Repurposing li- ion (18650S - 2500mAh) batteries?

[peterjd]

Hi, Can someone advise on a winter project please. Some time ago I had a 36V Li-ion battery go bad due to one of the sets of 4 paralleled batteries going open circuit. Eventually I stripped the battery down by which time several of the other paralleled sets had become iffy. So I am left with 7 x 4 apparently good batteries (all seeming to be holding ~4.2V over some months). Could I build those sets up into a 7S4P 24V 10Ah battery for my friction drive motors, perhaps using a 10S4P BMS (?) with 3 of the sampling wires unused) or if necessary a new 7S4P BMS. If so I might then invest in a spot welder or would it be safe to solder? Thanks Peter

PS I'll worry about a charger for such a battery later.

 

[Nealh]

Use a correct 7s BMS because the lvc and hvc will be correct.
Omitting three series balance wires won't work as the BMS will not see the correct voltage for charging. Again use a correct 7s 2a 29.4v charger.

One needs to spot weld , soldering not only being untidy on the cell ends also allows to much extended heat to be concentrate don the cell ends.

 

[peterjd]

Use a correct 7s BMS because the lvc and hvc will be correct.
Omitting three series balance wires won't work as the BMS will not see the correct voltage for charging. Again use a correct 7s 2a 29.4v charger.

One needs to spot weld , soldering not only being untidy on the cell ends also allows to much extended heat to be concentrate don the cell ends.

Thank you for your expertise. I don't want to spend more than I have to - bearing in mind that more cells may fail on use. Do you think the 7S BMS in ebay or similar and the mini spot welder might do the trick (I would buy latter from Banggood who have been reliable). Thanks Peter

 

[StuartsProjects]

So you have battery pack where quite a few of them have failed.

If that was a decent build battery pack, you would expect the batteries to be from the same manufacturing batch.

If so, then if some have failed already, its possible the others will be close to failure also.

 

[peterjd]

An inexpensive 18650 battery tester is on its way from China. I'll post the outcome in due course.

 

[harrys]

INR-18650S with green wrappers? I've seen them on Aliexpress for .88 £ per cell.

I had one fail in 2021 in a two year old battery. I went thru the trouble of replacing the whole cell group with new cells. Over the winter, another cell failed. I think they have a mean-time-before-failure around 2.5 years.

In another user group I follow, there's a repair guy in the UK who seems to see failed batteries with these cells on a steady basis,

 

[peterjd]

Hmmm. Those are they. I'm hoping that all the ones which were dodgy in my set have now fully or partially failed. Waiting for mini welder and new (different BMS for 28V nominal) to arrive before I make up another battery. Hoping that fault was largely in the BMS. The recovered 'good' batteries are all maintaining ca 4.2V. Peter

 

[peterjd]

At long last received items from China including battery tester. With one of the 'good' recovered batteries got a capacity reading of 2260mAh against rated 2500 mAh. Hoping the others are the same but its going to be a long job - test took several hours to check capacity (down to 3.5V from 4.2). Peter

 

[Nealh]

2260mah is pretty good down to only 3.5v, try 3.2 0r 3.3v which is often safe LVC level to see what MAH you get.

 

[peterjd]

The instructions recommend 3.5V so I think I will stick with that - I don't want to lose any more of them if I can help it. But thank you for the comment.

Since I have several 36V pedelec batteries now I thought it might be more useful to build several 14V sets, ideally with a Bosch charger compatible BMS. Does anyone know of a reasonably priced source of such a BMS - even recovered ie s/h, genuine Bosch ones, seem to cost ~£12 on ebay.

Peter

 

[Nealh]

If the sellers only recommend 3.5v as lvc then they don't have much faith in the product, 3.5v is well above that which a decent 18650 will operate at.
But then again as it is a crappy no name brand one can't expect to much.

 

[peterjd]

If the sellers only recommend 3.5v as lvc then they don't have much faith in the product, 3.5v is well above that which a decent 18650 will operate at.
But then again as it is a crappy no name brand one can't expect to much.

Probably but I want to see how much value I can recover from them without spending too much other than my time.

So have now tested most of the separate cells and it looks like I will have 7 sets of say 4p with about the right capacity and balance. I have sufficient other batteries to power my two hub drive pedelecs so it would be very useful to have a flexible back up battery for lighting (12V or existing small 100W 240V ac inverter) or to power the friction drive motors (24 -36V nominal).

My thinking at the moment is to weld up balanced sets of 4 parallel cells, and link the sets in series with thick wire via 2 terminal blocks located externally on the original discarded battery case so that I can wire up appropriately for different requirements, and monitor discharge.-

Since I don't have a suitable 24(28V) li-ion charger I was thinking of using the TP4056 single cell charger (can they be paralleled for more current?) to charge up the individual 4P packs? All time consuming of course but that's not really an issue for me.

There is also the issue of avoiding discharging any 4P pack beyond say 3.4V? That would require careful manual monitoring of all the parallel cell packs or a BMS (which I am trying to avoid)? Or is there any other simple way of achieving a LVC?

Thanks for any comments on feasibility. Peter

 

[Bikes4two]

The instructions recommend 3.5V so I think I will stick with that - I don't want to lose any more of them if I can help it. But thank you for the comment.

'The instructions' - what were this instructions from - your tester/charger or something else?

Seeing as you're building a battery pack with cells from a variety of sources (so the cells will have different discharge characteristics), using a BMS to prevent any one of your '4P' pack's from going too low, is even more important to consider.

Every time you look at a pack's balance with a volt meter means you're ever more likely to disturb something else. Below is a BMS from AliExpress at just over £10 which will deal with your LVC question and at the same time prevent over charging and excessive discharging - what's to loose? (Well, lots of things if your experiments go wrong).

