Speak EV - Electric Car Forums banner

1 - 20 of 48 Posts

·
Registered
Joined
·
3,441 Posts
Discussion Starter #1 (Edited)
Other cars have a battery health thread so I thought it was time to have one for these cars! ;)

To gather information on the battery health you'll need Canion with a suitable OBD adaptor. An alternative on the C-Zero/Ion is Diagbox using a Lexia 3, or on the i-Miev, the MUT-III tool. I don't expect many people to have a Lexia 3 or MUT-III however. (I do have a Lexia 3 as well as Canion)

There are two main things of interest when looking at the battery health - one is state of health or SoH, the other is cell balance after a moderately deep discharge that would flag up any weak cells. Another lesser factor is cell resistance but this can only be found with Diagbox/MUT-III and is not available from Canion.

Unfortunately neither Canion or Diagbox give a direct SoH figure - they only report usable Ah capacity calculated by the BMS. SoH can be derived from the Ah figure, but only if you know what Ah figure should correspond to a 100% SoH, and this is also not stated in either Canion or Diagbox! (n)

In short, nobody really knows for sure what Ah represents 100% SoH on these cars.

The raw capacity of the cells is 50Ah however they are operated over a reduced voltage range in the car (3.63 volts to 4.1 volts) than that used to specify their capacity, (2.7v to 4.2v I believe) so there would never be 50Ah usable even with brand new cells and making use of turtle mode range.

Charging cells to 4.1 volts instead of 4.2 volts reduces usable capacity by approximately 10% and from this and piecing together other information such as a single Canion reading taken on a warranty replacement pack on a US i-Miev, the generally accepted figure seems to be about 46Ah = 100% SoH, so that's what I use.

I have been recording the Ah figure vs date and mileage since I first bought the car, about 18 months ago first relatively infrequently using Diagbox (since it's a pain to connect and set up) and then more frequently using Canion. I definitely have some anomalous changes in the Ah figure over time so I thought I'd share these for discussion and comparison with others.

I have them recorded and graphed in excel, see below:

Battery 1.png


Battery 2.png


I do also have a graph of SoH vs date but as I do almost exactly 1000 miles per month the graph looks identical, so I have no real way to tell how much degradation is related to miles driven and how much the passage of time. I also stopped recording cell resistance when I stopped using Diagbox, although I could check it again if I needed to.

As can be seen, everything looked fine up until around 40k miles, with a drop of approximately 0.1Ah per 1000 miles, when there was a precipitous drop from 38.9Ah to 37.6Ah.

Because my measurements up until 41k miles were very infrequent (only every several months) I don't know exactly when that drop occurred and whether it was one large drop or a series of smaller drops.

I was a bit concerned about this drop and wondered whether I should do the dealer "Battery calibration" procedure in Diagbox, to measure the true capacity. This is a process the basically runs the battery down to a specified voltage under controlled conditions, charges it back up and does a calibration based on that.

Doing this calibration resulted in the figure dropping from 37.6Ah to 36.1Ah. Whoops! :ROFLMAO: This is the vertical downwards line on the graph...

Then after 10 days of normal driving it jumped back up to 37.1Ah all by itself. This just goes to show that while the real battery health never gets better over time, the BMS's estimate of battery health can sometimes rebound if a previous estimate was in error! Since then it has resumed a slow gradual decline.

The question is, did the battery suddenly get worse around 40k or had the BMS been over estimating the true capacity prior to that, and it had a calibration "wake up call" ?

I see roughly 0.1Ah decrease like clockwork every month/1000 miles or so (actually 838 miles when you calculate it) but it is highly likely that this decrease each month is predicted capacity loss based on the BMS's programmed in battery degradation model based on time and mileage.

Opportunities for the BMS to truly measure the capacity accurately don't come very often (as they need a relatively full discharge/charge cycle under favourable conditions, such as warm weather) so in between those opportunities it will follow an assumed degradation model programmed in by the manufacturer, occasionally corrected by actual measurements.

