Battery pack change plans

[alviseven]

My hunch is that your ‘brick’ charger is faulty and the EV-ECU of your 2012 car isn’t configured to accept a direct lead charge, therefore I suggest to plug the car into your house EVSE and see what happens…

you suggesting that I put back the original brick charger and try the house charge lead?

Not thought to do that stopped using the house "fast" charge as it seemed to drop out when the old pack was in car. Can charge the obc again dead easy now I've done it!

Still got to drop.the battery pack out and at least look at cmu09 surely?

Is it only the obc which can sort out the individual cells or will the chademo do this as well? Assume the chademo just charges even quicker and bypasses the obc but everything else works the same way?
Got distracted on imiev work today taking the two oldies out tomorrow to a show. Ended up tidying garage too.

 

[MickeyS70]

you suggesting that I put back the original brick charger and try the house charge lead?

No, just connect the car as it is to the house EVSE

Not thought to do that stopped using the house "fast" charge as it seemed to drop out when the old pack was in car.

Drop out would be an improvement to not charging at all, but best to verify functionality with another type 1 car (Leaf?)

Can charge the obc again dead easy now I've done it!

You’d want to verify that your house EVSE and cable is working beforehand

Still got to drop.the battery pack out and at least look at cmu09 surely?

Yes, even if unrelated, that still needs to be sorted (might get away without dropping it, see below)

Is it only the obc which can sort out the individual cells or will the chademo do this as well? Assume the chademo just charges even quicker and bypasses the obc but everything else works the same way?

Balancing will only take place after slow (AC) charge reaches 100%.

There may be an alternative, iirc, HobDrive has a ‘cell smoothing’ (or similar) function that can be started manually anytime (might take several attempts to run). Even if it works, it’s going to be a (very) slow process, take a reading of cell voltages every couple of hours to see if things improve.

If you want answers fast, drop the pack, measure the offending cells:
1. if they are lower, bench charge them to match the others.
2. if they are the same, check connections, if they appear OK, swap the CMU board with another inside the pack and reassemble.

 

[alviseven]

No, just connect the car as it is to the house EVSE

Even with the older style non brick type obc was not sure if the early obc was house charge 16amp compatible?

Drop out would be an improvement to not charging at all, but best to verify functionality with another type 1 car (Leaf?)

Not got another modern electric might ask friend ...

Balancing will only take place after slow (AC) charge reaches 100%.

So I need the obc to work to solve cmu issue possibly and chademo will not do this for me

There may be an alternative, iirc, HobDrive has a ‘cell smoothing’ (or similar) function that can be started manually anytime (might take several attempts to run). Even if it works, it’s going to be a (very) slow process, take a reading of cell voltages every couple of hours to see if things improve.

Will try this try to find it

If you want answers fast, drop the pack, measure the offending cells:
1. if they are lower, bench charge them to match the others.
2. if they are the same, check connections, if they appear OK, swap the CMU board with another inside the pack and reassemble.

This is a good idea. Not thought of this. Know I would be stumped if batteries measured OK and cmu looked OK.
I can swap the cmu boards over with an adjacent pack then issue will either go away or if reappears on same cell 09 block the cmu issue but if on swapped block its the pack...

Many thanks for all the advice could not even have started this work without the forum advice here

 

[MickeyS70]

After @coulomb ‘s post I think the 2015 OBC you swapped is stock and was not modified, at least not on a HW level. The distinction between brick/non brick charging seems be done on EV-ECU?
Your 2012 model therefore needs a working EVSE to charge, easiest way to verify this is to try with another car.

 

[kiev]

2. if they are the same, check connections, if they appear OK, swap the CMU board with another inside the pack and reassemble.

Did you mean to swap the cells to another location but leave the CMU board in the same "slot" location?

It was my understanding that due to the daisy chain link, the CMU numbering is tied to a fixed location (slot or tray) molded into the pack, and this number is stored in the eeprom on the CMU.

Unless you think or are sure that automatic CMU renumbering is a thing--but is it, without using a MUT3?

