Page 1 of 1

Feb 2012 Miev with Pack Trouble at 40k miles

Posted: Sat Nov 04, 2017 7:22 pm
by kiev
In some other threads i had posted about EVSE and charging issues due to dirt and debris in my port, and a nest of fireants in the EVSE handle. Thought i had all that cleaned up and working , but it only charged to ~10 bars even though the pack voltage was indicating 360 (which is 100% full).

So i went for a test drive and the car went into Turtle after only 5 miles. Connected CanIon and found cell #47 was ~0.2 volts below everybody else, so that would explain nearly all the weirdness except the ants, but now i have to wonder if there are more ants in the pack shorting out cell 47? And i to wonder if all the other 2012 folks having charging issues are also having defective cells making an appearance?

i had connected the clone version of the MUT3 tool and checked for DTCs before making the test drive, and none were present in the EV-ECU or the BMS, but i'm fairly sure that there will be some now.

i'll check it again with the MUT and if the cell is really out of line, then i will investigate if the 10 yr warranty is still valid...

Re: Feb 2012 Miev with Pack Trouble at 40k miles

Posted: Sat Nov 04, 2017 8:20 pm
by siai47
What does the CANION report when the vehicle reaches full charge before you drive it? What is the difference between the lowest and highest cell voltage? Drive the car with the CANION active and watch the cell voltages. Accelerate with full power and see if one cell drops dramatically vs. the others. One cell hitting cutoff voltage will trigger the turtle. The fact that you don't have all of the "bars" on your battery gauge suggests that you have a really out of balance pack that cannot be fully charged before some of the cells are exceeding the high voltage limit. When fully charged and balanced, all cells should be at or slightly exceed 4.105 volts. If you note one lower (.2 volts means you are seeing one or more under 4 volts) note the cell number(S) and see if they go very low when you drive. BTW--you are charging until the vehicle shuts down the EVSE---even if it seems to be taking several additional hours?

Re: Feb 2012 Miev with Pack Trouble at 40k miles

Posted: Sat Nov 04, 2017 10:07 pm
by JoeS
kiev - time for a trip to the Mitsu dealer, and presumably a new battery pack, under warranty.

Re: Feb 2012 Miev with Pack Trouble at 40k miles

Posted: Sun Nov 05, 2017 9:23 am
by kiev
siai47 wrote:What does the CANION report when the vehicle reaches full charge before you drive it? What is the difference between the lowest and highest cell voltage? Drive the car with the CANION active and watch the cell voltages. Accelerate with full power and see if one cell drops dramatically vs. the others. One cell hitting cutoff voltage will trigger the turtle. The fact that you don't have all of the "bars" on your battery gauge suggests that you have a really out of balance pack that cannot be fully charged before some of the cells are exceeding the high voltage limit. When fully charged and balanced, all cells should be at or slightly exceed 4.105 volts. If you note one lower (.2 volts means you are seeing one or more under 4 volts) note the cell number(S) and see if they go very low when you drive. BTW--you are charging until the vehicle shuts down the EVSE---even if it seems to be taking several additional hours?


Some answers for yesterdays events: The i909 MUT-clone was indicating pack at ~360 before i drove it. The delta V was about 0.2. When driving the cell 47 would rise up nearly to the other cell level under regen, and dropped dramatically under acceleration--i couldn't read the values while driving, but will try to get a helper to read the screen. i had plugged in the EVSE and let it run until it shut off.

Now this morning i pulled trouble codes using the i909 and found DTC P1A44 set in both the EV-ECU and BMU. Made a reading of the stream data for CMU07, module voltage at 32.02 and all cells were at 4.02 except cell C (#47) at 3.87. Other data indicated the board was Normal (Monitor IC, Voltage Sensor, Reference Voltages, CAN, ADC ref voltage, etc.). Compared data with CMU08, module at 32.21 and all cells A thru H were at 4.02.

So i cleared the DTC, and then read the TPMS data stream just for fun (expensive tire gauge), 42 41.40 42.80 42 0 were tire 1-5 pressures, warning threshold at 28.6 psi and release warning at 30.8 psi.

