Juanjo Madrid's cell replacement thread

Mitsubishi i-MiEV Forum

Help Support Mitsubishi i-MiEV Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I use 9.85 Diagbox, my dude is if I need wait until full discharge (0 bars) or only when start blinking (2 bars) when say "Wait for the charge warning lamp on the instrument panel to switch off then move on to step 2" previous middle year 2011 cars need three step, I need two because is 2012, the bigger difference is 3v83 to 3v91. I waiting to "electric station(LCD draw) or plug symbol lamps( red lamp) go off, and never happen, maybe need wait until not bars? only electric station lamp go to blink , if validate when blinked the error appears, afther say "disconnect from electric Network, put to off ventilation handler and the other knobs to off"
 
I use 9.85 Diagbox, my dude is if I need wait until full discharge (0 bars) or only when start blinking (2 bars) when say "Wait for the charge warning lamp on the instrument panel to switch off then move on to step 2" previous middle year 2011 cars need three step, I need two because is 2012, the bigger difference is 3v83 to 3v91. I waiting to "electric station(LCD draw) or plug symbol lamps( red lamp) go off, and never happen, maybe need wait until not bars? only electric station lamp go to blink , if validate when blinked the error appears, afther say "disconnect from electric Network, put to off ventilation handler and the other knobs to off"
In order to deep discharge the battery you need to put the heater on max, turn on AC and set the fan to boost.

This will take a long time if you start with cell voltages around 3.9V…
 
Para descargar profundamente la batería, debes poner la calefacción al máximo, encender el aire acondicionado y configurar el ventilador al máximo.

Esto tomará mucho tiempo si comienzas con voltajes de celda alrededor de 3,9 V…
I will try next time with only one bar (LCD lamp blinking), Today need use the car and finally charge without do it, show 88 Km in the RR when full, I feedback when do it
 
That seems pretty good for a 2012 I-MIEV, looks like the cells you replaced weren't the ones with the lowest capacity, they just couldn't handle high currents?

You may therefore not see a significant improvement of the remaining capacity after a battery calibration.
.

Something strange happens after my maintenance attempt, today morning the charger stopped with 78% SOC, not all the bars appear on the dashboard and I tried several times to fill it but it was not possible, today in the morning it shows 62 km RR and when I see the voltage before starting, many cells are at 4.1V, some at 4.089 and the two LEV50N at 4.03 and 4.04V but the 78% SOC is in the OBDZero application. Now charging in my job the SOC is 81,3% and 0,18A /100W (look like is finalizing soon)

I will probably try to do the procedure again tomorrow, something is now telling the vehicle that the capacity and the bars do not match when I made the failed attempt of maintenance.

Yesterday, I tested the faulty LEV50 cell with a RC Lithium charger/discharger (long time at 2A max) and the measured capacity is similar to the second LEV50 cell I replaced (this one was fine but I took it out so I can compare between the bad one and the others that are still useful using this sample one), so I had to use the two LEV50N I bought, I think it is necessary to test them with a 2000W 12V inverter with high current request, will 100A be a good test?

Driving this morning I see the turtle with high demand again (spend more time to see the turtle, it's on my daily route), I see three other cells now with similar voltage, 2V8 to 2V86V

I have two CATL 93Ah plus the LEV50 that I replaced without being damaged, so next time I open the package I think I need to evaluate with the inverter test or a 5000W UPS that works with 36V, maybe it would be better to test 8 complete cells without disassembling all, adding externally the 2 CATLs that I have in the series of 8.
 
Something strange happens after my maintenance attempt, today morning the charger stopped with 78% SOC, not all the bars appear on the dashboard and I tried several times to fill it but it was not possible, today in the morning it shows 62 km RR and when I see the voltage before starting, many cells are at 4.1V, some at 4.089 and the two LEV50N at 4.03 and 4.04V but the 78% SOC is in the OBDZero application. Now charging in my job the SOC is 81,3% and 0,18A /100W (look like is finalizing soon)

I will probably try to do the procedure again tomorrow, something is now telling the vehicle that the capacity and the bars do not match when I made the failed attempt of maintenance.
Charging stops once a cell reaches 4.1V, but this will only result in 100% SoC if the energy measured going into the HV battery equals the capacity value set in the BMU.
The only way to correct the discrepancy quickly is to to calibration.