The second main factor of interest in looking at battery health is how well balanced the cell voltages are after a deep discharge of down to say 30% SoC. To do this you need to fully charge the car on a Level 2 charger until the charging stops. At this point the cell voltages should all be balanced within 5mV, which they are on mine.

Now use the car until the SoC reported in Canion is about 30% then look at the cell voltages. Here are mine at 30%:

Cell voltages.png


As can be seen, Cells 25 and 70 are a little bit weaker than most others. Not enough yet to cause me concern, but a noticeable difference nevertheless that should be monitored to see if it worsens. (Which so far it hasn't)

Is this the cause of the sudden drop in estimated Ah capacity around 40k miles ? Unfortunately I was not checking the cell voltages prior to that so I have no way to know. For all I know it was already like that at 28k miles when I bought the car and is unrelated.

I would be interested to see other people's Ah figure for the mileage their car has done, and also see the cell voltage graph when discharged to about 30%. (Note: you must do a 100% Level 2 charge first to allow the cells to be top balanced or the result won't be valid as rapid charging does not balance the cells)
 

·
Registered
Joined
·
211 Posts
Have you tried EvBatMon iMiev? This app suggests my battery condition is at 83.54%, with 40.1Ah capacity.


Sent from my ONEPLUS A3003 using Tapatalk
 

·
Registered
Joined
·
3,441 Posts
Discussion Starter #3 (Edited)
Have you tried EvBatMon iMiev? This app suggests my battery condition is at 83.54%, with 40.1Ah capacity.
Thanks for that - no I haven't tried evbatmon, mainly because I feel it's a bit expensive for a program that basically provides the same information available from Canion and Diagbox. (Canion only had the ability to report Ah added about a year ago so before that evbatmon and Diagbox were the only options for retrieving the Ah figure)

The Ah figure reported, 40.1 on yours will be the same for all three programs. It's interesting that it looks like Evbatmon considers 48Ah to be 100% SoH if you divide 40.1/0.8354.

I'm a bit skeptical of this as being correct as charging the cells to 4.1 volts will lose about 10% of the usable capacity the cells would be capable of. The LEV50 datasheet says it can be charged to 4.2 volts and that its 50 Ah capacity is based on a 4.2v charge.

So 10% off 50Ah is only 45Ah, so even 46Ah might be a bit optimistic. I wonder what they base the 48Ah figure on ? Even the dealer diagnostic tool I have doesn't say what a new battery should be either in the diagnostic GUI or the built in documentation.

Yours at 40.1Ah is still very good, mine was already down to 39.9 at 28k miles. For completion sake, what mileage has yours done and what year is it ?

I hope to be able to build up a battery health reading vs mileage and age from as many cars as possible in this thread so we have a comparative resource as there is very little info out there on the battery health of these cars and what is "normal", or at least typical.
 

·
Registered
Joined
·
211 Posts
My car's covered almost 27k miles, and was first registered September 2013. I bought it March 2017 when it was on 8k miles.

I have been trying to keep a record of the battery condition over time.


Sent from my ONEPLUS A3003 using Tapatalk
 

·
Registered
Joined
·
3,441 Posts
Discussion Starter #6
OK, here's my datapoint: 2012 i-MiEV with 41,500 miles has SoH reported by hobDrive as 39.8 Ah.
That seems very good for 41k miles. Is yours one with the later LEV50N cells do you know ?

Also I haven't heard of hobDrive - what is that ?
 

·
Registered
Joined
·
3,441 Posts
Discussion Starter #7 (Edited)
My car's covered almost 27k miles, and was first registered September 2013. I bought it March 2017 when it was on 8k miles.

I have been trying to keep a record of the battery condition over time.
That sudden drop from 88 to 85% between October and November is very interesting - it looks like my sudden drop from 84 to 81% earlier this year after I'd had the car a bit over a year.

I almost get the feeling that these cars may do some sort of big battery health "recalibration" once a year, then during most of the rest of the year it just reports a gradual pre-programmed degradation rate, showing an update each time it is predicted to drop by 0.1Ah.