 

[alviseven]

Did you mean to swap the cells to another location but leave the CMU board in the same "slot" location?

It was my understanding that due to the daisy chain link, the CMU numbering is tied to a fixed location (slot or tray) molded into the pack, and this number is stored in the eeprom on the CMU.

Unless you think or are sure that automatic CMU renumbering is a thing--but is it, without using a MUT3?

no swap cells and keep the CMU in the same place that way if I get a CMU fault in the new position we know it is the cells but if we get a CMU fault in the same place then we can suspect the board is faulty.

You said that you cannot swap the CMU around

 

[MickeyS70]

I keep forgetting about the daisy chained auto-numbering function, the CAN bus doesn’t care in which order the CMU’s are. @kiev is right, moving CMUs around might not work, therefore keep them in place and swap the cells instead.

 

[coulomb]

My guess is that the daisy chained wiring is only used when the auto-renumber facility is used. [ Edit: Looks like I may have guessed wrong; see next post. ]

I assume that swapping two CMUs is a lot easier than swapping 16 cells. So I would give it a try just swapping the CMUs. If the car doesn't like it, it will soon let you know. Hopefully you don't have to reassemble everything to test.

 

[MickeyS70]

My guess is that the daisy chained wiring is only used when the auto-renumber facility is used.

It seems to do a function check and won’t go ready if it fails. One member reassembled a pack but forgot to label the modules initially, he couldn’t put them back in the right order and had to go to a dealer to re-ID the boards, but maybe that was just a fluke, see more below

I assume that swapping two CMUs is a lot easier than swapping 16 cells.

It’s almost the same process, open two 8 cell modules, swap CMU boards and reassemble. The only difference is that the complete modules go back to the opposite location thus keeping the CMUs in the original order

So I would give it a try just swapping the CMUs. If the car doesn't like it, it will soon let you know. Hopefully you don't have to reassemble everything to test.

If you have time for a quick experiment for the benefit of confirming a theory:
- swap the two 4cell modules as they are
- move the pack up so you can connect the data cables
- connect the 12V aux (but make sure the connections under the seats are made beforehand)
- turn on the key but don’t try to start

If the ‘fuel gauge’ populates bars, then the BMU seems happy with a CMU swap, verify cell voltages with OBD.

If it doesn’t, reverse the changes and try again, if that solves the issue then the CMUs have to be arranged in the correct order.

 

[alviseven]

So have dropped battery pack which is getting easy having done three drops in past 6 weeks!
Removed cmu 09 block and measured cells compared to what hobdrive read out said. Hobdrive was little variable with cells DEF being slower to get to the whole pack voltage
Measured all voltages with digital meter and all shown 3.97v and all seem to be even.
Removed cmu 10 as it was next to it and measured these cells too. All read 3.97v too.

Swapped these two cmu pcbs and replaced in the proper order so cmu number in the correct place.

Raised pack back to car and connected up all except the chademo HV leads and the fan cover back corner lead as cover was not on

Hobdrive still lights up with same cmu09 battery voltage fault so can only assume that I have a cmu board with issues and now need to advertise for one preferably with 09 register


About to try the cm12 swap with cmu06 four pack as outlined by mickey. Report back.

 

[alviseven]

About to try the cm12 swap with cmu06 four pack as outlined by mickey. Report back

Decided against disturbing the cell packs which were not giving me grief. As cmu09 and cmu10 were loose anyway swapped these over so the cmu were transposed cmu09 board was in cmu10 place.

Reconnected as per Mickey instructions and got to first key on ignition to find that all battery cells on the meter scale light up no issue.

Opened up hobdrive and cleared existing dtc codes and waited

NOTHING happened no further dtc codes brought up.

Looked at the screens for cmu 10 and 09 and battery voltages look pretty decent and similar to what they were before.

Not dared to turn car to drive to test can anyone confirm any issues with this dare not blow up stuff!

