Replacing the single cells in the battery pack?

[kojairvs]

We can agree that the BMS stops charging the cells when they (or at least one of them) reach 4.1V
My car this morning showed 95% and the graph showed 4.08V on all cells (as shown on the photo/graph). That is very positive because then they are all almost the same capacity, let me remind you that on the old battery all but one had 3.8V at 100% battery because the first cell reached 4.1V very quickly and all the others remained half empty.
When the BMS shows close to 0% battery then how much did the cells go down? 3.3V?

 

[kojairvs]

These are my self-made cables for the battery connection that sits next to the car, to check how good it is and will the car accept it before I start installing it. . The real DIY

 

[kojairvs]

Thanks for sharing your experience. I've learnt a lot from this whole thread.

I got your point about keeping the cells for yourself.

About cell compatibility:
As I wrote here: https://www.mykiasoulev.com/threads/replacing-the-single-cells-in-the-battery-pack.1906/post-15183

E-service in Hungary offers to replace old packs with used, newer packs from other cars (not necessarily Kia/HyundaI, but preferebly, whichsoever is available for them to purchase from the used/damaged cars market).
Since they don't touch the BMS's deep settings, the old BMS will charge the newer cells to the same Voltage level as the original SKI E375 limit, eventhough the newer cells have higher Voltage limits & could be charged higher, but the old BMS will stop the charging process sooner. According to e-service, this will help to keep your newer cell alive for longer period, as the old BMS will never reach the possible 100% charging level of the newer cells.

Also the displayed range will not be exact, as the real possible range will be higher, because the newer cells can keep more charge compared to our old geniune cells even though their full charge level will not be reached ever, but our old BMS is programmed to show only 210km as max. (or something like this, they said).

They are not mixing different cell types, but replacing the full cell-packs, which can fit into the old battery housing, keeping the old BMS.

I think, mixing the E375 & E400 cells would not be a good solution on the long run, especially I would not recommend to pair the different types in once circiut, but surely better than nothing.

So, in case they can do this with cells from batteries disassembled from other car types, maybe it's not a big deal, so maybe I can also do that as you did, but using cell packs from a crashed Ioniq (2nd gen.).

What do you think?

I have no clue about the batteries of the other cars, I’ve learned lots from this forum but no experience to back it up. I would not dare to give you un advice what to do, just trying to learn myself, use that knowledge on my car and share that experience here which might help others not to make my mistakes .
Sorry

 

[kojairvs]

How I see it, when the battery of my car becomes weak to the point that I need to do something about it, I’ll do things in certain order of actions.
First I would change whole module which holds the weakest cell/cells. That would be easiest and fastest way of doing it. As long as it’s not the module no 1, I have the modules 2-7. If that gives me more life for my car with acceptable range, keep repeating this action.
Second I would replace one layer where bad cell is situated. That means if no 1 goes bad, replace layer which holds no 1 and no 14. No need to weld, just unscrew and screw it back on. Do this as long as it’s useful.
Third, I would weld individual cells but by that time probably most of the cells would be so weak and not sure that it would be practical.
We will see in practice how this will pan out but I hope that will be not soon.

 

[ksoul2084]

Kojarirvs, I applaud your DIY approach, but please do charge this car away from valuable buildings, etc. So many unknowns in this approach, safety 1st (or at least 2nd!)!

 

[JejuSoul]

ksoul2084 I mostly disagree with your pessimism. This is the best DIY battery repair we have heard of.
Replacing the old pack with a complete replacement means all the cells are the same age and have the same wear. This is a better approach than 'mixing and matching' taken by some DIY projects and the re-manufacturing industry.
Two caveats to this.
1. We don't know the history of the replacement pack. if it had been involved in a crash that totalled the car there is a chance that the battery pack was somehow damaged. I think this would be visually obvious once the pack was taken out of the car, so this is probably not so serious.
2. I don't know if a DIY pack with E400 cells will be eligible for the 'fire-safety' recall. Maybe it has the BMS fix already. Maybe not.

Overall I think this is a great fix, although normally I would not recommend anyone to buy a DIY fixed EV.

 

[kojairvs]

This was the car, crashed front left side, I’ve got my battery in completely intact shell and all inside was in perfect condition.

 

[kojairvs]

Markinson, look into this link, I would contact them and leave my details to let me know when they get another Kia in this scrapyard.

https://www.evbreakers.com/

 

[ksoul2084]

ksoul2084 I mostly disagree with your pessimism. This is the best DIY battery repair we have heard of.
Replacing the old pack with a complete replacement means all the cells are the same age and have the same wear. This is a better approach than 'mixing and matching' taken by some DIY projects and the re-manufacturing industry.
Two caveats to this.
1. We don't know the history of the replacement pack. if it had been involved in a crash that totalled the car there is a chance that the battery pack was somehow damaged. I think this would be visually obvious once the pack was taken out of the car, so this is probably not so serious.
2. I don't know if a DIY pack with E400 cells will be eligible for the 'fire-safety' recall. Maybe it has the BMS fix already. Maybe not.

