BMS software update

Elmil · Sep 28, 2016

Yesterday I had my car in for 3:rd service after 45000 km in 17 months.

The SOH (State Of Health) value for the high voltage battery had changed. It is now 99.2%
I was hoping to discover a corresponding change in the BMS OBD2 data, but unfortunately there is nothing to find. It's annoying, because the SOH is the only value on the battery report that is not identified in the OBD2 data.
On the other hand, the deterioration values are not present in the battery report, so maybe the SOH is calculated on the fly using some algorithm involving the deterioration values.

My det.values are now: Max 11.5%, Min 10.0%.
Probably need more GDS/KDS readings to figure out the algorithm (if any..) I only had a battery report at 15000 km service, and at that time no OBD2 logging.

Anyone else having other than 100% SOH yet?

JejuSoul · Sep 28, 2016

Elmil said:
On the other hand, the deterioration values are not present in the battery report, so maybe the SOH is calculated on the fly using some algorithm involving the deterioration values.

SOH is calculated within the BMS itself. That value must be on the CANBUS. Can you post a full list of the hex codes returned from the BMS on the BMC Compare page - https://docs.google.com/spreadsheets/d/1YYlZ-IcTQlz-LzaYkHO-7a4SFM8QYs2BGNXiSU5_EwI/edit?pref=2&pli=1#gid=1174560131

See - http://www.skbattery.co.kr/what/bms.aspx : This link and image seem to be dead. Am looking for a replacement.

Also note there are two versions of the battery report. (possibly relating to the difference between GDS and KDS - or maybe relating to a difference in the English vs German software)
Elmil - what language is the software used in Sweden, and do you know whether it is GDS (windows based) or KDS (Android based).

In this German version deterioration values are not present in the battery report. Alkl who originally did the CANBUS decoding on the SoulSpy thread used this version which is why he did not find the deterioration values.

SOH (Gesundheitsstatus) nach 34'000 Km

This GDS (English) version does have the deterioration values. I found the deterioration values after Alkl had done all the hard work merely because I used the English version and he used the German. (Update: not necessarily true - there are also English versions that don't show the deterioration numbers.)

Elmil · Sep 29, 2016

JejuSoul said:
Can you post a full list of the hex codes returned from the BMS on the BMC Compare page - https://docs.google.com/spreadsheets/d/1YYlZ-IcTQlz-LzaYkHO-7a4SFM8QYs2BGNXiSU5_EwI/edit?pref=2&pli=1#gid=1174560131

Done!

JejuSoul said:
Also note there are two versions of the battery report. (possibly relating to the difference between GDS and KDS - or maybe relating to a difference in the English vs German software)
Elmil - what language is the software used in Sweden, and do you know whether it is GDS (windows based) or KDS (Android based).

First time I got the battery report they used GDS. This time I went to a new service provider, and now the report came from KDS. They are both in English and none of them include the deterioration numbers.

JejuSoul · Sep 30, 2016

-
Thanks Elmil.
I looked at your data and it seems the SOH is not one of the codes returned when we request values from the BMS.
I now suspect it is on the CANBUS all the time. Are you able to post the data from the C-Can tab.

I regret not being more thorough in photographing my own experience at Kia Service. Here is the only photo I have of my 100% SOH reading from GDS.
I didn't get a printed list of all the data.

Here's a picture from Hyundai Service upgrading my test drive I oniq EV with GDS. (Hyundai and Kia use the same GDS software)
To get the BMS data you click the BMS icon..

Elmil · Oct 1, 2016

I made a new tab in the spreadsheet, named "C_CAN_Compare".

Some ID:s don't respond, but it's a starter..

JejuSoul · Oct 1, 2016

Thanks Elmil. SOH is still not apparent from your data, but I'll continue to look. I think it must be somewhere, rather than a value that gets created on the fly.
It's interesting to work on this data anyway. There are values like the odometer reading there which I have never yet managed to get into Torque.

In PID 4F0 bytes 5, 6 and 7 your value 06FE0 corresponds to 45,667.2km driven. (although you write 45,640km - do you know why there is a difference?).

Elmil · Oct 1, 2016

It's accurate, I just had noted the odometer value before a trip today, then I made the final reading after I got home.

I actually already use some of these ID:s in my logger data:
200, 4F0 and 653 contain the DTE, Odometer and Outdoor temperature, respectively. It's very nice to have them together with the normal polled data, but it's a bit tricky (and slow). Also my PC software isn't optimized for speed, It's just a Python script...

