteries I have in mind use a naked 4-prong plug for the load side. With one plugged and the other unplugged, the prongs could easily short on anything conductive.

All the same, having diode protection at the cost of both wasting limited power and incorporating a component that can get injuriously hot doesn't appeal either.

Hey Chalo, this is something that we deal with on a very regular basis. Amberwolf is spot on in that the risk of charging via one battery while the packs are hooked up in parallel is that the battery that is getting charged from the discharge port typically (depends on particular BMS) won't have BMS overcharge cutoff. For instance if this battery was to have a parallel cell group short internally, then all the remaining cells could get significantly overcharged and the BMS would have no way to stop that.

In practice though, this is such a rare and unlikely event that we almost always recommend accepting the tiny risk and just leaving batteries parallel connected both for charging and discharging and treating the parallel batteries like a single pack. This is largely idiot proof for usage and it ensures that the batteries are always at the same state of charge level. When you start regularly unplugging and replugging the parallel joint and charging independently, you increase the chance of hooking up a fully charged battery to a flat battery at some point. That can result in hazardously fast 'C' rate of charging on the flat pack.

If you can find water bottle batteries that have a BMS with shared charge and discharge leads, then there is no concern at all. We just don't see too many BMS boards made this way since it has a greater mosfet expense.

Yeah, it is only a concern if they are unbalanced, for whatever reason.

But I try to think of the long view, later down the line, when that is probably going to happen, and the potential results of that (cells overcharging, possibly significantly, because of the incoming discharge-connector current that is potentially high enough that the balance resistors can't bleed off the overcharge fast enough to prevent the voltage from reaching some catastrophically high level on some already-full cells, since the BMS can't cut the current entirely like it normally would.

For myself, I don't even run a BMS, so it doesn't matter. But when I'm thinking of something for someone that I either don't know their level of expertise, or I know they don't know about (or care about) how these black boxes all work together, I like to work out "foolproof" (yeah, I know) methods wherever possible.

As Justin points out, it's not all that likely to be a catastrophic type of failure. But it *could* happen.



Regarding the plugging/unplugging, that's why I suggested using a different connector method for the charger, where it has a jumper in the one you unplug to charge (which is just a connector, no wires coming out of it), and no jumper in the one you plug in for charging.

The one you plug in only for running the bike (or whatever it is) can also be a theft deterrent key.

This method leaves no dangling battery wires, no worries about shorting anything, etc, as long as the charging connector used is keyed so it can only be plugged in one way, and cant' make a connection any other way.

If you want to ensure the packs can never be in different charge states (and use it for theft deterrent), then use a jumper for *each* pack that's paralleled, so they're both disconnected from the system with the jumper connector removed. It requires two more pins in the connector to do this, however, so six pins for one common charger, or eight pins for independent chargers (or if you want to be sure the charging ports are not connected while discharging.

Personally, I'd use the 8pin method for most of the people I know. The charger's connector (for just one charger) would have a jumper between the two positive charging pins, and another between the two negatives, and no connection or contacts in the discharge jumper pins.



Andersons are one way to do that, but there's plenty of connectors, including waterproof ones, that could be used for this.

I run 2 chargers in parallel quite often...anytime I want a faster charge. They don't seem to interfere with each other in any way, so the charger side of the equation should be a non-issue leaving only the battery side that Justin and AW mentioned.

chalo,I have been doing what you are asking about every day now for about a year or so.two 10ah 36v bottle batteries connected into a parallel discharge cable being charged with two 2amp chargers connected to the charge ports.never have had any unusual events and I use the CA3 for charge monitoring.I always seem to have equally charged packs and both chargers always seem to work independently of each other.
anyway thats been my experience.

Hi Folks,

I'm really glad that I found this topic / thread! I do have a question for the experts here - to confirm my understanding

I'm using two battery packs set up in parallel: 1) 16S8P 2) 16S10P

1. What happens when the 2 packs have different SOC (State of Charge) and you connect them in parallel?
My understanding is that the higher SOC pack will start charging the lower SOC pack, until they both have equal SOC?

2. When the difference in SOC between the 2 packs is low - then this should be fine (as the higher SOC pack would be slowly charging the lower SOC pack? The risk is when the difference in SOC between the 2 packs is high - as the higher SOC pack would be charging at a high rate to the lower SOC pack?

3. I have been charging the 2 packs separately all this time via the charging ports because of my understanding above (right or wrong) - and the 2 packs will not always have the same SOC all the time (we're talking 0.2 V difference at max) - before connecting both packs in parallel. Will this be an issue?

Thx a million as always!!!

In my experience 0.2v is too much if it's at the cell level, but if it's the global packs' voltages difference, it should be OK

I got advice on this subject on E.S. about 4 or 5 years ago. I run two 16S1P LiFeP04 in parallel. I charge them separately. Each pack has separate charge and discharge ports and each has it's own BMS. Each battery uses a key switch to connect to each other and the controller. After charging both separately and letting them sit overnight, I always turn on each pack individually and check pack Voltage on the CA3 before connecting them in Parallel. They are usually 0 - .1V apart (pack Volts). My charging is always done near the end of the day after a long ride so I have lots of time for the separate charge and for each to come to resting Voltage overnight.

If I needed to charge quickly for some reason, I would not hesitate to charge both packs in parallel up to about 85% charge. I would then finish charging each pack individually. This method would speed things up considerably. I see no issues doing this all the time.

Thanks for sharing!

That’s good advice. I guess the conclusion is as long as you don’t charge until full (which I never do anyways) charging the parallelled pack should impose a tiny risk