Battery pack made from several small modem packs.

I have a huge lot of modem battery pack consisting of (3) 18650 cells in each of the pack. Each modem battery pack is rated at 10.8v @ 2600mAh as each of the cells inside is rated at 3.6v, 2600mAh and they are connected in series, hence you get the 10.8v rating. The packs comes with built in BMS in each of them.

I am wondering if it is possible to wires these packs in series/parallel connections to be able to increase the voltage and the current rating to bring the battery to 54v for a 48v ebike with a much higher amps than what one pack would offer?
I have a total of 84 packs so after the series/parallel connections i would end up with a total pack of 54v with an estimated 43amps overall.

My question is, with this proposal each pack would already contain their own built in BMS, so therefore would it be possible to charge the overall pack with a 48v ebike Li-ion charger rated at 54.6v without causing any issues to each of built in BMS bearing in mind that the entire pack is made up of several smaller packs of 10.8v that are connected in series/parallel connections to get an increase in voltage and current.

Would there be any balancing issues or any other issue(s) that may or may not arise from this method?

Next question is if the charging is possible would each of the packs auto regulate the cells inside of them so they don't overdischarge during charging stage of the overall pack and they don't underdischarge during discharging stage when using the ebike?

Would i be able to can pull large amount of current from the overall pack for example 30-35amps without any adverse effect on the cells or each of the built in BMS?

Thanks in advance for the support.
Why i want this to work would be because it would be much easier and less time consuming to built a big battery pack with this method instead of having to scrapping out the cells out of each pack which would be a total of 84 modem packs then have to rewire them in series/parallel connections and then to add a 48v BMS to the overall pack. It would be a very time consuming task.

so what do you think guys would my method work perfectly without causing any issues?
what are your recommendations and suggestions or input on this regards?

Best approach would be to strip down to individual cells, leaving tabs attached. Test cells, then build 4 or 5 p sections which would then use a conventional bms.

The testing will tell what quality cells you have, but often the c rate is much less than the top quality cells used to make packs able to handle higher amps.

Not saying only way, but with scavenged cells, best, is always going to involve much testing and weeding out of weak cells.

If you choose to go no bms, then you could leave them in series, without the existing bms's, and in fact connect two packs in series for 6s, then parallel them to a size you can work with, and then can then charge with typical RC type chargers. Then the 6s modules can be series connected to run the bike, 12s, 18s, or even 24s. 12s easiest, and will work with typical 48v controllers.

what size pack are you building? 12s10p?

what is the brand of the 18650 cell ? Samsung 26f?

someone can help you better if you provide more information

These are used?

If so need to test each cell before using, best to atomize the triplets

modem packs are used, right? not new?
I would check every pack for weak cells
of course leave pieces of tabs welded to cells maybe like 1cm for future soldering.
packs are used and recycled for the reason. there is always weak cell or two in such pack.

fabieville said:

My question is, with this proposal each pack would already contain their own built in BMS, so therefore would it be possible to charge the overall pack with a 48v ebike Li-ion charger rated at 54.6v without causing any issues to each of built in BMS bearing in mind that the entire pack is made up of several smaller packs of 10.8v that are connected in series/parallel connections to get an increase in voltage and current.

Click to expand...

Sort of.

First off you can't do 54.6 volts. Since you have 3s packs your final battery has to be a multiple of 3. 12S would work as would 15S. You'd need a 50.4V or a 63 volt charger for them.

Secondly you will need a minimal BMS monitoring the inter-cell points. For a 12S pack, for example, you'd only need to monitor 3 points. A simple comparator circuit would likely suffice. And since you only need to monitor 3 points (really 4 with the B+ side) it would be a lot easier.

To balance you might just remove all the packs once in a while and charge them all individually. You could balance via the BMS but that's more work. You could also charge them as 4 separate 10.8V packs and balance that way. But you would have to do balancing at some point.

the cells are LG brand LGabb41865 these are made for storage and not for high drain.

However suppose i build a large battery pack with a lot of these in parallel would i be able to pull at least 25amps constantly from them without causing any issues????
I am planning to use 130 of them to build a 48v ebike battery packs. with (13) strings of (10) cells group together in parallel with a total amounting to 48v @ 26amp.

fabieville said:

the cells are LG brand LGabb41865 these are made for storage and not for high drain.