So it looks like Mickey was right and the cmu do not need assembling in order but assume this is different to the numbering of the cmu pcb boards. So if I think cmu09 is duff I still need a cmu09 numbered replacement but it does not have to.be in order within the pack just has to be the correct range of numbers not two the same ID within a car pack.


Dare I start car?

 

[MickeyS70]

NOTHING happened no further dtc codes brought up.

Looked at the screens for cmu 10 and 09 and battery voltages look pretty decent and similar to what they were before.

So if I think cmu09 is duff I still need a cmu09 numbered replacement but it does not have to.be in order within the pack just has to be the correct range of numbers not two the same ID within a car pack.

Thanks for volunteering to bring clarity to the CMU order question. The ultimate test would be to connect HV cables and see does it go READY and drive, but so far, so good.

You’re right, with matching cell voltages, the issue is with CMU09 or in particular its LTC chip.

 

[alviseven]

The ultimate test would be to connect HV cables and see does it go READY and drive, but so far, so good.

hv leads connected but not dared turn to drive. Shall l? What potential damage could I do.

 

[MickeyS70]

hv leads connected but not dared turn to drive. Shall l? What potential damage could I do.

None, if there is a problem it won’t go READY…

 

[alviseven]

car goes into drive no issues and still not DTC raised on hobdrive
cannot drive as battery pack separate to the car - best not to even get the car to start to move I think incase especially as the pack and car are 6ft up in the air
Cells all look pretty stable and decent too

not planning to move them about as they seem to work as they are - they are labelled for anyone else who goes into the pack after me.

Think I will assemble the battery pack and see if it will drive better.

Still does not charge on the early charger - this does drop out and goes to orange code - initially this was due to the battery coolant fan but this was cleared and not returned

Next job i think is to return the original OBC to the car now the battery cmu 09 fault seems to have gone away.

Hopefully under load testing the battery drop out fault may go away too

Perhaps today has been a good day.

 

[kiev]

It appears that you have an intermittent fault related to CMU 9, as it was faulting in original location, then with a module swap; and then no fault when the location was swapped. That hasn't solved anything or gotten to a root cause.

Inspect the daisy chain wiring for damage and connector contacts for corrosion. The CMU board contacts are gold plated but there could still be corrosion on the battery side.

What were the errant cell voltages reading before you "fixed" it?

 

[MickeyS70]

Still does not charge on the early charger - this does drop out and goes to orange code - initially this was due to the battery coolant fan but this was cleared and not returned

Came across another 2010 car with a 2015 OBC see post #115 and following below

https://myimiev.com/threads/the-tro...r-for-on-board-charger-obc-thread.4079/page-6

Starting to wonder if this model (9481A092) is direct AC supply only, does anybody know?

 

[kiev]

and post #119 is good too.

That OBC has the cover of the Gen 2 units, but inside it is still Gen 1 with the single waffle plate. Some of the early 2009-2010 models used a Gen 0 which had several mini waffle pads.

It would be an interesting test if someone with a Gen 1 or 2 OBC that normally uses a black-box EVSE, could try charging with a "plain" cable-only EVSE cord, to see if it would work. And vice versa. That would indicate that the code has been programmed into the units to accept either type.

 

[alviseven]

Started on car today
Put to DRIVE and moved an inch each way.
Instantly cm09 fault comes back into dtc codes.
Now does not go away.
However cells are still registered on the battery meter but once the cm09 fault is registered on the system the battery dash board meter goes blank even if it will still drive but with the orange fault code.
Guess once I have checked connections I might as well return the blocks back to their original positions and change battery packs back over sonthe pack is not confusing!

Anyone out there got a green CMU board which preferable is ID 09 call this an advert! Think my car is stuck till I find one.

 

[MickeyS70]

Anyone out there got a green CMU board which preferable is ID 09 call this an advert! Think my car is stuck till I find one.

Might get on to @martinwinlow, he has a similar issue/solution and may even have a spare LTC chip (post #29)
https://myimiev.com/threads/help-please-with-2009-i-miev-p1a33-p1a30-p1a39-p1a76-p1a52.5578/page-2