Drive battery degradation and the out of whack IMO BMU.

[Trex]

As part of my ongoing research into the drive battery as mine is showing in the 70s percentage of battery health I wish to run this past the members here what I have learnt so far.

This research is far from finished for me.

But lets bring some quotes from the I think German that has moved to Australia that I first saw on Youtube.

I've got a MY14.5 here in Australia, bought it end of 2015 from a dealer. Mid 2017 I lost range and could do only 35km (21mi) on a full charge. I complaint with another dealer and they did a battery cell smoothing procedure and a BMS reset. ........... The capacity was down to 80% state of health already with no more then 58000km (36000mi) driven.

After the procedure I had my full capacity back with 50km+ on pure EV range, but... well other problems occured.

Yes, the batteries capacity went back from 80.5% SoH to 99.7% with 38Ah. The first drive I did was 59km on pure EV, half of that on a highway. The very next day though I had already lost 0.1Ah again and until now (3 months later), I'm back to under 90% SoH...

Now I have left some of his quotes out here but according to Bert he was getting 35km at 80% state of health to getting 59km, half of that on a highway and 99.7% battery health just by doing a battery cell smoothing procedure and a BMS reset.

Now anko first brought up a refreshment procedure described by jaav as "A service offered by Mitsubishi balancing the cells" back in the other thread I started which I think is the battery cell smoothing procedure described above.

Here is anko's quote:

June, 2016: 32.5 / 33.4 (before / after refreshment procedure)

So a little lift for anko for "Mitsubishi balancing the cells" to a big lift for Bert for "battery cell smoothing procedure and a BMS reset."

So if I am right they are the same why would a BMS reset lift it up so far. For sure a reset should just "zero" the BMU but not get a increase in range like Bert.

So what is happening?

Well I think the BMU is getting out of "whack".

How do you measure battery capacity?

From my instruments for measuring cell capacity on my RC chargers and Ev pushbike they are measuring current and voltage and time to give a capacity reading ie top to bottom discharge.

But the PHEV never fully discharges its drive battery. So how does it work out its capacity properly?

By "guessing" using algorithms I think.

Now the BMU data list from my battery report from the MUT-III show:

No 400 Cell current capacity (measured) always 0.0 Ah on my PHEV

No 401 Cell current capacity difference always 0.0 Ah on my PHEV.

So I think Mitsi can do full capacity check with the MUT-III and will check with them next week to do so on my PHEV.

I am probably going to also ask for a BMU reset after that test.

Will let you know how I go.

Regards Trex.

Edit change name for searching purposes.

 

[AndyH]

I think you nailed it, Trex. I've seen the same thing in my ebike, smart, and (oddly enough) laptop computers.

When I kept my smart between 40% and 80% SOC, the range estimates would get stranger as the month went on. It got better when I charged to 100%, but it wasn't fixed until I drive the car down below ~10% SOC and charged to 100%.

It appears that 'cell smoothing' is just balancing the battery and saving the updated capacity in the computer. The range loss was real, and some of it was due to degradation, but some was also due to the management computer's 'guess' about state of charge diverging from the reality of the cells. 'Smoothing' can't restore capacity physically lost in the cells, but can 'restore' capacity the computer wouldn't let us use because it thought the battery was worse than it was.

Open the pod bay doors, Hal.
I'm sorry Dave - you've only got 1% charge left and I need that to run the intercom.
Dangit Hal - I know I plugged the ship in last night! (Dave floats around while digging for the end of the StarshipWatchdog cable...) :lol:

 

[Trex]

I think you nailed it, Trex. I've seen the same thing in my ebike, smart, and (oddly enough) laptop computers.

When I kept my smart between 40% and 80% SOC, the range estimates would get stranger as the month went on. It got better when I charged to 100%, but it wasn't fixed until I drive the car down below ~10% SOC and charged to 100%.

It appears that 'cell smoothing' is just balancing the battery and saving the updated capacity in the computer. The range loss was real, and some of it was due to degradation, but some was also due to the management computer's 'guess' about state of charge diverging from the reality of the cells. 'Smoothing' can't restore capacity physically lost in the cells, but can 'restore' capacity the computer wouldn't let us use because it thought the battery was worse than it was.

Open the pod bay doors, Hal.
I'm sorry Dave - you've only got 1% charge left and I need that to run the intercom.
Dangit Hal - I know I plugged the ship in last night! (Dave floats around while digging for the end of the StarshipWatchdog cable...) :lol:

StarshipWatchdog cable, took me a while to get it. :lol: :lol: I am tired.