Yesterday, I tested the faulty LEV50 cell with a RC Lithium charger/discharger (long time at 2A max) and the measured capacity is similar to the second LEV50 cell I replaced (this one was fine but I took it out so I can compare between the bad one and the others that are still useful using this sample one), so I had to use the two LEV50N I bought, I think it is necessary to test them with a 2000W 12V inverter with high current request, will 100A be a good test?
Max current (foot to the floor) is around 150A therefore a 2A discharge test is pointless, > 100A is certainly the way to go.
Driving this morning I see the turtle with high demand again (spend more time to see the turtle, it's on my daily route), I see three other cells now with similar voltage, 2V8 to 2V86V
Unfortunately that’s what happens if you replace the worst performing cells, others will take their place..
I have two CATL 93Ah plus the LEV50 that I replaced without being damaged, so next time I open the package I think I need to evaluate with the inverter test or a 5000W UPS that works with 36V, maybe it would be better to test 8 complete cells without disassembling all, adding externally the 2 CATLs that I have in the series of 8.
Use the CATL cells, don’t bother putting the LEV50 back, testing the cells externally is a good idea, you could even read out the cell voltages via CAN using the CMUs if you have an Arduino or similar..
 
La carga se detiene una vez que una celda alcanza los 4,1 V, pero esto solo dará como resultado un SoC del 100 % si la energía medida que ingresa a la batería HV es igual al valor de capacidad establecido en la BMU.
La única forma de corregir la discrepancia rápidamente es mediante la calibración.


La corriente máxima (del pie al suelo) es de alrededor de 150 A, por lo tanto, una prueba de descarga de 2 A no tiene sentido, > 100 A es ciertamente el camino a seguir.

lamentablemente, eso es lo que sucede si reemplazas las celdas con peor rendimiento; otras ocuparán su lugar.

Utilice las celdas CATL, no se moleste en volver a colocar el LEV50, probar las celdas externamente es una buena idea, incluso podría leer los voltajes de las celdas a través de CAN usando las CMU si tiene un Arduino o similar.
I was thinking of returning a LEV50 because it is better than others, I only have two CATL93 and 3 very clearly degraded cells, low voltage and high temperature, temporary solution of course. Yesterday I did the maintenance procedure, I confirm that the procedure is really possible with Diagbox V9.85 on Imiev cars. But after spending charging all the night at temperatures of 3°C, the charge indicator stopped at approximately 80% and a capacity of 20 Ah is registered in the BMU with the current Km, although it gave an error in step 2 and asked me to repeat Again, my mistake was not waiting for the end of the complete discharge, the complete discharge and charge take place on their own when the air conditioning is stopped, I thought that 20 Ah is not real, it indicates 72 km and three bars are still missing, the cells maximums do not reach 4V1 only 4V09 abd 4V08, now when starting the charge only enter a tiny current, I understand that it is balancing (again), but the indicator still wrong? I will try the process until I finish it without errors.

Traction battery total capacity with Diagbox show 2Ah more from 25.7Ah to 27.5Ah and 13.25 can be received.
 

Attachments

  • IMG_20241112_064450.jpg
    IMG_20241112_064450.jpg
    11.5 KB
  • Screenshot_2024-11-12-15-38-00-106_dc.local.electriccar.jpg
    Screenshot_2024-11-12-15-38-00-106_dc.local.electriccar.jpg
    46.5 KB
  • IMG_20241112_160601.jpg
    IMG_20241112_160601.jpg
    50.8 KB
  • IMG_20241109_042233.jpg
    IMG_20241109_042233.jpg
    49 KB
Last edited:
I was thinking of returning a LEV50 because it is better than others, I only have two CATL93 and 3 very clearly degraded cells, low voltage and high temperature, temporary solution of course.
Replacing cells in a degraded pack is like a game of ‘whack-a-mole’, regardless of how may you replace, different ones will show bad until you have swapped them all, I’m afraid..
, the cells maximums do not reach 4V1 only 4V09 abd 4V08, now when starting the charge only enter a tiny current, I understand that it is balancing (again), but the indicator still wrong? I will try the process until I finish it without errors.
Charging stops once one cell reaches 4.1V, but then balancing starts and the highest cells are ‘bled’ to reach the voltage of the lower ones, however because of the tiny balancing current this function eventually times out and this is what you see


Traction battery total capacity with Diagbox show 2Ah more from 25.7Ah to 27.5Ah and 13.25 can be received.
You need to finish the procedure without errors to get the indicator ‘fixed’
Btw one of the screen shots shows a 12V aux voltage of 11.6V (0%SoC) best to replace it altogether..
 
