EVBatMon Readings

[PV1]

In order to mirror the thread structure used for CaniOn, the main thread for the Android app EVBatMon was moved into the Instruments sub-forum, as that thread regards the app itself, which is an instrument.

This thread is to be used to post and discuss the data provided by EVBatMon. This thread deals only with the data, so anything regarding issues/changes to the app itself, please use the thread linked here (http://myimiev.com/forum/viewtopic.php?f=25&t=3022).

 

[pbui19]

So how does it work ? do we have to drive until the turtle shows up then charge to max capacity ? or the system has already been monitoring the capacity ?

 

[JoeS]

So how does it work ? do we have to drive until the turtle shows up then charge to max capacity ? or the system has already been monitoring the capacity ?

pbui19, we've had lots of discussions on this forum as to exactly how capacity should be measured, pointing out the errors associated with capacity calculations based on whether charging or discharging and to what limits, inability to control either charge or discharge rates precisely, differences if incrementally measuring at different SoC levels, and lots of other variables.

I suspect that EvBatMon's exact algorithm is their trade secret. Be interesting to see what the differences are between how EVBatMon does it and an 'official' Mitsubishi test.

BTW, I moved your post over to this thread as we'd like to keep the other thread focused on getting EvBatMon working in the car and issues with that.

 

[PV1]

The capacity showed up immediately after the app connected to the car. Apparently, they found where the car stores this information. I was under the impression that the MUT-3 requested this info, but apparently it's available on the CANBus.

 

[Malm]

The capacity showed up immediately after the app connected to the car. Apparently, they found where the car stores this information. I was under the impression that the MUT-3 requested this info, but apparently it's available on the CANBus.

I believe it does. So the number showed by EvBatMon is exactly the same as the MUT-3 at the Mitsubishi service. My data is in line with that idea. In the beginning of February MUT-3 showed 34 Ah in the Mitsubishi service. When I obtained for the first time the value with EvBatMon, in 13 of March, it showed 33,8 Ah, and now is at 33,7 Ah. It's always available on the CANBus. I always thought that it was there. My warranty is going to end next week, and I asked today for a new battery. I said I have a big drop in range compared to 2011, only possible to explain by one much more bigger drop in capacity that Mitsubishi suggested five years ago (20% in 5 years). There are many ways to see the actual state of my i-MiEV. Since almost nobody understands how does my i-MiEV is (but I do), I think I can have a chance. I have some astonishing values from 2013, like 175 km mRR and a trip of 201,4 km with 3 bars left, and now most of the days I only have something like 80 km of mRR, and that trip can't be done anymore, even arriving after 7 km in turtle. These are the facts that I will use to make my point. It sounds very bad, because I want to show the worst possible picture. The real picture, the true picture, I know very well and the 33,7 Ah says it all, almost 30% of loss of capacity, but not so much in relation to the loss of the inicial range (when it was new, it didn´t used a significant percentage of its capacity, so I never had avaliable all the capacity, stopped at 0,0% SoC but with the lowest cell with 3,6 V - now it goes to 2,7 5V).

 

[PV1]

I also like how the remaining Ah are shown. I've been tempted to put a feature request in for CaniOn to show Wh remaining, but Ah remaining is probably a better value to show.

I also pegged 45.4 kW of regen!

 

[kiev]

i may have posted this procedure from the FSM before,

miev capacity check

To just read what the stored capacity value presently is:

1.Connect the diagnosis connector to the M.U.T.-III with electric motor switch in the "LOCK" (OFF) position.
2.From the System Selection Screen of the M.U.T.-III, select the "BMU".
3.Select the "Data List" from the BMU Screen.
4.Check the BMU data list, item No. 23: Battery current capacity.
5.Turn the electric motor switch to the "LOCK" (OFF) position, and disconnect the M.U.T.-III.


It's just a matter to determine what the CAN id is for the BMU data list item #23, then it could be read by Canion or other CAN reader--it was on my to-do list once i got my hands on that mut...

 

[JoeS]

So, why the rigmarole the dealers subjected some of our members to in order to measure capacity (e.g., charge fully and drain battery using heater), not to mention a dealer telling me they couldn't do it?

