The cell to cell change is not very complicated but the LEV50 and LEV50N cells are not physically exact,
That depends on the batch of LEV50 cells, there are height differences between e.g. yellow and purple:
https://myimiev.com/threads/cant-mix-yellow-and-purple.5002/
achieve a perfect fit, raising the cell is not enough because the terminal height is different, the result is that the CMU remains slightly sunken and the floating PCBs remain below the main PCB.
The CMU underside needs to 'touch' the battery terminal for the cell temperature to be accurate
Keep in mind that as MickeyS70 says, then you have to leave a lot of time until the capacity is reprogrammed, so ideally you have to do battery maintenance that makes the process a complete discharge and charge, for this you must have the DiagBox software and Lexia OBD adapter.
Hobdrive (paid version) can do the same using a 'low cost' Bluetooth dongle on an Android device
Furthermore, the chemistry of LEV50N and LEV50 does not seem to be exact, so I must change the BMU if I want to do things right. I don't know what implications not doing so will have, but I think the small cost of the BMU is worth changing.
It's not clear what the exact implications of running LEV50N with a LEV50 BMU are, the degradation algorithm for one will certainly be more conservative. It's thought though that it's safe to use such a combination, unlike the opposite where a LEV50N BMU is 'in charge' of LEV50 cells
The idea was to use it temporarily to remove the next problematic cell if it exists, perhaps add 0.2V to the problematic one to reduce its indication but make it proportional, since the next one does not appear now because cell H of group 2 means that the vehicle with about 55Km traveled does not circulate anymore, I think you are telling me that it is not possible because of your fear that this cell will go into overdischarge and be destroyed or worse still catch fire, so: is the temperature a good indicator about the state of the most damaged cells?
There is very little capacity left once a cell is below 3V, the voltage drops very quickly into the 'red zone' if used further, an artificial offset of 200mV isn't going to make much of a difference in range.
There is going to be a delay in detecting the increased cell temperature (especially if there is a gap between the cell and temp sensor) and yes, the fear is that (worst case) a thermal runaway reaction cannot be stopped in time.
My friend from Hong Kong has requested a quote for cells filtered by internal resistance, we want to see if they can send us guaranteed cells with IR below 1 MilliOhm, it is expected that the used cells will be in a variable state and perhaps it is a dead end to use them as a traction battery for an electric vehicle with high current demands, where the internal resistance is decisive, they may only be used as a solar battery or UPS with demand currents over 30 or 50 A for each.
Yes, its highly recommended to use cells with an IR below 1mOhm
I'm already thinking about how to use the LEV50 batteries as support for these LEV50N but I don't like carrying another 100Kg of cells in the back of my cabin to get 360V and inject with a DC/DC converter
More like min 150kg, IMHO best to stop thinking about such a plan..
But I did part out lev50n cells from a couple of cars. When I was testing their capacities at 20A car A was getting 40-41Ah and car B 45-46Ah. Their internal résistances were 0.2-0.25 and 0.5-0.7 mOhms per cell respectively. On the other hand I believe the measurement of IR is somewhat variable depending on method and charge level so it is possible that comparing my values to yours is not valid.
Yes, that sounds more realistic, as long as you use a dedicated battery resistance meter (e.g RC356x), results should be comparable
Not really enough data points for a trend, but it might suggest that your lev50n are quite well used and likely to keep failing now that they have started.
There is more evidence of that; the wide spread of cell voltages at low SoC is an indication of a mixed capacity pack and unfortunately, the worst cell defines the performance of the whole pack.
There is one upside though, degradation rates of LEV50N are only about half compared to the original LEV50.