Overall I think this is a great fix, although normally I would not recommend anyone to buy a DIY fixed EV

Fair enough, I think I was confusing this poster with the more adventurous fellow swapping out non-matching cells and removing cells and trying to fool the BMS. That might not end too well.

 

[kojairvs]

When I took a closer look at the specifications of the model from which the new battery was removed, I began to realize that I got better than I thought at first.
The 2014-2018 model is advertised on the scrapyard website.
Kia is advertising the 2017-2019 year model for sale in the UK.
The battery that was used for that period has a specification of 192 cells that are 37.5Ah, not 200 cells of 40Ah.
Since I opened my new battery before mounting to check and compare everything, I confirmed that it is a 196 cell battery. And now I'm also pretty sure the cells are also 37.5Ah the same as mine from 2015.
This means that I will not have any doubts when I want to use cells from the old to a new battery.
This is very good news.
As for the current capacity of the new battery, I'm still gathering information, it will take me a couple of weeks...
But as of today, I know the following:
In order to charge from 35% to 100%, 21.5kWh was consumed.
The range showed 173 km.
It works almost like new if it took me 21.5 kwh to fill 2/3.
I will drive it until it drops to 15%, no lower than that and I will let you know how far it has gone.

 

[kojairvs]

I just replaced worst 28 cells from other battery to mine. Car now running in full power. About range I will know after few cycles, because I don't have any tool that I can recalibrate battery parameters. New cells are more charged than others. After charging new cells goes to 4.18V, older stays at 3.9V and charging finished, so I hope that battery will make self-balance after few cycles.

After this observation when battery is charging, I think BMS trying make all cells average voltage at ~4.05. So I think if we put some cells with more capacity than others, then bigger ones will have less voltage than smaller ones. Should It work, but bigger capacity in new cells will be not used.

It is normal that cells are charged to approximately 4,05-4,1V? Normally LiPo cells can handle 4,2-4,3V. How much voltage is rich in Yours battery with high SOH? My SOH is very low, because I cannot reset battery memory with any tool that I have. Maybe that's why BMS not charging to 4.2V ?

About battery from Ioniq 4, Anybody knows cell dimensions? They will fit to Soul? Ioniq has water cooling, so used battery can be in much more reliable to reuse.

Well done Dondahaka.
Keep us posted and do you have any videos?
I’m trying to merge few videos already made in January and make few more soon showing my old pack and post it on YouTube. And then post the link here.
Let us know every few weeks about the state of your car.
Great effort

 

[IanL]

Kojairvs - it is possible to calculate the SOH.

After charging, reset the m/kWh (shows "--" when reset). Note the mileage and the battery %age if not 100%. At the next charge note the mileage and the battery %age before charging commences and note the m/kWh. Divide the difference in mileages by the m/kWh to get the kWh consumed. Divide this by (difference in %ages x 0.01, i.e. a decimal fraction). This gives the kWh capacity at 100% charge. Divide that by 27. 30, 39.2, 64 or other value, depending on the nominal battery size, and multiply by 100 to get SOH.

I have tested the method against Spy, and it gives close agreement.

 

[kojairvs]

Hi,
Battery was fully charged, 100%, cells 4.12V, possible range indicated 173km.
Down to 10%, cells 3.36V, possible range left 13km.
I’ve managed to drive 175km, 25km was on the motorway, rest of it on the local roads, up and down hills, it’s Switzerland after all .
I’m happy with my second hand battery, now I just need to take care of it better than the previous one .
Maybe I didn’t mention it, but I didn’t know anything about electric cars when I bought my Kia in January 2023, got in side and drove it like any other car, I was reckless, speeding, discharging under 10/5% regularly, 3 times run it down to 0%. No wonder I’ve ruined no 1 cell.
And my wife will be happy that I won’t have to pay few hundreds francs annually on speeding tickets

 

[kojairvs]

I just replaced worst 28 cells from other battery to mine. Car now running in full power. About range I will know after few cycles, because I don't have any tool that I can recalibrate battery parameters. New cells are more charged than others. After charging new cells goes to 4.18V, older stays at 3.9V and charging finished, so I hope that battery will make self-balance after few cycles.

After this observation when battery is charging, I think BMS trying make all cells average voltage at ~4.05. So I think if we put some cells with more capacity than others, then bigger ones will have less voltage than smaller ones. Should It work, but bigger capacity in new cells will be not used.

It is normal that cells are charged to approximately 4,05-4,1V? Normally LiPo cells can handle 4,2-4,3V. How much voltage is rich in Yours battery with high SOH? My SOH is very low, because I cannot reset battery memory with any tool that I have. Maybe that's why BMS not charging to 4.2V ?

About battery from Ioniq 4, Anybody knows cell dimensions? They will fit to Soul? Ioniq has water cooling, so used battery can be in much more reliable to reuse.

Hi again Dondahaka,

Reading your post in more detail got me thinking and I have a few questions/conclusions that will be useful for my future cell replacement project.
Did you charge your old battery to 100% before starting the replacement process?
If so, what voltage were most of your best cells?
If not, what was the voltage on the best cells just before the bad ones were replaced?
What was the voltage on the new cells when installed?
Was it higher than the remaining old ones in your battery?

