Lindqvist method : DIY BMU Battery Reset

[MadTechNutter]

If I wait until 4 relays clicks, charger will not start charging.
If hold a second longer than the 4 clicks, there are 2 more clicks in the car in short succession and the charger's 'Charging' LED goes on.

Maybe the problem is a delay of the response time in the charger and we are using different chargers. Please compare your charger to mine and let me know.

I will be building my own charger soon.

 

[fprigge]

Here is my theory of what is happening... I don't have a deep knowledge of the phev, but I do design HD EVs (trucks, busses, etc) for a living...

The initial 12v depletion might have no effect at all (or maybe it has), I suspect that it is at the end of the charge that the magic happens. When the HV contractors open, at the end of the charge, the dcdc converter stops producing 12V instantly. Since the lead acid battery is not present, the 12V bus stop very quickly. The BMS (BMU) is writing in its non volatile memory the capacity at that moment, having just ended a charge. Either it never had time to complete a write, or it writes some corrupted data.

At the next 12V battery connection, the BMS boots and computes the CRC of those stored structures and realize they are corrupted. It then reverts to loading the default hard-coded values.

 

[MadTechNutter]

Here is my theory of what is happening... I don't have a deep knowledge of the phev, but I do design HD EVs (trucks, busses, etc) for a living...

The initial 12v depletion might have no effect at all (or maybe it has), I suspect that it is at the end of the charge that the magic happens. When the HV contractors open, at the end of the charge, the dcdc converter stops producing 12V instantly. Since the lead acid battery is not present, the 12V bus stop very quickly. The BMS (BMU) is writing in its non volatile memory the capacity at that moment, having just ended a charge. Either it never had time to complete a write, or it writes some corrupted data.

At the next 12V battery connection, the BMS boots and computes the CRC of those stored structures and realize they are corrupted. It then reverts to loading the default hard-coded values.

The 12v filter capacitors can usually store energy for a second after power down and mostly longer as MCUs usually crash below 2V.
Writing this small amount of data even with an ancient 4MHz clocked MCU would be a matter of less than 100µseconds.
The MCU's watchdog brownout detector will detect the power down well before the system can become unstable and this will prohibit corrupt data being written.

What I mentioned above was bringing the MCU of the BMU or the ECU in a state that the software designer may have not considered with his code.
So the MCU is running fine but in a state for example: 4 relays instead of the required 6 , yet charge has commenced , while the ECU or BMU are still waiting for the other 2 relays and one of them might not be aware of the charge actually happening.
In other words a typical software bug, which could as you said, then reset something to a default value.

As both the BMU and the car's ECU could be interacting together this could be a very complex matter.

 

[elm70]

Question

I normally have the switch for internal light set to have cabin light when door are open.

When I do this procedure I need to keep the trunk open .. so by default the light would be on.

When I did the long procedure, it was in the night, and I did switch off the internal light, else it would have been odd to leave the car with lights inside my garden

But .. now reading above, it might be a relevant point.

If the magic happen when the charging process is over, then having running light will quickly empty the eventual capacitor for protection.

How did you handle the lights while doing this procedure ?
Do you keep the rear trunk open or close ? (if got closed it is a bit annoying to climb on the back for reconnect the 12 battery)

PS: Our PHEV might have different firmware ... since firmware update happen at the service without our notice and not every service center do this all the time ... another reason why we get different result

 

[michalsarna]

If I wait until 4 relays clicks, charger will not start charging.
If hold a second longer than the 4 clicks, there are 2 more clicks in the car in short succession and the charger's 'Charging' LED goes on.

Maybe the problem is a delay of the response time in the charger and we are using different chargers. Please compare your charger to mine and let me know.

I will be building my own charger soon.

I don't have original charger it was not with a car.
I've got this one: https://duosidaevchargers.com/products/220v-level-2-fast-charger-16-amps

 

[MadTechNutter]

Question

I normally have the switch for internal light set to have cabin light when door are open.

When I do this procedure I need to keep the trunk open .. so by default the light would be on.

When I did the long procedure, it was in the night, and I did switch off the internal light, else it would have been odd to leave the car with lights inside my garden

But .. now reading above, it might be a relevant point.

If the magic happen when the charging process is over, then having running light will quickly empty the eventual capacitor for protection.

How did you handle the lights while doing this procedure ?
Do you keep the rear trunk open or close ? (if got closed it is a bit annoying to climb on the back for reconnect the 12 battery)

PS: Our PHEV might have different firmware ... since firmware update happen at the service without our notice and not every service center do this all the time ... another reason why we get different result

I switched off all the lights and have a bit of blanket blocking the lock so I can close the tailgate without it locking.