The key must be at ON in order to get readings even though the i909 powers up just plugging into OBDII.

So i plugged in the EVSE with i909 and key ON, and no charging occurred (no relays or water pump, etc). i read the BMU data and found Target Cell Voltage was 3.87, Battery Max Power Output 63.5 kW, Max Input Power 29.25 kW, Pack Voltage at 354 VDC.

Key OFF and plugged in EVSE again, this time charging started. Read the EVCU datastream: DC/DC f/b ON; charge contactor OFF; Main contactors (+/-) ON; charging permission ON; charging start OFF; Made a voltmeter check of aux battery at 14.35.

Read the BMU datastream: Battery Current 7A; Balancer Drives: CMU07 all ON except cell C OFF, CMU08 all ON; after several minutes of charging the BMU Target Cell Voltage was up to 3.94.

So the balancing drives appear to be ON for all cells except #47. Will see how it goes--maybe the cell will be okay.

Concerning the DTC P1A44, BMU Leakage Detection, when i cleared it with the i909 then the warning lights on the dash turned off. i wonder if i induced that fault by trying to externally charge the aux battery while the EVSE was plugged in yesterday?

Amazon has the i909 for $114, it is a nice tool to have in a situation such as this...It has a few features that could be updated and is missing some of the possible CMU data fields, and i intend to ask if they will make these changes in an update thru the USB port.

Re: Feb 2012 Miev with Pack Trouble at 40k miles

Posted: Sun Nov 05, 2017 7:35 pm
by kiev
Thanks to siai47 for the suggestion of a pack out-of-calibration. It got me to thinking that i haven't done a calibration charge in quite some time-- so my goal today was to drive around and drain the pack without getting stranded or stuck in Turtle. And this afternoon i was able to drive ~49 miles on 10 bars (RR was 37) taking it all the way to zero bars, blinking fuel gauge and --- showing for RR.

At the end Canion indicated it was down to 10% SoC with pack at 342. Most cells at 3.895 with #47 at 3.475. i909 reported cell delta V as 0.42 volts and the BMU had DTC P1A4B "voltage of each battery cell abnormal" which turns on the HV warning (car with !). i stopped and cleared it during the drive but it likely came back as i had ! showing for half the trip, but it wasn't in turtle--i could bury the needle at both ends, acceleration and regen.

i had various other DTCs that i cleared before departure: BMU P1A4D = IGCT Line Voltage Error, which gets triggered if the power supply control voltage for IGCT is either less than 8 or greater than 16V. IGCT is a gate-controlled turn-off switch which turns off like a transistor but conducts like a thyristor, abbreviation for integrated gate-commutated thyristor. This code can cause charging to halt and has several possible causes related to the Aux battery, the aux battery terminals, bad contacts or wiring with relay F in the front relay box next to the aux battery, or a failed BMU. i checked everything i could and demate/mated connectors and relays, then cleared the code and moved on.

i lost power one time under hard accel which threw BMU P1A44 leakage code. Pulled over and cleared it, then moved on.

Also when i got home there were 2 EV-ECU codes, P1A44 BMU leakage, and P1A2F Cell Malfunction, which is a generic code for when any of the diagnosis codes P1A4B, P1AC7, P1AC8, P1AC9, P1ACA, P1ACB, P1ACC, P1ACD, P1ACE, P1ACF, P1AD0, P1AD1 or P1AD2 is set to the BMU, the diagnosis code No. P1A2F will be set in the EV-ECU.

The P1A4B code was mentioned above where it was cleared, and those other codes P1AC7 - AD2 are related to each of the 12 cell modules when the "Difference in resistance values inside battery cells is great."

The other good news (besides not getting stranded) is that level 2 charging started on the first try (so none of the leakage issues that prevent charging have surfaced).

EDIT: it is still not balanced--have 0.15 V delta on cell 47. It is up to 360V with only 40 RR and 10 bars. Need the balancing drivers to turn on, but they were all off. Is there a trick to force balancing?