However suppose i build a large battery pack with a lot of these in parallel would i be able to pull at least 25amps constantly from them without causing any issues????
I am planning to use 130 of them to build a 48v ebike battery packs. with (13) strings of (10) cells group together in parallel with a total amounting to 48v @ 26amp.

Click to expand...

That is a very big and heavy pack.

But 1C rate should be OK, don't go much over

but they still might not last long

depends on how long and hard-used they were by the previous owner - what "modems"?

and only **if** you only use the healthy ones

The answer is yes and no.
I had, at one time, a whole pallet of two cell packs each with their own 8.4 volt BMS. I separated, tested and built several 12 S 12 P packs with the cells which worked fine but it was a lot of work. I decided it would be much easier and quicker to just put a bunch of the two cell packs together in series parallel and it worked. Unfortunately, the first time I let the pack get too low and the weakest sub pack shut off it turned into a cascading failure of all the other sub packs. The BMS's in all of the other packs shut down and none of them recovered. I had to go back to the original plan of harvesting the cells (they were all still good) and welding them into a conventional pack.
So yes it will probably work as long as you pay close attention and never go too low, but it is probably a beter idea to harvest the cells and build a pack with just one BMS from the start.

since you have the cell number, you can probably find the spec sheet for it, or realworld testing someone has done on it.

so to know how big a pack you need, you look at the continuous amp draw the cells were rated for, then derate that by some amount (50% is probably safe, with old cells).

then divide the amps you need at the controller by the above number. that gives you the minimum number of parallel cells you must have to prevent overloading the cells and heating them up too much / aging them faster / having too much voltage sag.

if you have enough cells to double that, it's better, because it gives you overhead room for aging cells, as they get less capable with time and usage, so you're still not pushing them hard.

my ebike is limited to 30amp draw.

SO if i use 12 of these in parallel with 13 strings in total for a 48v ebike would it be able to power the ebike with no adverse effect on the cells seeing that these are non high amp discharge cells?
I am just thinking that if i can pull 2.6amp from one of the cells constantly. then 12 of them in parallel mode i would surpass the 30amp limit of my ebike.
what do you think or suggest?

if you want ot know if they'll work, you *have* to know what the cells are capable of (or rather, what they *were* capable of when they were new).

your basic options are:

-- simply look up the cells based on the numbers you have on them (this is free)

-- spend more money, and more time, and buy or build testing equipment, and then do the necessary testing at different rates

-- build a pack however you feel like and then test it by riding it, potentially damaging the cells if they can't handle what you're after, and then rebuilding it bigger if it can't, spending more time and money, etc.

personally, i'd look up the information first, since it's free, except for the minimal time it will take. then do the math, and see if it can even come close to working before starting on the project.

i'd assume that since they're eitehr old or well-used or both (you have no way to know), that all specs regarding capacity or current delivery ability or charge rate should be cut in half, to be safe and try to ensure the cells all react more or less the same (thus staying balanced better).

then i would test a handful of randomly chosen cells to be sure they can take the load you expect without severe voltage sag (and if possble test for capacity, if you already have the stuff to do it). the voltage sag / load test can be done with simple resistors or light bulbs in parallel, etc., if you have to.




but it's up to you to decide whether to proceed on hopes and guesses or to start with actual data.

here you go, first hit on google
https://secondlifestorage.com/showthread.php?tid=1789
spec sheet
http://www.batteryspace.com/prod-specs/5457_B4.pdf


below is the critical bits from the pdf, and the really critical bits are bold italic underline

2.1 Capacity Std. charge / discharge
Nominal 2,600mAh (Cnom)
Minimum 2,500mAh (Cmin)

2.2 Nominal Voltage Average 3.6V

2.3 Standard Charge
(Refer to 4.1.1)
Constant current 0.5C(1250mA)
Constant voltage 4.2V
End current(Cut off) 50mA


2.4 Max. Charge Voltage 4.2V

2.5 Max. Charge Current 1.0C(2500mA)

2.6 Standard Discharge
(Refer to 4.1.2)
Constant current 0.2C(500mA)
End voltage(Cut off) 2.75V

2.7 Max. Discharge Current
-20 ~ 5℃ 0.5C(1250mA)
5 ~ 45℃ 2.0C(5000mA)
45 ~ 60℃ 1.5C(3750mA)

Click to expand...