I can confirm that there is NO voltage left in the 12V bus after removing the battery negative.
In the time it takes to walk to the front and measure the voltage on the jumpstart connecter in the already open fuse box the voltage had dropped below 0.005V within the error margin of a standard multimeter.
No electronics can be active. A volatile CMOS type storage memory requires 2V maybe still 1.8V.
The lowest voltage any electronics can operate is 0.2V. These are special joule thief circuits that use an inductor to pump the voltage higher.
Old germanium transistors(definitely not used in a PHEV) could run directly on 0.5V. Silicon transistors might work at 0.9V under circumstances, a bit lower only inside cores of PC CPUs but not MCUs that would be used here.

The moment the ESVE charger switches off, it is the same as disconnecting the battery from the 12V bus.
Different is if the BMU gets it's voltage internally from the AC/DC converter from the charge circuit. This has nothing to to with your Auxilary battery 12V bus as it is likely separated with diodes. The voltage can not discharge through the car and filter capacitors can hold it for a longer time. The backup capacitor power for volatile memory is separated with another diode again from a 3.3-5V regulator that feeds the logic and MCU.
Depending on the type of capacitor, data on volatile memory can stay resident from hours to months.

So you see it is all not that straight forward.

Those who had their service done at a non Mitsubishi dealer might have missed out on software updates but I am not sure how this affects warranty.
This could be from country to country different.
I will have to go through the service book that came with my vehicle and look up the companies. I noticed they were not all the same.

 

[michalsarna]

Question

I normally have the switch for internal light set to have cabin light when door are open.

When I do this procedure I need to keep the trunk open .. so by default the light would be on.

When I did the long procedure, it was in the night, and I did switch off the internal light, else it would have been odd to leave the car with lights inside my garden

But .. now reading above, it might be a relevant point.

If the magic happen when the charging process is over, then having running light will quickly empty the eventual capacitor for protection.

How did you handle the lights while doing this procedure ?
Do you keep the rear trunk open or close ? (if got closed it is a bit annoying to climb on the back for reconnect the 12 battery)

PS: Our PHEV might have different firmware ... since firmware update happen at the service without our notice and not every service center do this all the time ... another reason why we get different result

I was running procedure in garage so no problem with open tailgate. When first 12V disconnect period lights go off. After "short" 12v connection light go on, but will go off eventually - I think it can be setup in settings menu. In the morning when I was disconnecting the charger and reconnecting 12v battery, lights ware off but I haven't touch any doors. I'm not sure if (P)ark was lighten up before reconnecting 12v battery haven't checked. I'll check FW version and get back with answer

 

[Regulo]

Right, I'm stupid so indulge me, please! If my 5 year old battery has degraded to, say, 85%, how can this method restore it to full capacity? Or is it a case of the electronic gubbins thinking my 85% is now 100%? What does this mean in the real world, not the electronic one? Will I get back any actual EV range? Keep the answers simple, I told you I'm stupid!

 

[MadTechNutter]

Right, I'm stupid so indulge me, please! If my 5 year old battery has degraded to, say, 85%, how can this method restore it to full capacity? Or is it a case of the electronic gubbins thinking my 85% is now 100%? What does this mean in the real world, not the electronic one? Will I get back any actual EV range? Keep the answers simple, I told you I'm stupid!

It will not restore a battery that has truely degraded to lets say 85%.

It MAY correct an error that the system is reporting by resetting to original factory settings.

A degraded battery will show it's true face very soon again, based on performance.

A 5 year old battery showing still 85% is surely real degradation and that is pretty good.
I would be very happy if mine only had that and I would leave things alone.

 

[michalsarna]

Right, I'm stupid so indulge me, please! If my 5 year old battery has degraded to, say, 85%, how can this method restore it to full capacity? Or is it a case of the electronic gubbins thinking my 85% is now 100%? What does this mean in the real world, not the electronic one? Will I get back any actual EV range? Keep the answers simple, I told you I'm stupid!

My battery showed 84,7% (32.2Ah) capacity (after 5 years). With this method just after resetting car showed 105,3& (40Ah) of capacity but with Error.
After driving for ~9 km I've stopped the car, turned it off and on again. Error was gone but so was 2Ah. I was back @ 100% (38Ah).
After next 40-50 km I've checked and it was 99.7% (37.9Ah) and after next ~30 km I'm @ 99.5% (37.8Ah).
So the procedure works, but this first drop is significant. I think that it will stop dropping after some time and will keep higher rate then I initially had that was this 84.7% (32.2Ah). But time will tell.

In simple words I can confirm that range is greater than it was initially but we will see for how long.

 

[Regulo]

Thanks for those explanations. I really struggle with getting my head around all this. I'm used to small rechargeable batteries in torches!!!