 

[anko]

Hi Trex and others,

There are 10 slots in the BMU ECU for storing "Automatic Capacity Measured Result":
- 402.Auto capa measured result his.1
- 403.Auto capa measured result his.2
- 404.Auto capa measured result his.3
- 405.Auto capa measured result his.4
- 406.Auto capa measured result his.5
- 407.Auto capa measured result his.6
- 408.Auto capa measured result his.7
- 409.Auto capa measured result his.8
- 410.Auto capa measured result his.9
- 411.Auto capa measured result his.10

To get data from these, send this:

ATZ
ATL1
ATH1
ATE1
ATSH761
ATFCSH761
ATFCSD300000
ATFCSM1
2114

For me, this brought back:

2114
762 10 16 61 14 01 20 01 4C
762 21 01 90 01 90 01 90 01
762 22 90 01 90 01 90 01 90
762 23 01 90 41 74 2F EC 41

When you take bytes A+B (0120), convert to decimal (288) and divide by 10, you get 28.8
When you take bytes C+D (014C), convert to decimal (332) and divide by 10, you get 33.2
When you take bytes E+F (0190), convert to decimal (400) and divide by 10, you get 40.0
Same for G+H, I+J, K+L, M+N, O+P, Q+R, S+T you get the same.

So, 28.8, 33.2, 40.0, 40.0, 40.0, 40.0, 40.0, 40.0, 40.0, 40.0.

A few years back, before the first time the procedure was applied to my car, all slots all showed 40. But now, the procedure has been applied twice, first time it came back with Current Capacity 33.4, the second time with current capacity. A 'misalignment' of 0.2, but still.

 

[Trex]

Hi Trex and others,

A few years back, before the first time the procedure was applied to my car, all slots all showed 40. But now, the procedure has been applied twice, first time it came back with Current Capacity 33.4, the second time with current capacity. A 'misalignment' of 0.2, but still.

Hi anko,

Are you saying No 400 and 401 have now been populated from the refreshment procedure or as jaav called "A service offered by Mitsubishi balancing the cells".

Regards Trex.

 

[anko]

Hi Trex and others,

A few years back, before the first time the procedure was applied to my car, all slots all showed 40. But now, the procedure has been applied twice, first time it came back with Current Capacity 33.4, the second time with current capacity. A 'misalignment' of 0.2, but still.

Hi anko,

Are you saying No 400 and 401 have now been populated from the refreshment procedure or as jaav called "A service offered by Mitsubishi balancing the cells".

Regards Trex.

I believe it id's merely a recalibration procedure, but yes.

 

[Trex]

Hi Trex and others,

A few years back, before the first time the procedure was applied to my car, all slots all showed 40. But now, the procedure has been applied twice, first time it came back with Current Capacity 33.4, the second time with current capacity. A 'misalignment' of 0.2, but still.

Hi anko,

Are you saying No 400 and 401 have now been populated from the refreshment procedure or as jaav called "A service offered by Mitsubishi balancing the cells".

Regards Trex.

I believe it id's merely a recalibration procedure, but yes.

Ok thanks anko.

 

[Trex]

Spoken to dealer and Mitsubishi Australia, who were extremely helpful, to start the ball rolling.

Dealer says I will probably get a new PHEV. I laughed but said I would just like you to reset the BMU please.

Will not do it until further checks are made and given approval from Mitsubishi Australia.

And so it begins. Thursday 1st March the big tests on the battery commence.

I cannot wait.

I'm so excited and I just can't hide it............

 

[Regulo]

I'm so excited and I just can't hide it............

Are you about to lose control and like it though?

 

[anko]

Now sending 2115 to 761 (BMU) in my case gives back:

762 10 16 61 15 50 07 31 15
762 21 00 00 00 00 00 00 00
762 22 00 00 00 00 00 00 00
762 23 00 00 41 74 2F EC 41

Last BMU calibration procedure was done at 131.115 km. Dividing this number by 6.4 (*) yields 20.487 (after rounding). Converting that value to hex yields 0x5007, or bytes A+B. it seems bytes A-B have the odometer for last time procedure was performed.

The BMU calibration before that was done at 80.417 km. Dividing this number by 6.4 yields 12.565 (after rounding). Converting that value to hex yields 0x3115, or bytes C+D. it seems bytes A-B have the odometer for the previous time before that procedure was performed.

* It was a wild guess, but it gives us the hexadecimal representation of odometer in 0.1 km, but then shifted 5 bits to the right, to save space. This approach allows 'them' to register odometer reading up to 419.424 km, with an accuracy of 6.4 km, while using just two bytes

 

[Trex]

I'm so excited and I just can't hide it............

Are you about to lose control and like it though?

:lol: :lol:

No I do not think I will lose control but Hal the BMU computer (sorry AndyH for pinching your quote) may lose control when I hopefully reset him.

BTW how are you going Ray? With that signature about the nuclear-powered flying car always makes me remember you.

Regards Trex.

 

[Trex]

Now sending 2115 to 761 (BMU) in my case gives back:

762 10 16 61 15 50 07 31 15
762 21 00 00 00 00 00 00 00
762 22 00 00 00 00 00 00 00
762 23 00 00 41 74 2F EC 41

Last BMU calibration procedure was done at 131.115 km. Dividing this number by 6.4 (*) yields 20.487 (after rounding). Converting that value to hex yields 0x5007, or bytes A+B. it seems bytes A-B have the odometer for last time procedure was performed.

The BMU calibration before that was done at 80.417 km. Dividing this number by 6.4 yields 12.565 (after rounding). Converting that value to hex yields 0x3115, or bytes C+D. it seems bytes A-B have the odometer for the previous time before that procedure was performed.

