Switching on the balancer for an individual cell

[mani9876]

Hello!

Does anyone know if it's possible to switch the balancing circuit on for an individual cell?

I have one cell which has a lower voltage then all the others, the balancing circuit during the charging can't balance the difference (I think I would have to charge about 20 - 50 times that it would be equal). I think the problem is, that I didn't drive the car very often, only about 1000 miles / year.

So is there an option to balance all the other cells, or an individual cell, or do I need to put out the battery of the car, to maintain it?

Thank you

 

[kiev]

Howdy mani and welcome to the forum,

How much is the difference in voltage between the low cell and the rest?

i had a similar problem and tried driving to empty the pack below 2 bars on the fuel gauge, and then re-charging and letting it stay until the charging quit on it's own (after balancing). I did this at least a dozen times over 2 weeks, but it didn't help.

You likely have a weak or damaged cell and if several charging sessions doesn't bring it back up, then the pack or cell will need to be replaced to get your range back. The charging stops when the cells reach the upper limit and there is just not enough current during balancing to raise a low cell. Balancing can only bleed down high cells, what you need is to turn on balancing on all cells except the low one. i wish it were possible but haven't figured out how to do this. A fellow on diyelectriccar forum claims to know the command but hasn't shared it yet.

 

[Aerowhatt]

The balancers do dump only a small portion of the charging current which makes a larger imbalance less likely to quickly be remedied. There are a couple of ways to improve your odds. As near as I have been able to tell (with the I-909 scanner) balancing starts at 75% or 80% SOC during the charge. It's hard to be sure where exactly it starts because the scanner only shows you three balancers at a time.

So there doesn't seem to be a need to run the battery way down to work towards a better balance. Anything around 70% to 80% SOC as a staring point for charging will yield the same result as far as balancing per charge is concerned. To reduce the number of charges needed, reducing the charging amperage would leave the high voltage cell balancers running longer and dumping a higher percentage of the overall battery charge current. So L1 for sure, at minimum setting if you have a choice. Depending on your situation you may be able to safely run some accessories during this critical charging time. If the car is already charging you can turn the key on and run the headlights or something. The power being used by the accessories is not charging the battery and therefore gives the balancers an even higher percentage of the overall charge current to dump. You would need to monitor it reasonably closely. When the charger finishes the accessory loads would then be draining the 12 volt battery under the hood. Not a good thing for very long.

Ideally one would want to charge the battery with an amperage equal to balancer dump amperage. Low cells would trickle charge and High cells would not end the charge by going too high. This is not possible with the cars systems but the techniques above would speed things up as much as is possible without hacking the system.

Aerowhatt

 

[kiev]

With the i909 i select many pages of data items, such as all the balancers, then turn on the capture and scroll thru the pages to see all the selections. It does display just 3 or 4 items per page, but you can scroll forward and back.

There is a balancing set point voltage that is displayed, but i've never found any balancing to occur until it reaches near the end--the set-point keeps getting adjusted and moving higher as the charging progresses.

i write down some of this data that looks interesting and try to figure out what it is doing...

 

[mani9876]

Hello!

Thank you all for your hints.

I have an imbalance of about 100mV, so that's very much, and is about 30 - 35% if I can trust the display.

With my diagnostic tool I can see the balancing starting at about 20% SoC, but at this point, the good cells are at about 3.900V,so maybe it's more the voltage which starts the balancing, than the SoC level?

Do you think I have to replace the battery, or is there a way to get it back up working?
But it doesn't make any sense to me, this cell is the first which is at low voltage, and the voltage don't get higher than 4.000V at the moment. So I don't think the cell is damaged, I rather think it has a higher self discharging current, then the others, and because I don't drive the car in the winter that yields to this problem. Or am I completely wrong?

 

[kiev]

i had a very similar situation in which one cell was lagging all the rest. i didn't want to believe it could be a bad cell, but there was no easy way to access the cell in order to troubleshoot and test. The cell "seemed" to be working in that it would charge and discharge, so maybe if it had a little help to get up to the level of the rest, it would be okay?