 

[elm70]

Right, I'm stupid so indulge me, please! If my 5 year old battery has degraded to, say, 85%, how can this method restore it to full capacity? Or is it a case of the electronic gubbins thinking my 85% is now 100%? What does this mean in the real world, not the electronic one? Will I get back any actual EV range? Keep the answers simple, I told you I'm stupid!

I think this point has been discussed in 2 or 3 or even more posts in this thread ... check page 2 and 3

About the real SOH .. this is an interesting topic

I believe .. the only way to test real SOH is:

Fully charge the PHEV
Wait 2h for battery to rest and avoid any BMU madness with SOC above 100%
Drive untill EV range is zero.
Check with the WatchDog Apps the SOC
Let the car rest for 2h or more (especially in winter rest need to me longer)
Check with the WatchDog Apps the SOC ... if SOC has gone up ... it means the SOH is probably better then what is defined in the BMU ... if SOC has gone down .. possibly the SOH from BMU is an optimistic guess

The usage of conditional above is beacause SOC from BMU after a long rest is function of a BMU voltage to SOC mapping table .. which I believe change with time .. which normally expect 30% SOC with 3.82v per cell or above (in my car is looking is almost 3.85v per cell, which is "madness")

Actually the only real way to check SOH .. is to disconnect the main battery, take a couple of battery as sample .. and allow a full charge to 4.1v and a full discharge down to 2.75v .. using a proper charger/discharger equipment (which is cheap and common, especially for RC Model Hobbyist like me) .. the challenge is to disconnect the main battery, open the massive cover .. and get access to the individual cell ... but I would expect Mitsubishi to do this for car which got a new battery in the warranty replacement, and it might have provide them some good input for improve the BMU firmware .. but .. I doubt they want to share their knowledge in this area .. too much sensible information .. and I'm not sure how much they care if the BMU has pessimistic SOH, since for them does help on real battery degradation and less warranty to be cover ...

 

[elm70]

Thanks for those explanations. I really struggle with getting my head around all this. I'm used to small rechargeable batteries in torches!!!

I'm using Lithium battery (including self make battery pack), with charging / discharging in RC Model Airplanes since 2001

I also experience in the hard way what is the result to store for 5/6 years battery left in fully charged status .. battery capable to deliver over 10A current, now capable to handle no more then 1A when used on full load on an airplane (10 reduction factor in a RC airplane, it means there is no more power for fly)

Still .. there are multiple chemistry of lithium battery ... so what can be verify in one case not necessary is applicable on a different case.

Anyhow .. there are Yuasa official documents ... these are the best reference for our battery .. I have seen even a snapshot of a Yuasa document for the LEV40 (the battery cell in the PHEV) .. but was only in Japanese .. still the picture in the snapshot was self explaining.

Anyhow .. what is never properly tested in these battery is the effect of time ... Yuasa can easily model thousands of charge and discharge in a lab ... but consider the aging battery degradation factor ... can't be tested in a lab in a short time ... it needs keeping "busy" a lab for years ...

As well ... their test for battery degradation are mainly based at 25c and 45c ... this does not cover what happen in our PHEV which is from -20C to +35C (over 35C there is an active cooling) .. below -10/-20C charging and usage of the PHEV should be blocked or massively slow down .. but personally I never seen lower than -4C in my PHEV which has no garage access (garage was first used by my lawn mower tractor .... and now by an old Alfa Spider S2, also tractor now sit "sad" in the garden)

 

[Pomst]

Hi All,

Im new here.
Just joined because I got my PHEV 2014 with 44000km on the clock just 4 weeks ago.
I have been following the Youtube videos since my purchase and think I might be able to contribute to this battery issue since I work with vehicle lithium batteries on a daily basis.

First off.
All the discussion of degraded batteries needs to be ended with a simple test.
Some one can go get a used battery from the scrap yard, hook it up to a battery tester and run a test.
This will give you a base line of the actual degradation.

If this test would be done then you will have the fact to have the objective discussion about how correct the SOH is.
If the test shows that the actual capacity of the battery is near new then the calculated SOH is in error.
Also, the data could be used to confront MMC.

I am going to go have a talk with my friends at the junk yard to see if I can get my hands on a pack.

 

[michalsarna]

...
If the magic happen when the charging process is over, then having running light will quickly empty the eventual capacitor for protection.
...

And that might be the point. As You think of it. When charge is completed from BMU/ECU point of view it might try do move it's internal operations to 12v battery (since it should be already fully charged).
If above happens BMU & ECU looses power (12v battery is still disconnected) but internal charger might be still powered up by wall charger. So it will pump some more juice to battery and only onboard charger can (and will) send disable charging signal to wall charger.
So we can assume that onboard charger is set for 40Ah as default but BMU is telling it to stop early on 38 Ah for 100% or any other value it calculated.

