Single LEV50-cell wanted for our Citroen C-Zero

me68 said:

Hello,
again: one cell does not reach end of charge cell voltage anymore.

Click to expand...

Hi,

I notice your screenshot is at 85% charge - does it have the same voltage as other cells when the battery is fully charged, or is that as far as it will charge up?

Looks like the cell is faulty, these cells are hard to get hold of, your only option is usually salvaging them from a dead car, but then they can be low capacity as they are used.

I may have some cells in the next few weeks but I don't know if they are any good, will need to test them once they arrive.

Cheers.

Low voltage of one cell at or near full charge that the balancing system can’t eliminate even across multiple charge cycles could be a fault in the cell balancer for that cell such as a shorted transistor for the balancing resistor.

Unless it’s so bad that the car can barely drive any distance a weak cell would not normally cause a voltage imbalance at full charge - instead it would cause an imbalance at a low state of charge as the weak cell discharged and reached minimum voltage before the others.

I'd like to point out to those new to this forum is that Martin (me68) is the developer who created the CaniOn app and was the first (on this forum) to successfully replace a battery cell, which he documented for us earlier in this thread.

Can anyone suggest a specific test to perform when the pack is opened up on the bench to verify the cell balancer is operational?

Hope someone can provide Martin with a good usable cell. His free CaniOn app has been such a great contribution to our i-MiEV community and I would hope that we can help him out.

JoeS said:

Can anyone suggest a specific test to perform when the pack is opened up on the bench to verify the cell balancer is operational?

Click to expand...

Before even opening the pack I would try running the "cell smoothing" diagnostic operation to attempt to force the cells back into balance.

This can be done with Diagbox (which I suspect Martin has access to) or it can also apparently be activated by recent beta versions of Hobdrive.

Although I have not used this diagnostic mode before (my cells balance perfectly during a charge) I believe what it does is attempt to balance the cells without charging. Normally balancing occurs towards the end of the charging cycle however if the imbalance is too large it could take multiple cycles to balance.

The smoothing operation will work at any SoC and activate the balancers on all high cells to bleed them down to the lowest cell's voltage without charging, and as a result should be able to correct even a large imbalance. As the bleeders are small this could take many many hours but in theory should bleed the high cells down until they match the low cell and then a charge could be performed to charge all cells back up again.

Assuming this succeeded I would then watch to see whether the cells get out of balance again.

A persistent cell voltage imbalance at a low SoC indicates a difference in Ah capacity between cells, however at a high SoC it indicates either a problem with the balancing circuitry, or a parasitic load that is discharging one cell more than others. (Such as a fault on the CMU board) In theory a cell with high internal leakage could also cause this but it is unlikely.

We've seen a failure of quite a few LTC chips on the CMU boards now - a faulty LTC chip could easily cause the balancer for a cell to remain on all the time, or the LTC chip itself could present a parasitic load since it is individually connected to each cell.

If one cell is subject to a parasitic load/discharge path which exceeds the discharge rate of the cell balancers, then the balancing system will not be able to maintain balance of the cells. There has been a reported case of this happening on a Leaf over on Speakev.com.

Not that it will do much good in this case, but here's an eBay ad for a single cell out of a 2015 i-MiEV (I didn't know there was such a model year in USA) so it must be LEV50N and they're asking $40USD with free shipping from Texas.

https://www.ebay.com/itm/2015-imiev...lf-Cart-Off-Grid-Solar-EV-i-miev/173781412444

JoeS said:

Not that it will do much good in this case, but here's an eBay ad for a single cell out of a 2015 i-MiEV (I didn't know there was such a model year in USA) so it must be LEV50N and they're asking $40USD with free shipping from Texas.

https://www.ebay.com/itm/2015-imiev...lf-Cart-Off-Grid-Solar-EV-i-miev/173781412444

Click to expand...

For someone in the UK or EU you could also try this UK based company:

https://www.secondlife-evbatteries.com/

I've just ordered 4 cells for my car yesterday and expect delivery tomorrow. I can't vouch for the quality (eg remaining capacity) of their cells yet however I've been assured that the cells they supply should have a capacity of 40Ah or higher and that they are happy to fully refund the purchase if my testing proves otherwise. My weak cells are now at about 33Ah and the best cells in my pack are around 40Ah so this should be fine.