The cell balance graph at 93% SoC you posted doesn't really mean anything as the charger will have balanced all the cells at 100%.

For the cell balance graph from Canion to be useful you'd need to run the car down to about 30% SoC then check it again - mine are in almost perfect balance down to about 50%, it's only when mine gets towards the bottom that two cells start to diverge somewhat from the rest.

The basis of the test is that you start with all cells balanced at 100% charge then discharge the traction battery most of the way down, once they're mostly discharged any capacity imbalance between the cells will result in individual cells being at a different SoC relative to each other - which is reflected by their individual no-load cell voltages. So the test should also be done with the key on but the car not in READY mode so that there is no load on the traction battery to skew the voltage measurements.
 

·
richi.uk
Joined
·
60 Posts
That seems very good for 41k miles. Is yours one with the later LEV50N cells do you know ?
Not as far as I know. It's an early 2012 example. But I suppose it's possible it had a battery replacement by a previous owner (I only bought it a few weeks ago). How does one tell?

Also I haven't heard of hobDrive - what is that ?
It's a Russian app, that's something of a work in progress. Find it on the Google Play store (get the free "demo" version), sign up for the Beta channel, install the beta, then to configure it for an i-MiEV or clone, go to Screens|Settings|VehicleSettings|ECUsettings and set Vehicle ECU type = "Mitsubishi iMiEV"
 

·
Registered
Joined
·
2,248 Posts
Last couple of discharges down to 30% or so indicates that the cell balance is within 0.005 volts across all cells. That was after a charge to 100% complete with cell balancing the night before. Vehicle has 13,500 kms approx.

With regard to measuring SOH using CanIon, any chance of someone posting a step by step guide on how to do this: bearing in mind the Android device I'm using at the moment is just my phone so not permanently present/powered inside the vehicle.

Many thanks
 

·
Registered
Joined
·
3,441 Posts
Discussion Starter #10
It's a Russian app, that's something of a work in progress. Find it on the Google Play store (get the free "demo" version), sign up for the Beta channel, install the beta, then to configure it for an i-MiEV or clone, go to Screens|Settings|VehicleSettings|ECUsettings and set Vehicle ECU type = "Mitsubishi iMiEV"
Thanks for that.

I did manage to get the beta version working - very, very buggy and doesn't seem to want to save any settings (so I have to re-select ECU's etc every time I launch it) but it does provide some additional PID's that Canion doesn't expose, such as motor coolant loop temperature which is quite neat. Although nothing that I can't also access via Diagbox. Good to see additional 3rd party support for these cars though if it does eventually mature into a stable release.
 

·
Registered
Joined
·
3,441 Posts
Discussion Starter #11 (Edited)
Last couple of discharges down to 30% or so indicates that the cell balance is within 0.005 volts across all cells. That was after a charge to 100% complete with cell balancing the night before. Vehicle has 13,500 kms approx.
A maximum differential of 5mV across all cells at 30% is extremely good. I'm seeing about 30mV difference at 30% SoC with two cells in particular weaker than the rest. The other cells are all within about 10mV of the average. Will need to keep an eye on those two cells I think...
With regard to measuring SOH using CanIon, any chance of someone posting a step by step guide on how to do this: bearing in mind the Android device I'm using at the moment is just my phone so not permanently present/powered inside the vehicle.
There is no direct SoH figure available from the car, the only thing you can get is the "Ah capacity" of the battery.

This is visible on Canion on the very first "My Trip Timer" screen, however as the lines in this first screen are all customisable it's possible the Ah line is not there. If so, click on the beginning of one of the lines of large text and you should get a scrolling pop up - look through that list and near the end is "Ah". Select that and it should show you the usable Ah capacity.

I've highlighted it in red in this picture I grabbed from the Canion Play Store photos:

Canion Ah.png


There is also another screen that shows "Ah history" which will graph the Ah figure each time it changes so you can see at what date and mileage the figure changed, however the bar graphs are fairly small so you can't get the exact figure from the graph.

It's probably best if we all report and discuss the Ah capacity of our batteries rather than SoH, as there is not agreement on what Ah capacity = 100% SoH between different sources which will tend to confuse things, however the Ah figure will be directly comparable across all cars and all software which report it.
 

·
Registered
Joined
·
3,441 Posts
Discussion Starter #13
Brilliant, will edit the page as advised.

Is there any 12volt reporting available? Eg voltage and load?
Not in Canion. Traction battery voltage and current yes, 12v battery voltage, no.

The car is capable of reporting the 12v system voltage - I can measure it using Diagbox as several of the ECU's report the 12v supply voltage. Not sure about current drain of the 12v battery though, possibly not.

It's also possible to measure the voltage at the OBD-II socket as the STN chipset in the OBDLink adaptors can report the battery voltage with an appropriate AT/ST command, but Canion doesn't seem to make use of this either.

I wouldn't worry about the 12v battery too much though - unlike a Leaf the charging system in these seems to work well, I've had no problems with my 12v battery and when I did an overnight rest voltage test it was 12.7 volts - which is just a fraction above the expected 12.6v of a fully charged battery at rest.

The only thing to keep in mind is that the 12v battery is only charged either while the car is in READY mode, or while it is plugged in and actively charging the traction battery.

Once the traction battery is full and charging stops, charging the 12v battery also stops despite still being plugged in. So if you go away on holiday for 3 months leaving the car plugged in won't stop the 12v battery going flat.

For leaving the car for a long time you'd want to discharge the traction battery to be between about 40-80% for minimal degradation and then put the 12v battery on one of those automatic battery conditioners which will keep the battery topped up.
 

·
Registered
Joined
·
3,441 Posts
Discussion Starter #15
Well today I am not happy. Another sudden and significantly large drop in Ah capacity. Something is definitely up with the battery in this car. Due to the ambiguity of SoH calculation on these cars I've updated my graph to show Ah instead, here are my updated figures:

Ion battery.png


As can be seen there has been another sudden drop of 1Ah in the last month and only about 900 miles.

Referring to last year it can be seen that I did over 10,000 miles over about 10 months for the first 1Ah drop in capacity, so the same drop in 900 miles and 1 month is just not normal. Given that two cells seem weaker than the rest its a fair assumption that it's these two weak cells that are driving down the usable Ah capacity at a rapid rate and that those two cells could potentially be on their way out soon.

Time to get in touch with Peugeot I think to see if their 8 year/60k mile "battery warranty" is worth the paper it's written on. Has anyone ever contacted them formally to see what the exact provisions of the warranty are, such as "acceptable" SoH loss, and whether any allowance is made for individual weak cells ?

I have a feeling they're going to say not covered but it's worth a try. :(
 

·
Registered
Joined
·
2,248 Posts
It might not even be written down on paper (the battery warranty).

Is there any noticeable reduction in range? the the:"
 

·
Watford FC Supporter
Joined
·
103 Posts
Eventually got an OBD dongle so this is the battery health of my two C Zeros. They were registered in November and December 2012 and have similar mileage but one has 2 Ah better health. We bought one in March 2017 and the other June 2017.

pict_2018-12-01_13-44-51.png
pict_2018-11-08_08-08-41.png
 

·
Registered
Joined
·
3,441 Posts
Discussion Starter #19 (Edited)
Thanks for the data points @Timmc :) You "mileage" figures are actually in km I presume ?

At 30k miles mine was at about 39.7Ah, so half way between your two cars, so all three cars have comparable degradation at that point. This was before mine started it's more rapid downward slide mind you.

Cell balance looks very good at 40% but you won't really start to see any imbalance until you get down to about 10-20% as 40% is within the flat "plateau" voltage range for the cells where the voltage hardly changes with SoC.

If you run it down to 20% and take a voltage graph snapshot that would give data that was directly comparable to the snapshots I've posted.
 

·
Watford FC Supporter
Joined
·
103 Posts
Yes the Odo is in kilometres. The 40% is after my usual daily commute charging to 100% the previous night. Will post again at a lower SoC.
 
1 - 20 of 48 Posts
Top