* It was a wild guess, but it gives us the hexadecimal representation of odometer in 0.1 km, but then shifted 5 bits to the right, to save space. This approach allows 'them' to register odometer reading up to 419.424 km, with an accuracy of 6.4 km, while using just two bytes

Hey anko. Want to reset your BMU as well? It might be worth a try.

 

[anko]

You are saying you know how?

 

[Trex]

You are saying you know how?

Yep see your dealer and ask nicely. :lol:

But seriously, after asking mine what can be done to the BMU on the MUT-III he said Battery smoothing, BMU reset and BMU replace.

Tomorrow I hope to get mine reset and may be smoothing.

 

[Trex]

Now in my first post in this thread I stated:

But the PHEV never fully discharges its drive battery. So how does it work out its capacity properly?

By "guessing" using algorithms I think.

So what data does the algorithms use for "guessing" the drive battery capacity in the PHEV?

I think they are using age and cycles and amount of charge and discharge of the drive battery as these are stored by the PHEV computers and show up in the reports.

Now Bert who I quoted in my first post here said:

Yes, the batteries capacity went back from 80.5% SoH to 99.7% with 38Ah. The first drive I did was 59km on pure EV, half of that on a highway. The very next day though I had already lost 0.1Ah again and until now (3 months later), I'm back to under 90% SoH...

Now notice that last bit about "The very next day though I had already lost 0.1Ah again and until now (3 months later), I'm back to under 90% SoH... "

How come he is, after a reset of the BMU, getting such a quick drop in battery health?

Well I think anko explained it when he said to me in another thread:

But my own tools tell me about 4900 or so. But these are also many partial charges.

Does it have an BMU data list number.?

Got a feeling I cannot find it because I am looking too hard. :lol:

No, it doen't. But it has a OBC data list number Kind a make sense as the OBC takes care of charging (at least A/C charging).

So the number of cycles charged is stored under the OBC data list and unless Bert got the OBC computer reset as well as the BMU computer reset he would be getting a quick reduction in battery health as this is ie cycles charged, one of the data used for "guessing" the capacity I think.

Tomorrow my PHEV goes in for drive battery tests and I hope to get the BMU and the OBC computers reset.

For people not knowing:

BMU Battery Management Unit
OBC OnBoard Charger

Regards Trex.

 

[Hypermiler]

Any news Trex?
Cheers
H

 

[Trex]

Any news Trex?
Cheers
H

Hi Hypermiler,

Only have a minute but the dealer cannot get his MUT-III or Multi Use Tester (Mitsubishi diagnostic device) to "talk" to my PHEV's BMU to do reset or Cell smoothing or Auto capacity measured procedure.

I could have a "faulty" BMU or the dealer may have a problem with his MUT-III. He can read data but not "write" data (for control purposes) to the BMU in my PHEV. The small dealer I am using has only 1 MUT device.

Now from what I have seen in the Workshop manuals the BMU is small box down the back of the PHEV near the right rear wheel-well and a person in Western Australia has had his BMU replaced on his PHEV.

He did NOT have battery problems but they replaced his BMU for a "EV system service required" error message.

This is what he quoted from another forum.

"We've had no further problems with our PHEV since that time. The fascinating sequel is that our range on battery has increased by about 20% since that repair! I regularly travel from home to our yacht in Fremantle, arriving there with between 5 and 12km residual range on battery (having driven the whole dist ance on battery alone), plug in at the Yacht Club and recharge (usually fully) and then drive home again, in the past arriving home with 10-15km left in the electric tank. Since the BMS was replaced, my fully charged, startup battery range is usually reported as 62km (up from 45km) and I now arrive at the Yacht Club with 20-25km left on battery range and arrive home with 25-30km left on battery range.

A week or so after getting the vehicle back from Melville Mitsubishi, I called them to tell them that my battery range had increased and to ask what they had done to achieve this. They couldn't explain my observations. I asked if perhaps my 2015 PHEV BMS had been replaced with a 2017 unit, but they were adamant this wasn't the case. I suspect that, notwithstanding the dealer's denial, this is what has happened. I understand the 2017 PHEV has better battery range compared with previous models and that this was achieved with better BMS software so in the absence of any other explanation, I'm going with this one."

So in his case a new BMU has helped get increased range so I am still thinking the BMU is getting out of "sync" with the real capacity of our batteries.

I have my PHEV back and am organising to take to another dealer towards the end of the month.

Will keep you all informed.

Regards Trex.

 

[Trex]

Taken from Workshop manual showing the BMU location.

 

[Trex]

This is from the same website as above but from their Technical information manual.



Please note the Calculating total amount of battery (including capacity adjustment) using History of battery usage and History of discharge and charge
on the bottom left of the BMU.

 

[Bert]

Guys, I've posted mi findings already in another thread here and explained it also in some of my recent videos. After the procedure the total cycle count has not changed and the cars is not digging deeper into the lower 30% than before. Neither has the upper voltage changed and is still at 4.1V with a fully charged battery. The range though was incredible, so further assumptions need to be made what exactly happened... reset, calibration or smoothing? So far we have 3 cars here from Australia which have done this same procedure and went back to 100% SoH. I'm in email and forum contact with them to follow this up...