In the end i was lucky to still be under the 10-year battery warranty, so i let the dealer have it and they swapped out the pack. If you are under the battery pack warranty period, then you should get a new pack from Mits.


If i had no warranty then it seemed to be 2 options:

1. open the pack and remove the lagging cell, then test it on the bench to determine capacity, self-discharge, impedance, etc. If it checks okay, then charge up the lagging cell to the level of the rest and re-install.

2. open the pack and replace the cell with either a new one or a used one from a salvage pack that had been tested on the bench.

If you just open the pack and charge the one cell without testing it, then it is a lot of work to repeat if it turns out the cell really could not be saved.

Did you see this thread? http://myimiev.com/forum/viewtopic.php?p=23601#p23605

i have to think why would a cell be lagging? The only reasons would be some extra load on it from the CMU board (defective CMU), or it just has an internal defect. If it has a self discharge, then it is defective (internally shorted), and it can't be repaired by charging.

Keep us posted on what you find out about your cell.

 

[Aerowhatt]

With the i909 i select many pages of data items, such as all the balancers, then turn on the capture and scroll thru the pages to see all the selections. It does display just 3 or 4 items per page, but you can scroll forward and back.

There is a balancing set point voltage that is displayed, but i've never found any balancing to occur until it reaches near the end--the set-point keeps getting adjusted and moving higher as the charging progresses.

i write down some of this data that looks interesting and try to figure out what it is doing...

Thanks for the tip on scrolling with the i909! I will have to give that a try. I think I have the balancing algorithm figured out (key word is "think")

It looks to me like the voltage set point is where the balancers of cells at or above that voltage turn on. Either the set point rises when any high cell reaches the next target set point. Or, it rises when a plurality of the lower cells have reached the current set point. Still fuzzy on that one. With a Canion type graph display that changed the color or highlighted cells with balancers currently "on" it would be very easy to divine what is going on. (I wonder if that is possible to add to the Canion App??). With the i909 and it's limited display capabilities it's more of a mystery rapped up in a conundrum.

Aerowhatt

 

[Aerowhatt]

Hello!

Thank you all for your hints.

I have an imbalance of about 100mV, so that's very much, and is about 30 - 35% if I can trust the display.

I agree with you that as long as the car sits the imbalance could be internal discharge. However if you think about it, one cell out of 80 or 88 is a red flag. Cell compatibility is pretty important to these high voltage packs. So it's always going to be an issue in this pack.

You should be showing less than 16 bars when the charger shuts off right? If it were mine and it was under warranty I would take it to the dealer and let Mitsu figure it out. Guessing that they will have to replace the battery pack. If it's out of warranty then I would try what I suggested before to maximize the balancing per charge and see if it makes progress towards the higher cells. It could normalize and then charge and drive the car some in the off season to reduce the self discharge effect from this cell. Could buy you some time anyway.

Curious do you have the number of the cell with is misbehaving?
Aerowhatt

 

[PV1]

Mani, how are you charging the car? Are you using a cord plugged into a 120 volt outlet or a 240 volt charging station? What may help is running the car down to 2 bars and charging uninterrupted using the slowest charge method available (in most cases, the stock cord at 8 amps, 120 volts). One member reported successfully recovering a low cell by doing this repeatedly.

What this does is force a calibration/balance charge by being at or below 2 bars, and the slow charge rate lets the balancers work longer than on a faster charge.

Of course, this will likely only work if the difference between the low cell and the others isn't growing. When I had a cell go bad, I went from 12 bars to 4 over the course of a couple weeks.

 

[kiev]

A fellow contacted me about reading the CAN buss between the BMU and the CMUs, and from looking at the data has figured out how to turn on the balancing on one of the 4-cell CMU boards.

Here are couple of videos showing the pulsing resistors with a thermal imaging FLIR camera. The visual overlay is slightly offset from the FLIR, but those big "820" resistors are the loads that are glowing.






This image is rotated 90 CW with respect to the video orientation, showing the FETs in blue and resistors in red: