Losing Charge When Parked Up

[miggols99]

So I don't know if this is just a side effect of my very tired battery (24 Ah, or around 50% SoH), but I've found my SoC to be very unstable above 70%. Here's what I'm seeing recently:

- Charge to 100% on level 2 (RR @ 80~90km depending on the driving style)
- Drive around 10km back home from a charging station, RR @ 70km, 90% SoC (I rent, so charging at home involves running a cable out the window...)
- Come back to the car after a couple of hours to find the RR has gone down around 6-7km, with the battery gauge also losing a bar - I can literally see the bar disappear right before my eyes! SoC has also dropped 7-8%. This happens even after just 5~10 mins parked up, so it doesn't seem to be parasitic drain.

Going below 70% the battery gauge/RR behaves as expected.

Has anyone else experienced anything like this?
I have charged to 100% then driven longer distances (30~40km), but when I do that the meter doesn't seem to drop as quickly for some reason.

This in addition to the battery degradation gives me a real-world range of maybe 60km which means charging at any possible chance...
I do a lot of short trips so if not for this it would be perfect for my needs, but at this rate I may have to say goodbye to this one and look for another with a healthier battery

 

[dopey]

If I am not mistaken the traction battery is physically disconnected when the car is parked, making any sort of actual consumption 'impossible'. -- one reason heater, etc does not work until the car is in 'Ready'.

The range gauge is some sort of estimate based off consumption+previous driving.

Any chance when you first get in the car, something like A/C or heat is coming on that you had switched off when you parked? Sitting in the car and toggling these things you can see an immediate change to the range estimate on the car.

50% SoH is quite low though -- and quite shocked you get 80-90km RR out of that. If you are able to check your SoH, can you check your cell voltages? Just speculating, but perhaps if you have cell(s) which are way below others that it causes some funkiness and effectively triggers the car to reduce it's Ah after sitting?


My 63% SoH range display varies wildly, some full charges i'll be at 109km other's at 80. Real world anywhere from 45-80km depending on speeds we drive..


(Very new myself to this car, so may be waaay off but have not personally noticed such reduction other than 'estimate' range due to A/C or heat being on vs off).

 

[JoeS]

@miggols99, when the aircon is activated, the RR (Range Remaining) drops by ~10%. When the heater is activated, the RR drops by ~20%. Perhaps this is a reason for your 'inconsistent' RR drop?

 

[miggols99]

Thanks for the replies. The weather has been quite mild recently, so no aircon/heater. I'll have to see if I can just sit in the car for a bit and see what exactly is going on using OBDZero when the SoC drops...

 

[PV1]

Given that you have a very early i-MiEV (2009), it may not behave like most of our cars.

The drop in SoC wouldn't be caused by heat or A/C. My guess is that the instrumentation isn't calibrated to the actual capacity of the battery. The car is expecting less voltage sag (from assuming more capacity than actual), but after sitting, it reads the cell voltages and adjusts down to match the actual charge of the battery. My one car, Koorz, exhibits behavior like this from time to time, but the opposite. It will gain 4-5% charge if parked after a short drive (5 miles or less). I have decent capacity, but higher internal resistance in that pack.

To fix this, it might take a dealer calibration. Newer i-MiEVs like many here on the forum can do this on their own, but older ones need the diagnostic tool to perform capacity calibration.

 

[miggols99]

So funnily enough, after getting in the car this morning I found I had recovered a bar since the day before, from 7 bars, 1 below half, to 8.
Since my SoH is so low, perhaps the battery management system is just not able to keep up?

A perk of owning an i-MiEV here in Japan is that Mitsubishi's charging card gives you a free dealer battery calibration every year. I talked about this in my other thread, but I actually had this done around 4-5 months ago, so I would assume the battery is calibrated as well as it can be. I do probably use DC charging a lot more than is good for the battery though, mostly due to not being able to charge at home..

 

[miggols99]

So I was able to capture this loss of range using OBDZero.
Before turning the car on, the SoC and range remaining show the same as when I parked up.

However, after putting the car into "Ready", I can see the SoC drop from 90.5% down to 83.5%, with a consequent drop of 6km of estimated range!

I wonder what could cause around 0.5kWh to disappear into thin air?
Or is this simply the car's way of recalculating the true SoC?

 

[CZeroOwner]

Hi Miggols99

There are a number of curious things about the two screenshots. Look at the last line that shows the capacity and the Ah remaining. The car computes the SoC by dividing the Ah remaining by the capacity and multiplying by 100. To find the Ah remaining the car sets the Ah remaining to the capacity when the car it is charged to 100%. After that it keeps an account of the Ah removed from the battery using the amps measurement. This is called coulomb counting. OBDZero also keeps an account of the Ah removed from the battery starting with the capacity at 100% charge. I have not found a direct read out of the cars Ah account but computing backwards from the capacity and the SoC the car's account is 21,27 Ah in the first shot and 19,62 in the second shot. On the other hand OBDZero's Ah account is much lower 15,02 in the first shot and 14,56 in the second. OBDZero believes that almost 9 Ah have been used since the car was charged to 100% while the car "believes" there has only been 2 to 4 Ah used. In my cars the computed car Ah account and OBDZero Ah are always very close to each other but they are not here. So something is not right. Could you please send me the data collected by OBDZero. Thats the text files in the OBDZero folder. I may be able to answer your questions after I have seen the data.

Cheers
David

 

[miggols99]

@CZeroOwner: Sorry for the late reply - I was away travelling. I'll send over the text files in a PM in a bit.

Perhaps unrelated to this problem, I took the car in to a dealer as I have been having problems using certain public chargers.
This owner of a 2010 i-MiEV goes into the details in their blog post below:

https://ameblo-jp.translate.goog/studio-prato/entry-12585109117.html?_x_tr_sl=auto&_x_tr_tl=en&_x_tr_hl=en-US&_x_tr_pto=wapp

(my car shows the exact same symptoms...i.e. a dying OBC)

They told me that error codes P1A2F and P1A4B (Malfunction of the BMU traction battery cell system) are showing up, and I would have to pay over 1 million yen ($7,000 USD at the current rate) to replace the traction battery, which is obviously unrealistic for a car I bought at more than half of that :roll:

If the OBC fails completely, it would qualify for a service campaign they say...


I have been getting "cell voltages difference too large" errors as shown in Hobdrive, so I may need to do a few "calibration" charges.
The charging stand I use most frequently has an unusual charging curve, as it doesn't seem to taper off much at all when reaching a higher SOC, so I am not sure cell balancing is taking place.

 

[CZeroOwner]

I would have to pay over 1 million yen ($7,000 USD at the current rate) to replace the traction battery.

I was quoted twice that by the local Citroen dealership. I'm wondering how much it would cost to ship a battery from Japan to Denmark. However I only payed the Japanese price for a new battery, for my car. So I'm just curious.

 

[kiev]

...
(my car shows the exact same symptoms...i.e. a dying OBC)

They told me that error codes P1A2F and P1A4B (Malfunction of the BMU traction battery cell system) are showing up, and I would have to pay over 1 million yen ($7,000 USD at the current rate) to replace the traction battery, which is obviously unrealistic for a car I bought at more than half of that :roll:

If the OBC fails completely, it would qualify for a service campaign they say...


I have been getting "cell voltages difference too large" errors as shown in Hobdrive, so I may need to do a few "calibration" charges.
The charging stand I use most frequently has an unusual charging curve, as it doesn't seem to taper off much at all when reaching a higher SOC, so I am not sure cell balancing is taking place.

We haven't seen or heard much from any Japanese owners so this was very interesting to see the various types of charging stations available there, and also the "mode1" charging cable.

Before the OBC is activated there is a communication signal between the EVSE and the car. If there is anything sensed as being incorrect in this signal, then the session will fail and no charging occur. This is not the same as an OBC hard failure due to blown components. The various EVSE manufacturers may have implemented the protocol in different manners such that some don't work with the imiev. This happens with the Nissan Laef also, some work and others don't.

About the DTCs:
If the BMU gets a voltage reading from a CMU that a cell(s) have more than 0.2 V difference from the rest, then it throws the P1A4B code to the EVCU, which then throws the P1A2F, so those 2 codes are just the same fault. So there is likely one or more cells that are weaker than the rest, and most any of the OBDII scanners or phone apps can read and display the cell voltages to determine which are the culprits.

The pack and cells were designed and assembled to be easily serviceable, but the Mitsubishi Service Manual does not pursue that repair path, it just indicates to replace the entire pack. That is quite a wasteful approach, but it is a safe repair strategy from both a technical and legal liability view .

 

[CZeroOwner]

Thanks to miggols99 for sending the data from the trip shown above.

First of all, changes in SoC2 when starting the car after an idle period do occur. However, they are not common. I have found about 10 instances in about 400 hours of data collected from my CZero. SoC2 is the SoC shown in the screen shots in miggols99's post. It is also the SoC reported by e,g, CaniOn. Changes in SoC1 are more common. SoC1 is reported by the car in the same PID as SoC2. It is shown on the OBD screen in OBDZero.

The BMU uses idle periods to compute the true SoC based on the low current voltage of the battery. It is the same process that occurs during the 10 min. pauses while the car is slow charged. As a rule only the SoC1 is adjusted to the true SoC but sometimes both SoC1 and SoC2 change. I believe that SoC2 only changes when the difference between the true SoC and SoC2 is large.
This change in SoC2 does not mean that there is a real change in the Ah or Wh in the battery. It only means that the BMU’s estimate of the Ah in the battery changes. As a rule the BMU keeps track of the Ah using coulomb counting. However errors in this count accumulate so the BMU needs to check its Ah account once in a while. A new value for the Ah in the battery is computed using the BMU’s present estimate of the battery 100% capacity and the true SoC based on the low current voltage.

Looking at the data miggols99 sent I can see that the SoC2 and the battery’s Ah estimate changed twice on the day in question. This graph shows the changes.

The first change came just as OBDZero started collecting data. Both SoC1 and SoC2 shot up to 100% from about 75%. The second change came after the idle period when SoC1 dropped to 73% and SoC2 dropped to 84%. I looked at the cell voltages for the battery's good cells at the times for these two changes and I believe the true SoCs were 85% to start with and 73% after the idle period. I also believe that the BMU overestimated the SoC to start with because the poor cells in the battery had a higher voltage than the good cells.