The store where I bought two cells (by air) has stock of LEV50N cells dismantled in China more than 100, claims they have around 40 Ah each, does anyone know how to import "used" lithium batteries from China to the EU by ship ? The price is around €29 door to door buying 100 units, 160 Kg at least, a friend who works importing new batteries and power banks tells me that it is not possible, if I do not have some documents: MSDS, DG CERTIFICATE and ROHS, but I think this It is for new cells, he does not know about used ones.
I tried new CATL 93Ah and it is always out of stock and costs more than €45 each plus shipping, if anyone knows where they are available please tell me where.
Maybe the best option is a used package from a wrecked car, but I think it's a lottery.
 
The store where I bought two cells (by air) has stock of LEV50N cells dismantled in China more than 100, claims they have around 40 Ah each, does anyone know how to import "used" lithium batteries from China to the EU by ship ? The price is around €29 door to door buying 100 units, 160 Kg at least, a friend who works importing new batteries and power banks tells me that it is not possible, if I do not have some documents: MSDS, DG CERTIFICATE and ROHS, but I think this It is for new cells, he does not know about used ones.
The same applies for any batteries, new or used doesn’t matter. IMHO there is a good chance that you’ll end up paying €€€ and receiving nothing in the end.
I tried new CATL 93Ah and it is always out of stock and costs more than €45 each plus shipping, if anyone knows where they are available please tell me where
…there seems to be a shortage of these cells, currently.
Maybe the best option is a used package from a wrecked car, but I think it's a lottery.
Not really; LEV50N cells are degrading at around half the rate of the original LEV50.

If you know the age and mileage of the car you can pretty accurately predict the remaining capacity.

Best to swap the BMU also as it’s optimised for use with newer chemistry; that way you can even use a 80cell pack with an 88 cell IMIEV.
 
I have news, and I need opinions to decide, I have many doubts to clear up:
-I have the possibility of bringing 88 LEV50N cells from China, it is the warehouse that sent me the first two, it has a lot of stock and it tells me that they reach 40Ah, uncertain, but the first two appear strong in my package, 1mOhm measured by me and they do not go down Like the others, they do not increase their voltage like the others, reaching a maximum of 4V03 when the rest reaches 4V1 and on the way down there is even more difference, my cells theoretically have 27.5Ah and 2.5mOhm measured. The amount is €2700 at the door.

-I have some complete C-Zero 2018 crashed packages available, there are several, it would greatly facilitate the assembly speed, but I would limit myself to 14.5Kwh and I still have to get the BMU, I hope to be able to replace mine with mine easily plug in under the back seat? Can any BMU from another c-Zero or ION work for me? And you don't have to code VIN or anything via software? (edit: reading, see the 2017 cars have the BMU inside the pack, it is a barrier to swap 16-14,5Kwh? and what about my OBC 2012, he work with different charging voltage, support 80/88 cells?)
The first price is €2500 + shipping from Spain to Spain, I hope around 200€ more.

As a last option I am considering making the change to CATL, I don't know stock dates, I can´t I imagine the time/money invested in busbar modifications, copper bars, it sounds wonderful to be able to reach 200Km of autonomy in the future, but what is the real cost? the time will be spent by me and my son.

Apart from this, yesterday I tested my damaged cells in an 800VA UPS, the damaged cells make a sound (50 Hz very clear) when the UPS undulates, it is very audible, while the 2 Catl that I put in series remain silent and clear their voltage in the oscilloscope, I know temperature/voltage is a way of detecting defective cells, but with the ear? Busbar are ok, load was small, the UPS is square wave form. Some experience with this?
 