I will of course be doing my own Ah capacity measurement when I get the cells. Unfortunately the discharge rate on my tester is quite low as it's designed for radio controlled cars and quadcopters, so it will take approximately 2 days per cell to do a discharge capacity test, so I should know how good the cells are in a week or two... :lol:

One question I haven't decided on yet is how low in voltage to do the discharge test. From memory the lowest my tester will let you discharge a Lithium Ion cell is 3.0 volts, and the highest is 3.3 volts. Depending on what end voltage you choose you will of course get a slightly different result. In the car the cells are only ever used down to approximately 3.6 volts before the car shuts down to avoid a low voltage situation under load, so I'm inclined to test them only down to about 3.2/3.3 volts even though that will give a lower Ah figure.

I should add that the account signup process on their website seems to be broken (I've let them know) however I was able to place an order without creating an account.

My battery have 32Ah (71 000 km) and I start thinking about replace whole LEV50 cell instead something with better capacity.
I dont want buy used LEV50. Too much monny and result nothing special.
18650 cells dont like 2C charging. This is not good way.

Whot you mean about this cells? It looks similar original LEV50. But only 50Ah.
https://www.aliexpress.com/item/32888613944.html

This cell looks good, but I mean too big.
https://www.aliexpress.com/item/32891711059.html


Another cells maybe can use in imiev
https://www.aliexpress.com/item/32968678485.html
or 60Ah from BMW i3?

the first one looks exactly like the cells used in our cars. Not sure if it is a real Yuasa cell, or a clone.

the second one looks like a Samsung SDI cell used in Fiat and BMW packs, except it has a threaded stud welded to the terminal caps, and they just weld straps to connect each cell. there is no threaded holes for the CMU card to connect, but it might be made to work if it could fit somehow.

the third one, a cylindrical 55x360mm cell looks promising, but unknown manufacturer and lots of work to make it fit and connect.

that's some good research on cell options, thanks for sharing.

DBMandrake said:

Before even opening the pack I would try running the "cell smoothing" diagnostic operation to attempt to force the cells back into balance.

This can be done with Diagbox (which I suspect Martin has access to) or it can also apparently be activated by recent beta versions of Hobdrive.

Click to expand...

That wasn't true for Hobdrive until last month (Jan/Feb 2022) but it's there now and tested on my 2012 Peugoet Ion (thanks guys!!)

If the procedure doesn't work for you then they'd really appreciate feedback. The programming varies a little between model year to model year so they need some testers

NB: this test ONLY works when plugged into a slow charger and when triggered the test first of all dischargess the battery using heater and AC as loads with the car switched off. Heater and fan controls need to manually be set to max and windows open to get rid of the heat. Following this it will run a charge cycle at 0.1C (5A), so exoect it to take a long time

On a slightly different note: does anyone knoe what the difference between Yellow and Blue lev50 cells is?

I have blue ones from a 2011 Imiev, my 2012 Ion has yellow ones, but LEV50N appear to be blue

kiev said:

the first one looks exactly like the cells used in our cars. Not sure if it is a real Yuasa cell, or a clone..

Click to expand...

BEV2 format is a standardised size format like AA or AAA, but voltages and chemistry are not

LEV50 and LEV50N are LIMnO. These have different curve to newer cells and the BMU needs reprogramming or they won't work (or at least, will work HORRIBLY)

Because the chemistry is different, you can't just swapout one cell if you take this path - it's an all or nothing proposition and this applies to 18650 swaps too

Oz Electric Vehicles have unveiled a couple of prototype LiFePO4 conversions but they're $12kAU apiece (about £6k or $8500US)

DBMandrake said:

For someone in the UK or EU you could also try this UK based company:

https://www.secondlife-evbatteries.com/

Click to expand...

As an update for this: They no longer sell LEV50s or LEV40s

having just replaced a bunch of really bad (and aging) cells in my 2012 Ion, it's obvious that the rest of the pack isn't actually much better and the "new" (old) cells are streets ahead of them

LEV50s are now like hen's teeth

I suspect the only viable way forward is the Oz Electric Vehicles Li-Ion repack and BMS reprogramming path