so they're only meant for 0.2c discharge, or 500ma (0.5a) so if you need 30a out of them, you should really use a minimum of 60 of them in parallel, meaning for your 48v pack, that'd be 13s 60p. gonna be a little on the large side.

when they were new, they were capable of up to 5a discharge as long as you kept them within 5degrees c to 45 degrees c, but they're not new now, and that's a max current. it doesn't say how long they can maintain that current, but the max is not usually meant for continuous.

it does say that a fast discharge is at 0.5c and after that the cells should rest for 20 minutes, so i imagine at 2c, or 5a discharge, they probably get pretty hot, especially inside a pack that's inside a battery box or case.


i missed the dcir if it's in there, but the ac impedance is 70mohm. i couldn't find anything that shows a direct relationship between dcir and aci, because they measure different things, but the few things i found that show both a dcir and aci for the same cell show dcir being antyhing from 1.2x to 2x the aci.

if that holds on this cell, too, then it's dcir could be 100-140mohm, which is incredibly terrible.

What if i only draw max 30amp for several secs but keep my constant current to 15amp or below would building the pack with a 9cells in parallel or 12cells in parallel be ok?
Also suppose I keep a fan in the battery pack to keep them constantly cool would that be a good idea?

Or better yet suppose I downgrade to 24v with 15 cells in parallel with 7s in series and still limit my draw to around 15 to 20amps constantly would that be ideal?

These really are the wrong cells. But parallel enough of them and you can do something with them.

Really suggest you break them up into individual cells, with some tab attached if at all possible.

Once you have the suitcase size thing built, use voltage sag under load to tell you what your real world max amps will be. Anything over 5 or 6v sag under load is pretty unusable, unless you lower the amps enough to get sag to be less.

Cell spec is nice, if cells meet spec in the first place. Then you got all those zillion connections, each adding more resistance. Performance of the assembly can be very much less than your already pretty weak spec.

Gonna be a lot of work, likely ending in a disappointing result. But it should run a bike, just not at much amps.

Is there a way to limit the max draw amps to 15amp?

A fuse or CB of course.

But ideally your controller's job.

CycleAnalyst v3 can do it.

fabieville said:

What if i only draw max 30amp for several secs but keep my constant current to 15amp or below would building the pack with a 9cells in parallel or 12cells in parallel be ok?
Also suppose I keep a fan in the battery pack to keep them constantly cool would that be a good idea?

Or better yet suppose I downgrade to 24v with 15 cells in parallel with 7s in series and still limit my draw to around 15 to 20amps constantly would that be ideal?

Click to expand...

Figure out C-rate vs your # in parallel.

Voltage is set by the motor + controller, higher is better but really just determines speed.

Amps is the critical part for power, especially torque at low speeds, that is the stressful part.

Fan should not be needed, design the pack / case to shed heat without it, expose to airflow.

You may get more best response if this was in the battery section. Plus look up DrAngle here on E.S. lots of info. Do you have a spot welder ?

fabieville said:

What if i only draw max 30amp for several secs but keep my constant current to 15amp or below would building the pack with a 9cells in parallel or 12cells in parallel be ok?
Also suppose I keep a fan in the battery pack to keep them constantly cool would that be a good idea?

Or better yet suppose I downgrade to 24v with 15 cells in parallel with 7s in series and still limit my draw to around 15 to 20amps constantly would that be ideal?

Click to expand...

you should probably try doing the math for stuff. if you can't (or won't accept the results the math gives), you can do whatever builds you want with the batteries, but i don't think you'll like the results.

i'm assumign you're not going to test the cells before you build them into a pack, so you should at least assume the cells aren't anywhere near as good as they used to be, and they were never very good at all. (they weren't designed for the purpose you want to use them for).


so, let's try this again.

they're 0.2c cells.

no..actually, they *were* 0.2c cells when they were *new*.

i'd guess they're really 0.1c cells now. (you can test them, per the info provided in that data sheet, and see what results you get--that will help you see what they are really capable of *now*--they could actually be worse than I'm guessing). they could be better than that...but i doubt it. it's much safer to assume the cells are not very good, and then get lucky and find they perform better, than to assume they're good, and then find out the hard way they won't.

that means that whatever ah you have, you get to draw a current that is one tenth of that, if you want to keep them happy.

so if you have 15 in parallel, that's 15p x 2.5ah = 37.5ah. so you could safely draw 3.75a from that pack continuously.

if they were *new* cells, you could safely draw 7.5a continuously.