 

[peterjd]

Thank you for that comment. I guess I should follow your advice and fit a BMS. After my previous experience where I wondered whether the original BMS contributed to the loss of more cells than the original failure in one paralled set of cells. I am a bit mistrustful of cheap BMS circuits and not sufficiently knowledgeable of BMS design to distinguish good and bad designs.

I might fit a BMS and monitor myself the voltages on the individual paralleled cells.

Thanks Peter

 

[Bikes4two]

...... After my previous experience where I wondered whether the original BMS contributed to the loss of more cells than the original failure in one paralled set of cells. I am a bit mistrustful of cheap BMS circuits and not sufficiently knowledgeable of BMS design to distinguish good and bad designs ..........

I entirely understand your feelings towards cheap BMS and the problems they might cause - I had a BMS failure on my 10Ah bottle battery - it was a PITA in that I'd be out riding and .... no power. Sometimes a power off/on would work and sometimes not.

The one I linked to was one recommend by someone on here (@Nealh possibly?) and I've just fitted one in to my first battery build, so time will tell!

Some BMS have a bluetooth facility so you can via a smartphone, see what's going on inside the battery, but they are relatively costly as far as I can tell.

There's a particular one used in electric skate boards (aka ESKs) by Daly, but even they seem to be problematic if postings about their problems are anything to go by.

 

[Nealh]

I did post regarding the ' BMS' branded one but can't recall recommending it as I have never tried it out until yesterday and it has no balance features. I know I mentioned it as it was of a particularly small footprint ideal for my 1l water bottle battery,
For sure I can't recommend the Red Daly BMS, my water bottle suffered it's second Daly failure this week so have now fitted the 'BMS' branded one, utilising the same port charge discharge and replacing the charge port under the screw on cap with a on/off button switch to turn the new BMS on and off.

 

[Bikes4two]

I did post regarding the ' BMS' branded one but can't recall recommending it as I have never tried it out until yesterday and it has no balance features.

Hi Neal,
...has no balance feature... that caught me by suprise as I didn't know that but on looking further... well well!

As said, I've just used that particular BMS in my first battery build but now I'm thinking I should look for a BMS that does balance - I like the common port charge/discharge feature but that's not essential, but the small footprint is neat.

I need to look around for another BMS now - knowing that you've spent a lot of time researching this sort of thing, do you have any pointers to sources of alternative BMS (or how to cope with the non-balancing 'feature' of the BeMuchSafer one)?



Cheers, B4t

 

[Nealh]

BeMuchSafer do sell a BMS with Balance but it is not same port and has a larger footprint.
The without Balance same port models still have sense wires to monitor HVC (4.22v) & LVC (3.0v) for each cell group during the charge /discharge process. If the cells are very good evenly matched cells then Balance shouldn't be a big issue.

The BMS I have seen issues with have all had the balance feature, one 48v battery with the Panny PF cells died because of one Balance BMS , likely because a bleed mosfet failed. One week it was fine gave it a top up charge and then 3 weeks later no good.
One small component on the balance side goes awol and the whole battery can be lost or it's use curatiled until the bms is replaced.

The option is to use a same port BMS without balance and then also add a balance feature only board to the battery as well, this acts independently of the BMS and monitors cells all the time. Ultimately these boards cost more then ( approx. £15 - £20) a BMS putting up the over all cost, they appear or say balance is maintained to within 0.03v and contantly monitor. This means low current is active all the time though so possible could eqaully drain a battery is left standing a while.

When one thinks a cheap Balance BMS is about £10 or so , one has to wonder how good the Balance feature actually is with reliabilty.

 

[Nealh]

My bottle battery I made using the LG hg2 cells has been great for a lighter weight battery giving still 25 - 30 miles range from it's 6ah footprint circa 5.2a theorectical footprint with the current I use in PAS1.
Twice now in 6 months I have had to replace the Daly so I shall no longer use or can recommend the brand /type of BMS. I see also on ES the smae issues being reported their on poor reliability.

I have one on my Moli p42a pack so am keeping an eye on it and when/if it fails have a smart BMS to try out on it with the added uart to adjust parameters.
The Daly doesn't overly appear to drain a single cell to lvc, but a/some components appear to suddenly fail . It is suggested on ES it is due to heat or an overload and suddenly as on Sunday I rode to work fine and disconnected my battery as I don't leave it on the bike, re-attaching it to the bike and the bike was dead.

At home I metered the battery and all I got was 0.375v, checked the pack and directly read 47.6v . Next day rewired with the No Balance BMS with sw and got the expected 47.6v on the discharge.

 

[Bikes4two]

• OK @Nealh , thanks for that very useful backstory to your BMS experiences.
• I had spotted the BeMuchSafer Balance BMS and as you say it is slightly larger and needs a separate charging port. I was happy with both those aspects as it'd fit my battery pack and running out a charge port connection is no biggy.
• However, your comments about the BMS balance circuitry reliability issues etc has made me pause on the purchase/installation of one, so for the moment I'll stick with the more basic (no balance) BeMuchSafer BMS.
• You will recall from earlier postings that I was building (have now built) a 10s 2p using the LG M50LT (21700/5,000mAh) and I've done a couple of charge/discharge cycles on my home made testing board, and at 3 amps ish, and with an LVC set to 31.5v, I'm recording 6.8Ah battery pack capacity.
• I do routinely record pack capacity at various discharge levels just as a way of determining any battery performance changes, so I think I'll stick to that regime for the LG M50LT pack and not worry about checking balance unless something changes.
• Further, my TSDZ2 controller is set to an LVC of 30v and my 42V charger is set to an output of 41.4v, to minimise the chances of over charging/discharging the cells.