Elmil · Nov 3, 2016

Today I had my car to the workshop for the steering recall (which I think most of the world did last winter :? ).

Took the opportunity and asked for another battery report. The SOH had changed from 99.2% to 99.5%. So I checked the deterioration numbers from these two occations and think I may have found a feasible formula for the SOH computation!

Here's the data:

Code:

Date	SOH		MaxDet	MinDet
27/9	99.2%	11.5%	10.0%
3/11	99.5%	11.2%	 9.7%

We know that the SOH stays at 100% quite long, and the deterioration values are changing frequently when triggered by the right charging event. So I made the assumption, that if the SOH is computed from the deterioration numbers, it must actually start a bit above 100% for a brand new car. The instrument could display 100% though.

To get my two datasets fit in a common formula, I came up with this:

Code:

SOH% = 110% - Average(deterioration)  =  110% - (MaxDet+MinDet)/2

If the SOH is calculated like this, it also means that in a situation of battery warranty at 70% SOH, the average deterioration would be 40%..

Side note:
The 110% baseline would correspond to a total energy capacity of 27kWh * 1.1 = 29.7kWh, or Amp capacity of 75Ah * 1.1 = 82.5Ah. This is just below what has been measured by the AVT test lab. Actually I think the cells were slightly over-speced by SKI for the test (4.3V max and 2.5V min), probably using the ABSOLUTE maximum and minimum voltages. I'm quite sure that SKI would never allow these extremes in any situation. For example, when I ran the battery flat this summer (on purpose), the HV battery was at 297V, which corresponds to 3.09V/cell.

Again, all this is still very speculative because of lack of data, but I think it makes sense as a starter point. The SOH formula should be easy to implement in Torque Pro too. And more data needed.

SoulEV2016 · Nov 3, 2016

Oh yeah, i like that.
Now, i can integrate this in the Torque Pro 004_CALC CSV.

JejuSoul · Nov 4, 2016

Elmil, am glad your SOH has improved. This is not what I expected. But clearly it is a fact. What was your odometer reading at the last test?
I would like to wait for more data before being confident that there is a simple formula linking Max and Min deterioration to the the SOH.

Elmil said:
The 110% baseline would correspond to a total energy capacity of 27kWh * 1.1 = 29.7kWh, or Amp capacity of 75Ah * 1.1 = 82.5Ah. This is just below what has been measured by the AVT test lab. Actually I think the cells were slightly over-specced by SKI for the test

No, the AVT tests we are using to state total capacity is 30.5kWh are from the 4 actual Soul EVs that are being tested at Phoenix. There is very little probability that these are over-specced. I ignore the value published in Vehicle Specifications and Testing Results 32.5kWh because that is greater than in the actual cars. It's probably just a typo because the Ah value is the same.

Compare the batteries of the 4 cars tested at https://avt.inl.gov/vehicle-button/2015-kia-soul

Elmil · Nov 5, 2016

The odometer was at 48978 km, previous SOH reading was at 45267. Between these readings I had several deterioration changes too, ups and downs.

Code:

Date  Odo (km)	 SOH (KDS)    MaxDet   MinDet
27/9  45267  		99.2%   		11.5%   10.0%
3/11  48978  		99.5%   		11.2%    9.7%

I will certainly ask to get the SOH-report every time I have the car to the workshop, next service will be around March if nothing else calls for it.

JejuSoul · Nov 5, 2016

Thanks Elmil, Your car is at number 2 in the deterioration race with 187,476km
The numbers I have always used in this race were a simple average of the deterioration numbers, and then follow a linear descent to 40% loss.
This was never intended to be an accurate assessment of the SOH.
Even if it turns out that this formula is the exact one that Kia is using, I don't want to put any faith in its predictions.

Elmil · Nov 5, 2016

JejuSoul said:
The numbers I have always used in this race were a simple average of the deterioration numbers, and then follow a linear descent to 40% loss.
...
Even if it turns out that this formula is the exact one that Kia is using, I don't want to put any faith in its predictions.

So you already predicted this? Really nice!

Well, it's most often the simple explanation that is the correct one..

And of course, what is important is whether or not this curve will be linear. Future will tell.

ZuinigeRijder · Nov 9, 2016

I had my car in for service (40.000 km) and the SOH was measured as 96.4%.

Torque Pro showed Min Det 12.1% and Max Det 14.1%, which gives an average of 13.1% just after the measurement.

