Paralleling BMS protected Li-ion packs.

[jGecko]

Edit: The underlying fundamental question is...can the charging port of an ebike battery pack be used while the battery pack is being used to ride about?

What do you think of this idea? I make a small current limiting circuit and put this current limiter on the output of 36V pack#1 (say...limit to 2 to 3 Ampa), and then use freshly charged Pack#1 to recharge 36V Pack#2 through it's normal coaxial 2Amp charging port WHILE it is in use on the ebike. This is not putting the packs in parallel...this is using one pack to charge the other while it is in use.
The problem is that I am just beginning to study ebike BMS/charging/operation circuits and don't know if they are designed to prevent charging while a battery pack is in use.
I also don't know what (if anything) REGEN designs disable while charging the battery pack with the bike in motion. So would the viability of what I'm suggesting change depending on whether or not the bike uses regen?

Or, I suppose I could even use a higher voltage battery for Pack#1 and just use a current AND voltage limiter set to 2A/42V.


Edit to add: Explaining the whole reason for thinking about this, is that the existing battery pack uses an integrated output connector (no harness) that slots into a mating connector integrated directly into the bicycle frame. This is an attempt to avoid having to cut open the original battery pack to substantially modify it in a way to allow paralleling the outputs of the packs.

[paxtana]

Charging the battery resets the bms. Meaning that if you are charging the battery while using it, if anything happens on the discharge side or internally like a short, it may not be able to provide short circuit protection.

I can not say for sure as it may vary depending on the battery, but in the event of a short it may attempt to shut down, then immediately wake up from the charging, and short out again and just keep repeating this dangerous cycle.

The only battery I would feel safe doing this on would be a Luna Wolf because you would be stripping it of some protections. You can fuse on discharge but you can't fuse at the cell level unless the battery was designed that way, so with any other battery you are losing some protections.

[SmokerX]

Q: Can I use a large pack and a small pack if they are at the same voltage?


I have read other places this is ok. But I think this is incorrect and the original posters explanation of why this is a bad idea possibly makes sense. Considering the opposing argument goes into no detail on why this is OK. Im no expert on this, have no idea what book would go into a detailed explanation and have to default to gut instinct there and I guess whoever has more words wins in my mind.

If a pack is wired internally in parallel between cells and Resistance increases with length of the wire, then the same ohm's law could argue that the battery closest to the packs output would suffer a similar effect to the "weaker pack" in the original explanation. Or vice versa(not sure considering path of least resistance). Im really not even sure if the "stronger battery will discharge through the weaker battery" Suggests an effect on the rated lifetime of charge cycles of individual cells, considering "the weaker battery" still only goes through one charge cycle. Even though it would be simultaneously being charged and discharged through the other cells in the pack and the current load.

However. "properly configured and properly integrated BMS" "operated outside the manufacturer’s specifications" "cell matching" suggests tapping into series strings and using 2 BMS "properly configured" to each individual pack at least voids the warranty. It would also appear that you'd be throwing the concept of cell matching into the wind. Whatever any of this means to someone with a better understanding of lithium pack configuring.

I think there's a work around here to isolate the 2 packs and keep the warranties without mucking about in the dark without fear of the unknown and doubt.

1. The lowest tech. solution would be to swap the two packs using XT60 connectors or similar. Like swapping the battery pack on a battery powered drill. Except possibly only swapping out the connection. Not the physical pack. When one pack runs dry. This is also the cheapest.

2. 2 Knife switches used to isolate the currents of packs. There's no point in that unless you want to make some steam punk art project. It would work but several drawbacks like exposed metal parts leading to shock and burns. or operator error closing both switches.

3. 2 double pole single throw battery disconnect switches. That would remove the exposed metal parts from 2's equation but still have the drawback of operator error of closing both switches.

4. Battery selector switch. Used in marine application to solve this problem. The only drawback beyond cost is most contain 4 positions. battery 1, 2, off and 1&2 or all. I think it's possible to open some models up and tamper with that "1&2" positions contacts to destroy that switch position and remove operator error, find a switch that doesnt contain that 4th position or a switch that doesn't have to cycle through that position to cycle between the off,1 and 2 positions and just never turn the switch to that position. The problem I see with some of these is parking unattended and the possibility of tampering throwing it into that "1&2" position. The benefits of #4 and #1 is not only the off switch but allowing 1 battery to be held in reserve. for example a buck converter to radio thrown into this. Switched to battery 1 with battery 2 being held for reserve to power the engine. The main reason marine application uses this solve.