9p means 9 x 2.5ah = 22.5ah. 22.5 * 0.1 = 2.25a. so drawing 15a continouously from it is about six times as much as it cna probably realistically handle. it's three times what the cells were desigend for way back when they were new.

12p means 12 x 2.5ah = 30ah. 30 * 0.1 = 3a. and so on.




regarding "downgrading" voltage: if you decrease the voltage, then to get the same watts (same motor power) you have to increase the amps proportionally, so it will be *harder* on the cells at a lower voltage.



to limit current draw to 15a on a 30a controller, you change the controller settings for it's current limit.

keep in mind that 15a (assuming 15p) is still twice as much as those cells were meant to draw way back when they were new, probably a decade ago. and they're not new anymore, so it's probably like four times what they would like now.

if it is not programmable, then you either have to modify the shunt or the electronics that read the shunt. if there are two shunts, you could remove one, and that would cut the controller current limit in half, from whatever it is now. if ther'es three, removing one cuts it by one-third. etc.


alternately you can get a much much smaller controller.


but if you need a certain amount of power from the motor to do what you're after while riding, then the answer isn't crippling the system like that, it's giving the system a battery that can provide the power needed to do what you're after.

how much power do you actualy need?

what kind of riding do you do? do you do a lot of stops and starts? or more continous at one speed?

what terrain do you ride on? hills? steep? long? flat roads? gravel? sand? ??

how much do you and the bike and any cargo/etc you'll carry weigh?

what speeds do you want to go?

what kind of bike is it?

etc.

with that infomration we can help you figure out the kind of power you actually need to do those things, which will help us help you figure out what you need for controller and battery (and motor).

What I was trying to say as well. they are the wrong cells when new, and now they are at least aged, if not used, if not used up.

Go for it if you want, but be expecting spectacularly poor results, after a lot of hard work.

Throttle can limit amps,, don't be surprised when the system shuts itself down due to low voltage, ( sagging like mad under tiny loads) with tiniest touch of that throttle.

Others have got usable batteries, out of .5c cells, by making the pack very very large. Like 30 amp hours. I'm not sure if that's 10p or 20 p, or what with your cells. But typical 4p or 5p is not even close to going to work. I'm saying they needed 30 amp hours size, you will need more like 60 ah with cells that weak.

NOT worth it. But it could make a great pack for a solar charged lighting setup. One that will draw 2 amps or so.

I'm working on a pack with the identical cells and my cells are all testing at 2400-2800mah(bt-c3100 tester). I'm testing them at one amp discharge. Under one amp load they go from 4.20v down to 4.12-4.09v. They do become warm but not hot. My end goal is a 20s 20p 20 amp discharge for my cargo bike. My test results are quite promising. However after reading this thread I'm questioning it. The seller did claim they where never cycled. Time will tell I still have a LOT of cells to test before I start spot welding.

Plenty of cell models would not be suitable even when 100% perfect brand new Grade A cells.

The energy and power density required for a given use case should be calculated before selecting the shortlist of cells to consider

long before you actually start shopping.

im doing a lot of learning/reading on this forum

Sidotian

20s20p is a 50?ah pack



heres my question, they suggest standard charge rate is 0.5c/1250mah and max charge of 1c/2500mah

why cant you discharge this cell safely at 0.5c or 1.25a

20p is 25amps at 0.5c

Like I said in my previous post my test of these cells is looking promising. 20p in my case is working out to be around 53.5ah. I suppose if I need to I can change my final build to 13s30p for 70ish ah. I don't expect to pull crazy Amps from this pack, I have lipos for that. This battery is going on a cargo bike but I could imagine it being difficult to fit 400 cells on your average bike.