Malm, since you have a good handle on this, how do your EVBatMon (CAN) numbers compare with various other techniques for determining capacity? Incidentally, wish you all the best in getting Mitsubishi to replace your degraded (but not broken) pack. The magic number for Leaf and Kia, IIRC, is 30% capacity loss; i.e., down to 70% of original capacity.

 

[zzcoopej]

So, why the rigmarole the dealers subjected some of our members to in order to measure capacity (e.g., charge fully and drain battery using heater), not to mention a dealer telling me they couldn't do it?

There is not much to service in an EV, so the dealership has to be useful for something?

It is true, there are 2 battery capacities - true, and calculated (eg PMC, SOH). The actual true capacity can only be determined by a complete discharge from full under controlled conditions. However the capacity can also be calculated based on driving. Over time with a healthy battery this calculation can become quite accurate and is what EvBatMon reports as PMC (Percentage of Manufacturer's Capacity). However with an unhealthy battery, degraded capacity is a moving target and the calculation may not be as "scientific" as it is not based on a precise and constant load (however it could be argued that it is more accurate as it represents the actual driving experience).

When you get a normal service at the dealership, they report the same BMU calculated capacity as EvBatMon does. If this calculated capacity (or the cell voltage range when the battery is flat) indicates a problem, you should still get the full tested capacity.

 

[Malm]

So, why the rigmarole the dealers subjected some of our members to in order to measure capacity (e.g., charge fully and drain battery using heater), not to mention a dealer telling me they couldn't do it?

Malm, since you have a good handle on this, how do your EVBatMon (CAN) numbers compare with various other techniques for determining capacity? Incidentally, wish you all the best in getting Mitsubishi to replace your degraded (but not broken) pack. The magic number for Leaf and Kia, IIRC, is 30% capacity loss; i.e., down to 70% of original capacity.

Determining capacity by discharging it to 20% and then charging it to 100%, in my opinion, doesn´t work. Why? 1 - Because SoC can sometimes jump for another number in a millisecond. 2 - Because 20% of SoC can be 35% of real charge of the battery (specially in a new car, at this stage the car can stop at 0,0% SoC with still 3,6 V in the weakest cell), at 0,0% SoC the car will be 15% full. 3 - A 100% charge is not always to 100% full, sometimes the charge stops at 359/360 total voltage, and not at 361V.

Before I had EvBatMon I had my own method. See the voltage at 0,0% SoC by Canion and how much was a bar of 5% in energy (50-55%) when charging. If the car can go to 0,0% SoC and one bar of 5% is about 630 Wh, then I think the car has a range of a new one and the remaining capacity will be over something like 42 Ah. I tested some second hand i-MiEVs for some of my friends with this method and it worked well.

For me, the best way to know the capacity in an i-MiEV is this number given by the EvBatMon. Yes, it can be not very well calculated in a old battery with cells much with more degradation then the others (I agree with zzcoopej).

 

[Malm]

I know that magic number. 70%. I am slightly over 70% at this moment (33,7/48=70,2%). So, with a Nissan Leaf, I would dye at the beach. I dream that in Mitsubishi Portugal they have 5 packs in storage for use in warranties, I dream that they will not be very scientific in giving me one, since the batteries are there and must be used after sometime. So, I dream that they will give me one of that 5 batteries without problems, since there is only one more owner(I believe), that asked for a new one in the warranty. He says he is getting a new one, and I know that his battery is in fact a little better then mine.

So, not smart to not try.

But I will be happy if they don´t give me the battery. This one is working well, and for my daily use has more then the sufficient capacity. I am now loosing 1 to 2 Ah an year, so it will work for me at least for more 5 years. And we all can continue to follow this "old" battery and how it behaves trough time.

 

[Lic]

EVBAtMon shows 37.4Ah for my 2012 i-miev (51k km) and car usually charges only up to 97.5%, so real available capacity is only 36.4Ah. And it's sad that Mitsu doesn't exchange batteries with this high degradation, because when I was looking for this car I thought If Mitsu tell that they expect 20% loss for 5 years - it is part of warranty promises, but right now it's look like it's just words, nothing real behind them.