5. Remoting into a #3 using 2 contactors (or relays), particle electron board and cell phone. Im thinking you would need to run an arduino somewhere and step the voltage down to whatever voltage these things want to control the smart circuit(5v? 12v?) then write out some software and pay a monthly cell plan. Maybe you need to remote into your battery selector from a distance. Or want everything completely sealed up in a waterproof project box. Or don't like the hassle of the manual nature of a #3 or #1 solve. The downside here for me beyond complication is the slow drain on the battery from powering relays and particle electron boards. The upside is you can GPS this battery around in case it gets stolen. It cant work without giving away it's position.

Moving back to #1 i think you could create a jumper wire. 3 XT60 connectors sitting flush to a project box. labeled 1,2 and motor or load. These 3 being female ends. Then create a jumper wire with a piece of wire and 2 male ends to jump from positions. This way you could take that jumper wire with you to prevent tampering and act as a theft deterrent since you would need this jumper wire to jump back to the motor. Without it the system would be powerless. It would also look a little nicer than just a basic #1 solve.

[calfee20]

I have done it and have not had a problem. I have a 17AH Luna GA triangle connected to a 14 AH Luna GA rectangle. I matched the voltages and plugged them into a harness that I made a year and a few thousand miles ago. They both charge at the same time to. Once they are connected together the voltage of one has to be the voltage of the other. They share the load so there is very little strain on either battery. I don't know, do what you want. All I can do is tell you what I have done. Personally I think everyone is over thinking this issue. Vector ebikes was building a pak for their bikes that used 2 different chemistries in one pack so they could get the benefits of 2 different cells. One was longer range and the other was for high amps.

[SmokerX]

Thanks for the input on paralleling different size packs working out over the long run calfee. "They both charge at the same time" my packs have 2 charge ports. 1 for each pack. This makes sense that the charge ports would jump over to the next pack over the paralled out put bridge. Over a single bms. I like the "i have no idea what im doing... This should be ok" aspect of a modified #1 solve with the jumper wire. Considering I have a handful of XT60's. Considering I have 2 chargers, 1 for each pack, hooking up both chargers should half the time of charging over a parallel wiring solve on 1 charger. I've also wired 2 chargers to double the amperage output after consulting the max charge rate, just for funsies, like a gamblers high of risking the house fire and it halfed the charge time this way. Although Im personally not overly concerned with charge times.

Comparing solves, the modified solve #1 over a parallel solve Im having trouble identifying any advantage other than seamless transferring of amps of the packs. IE, Vs. having your power run out and manually having to switch over to the next pack. This is the only real world advantage Im able to come up with. The other advantage of the modified #1 is the 2 outputs meaning the possibility of the option of simultaneously powering 2 motors. Either a motor powered trailer. Front and rear wheel motors, or run an extension cord to your friends batteryless electric bike(lol).

[calfee20]

I should have mentioned that you should do a search for "bms wiring diagram". Look at images and you will see that the charge wires are connected to the supply wires.

[Sebz]

I wrote that article a while back and had only done basic testing on paralleling different chemistry and capacities (14S 17.5ah GA with a 14S 30Q 6ah) IT was a bench test and I was measuring currents flowing between the 2 batteries. At some point after removing a 30amp load applied the current flow from the large pack to the small one was over 10amps and then the BMS of the small one fried. Since then I have not paralleled these batteries together and put a warning on this.

I do not know exactly what happened but if there was a current flowing from the large to the small after I removed the load from the circuit that means there was a voltage imbalance between the 2. Could be the BMS of the big one that tripped and then the small took the load and then the big one woke up with a higher voltage and pushed a current to the smaller one once I removed the load.

But one thing is for sure if they are connected in parallel the voltage will be equal, what you don't want is current flowing from one to the other through the unprotected discharge port.

Next time I get my test rig out, I will log current flowing between my 13.5ah against the mini 6ah pack and post the results.

[tklop]

Thanks, AZguy :-)

I'm not real certain about all the various complexities--as I said... "my ideas might be utter garbage" --and I stand by that assertion!

But I'm also not so sure that the voltage can't sag on one battery in this scenario...

What you said about resistance makes sense though... Still, I guess I just don't understand.

I'm thinking that the two batteries' cells themselves are not actually in paralell; only their BMS's are. It's not just "a bigger battery" that Sebz has created--is it? Seems to me, that it's more like two separate and complete power-systems, which are attempting to work in cooperation... And that seems to be what it is, that makes things complicated... Or at least that seems to be the concern...

Sebz said he observed current-flow between the two--which (if I'm not totally oblivious to Ohm's Law) implies that indeed imbalances in voltage can in fact exist... Were measured... Again--maybe that's my ignorance speaking--but seems to me that otherwise, Sebz could not have observed current-flow between his batteries in his test-setup...