If you can answer these questions, it will be very useful for us.

You wrote to us that after charging your refreshed battery the voltage is as follows.
New cells 4.18V
Old cells 3.9V

I think we have a small problem here, if I'm not mistaken.
If we consider that the old bad cells were weakened, much lower capacity than most of the remaining ones, your situation was probably such that a few cells reached their maximum voltage earlier than all the rest.
You replaced the 28th.
You have 68 old ones left, which are now only at 3.9V.
Since most of the cells are at 3.9, imagine if they don't balance over time, which is very possible because they have less capacity than the new 28.
It means that you lose the chance to charge another 0.2V on 68 cells and thus add more energy and range to your car.
I think the point is that the majority of cells reach their maximum possible capacity and not the minority.
Maybe you should have discharged the new 28 cells below the current voltage of the 68 you kept. Then they would be charged to a lower voltage and 68 old ones to a full 4.1v (4.18V?).

Mathematically, this makes sense to me.
Practically it should be checked, but it is a bit more complicated.
I only know that if I charged my old battery (in which the weakest cell at 0% shows 3.3V) at a rate of about 1.8kW and it charges for 5h until one or more cells reach 4.1V, we can put only 9kWh
All others stop at 3.9V.
If we "unlock" most of the cells by replacing the bad ones and manage to charge most of them up to max 4.1V, the time required to charge the battery is extended and thus the capacity is increased, example 1.8kWh times 10h equals 18kwh.

I would like to find out more about this theory, all comments welcome

 

[Dondahaka]

Did you charge your old battery to 100% before starting the replacement process?

no, and this was mistake. I thought that any charging will balance each other. But it's not. Last time I found information that this can happened when battery is discharged more than 20% and then after fully charged makes some "calibrations". When I tried discharge while driving, at 50% older cells goes to 2.7V and car starts crawling.
Now I already opened battery again and I'm charging old cells manually to new cells level. I will keep differences in ~0.04V. After that I will drive under 20% and try to charge again, maybe cell balancing will work.

If so, what voltage were most of your best cells?
If not, what was the voltage on the best cells just before the bad ones were replaced?

Before repair best cells gets around 3.95V after charging. One totally broken cell was at 4.2V. But I think that most important think is how much cells goes down at fully acceleration of car, and which cells keeps voltage after distance.

After repair (but before charge) new cells was at around 3.9V, old ones was at 3.6V. And this differences keeps after few cycles. But I couldn't drive under 50%, so every time I was charging above 50%. From 100% to 50% I could make around 50km with heating (with resistance heater, because heat pump is empty of gas). Average consumption is 20kwh/100km

To this time after charging, new 28 cells can charge to 4.15v, older ones stays at 3.85. So I'm correcting now these differences (around 0.3v).

When I finish, I will report results

 

[JejuSoul]

Balancing and Calibration are two different things.

See - Battery Calibration

Battery calibration is entirely neutral as regards the battery. All it does is adjust the BMS SOH data so that the GOM has a better estimate of the range.

It is Calibration that needs a full charge cycle.

 

[kojairvs]

Thanks for the info Dondahaka,

I think that you should charge old ones to maximum voltage (4.1v or higher if it’s possible but no more than what BMS think it is maximum) but keep new ones around 3.9V.
This should give you max charge/discharge range for majority of cells (68 old ones will go from 4.1V down to 3.3V and back up). Newer ones will go from 3.9V down to 3.4V (just un exemple) and back up to 3.9V. This will optimize use of older ones which they are weaker and newer ones will be charge/discharge same amount of kWh but in smaller voltage range.
And this should stay like this, balancing is not possible/necessary as long you optimize majority of your cells to go max voltage up and min voltage down on your cycles.

 

[Markinson]

Who would have thought, science behind charging is such complicated?!

Meanwhile all hints & trick I've ever read about this topic are logically rooted in the system how cells are connected & in the rules & methods how BMS is developed to do its duty.

I wish there were a detailed and logically built-up 'battery manual' issued by the manufacturers, which gathers all these tipps and trick for us EV-geeks.

That would be much more precise & which eliminates rumors and speculation from novice/advanced users.

 

[kojairvs]

I’ve made short video about what information BMS gets from one module.

 

[kojairvs]

This confirms to me that BMS reads only voltages from individual cells directly.
Doesn’t matter what capacity they have and if they differ slightly or massively.
Which means that we can put higher capacity cells as long as they respect 3.3V-4.1V range which is BMS programmed to monitor. Problem lies in fact that the weaker ones dictate the BMS, not the stronger ones. One week one can cripple the whole system, like in my case and Dondahska’s.
I could find somewhere 96 cells of any make and completely swap them all as long as we connect each cable to measure voltages and temperature readings and security switches. They need to be sized enough to fit into the shell of the whole pack and that they can trigger security switches if they get swollen. They can be 200Ah capacity, BMS will let them charge between 3.3V and 4.1V. But higher capacity for same voltage range, longer charging times, longer driving range.