This next graph shows the Ah in the battery.

The blue line is the BMU’s estimate and the green line is my estimate based on the true SoC shown above. Note that there is no drop in the true Ah after the idle period. 23.5 Ah is the BMU’s present estimate of the batteries 100% capacity. I would guess that the best cells have a capacity of 26 Ah while the poorest cell is much lower than 23.5 Ah

Unfortunately the instability of the voltage and the BMU's Ah estimate are more signs that the battery is in poor condition. This last graph shows the highest (red) and lowest (blue) voltages of each of the 88 cells in the battery. This is based on data miggols99 sent previously as well as the data shown in the graphs above.

For comparison the green line is the lowest voltages for the cells in my car. It looks to me that at least half the cells must be replaced before the battery would have a capacity that would be worth the effort and cost. I would guess that replacing the whole battery would be better. Another thing, the 6 poorest cells have a minimum voltage below the design 0% SoC of 2.75 volts. This may be enough to claim that the battery has reached its useable life.

Btw it is now easier to analyze OBDZero data because the newest version (3.90) provides a initial value that limits the amount of data recorded. This makes it much easier to import this data into e.g. Excel.

Cheers
David

 

[miggols99]

After multiple "calibration charges" I didn't see any improvements, but when charging at my usual 200v charging station from around 7 bars (just below halfway) today, I think I saw the car finally calibrate itself!

Usually I would see the charge rate stay fairly high (around 1500w) until SOC 1 reached 100%, with SOC 2 around 5% higher, but today the charge rate tapered off quite a fair bit to around 0.3 amps, with a much smaller difference between SOC 1 and SOC 2.

It did stop at around 97% for some reason, but after doing my usual errands with a lot of short trips, the RR is no longer readjusting/dropping!

Fingers crossed this isn't just temporary...

 

[miggols99]

So the car calibrating itself seems to have been a fluke, as I am back to the same again...
I do get ridiculously unrealistic RRs at 100% though (low 90s) which the BMS readjusts for after multiple short trips.
Driving relatively economically, I am getting around 70km range from 100% to 20% (2 bars flashing).

Time to consider looking for a younger car, or maybe even the successor?

 

[MickeyS70]

Driving relatively economically, I am getting around 70km range from 100% to 20% (2 bars flashing).

Time to consider looking for a younger car, or maybe even the successor?

As long as 70km on your bad battery gets you trough the day I wouldn’t be too bothered with a fluctuating RR…

 

[miggols99]

As long as 70km on your bad battery gets you trough the day I wouldn’t be too bothered with a fluctuating RR…

That would be true if I had access to a home charger...since I'm renting my only choice is a cable going to a 110V outlet which I have to run out the window from the second floor
The management company wouldn't even give me the time of day when I asked about them installing a simple plug socket, so that leaves me reliant on public charging.

If Chademo wasn't so slow (I only get around 3-4 kWh in 30 mins, plus the majority charge per the minute rather than the kW), it would be more tolerable. Since it was so cheap and I'm saving so much money anyway, I'm thinking I will just put that towards a better car.

 

[MickeyS70]

If Chademo wasn't so slow (I only get around 3-4 kWh in 30 mins, plus the majority charge per the minute rather than the kW), it would be more tolerable. Since it was so cheap and I'm saving so much money anyway, I'm thinking I will just put that towards a better car.

Now I see where you’re coming from. I think you have already answered your question in a different discussion: as Chademo is your main means of charging then the I-MIEV M (SCiB technology) is the obvious choice even if that means spending a bit extra…

 

[miggols99]

Yes, you're right that the M type i-MiEV seems more suitable for my use case.
Would definitely prefer a model past 2014 with the heat pump and keyless start, but these of course come at a premium compared to older models with the resistive heater. Going to keep my eyes peeled for a good deal...

 

[MickeyS70]

Would definitely prefer a model past 2014 with the heat pump and keyless start,

Are you sure? I have come across this heat pump rumour for JDM but I don’t think that was ever confirmed; even the latest service manual still shows the PTC heater.

Another clue Might be in the link regarding the heating/cooling from 2010, it appears that Mitsu was trying to improve the size/efficiency of the heater, not replace it with a heat pump?

https://docplayer.net/54900352-Air-conditioning-system-for-electric-vehicles-i-miev.html

 

[miggols99]

Would definitely prefer a model past 2014 with the heat pump and keyless start,

Are you sure? I have come across this heat pump rumour for JDM but I don’t think that was ever confirmed; even the latest service manual still shows the PTC heater.

Strangely enough I wasn't able to find any concrete information on English language internet sites, but a quick search in Japanese clears things up.

https://www-nextage-jp.translate.goog/carcatalog/mitsubishi/i-miev/zaa-ha4w/?_x_tr_sl=auto&_x_tr_tl=en&_x_tr_hl=en-US&_x_tr_pto=wapp

It was actually added on the late 2013 model, but if you go 2014 or later you're guaranteed the heat pump.