Attachments

  • Celdas YUasa12 case .jpg
    Celdas YUasa12 case .jpg
    97.3 KB
Last edited:
As a last option I am considering making the change to CATL
From the forum and community point of view this would be by far the best solution. Also for you probably the most interesting one. I read you are from spain. So its not that cold over there and the NMC cells would always run in good temperature conditions.
Why this would be the best outcome for the forum:
To my knowledge all of the upgrades done on this forum used a CAN bridge to fake the CAN traffic. This was until few weeks the only solution when you do not want to develop your own understanding of BMU modding or bring the car to Novosibirsk and meet kolyandex in person there.
But now this changed. For read/write the BMU firmware you can get the software from kolyandex and at first make a backup of the original BMU firmware/configuration on your own.
Then let kolyandex make the remote BMU firmware modification. Kolyandex would probably use teamviewer to get on a computer you have installed the software to read/write the BMU connected with OpenPort 2.0 to your car. Then he read out the BMU firmware, modify it probably on his computer and then copy the newly modified BMU firmware back to your computer and write it on the BMU. I expect that he would then delete the modified BMU firmware from your computer, but this does not matter. Just read it out from the BMU again.
Please upload then to the forum the original/old BMU firmware and the newly modified one so that the people would finaly be able to find out what have to be changed in the BMU firmware to be able to configure those settings on your own:

Capacity 0-120 Ah
Charge voltage 4.0-4.3
https://forum.kolyandex.su/viewtopic.php?f=22&t=869
 
I have the possibility of bringing 88 LEV50N cells from China, it is the warehouse that sent me the first two, it has a lot of stock and it tells me that they reach 40Ah, uncertain, but the first two appear strong in my package, 1mOhm measured by me and they do not go down Like the others, they do not increase their voltage like the others, reaching a maximum of 4V03 when the rest reaches 4V1 and on the way down there is even more difference, my cells theoretically have 27.5Ah and 2.5mOhm measured.
Although your 2012 BMU is optimised for LEV50 cells, it’s regarded as safe to use it with LEV50Ns (different story the other way round)

It’s a known fact that if you run a pack with mixed capacity cells, equal balancing isn’t possible.
I have some complete C-Zero 2018 crashed packages available, there are several, it would greatly facilitate the assembly speed, but I would limit myself to 14.5Kwh and I still have to get the BMU, I hope to be able to replace mine with mine easily plug in under the back seat? Can any BMU from another c-Zero or ION work for me? And you don't have to code VIN or anything via software? (edit: reading, see the 2017 cars have the BMU inside the pack, it is a barrier to swap 16-14,5Kwh? and what about my OBC 2012, he work with different charging voltage, support 80/88 cells?)
A 2018 C-Zero will most likely have the BMU inside the pack, it’s not VIN coded and the OBC doesn’t seem to care. Some wiring modifications will be necessary to use the 80cell BMU in the IMIEV (if the units are indeed in different locations)

May I suggest a 4th option:

Buy a C-Zero pack locally and get 6 more from China then swap all ten 8 cell modules into the I-Miev, no wiring modification necessary.

As far as @hjdlsnbc suggestions regarding remote BMU upgrade, be wary; kolyandex has promised to do as described to other members on this forum but never actually delivered??
 
The remote BMU upgrade is offered just since few weeks on his forum. Before it was only offered in Novosibirsk.
But do not forget that the BMU configuration upgrade is just the cleanest option to not have to do CAN data modification. But you still can do CAN data modification and not do BMU firmware modification. You have both options.
And if you do not do either BMU upgrade or CAN data modification, you would still get higher range then with new LEV50 cells. The BMU is doing calibration of the capacity all the time when usng. It does up to something below 50Ah with the default BMU firmware settings you have now in your BMU.
So you have all time to drive with the car with the new cells for months/years. Because the default LEV50 BMU configuration is not charging up the NMC cells to max possible voltage, you even extend their life.
 
.
And if you do not do either BMU upgrade or CAN data modification, you would still get higher range then with new LEV50 cells. The BMU is doing calibration of the capacity all the time when usng. It does up to something below 50Ah with the default BMU firmware settings you have now in your BMU.
So you have all time to drive with the car with the new cells for months/years. Because the default LEV50 BMU configuration is not charging up the NMC cells to max possible voltage, you even extend their life.
That’s one of the assumptions that has turned out to be wrong, a stock BMU cannot ‘learn’ the capacity of NMC cells.

Best you can do is to set it manually to 45Ah using MUT3/Diagbox (battery replacement function)

You’re right in the sense that this will indeed extend the life of the NMC cells.
 
That’s one of the assumptions that has turned out to be wrong, a stock BMU cannot ‘learn’ the capacity of NMC cells.
It was written here somewhere in the forum that the BMU could rise its capacity expectations to a certain level on its own. Thanks for the update on that.
 