EDIT2: So i drove it around monday morning to drop the pack to 354, then tried level 1 charging--and the balancer drives are now showing ON for all cells except #47. i left canion and the i909 plugged in at the end of the drive, pulled the key out and plugged in level 1 EVSE, and left it for the day.

Later this evening when i got home the EVSE was blinking and done, the pack was at 66%SOC, 11 bars, 360 VDC, and cell #47 was still 0.150 V below all the rest. Since the balancers only drain at 0.1 Amps, it may take several days to bleed all the high cells down to the 3.95. At least i gained 1 bar back, so i'll repeat this procedure of draining the pack and charging with level 1 hopefully with balancing active.

Re: Feb 2012 Miev with Pack Trouble at 40k miles

Posted: Tue Nov 07, 2017 9:51 am
by ed5000
kiev wrote:Later this evening when i got home the EVSE was blinking and done, the pack was at 66%SOC, 11 bars, 360 VDC, and cell #47 was still 0.150 V below all the rest. Since the balancers only drain at 0.1 Amps, it may take several days to bleed all the high cells down to the 3.95. At least i gained 1 bar back, so i'll repeat this procedure of draining the pack and charging with level 1 hopefully with balancing active.

Any luck so far? This is an interesting way of balancing the pack - by bleeding down all the good cells but I guess there's no other easy way do it short of lowering the pack and putting cell #47 on a charge. Sometimes I wish the battery cell access was easier. Good luck!

Re: Feb 2012 Miev with Pack Trouble at 40k miles

Posted: Tue Nov 07, 2017 10:10 am
by kiev
Not yet. This morning i drove 25 miles to drain the pack again (down from 11 to 3 bars, 10 A-Hr remaining). When i connected level 1 EVSE the charging began but the balancer drives were all OFF.

i was hoping they would all come ON except #47 in order to let it catch up. Maybe that will happen later today when the pack gets up closer to the top.

i'm guessing that i need to bleed about 3 to 6 A-Hr from the high cells, which would take 30 to 60 hours with the balancer drives ON.

Another option is to take it to dealer and do the special "cell voltage smoothing" procedure with the MUT3, which is supposed to take 2 hours. But i don't see how it could drain any faster with 41 Ohms in the bleed circuit for each cell.

Re: Feb 2012 Miev with Pack Trouble at 40k miles

Posted: Sun Nov 12, 2017 2:12 am
by matija
Hey, doesn’t the BMS trigger cell balancing, when the pack is down to 2 bars ? That would make it logical that if was OFF with dischage beeing done to 3 bars.

And if that is perhaps percentage based 2/16 would mean roughly 1/11 in you case, needing you to drain it to 1 bar.

Re: Feb 2012 Miev with Pack Trouble at 40k miles

Posted: Sun Nov 12, 2017 10:09 am
by kiev
From what i could tell by monitoring with the mut-clone, the balancers come on during the final slow stage of balancing ("CV") and not during the faster bulk charging ("CC").

i have a second kiev (white/blue 26k miles) that is showing the same cell module CM07 with #47 slightly lower than all the rest, by ~0.015V, but it is charging to 16 bars, shows 100% SoC and 67 miles on RR. i have done both level 1 and level 2 charging with that car and the imbalance remains.

It seems too coincidental for 2 cars to experience the same low cell. CM07 is the first module above the midpoint of the pack, and cell C (47) is the third cell above the bottom of that module. my guess is that cell C is likely providing power for the microcontroller in that module. Over a long period of time that little bit of current drain has noticeably depleted cell C.

Considering the very low balancing current (0.1 A) it may be that it takes 20 to 30 charging cycles to allow the low cell to catch up.

Edit: Looking back at some canion screen shots, this same cell #47 was the laggard of the pack back in June 2016 by 0.015 V. It may just be a weak or bad cell that is now showing itself to a greater degree. i'm going to try several dozen balance charging sessions before throwing in the towel...