I bought my car with 10.000 km on the clock and the first measurement was at 24454 km, Min Det 9.6%, Max Det 11.1%, average 10.4%.
From the CED value, the consumption was 0.20 kWh/km at 24454 km. I assume the CED value is including regen?
From 24454 km till 39901 the consumption was 0.16 kWh/km. So the demo car has been driven much faster/less economic. Unfortunately I do not know the figures at 10.000 km. If you look at the average deterioration, you see a rise in deterioration lately.

So would be the formula for deterioration: 109.5 % - 13.1 % = 96.4%.
Or another algorithm indirectly related to battery deterioration.

SoulEV2016 · Nov 9, 2016

i really think that i can build the 004_CALC in this way now (with 110% from the first measure). :idea:

Elmil · Nov 9, 2016

ZuinigeRijder said:
I had my car in for service (40.000 km) and the SOH was measured as 96.4%.
..
Torque Pro showed Min Det 12.1% and Max Det 14.1%, which gives an average of 13.1% just after the measurement.
..
So would be the formula for deterioration: 109.5 % - 13.1 % = 96.4%.
Or another algorithm indirectly related to battery deterioration.

Thanks for the data, ZuinigeRijder!

So, this shows it's not just the average of the extremes, but probably the average of all 96 cells. This in turn means that a big PID structure (like the cell voltage structure) should be present somewhere from the BMS, but we still don't know how to find it. Until it's found, I think the average of the min/max values is good enough as long as they don't diff too much.

SoulEV2016 · Nov 9, 2016

Yes, the BMS manage all cells individually ... and use discharge stage to manage the unbalanced cells.

http://www.mykiasoulev.com/forum/viewtopic.php?f=6&t=471&start=50#p3240

and

http://www.mykiasoulev.com/forum/viewtopic.php?f=6&t=471&start=140#p4570

balance stage is not very powerful, so a strategy is used to "mark" the cell ... the DET is based on this and the experience of the SK-Innovation (of destruction of cycled cells on benchmark).

if the cell rise not the 4,15-4,16v max voltage ... the DET indicator evolves with it.

ZuinigeRijder · Nov 10, 2016

Not the whole 30.5 kWh capacity is used, otherwise the total theoretical capacity would be 30.5/27 = approx 113%.
I think that it is around 110%, because there is a lower limit capacity never used, to protect the battery.
If the SOC % is 0%, the BMS % is around 2% and there is always a 1% left?

Elmil · Apr 4, 2017

.
My SOH has changed. But in the same time, it has NOT! :?

60,000 km service today. Went to the same KIA-authorized workshop as last time, and expected to find an SOH value of around 99.3% (matching Det 11.4/9.9) on the battery report.

When the car was ready, the battery report from KDS showed this:

Code:

SOH (State of Health):			   100.0 %
Accumulative Charge Current:		0.1 Ah
Accumulative Discharge Current:	0.6 Ah
Accumulative Charge Power:		  0.0 kWh
Accumulative Discharge Power:	  0.2 kWh
Accumulative Operating Time:		675 Sec

It seems that they have managed in some way to reset the BMS and clear all the history counters, probably not knowing what would happen (and still did it..)

My deterioration data is also cleared: MaxDet: 0.0, MaxCell: 0, MinDet: 0.0, MinCell 0
The BMS is in a virgin state, as if it was brand new... but the battery is the same, with 61000 km in it.

Seeing that this actually is possible, one should be careful when buying a used Soul EV from a dealer (or anyone). And what about warranty claims? Would an actual battery test be done, or just reading the SOH value.. A complete reset like this should be quite easy to discover though: There must be reasonable values on the energy and time counters compared with the odometer.

I expect my deterioration numbers to start showing values again soon - maybe tomorrow, as the c harger goes from 16% to 90%+ tonight.

SoulEV2016 · Apr 4, 2017

probably BMS firmware update to correct :
- kWh/Ah counting deviation
- and operating time deviation.

BMS software update

Help Support Kia Soul EV Forum:

Well-known member

Moderator

Well-known member

Moderator

Well-known member

Moderator

Well-known member

Well-known member

Well-known member

Moderator

Well-known member

Moderator

Well-known member

Well-known member

Well-known member

Well-known member

Well-known member

Well-known member

Well-known member

Well-known member

Similar threads

Join the conversation!

Join today and get all the highlights of this community direct to your inbox. It's FREE!