Compre un paquete C-Zero localmente y obtenga 6 más de China, luego intercambie los diez módulos de 8 celdas en el I-Miev, sin necesidad de modificaciones del cableado.

May I suggest a 4th option:

Buy a C-Zero pack locally and get 6 more from China then swap all ten 8 cell modules into the I-Miev, no wiring modification necessary.
Good!
 
Buy a C-Zero pack locally and get 6 more from China then swap all ten 8 cell modules into the I-Miev, no wiring modification necessary.
I'm worried with the different batteries, mix 8 of 34Ah and 80 of 40Ah, result in a 34Ah battery, 34Ah working100% and 40 Ah only around 80%. I think the experience here is need change all cells or have problems in future.

The local option look well for my, when finish have some CMU for parts. I could be use 8 modules (8 batt each) inside my pack if is possible extract the BMU module from the pack and put in the rear seats place?, not necesary wiring. But I unknow if have the pinout connector of rear Seat modules, do this (if possible) and put some cooper shortcircuit busbar what avoid use of 6 and 12 battery packs (4 cells) in my pack. All this to avoid mix cells, give me 14,5 KWh finally with the correct BMU for LEV50 N.

If this last option is possible, why not reduce my pack from 88 to 80, extracting 8 selected bad cells and changing BMU module, or only the firmware of my BMU car for 80 cells BMU or firmware module. Them I need a BMU 80LEV50N what work well with 80LEV50 chemical.

Many maybe!
 
I'm worried with the different batteries, mix 8 of 34Ah and 80 of 40Ah, result in a 34Ah battery, 34Ah working100% and 40 Ah only around 80%. I think the experience here is need change all cells or have problems in future.
Isn’t this the other way around? Btw a local 2018 LEV50N pack (80 cells) should have way more capacity than 34Ah and therefore be a lot closer to the 8 40Ah cells from China.
The local option look well for my, when finish have some CMU for parts. I could be use 8 modules (8 batt each) inside my pack if is possible extract the BMU module from the pack and put in the rear seats place?, not necesary wiring. But I unknow if have the pinout connector of rear Seat modules, do this (if possible) and put some cooper shortcircuit busbar what avoid use of 6 and 12 battery packs (4 cells) in my pack. All this to avoid mix cells, give me 14,5 KWh finally with the correct BMU for LEV50 N.
If you mean 10x 8 cell modules, then yes, that should work, there may be a mod required for the battery fan iirc. Check out circuit diagrams below

http://mmc-manuals.ru/manuals/i-miev/online/Service_Manual/2017/index_M1.htm

Just bear in mind that the 14.5kWh is only a theroretical value..
If this last option is possible, why not reduce my pack from 88 to 80, extracting 8 selected bad cells and changing BMU module, or only the firmware of my BMU car for 80 cells BMU or firmware module. Them I need a BMU 80LEV50N what work well with 80LEV50 chemical.
A stock 80 cell BMU is optimised for LEV50Ns which have a wider tolerance/better performance, using it to run 1st gen LEV50 can be dangerous!

Apart from that, even removing 8 additional cells probably won’t improve things for long.
 
Last edited:
If you mean 10x 8 cell modules, then yes, that should work, there may be a mod required for the battery fan iirc. Check out circuit diagrams below

http://mmc-manuals.ru/manuals/i-miev/online/Service_Manual/2017/index_M1.htm


Interesting read at

http://mmc-manuals.ru/manuals/i-miev/online/Service_Manual/2017/00/html/M100102100002203ENG.HTM

you can read the old data learned from the BMU and pass it all to the replacement BMU, isn't this actually a way to hack the BMU parameters? Do you know if anyone analyzed these files? If we have several dumps perhaps we can draw conclusions, cross BMU LEV50 and LEV50N, etc... and in the best of cases understand it, but it seems that everything the BMU needs is passed in this file and it is not its firmware but data of the actual battery installed.

Also very interesting is the writing of manual data where we indicate the date of manufacture of the battery and km traveled, it seems that we are using a crossing point of two tables as a starting point, something that would take many stress tests to Mitsubishi to parameterize the standard behavior of these YUASA cells with their aging and use. It seems that we can even go back in time to the BMU, all safely because we will have saved our real file first.

I will focus first on my battery change, for now I think that winter will eat up my possible time and it will be for spring, while I will use the car less than 40 km
 
Back
Top