If You are going to run this procedure check before point:
7 - Disconnect Charger
if there is anything happening inside car. Is there light on (P)ark or dash and if it's powering internal lights on door open.

This will give us some more info on how it's behaving.

 

[michalsarna]

Hi All,

Im new here.
Just joined because I got my PHEV 2014 with 44000km on the clock just 4 weeks ago.
I have been following the Youtube videos since my purchase and think I might be able to contribute to this battery issue since I work with vehicle lithium batteries on a daily basis.

First off.
All the discussion of degraded batteries needs to be ended with a simple test.
Some one can go get a used battery from the scrap yard, hook it up to a battery tester and run a test.
This will give you a base line of the actual degradation.

If this test would be done then you will have the fact to have the objective discussion about how correct the SOH is.
If the test shows that the actual capacity of the battery is near new then the calculated SOH is in error.
Also, the data could be used to confront MMC.

I am going to go have a talk with my friends at the junk yard to see if I can get my hands on a pack.

Hi.
That test might ba a proof but as seen in many videos when Mitsubishi was doing triple procedure on battery it went up and stayed up.
https://youtu.be/9Pw3mWsFxh0?t=343
Whole problem and point of finding how to DIY this procedure is that Mitsubish Japan is not allowing they Service Centres to do it any more - spill from UK.
I would love to go to my local service centre (which by the way is great) and tell them to run this but they are not allowed any more.
Funny thing is that I've purchased my car few weeks before Andy talked about this ban. Just after watching His video I've called my SC and they didn't confirm that but also couldn't set up this triple procedure for me. From other PHEVs users in region I know they ware doing this before. So for me it was a proof-of-ban.

 

[Pomst]

Hi All,

Im new here.
Just joined because I got my PHEV 2014 with 44000km on the clock just 4 weeks ago.
I have been following the Youtube videos since my purchase and think I might be able to contribute to this battery issue since I work with vehicle lithium batteries on a daily basis.

First off.
All the discussion of degraded batteries needs to be ended with a simple test.
Some one can go get a used battery from the scrap yard, hook it up to a battery tester and run a test.
This will give you a base line of the actual degradation.

If this test would be done then you will have the fact to have the objective discussion about how correct the SOH is.
If the test shows that the actual capacity of the battery is near new then the calculated SOH is in error.
Also, the data could be used to confront MMC.

I am going to go have a talk with my friends at the junk yard to see if I can get my hands on a pack.

Hi.
That test might ba a proof but as seen in many videos when Mitsubishi was doing triple procedure on battery it went up and stayed up.
https://youtu.be/9Pw3mWsFxh0?t=343
Whole problem and point of finding how to DIY this procedure is that Mitsubish Japan is not allowing they Service Centres to do it any more - spill from UK.
I would love to go to my local service centre (which by the way is great) and tell them to run this but they are not allowed any more.
Funny thing is that I've purchased my car few weeks before Andy talked about this ban. Just after watching His video I've called my SC and they didn't confirm that but also couldn't set up this triple procedure for me. From other PHEVs users in region I know they ware doing this before. So for me it was a proof-of-ban.

From what I understand this is conjecture since its not reproducible.
If people are able to reproduce the "triple procedure" the same as MMC does it at home then it will be proof as well.
But as it sits there is no proof that is reproducible in a reliable way.

Probably, if anyone is messing around with the CAN you should be able to do some magic.
https://github.com/tomdebree/OutlanderPHEVBMS
I'm not brave enough to run this code on mine just yet.

 

[MadTechNutter]

@michalsarna,
do you have the same charger as mine in the image above?

 

[MadTechNutter]

Actually the only real way to check SOH .. is to disconnect the main battery, take a couple of battery as sample .. and allow a full charge to 4.1v and a full discharge down to 2.75v .. using a proper charger/discharger equipment (which is cheap and common, especially for RC Model Hobbyist like me) .. the challenge is to disconnect the main battery, open the massive cover .. and get access to the individual cell ... .

If I am THAT FAR, that I have taken it all apart to access the individual cells, I am not going to waste one further second on those old cells and I will replace them all with brand new LEV50s.

 

[Pomst]

Actually the only real way to check SOH .. is to disconnect the main battery, take a couple of battery as sample .. and allow a full charge to 4.1v and a full discharge down to 2.75v .. using a proper charger/discharger equipment (which is cheap and common, especially for RC Model Hobbyist like me) .. the challenge is to disconnect the main battery, open the massive cover .. and get access to the individual cell ... .

If I am THAT FAR, that I have taken it all apart to access the individual cells, I am not going to waste one further second on those old cells and I will replace them all with brand new LEV50s.

How do you plan on making the car use the extra 10Ah?
As I see it the pre-2019 models are hard-coded with 40Ah as a fixed peak limit.