 

[JoeS]

...when I was looking for this car I thought If Mitsu tell that they expect 20% loss for 5 years - it is part of warranty promises, but right now it's look like it's just words...

I'm curious if you had received anything in writing, as in the US I don't recall anything about capacity loss? Lic, refresh our memory, whereabouts are you located?

 

[ICEM4N]

41.3ah for me with a 2011 car and 30k km, not bad

 

[Lic]

Nobody promises me exchange battery if degradation will more than 20% for 5 years. When I was thinking about buying I -miev I saw this statement very often and before buying car I didn't check if warranty documents has anything about it (my fault). But in same time I think if Mitsu doesn't exchange batteries with high degradation then the shouldn't tell anything like this, because people expect it should like this and if real things going in worst way - people feel your self flamed.

 

[jsantala]

C-Zero 2011 28212km 39.5Ah 82.3%

I think the procedure to drain and fully charge the battery is to make sure that the BMU has an accurate picture of battery capacity. When a random customer comes in there's no way to know (or trust) the customer on how the car has been used. Thus a full discharge-charge is needed to train the BMU.

 

[DonDakin]

Hey kiev,

From the procedure you posted it looks to me like the car is off and you plug in the mut 3 and select the capacity parameter on the mut 3

Does that mean perhaps that the capacity is not in the regular data update that is transmitted when the car is on ? Maybe the mut 3 sends a command to interrogate the car and get the battery capacity number. If so that go against the philosophy of Canion not to write any data to the car to get this information.

Its just something i was thinking about when i read your post.

It would be great if canion could read and display this information.

I have not tried the evbatmon app yet. Other then the capacity number is there any other benefit over canion which has been very good to us ?

Don.....

 

[zzcoopej]

I have not tried the evbatmon app yet. Other then the capacity number is there any other benefit over canion

Another benefit of EvBatMon is that you can customise the screens to display the data you want, in the format you want (value, bar, dial or histogram). Eg what PV1 is doing here http://myimiev.com/forum/viewtopic.php?f=25&t=3022&start=34

You can also automatically compare your battery capacity to all the other users on graphs eg http://www.evpositive.com/battery-history.html for an idea of what EvBatMon users see via our http://www.EvBatMon.com website (only EvBatMon App users can see graphs sorry!)

 

[Malm]

Another number that we can't see in CaniOn, but shown on EvBatMon it´s SoC Real. CaniOn only shows SoC Display (or SoC Dash). Apparently, the car calculates to different SoCs, using different methods to calculate each one. I believe sometimes my car changes its SoC Dash to be more close to the SoC Real Number.

 

[kiev]

...
Does that mean perhaps that the capacity is not in the regular data update that is transmitted when the car is on ? Maybe the mut 3 sends a command to interrogate the car and get the battery capacity number. If so that go against the philosophy of Canion not to write any data to the car to get this information?..

Howdy DD,
i think you are correct that the mut3 must query the BMU to get item #23. Here are some notes of what the system looks like to me.

The FSM shows the following components connected in the CAN bus line:

ETACS-ECU
ASC-ECU
EPS-ECU
Combination meter
Steering wheel sensor
G and yaw rate sensor
Battery management unit (BMU)
On board charger/DC-DC converter
Compressor/heater controller
Electric motor control unit (EMCU)
EV-ECU

The mut3 connects to the CAN buss thru the diagnostic connector.

The BMU talks to the Cell Monitoring Units (CMU boards mounted in each 8- or 4-cell module) but it appears that the CMUs are not on the CAN. The mut3 goes thru the CAN to the BMU to get CMU data.

There is a BMU data reference list of mut3 data values and trouble codes that are issued, and there is a BMU monitoring list with ~400 CMU-related data items that are stored of measurements or calculations related to cell and pack conditions that are available to the mut3 for monitoring the CMU control status.

Are all of these items sent over the CAN bus all the time?--probably not, since they are not needed for normal operation, but with the right command string on the CAN bus they probably could be sent.

There is an interesting note related to these data lists,

"These items are difficult to clearly show the specification value because the values much vary depending on the slight difference in the measuring conditions, in the environment or in the chronological change of the vehicle.The check conditions, displayed ranges and value variations are described."

i'm curious about how this all works and of course want to get a mut3 to figure it out...