When connecting individual battery-cells in parallel with each other to build up series-groups, the moment you've connected their tabs, the voltage-readout for all the cells in that particular series-group will instantaneously be equal. Their tabs are bonded--so of course the meter cannot possibly read anything else. But internally, between the individual cells in that group which you've just bonded together--there are still differences, and there is indeed current-flow--it's just not happening where your meter can read it. It takes a little while for those cells themselves to finish equalizing. In the mean-time, if you take them back apart again, their "different" voltages will again be readable on a meter. The cells in each series-group of course do eventually equalize, and remain that way (though they'd drift apart again if the series-groups were separated back out into individual cells again)--but from what I understand, that's also a function of the fact that they're all of equal capacity, chemistry, etc. It seems matching battery-components are happy battery components...

Now: With two different batteries--once their two BMS's outputs are bonded, their voltage-readouts too will match--instantaneously. But also--just like when bonding cells in a series-group, there would also be some current flow--no matter how closely matched--when you first connected the two batteries (assuming a non-discharge state). I expect too, because each BMS might have its own balancing program, there might be still more back-and-forth with current-flow during charging, or even when static--however minimal that might be.

But it seems to me, that once you began to discharge--because the two different batteries would have different internal-resistances, they'd be expected to discharge at different rates--and then what were subtle differences might become amplified. At any rate, any current-flow between the two batteries which resulted (as the one battery attempted to equalize the other) would have to be evidence of an actual difference in voltage... Wouldn't it? Ohm's law seems pretty inflexible on this one: You cannot have any measurable current flow, unless you also have a measurable voltage... Can you?

Gosh... I think I think too much--and ultimately, I'm afraid I'm only really showing my ignorance-level here. I'm pretty-much your run-of-the-mill basic-electricity education-level-type-guy when it comes to this stuff. My background in aerospace maintenance didn't require much more than that. In any case, I'm obviously not a battery-chemist! ;-)

Anyways... I remain intrigued by the topic--and the ins and outs of it. I've got potential to take advantage of the concept--esp. if there's a simple, safe and reliable way to get it done.

I'll certainly keep following the thread!

[AZguy]

If connected in parallel and charged in parallel their inernal voltage will remain the same but the lesser battery will have more internal resistance during discharge so during discharge less will come from it and their voltage will in effect remain the same... to a point... If the smaller capacity battery has a lower internal resistance due to being made up of lower internal resistance cells or being new then after being discharged for some time it's very likely current will flow from the larger capacity battery to the lower capacity one after the load is removed. If the load stays constant (which never happens in e-bikes) then things should work out until the load is removed and at that point there would be the case of one battery discharging into the other

I understand folks run different batteries in parallel and anecdotally don't run into trouble

Me? No way. I might consider two equal *new* batteries with cells preferably from the same lot but only with a *very* compelling need to do so... Otherwise much, much better to have a battery built to the size needed... way lower down would be to have two equal batteries with a battery management unit taking care of pulling power from them to achieve higher currents which would be the only use case I can see since getting more range it would be much better to just switch between batteries... at the bottom is just connecting two unequal batteries in parallel and expecting it to work well...

YMMV

[tklop]

Thank you, AZguy for the expanded explanation! That helps :-)

[philipgaines]

Sebz: I've read everything you've written about paralleling batteries, and it's pretty overwhelming for someone who's really handy but new to this. Let me ask a bottom-line question: If all the planets are aligned in terms of battery chemistry, voltage, amp hours, BMS priorities, connection details, etc., would I be able to 1) add a second battery to my current 48v 8.7ah battery and run the whole thing as one battery which will discharge equally between the two units and 2) charge both units at the same time? (I would be making a 3P x 13S out of 2.9ah 18650s.)

[paxtana]

Due to liability we can't say yes, there are obviously some risks. But even some of our own guys, myself included, do put batteries in parallel, sometimes even batteries that do not have the same chemistry and so on. It is important to understand the risks though.

Don't charge them when they are connected in parallel, that's a lot more risky

[philipgaines]

Thank you, paxtana. Will the second battery require its own BMS? If so, how do I choose one, price being no object? And here’s a super-naive question: Again with all the appropriate caveats, is this basically just a matter of connecting the black and red wires leading to the first battery to the same wires of the new battery? I’ve watched Micah’s videos, and I’m confident I can build the 3P13S myself; what are the really important things I have to keep in mind?

[paxtana]

The comments section of this knowledge base article is not really the place to get in-depth with info on how to build a battery. I would recommend to start a new thread asking those questions. Generally you do want a bms and there's even a parallel adapter you can use to connect up the battery as well as pics showing how this works right in this same thread

[BobMc]

I’ve read through the whole forum and can’t find an answer to my question. I have two 48V 13Ah batteries, one original on my Rad Runner. Can I wire a three way battery switch, #1, off, #2, to my two batteries? The USAF sent me to a 10 month electronics school in 1968, in the last class they said there is a new device coming out called a transistor. Things have changed. I was thinking of using a boat battery switch, are there smaller switches? Thanks.

[AZguy]

The power that a passive switch has to deal with is quite high and switches rated for it are very expensive and on the large side of things - it better be break-before-make!

I carry two batteries mounted to my bike and switch between them by very simply changing which [XT90s's] connector the controller is plugged into... easy peasy, takes seconds, looks very clean, they handle the power and are very inexpensive

IMO the only good reason to parallel batteries is if you need more current than one can provide, and even then, not for me, I'd just budget a proper battery for the application

I was working on a solid-state solution a couple of years ago and recently dusted it off because my business slowed to a crawl and I prefer to stay busy... It's non-trivial to pull off a cost effective approach... I may look into a load share solution but that's for another day...

[AZguy]

Battery 1:



Battery 2:

[calfee20]

I’ve read through the whole forum and can’t find an answer to my question. I have two 48V 13Ah batteries, one original on my Rad Runner. Can I wire a three way battery switch, #1, off, #2, to my two batteries? The USAF sent me to a 10 month electronics school in 1968, in the last class they said there is a new device coming out called a transistor. Things have changed. I was thinking of using a boat battery switch, are there smaller switches? Thanks.

I have not done this but here is a thread that touches on this. This thread goes back to 2016.

https://electricbike.com/forum/forum...eed-more-range

[philipgaines]

I just completed what so far appears to be a successful addition of a second battery in parallel. B1 is 48v 8.7ah; B2 is 38v 12h. In the first test, B2 would want to charge B1 when not under load, e.g., while coasting—which is not a problem electrically or safety-wise. But my controller would start reading charge mode, and when the balancing between the two batteries was done after a few seconds, it would revert to assist mode at the zero level, and I would have to reset the assist level to what it was before the charging happened. Unacceptable. So I installed a 30amp diode on the discharge wire of each battery to block the B2 to B1 charging. The charging stopped, so everything seems to be operating just like before the B2 addition.

However, I haven’t done a long ride yet, and something has me nervous: On my short test drive (3 miles) the range readings at the various levels did not go down at all, so I’m wondering if my set-up is confusing my controller. The range readings on full charge are still the same as they were with one battery, which doesn’t surprise me since the controller is only wired to B1 and I know that B2 is discharging under load, so it’s not clear to me what’s going on. A long ride might tell me more.

[AZguy]

I was thinking the same, connecting up a 48V to a 38V might even make the news - and as my primary flight instructor said to me, we aren't the kinds of guys that want to end up in the news

For what it's worth, those aluminum electrolytics "fragility" tends to be how they're soldered to the PCB and it looks like there's potting all over the bottom of them where they attach so not so fragile

However, as an aside I personally try to avoid using aluminum electrolytics anywhere - along with the mechanical issues of how they get attached to the PCB they don't age that well, especially cheap ones and in high temperatures, but if you need a lot of bulk capacitance at higher voltages, there's little other choice... but potting them, even if just the lower half, will make them rugged mechanically

That device looks like it may serve the or'ing diode function - hopefully with MOSFET's actively driven instead of diodes

[ncmired]

Hi AZguy, on those aluminum electrolytics what would happen if the exposed part of the can was dented?

[AZguy]

Wouldn't advise it - worst case it could short the capacitor... they're just rolled up layers of metalized dielectric... but they're not much more susceptible to damage than most any other electronic component outside the points I made

If their voltage is exceeded they tend to explode and the little scores on the top of the can is to make that not happen catastrophically (i.e. shrapnel) which is a good reason not to have the tops under a hard potting - I don't know what kind of potting that is

I've seen a lot of vehicle products with them and I've seen them shake off the PCB's when they aren't attached well which is by far the greatest issue... I don't like their temperature and aging concerns either, especially cheap ones - I usually go for tanatalums but they are too expensive for high capacitance in the higher voltages like we see

[Fred]

I’ve read through the whole forum and can’t find an answer to my question. I have two 48V 13Ah batteries, one original on my Rad Runner. Can I wire a three way battery switch, #1, off, #2, to my two batteries? The USAF sent me to a 10 month electronics school in 1968, in the last class they said there is a new device coming out called a transistor. Things have changed. I was thinking of using a boat battery switch, are there smaller switches? Thanks.

Here is one that I was considering, but decided not to use:
https://www.amazon.ca/dp/B01J31PRBS/ref=dp_cerb_1

It is rated at 12V 25amps. Obviously most of us are at 48V or more, but I believe there is a good chance it would suffice as long as your drawing less than 25 amps. I have one, but as I say, have decided to go the unplug/replug route for now. After all, I will be switching infrequently -- only if I run out of juice on a long trip.

[philipgaines]

Ar-r-rgh! Yes, I meant 48v for B2 also.