Drive battery degradation and the out of whack IMO BMU.

[elm70]

BMU thinks SOC = X and Capacity is Y. Based on X and Y, BMU thinks a certain amount of charge should be added before reaching the target cell voltage. If that cell voltage is reached earlier or later, then apparently either the SOC or the Ah number was incorrect -> Error.

Is that the idea? Maybe I should have said "coulomb count" instead of kWh?

I expect charging process to not consider any SOC

Charging a lithium battery is always about constant current charge, (more for convention/convenient then for anything else) , follow by a constant volt charge (4.1v per cell should be never exceeded)

So .. charging should be about a flow of current (variable at end of charge process), that has nothing to do with SOH and SOC ... still measuring the flow, the BMU could use this parameter for check and re calibrate SOH of the battery

Only issue is that by definition, charge from 30% up to 100%, needs more Ah, then the current available at discharge 100% to 30% ... so .. there must be some correction factor if this value need to be used by the BMU ... actually checking some diagram here: (https://lygte-info.dk/review/batteries2012/Sony%20US18650VC7%203400mAh%20%28Green%29%20UK.html) slow charge and slow discharge .. shows almost the same current in mAh ... so .. maybe it is used directly without any factor (clearly this is different then the kwh from AC charging side , since the inverter AC-DC has a efficiency impact to be consider, especially at 16A charge, )

 

[anko]

still measuring the flow, the BMU could use this parameter for check and re calibrate SOH of the battery

I 'think' this is exactly what I meant earlier: before charging commences, the BMU calculates how much energy should flow into the battery before it reaches 4.1 cell voltage, taking into account its notion of capacity and SoC. If less (or more) energy flows into the battery before 4.1 volt is reached, either the notion of SoH was wrong or the notion of SoC was wrong. I thought you considered the deviations 'errors'.

 

[anko]

Did my "what happens after two hours rest" test: upon arrival 29,17% SoC. After 2.5 hours: 29,12% SOC. No significant difference, I guess.

Edit - come to tink of it, I checked SOC reported by BMU using my own tools, not current capacity Ah using the Dog. Could it be possible that, after reading, current capacity / remaining capacity is not the same as SoC%? To late to check now.

 

[elm70]

Did my "what happens after two hours rest" test: upon arrival 29,17% SoC. After 2.5 hours: 29,12% SOC. No significant difference, I guess.

Edit - come to tink of it, I checked SOC reported by BMU using my own tools, not current capacity Ah using the Dog. Could it be possible that, after reading, current capacity / remaining capacity is not the same as SoC%? To late to check now.

29.17% down to 29.12% .. yes it is insignificant

But, these are strange numbers ... SOC in & is equal to 100 times the estimated capacity divided by Estimated Max capacity .. both values are known with 0.1Ah accuracy .. at least using the DOG or EvBatMon ... your tool apparently is able to read the BMU with higher accuracy .. or it is reading SOC in a different way ...

 

[anko]

Did my "what happens after two hours rest" test: upon arrival 29,17% SoC. After 2.5 hours: 29,12% SOC. No significant difference, I guess.

Edit - come to tink of it, I checked SOC reported by BMU using my own tools, not current capacity Ah using the Dog. Could it be possible that, after reading, current capacity / remaining capacity is not the same as SoC%? To late to check now.

29.17% down to 29.12% .. yes it is insignificant

But, these are strange numbers ... SOC in & is equal to 100 times the estimated capacity divided by Estimated Max capacity .. both values are known with 0.1Ah accuracy .. at least using the DOG or EvBatMon ... your tool apparently is able to read the BMU with higher accuracy .. or it is reading SOC in a different way ...

Indeed. As far as I know, there are three ways:

1 - Taking item 1 rom the BMU Data List.

This gives an granularity (I think this is the term to use, instead of accuracy?) of 0.5, as the formula is byte A / 2 - 5.

2 - Take Current Capacity (item 17 from the BMU Data List) and divide by Remaining Capacity (item 18 from the BMU Data List)

I think this is what the Dog does. Both inputs have granularity of 0.1.

3 - Retrieve from traffic broadcasted on CANBUS

Frame type 375 with formula (D*256+E)/100-5 gives SOC at a granularity of 0.01 and at a much much higher refresh rate

Apart from the granularity and refresh rate, methods 1 and 3 yield the same value. After last night, I am questioning whether method 1 and 3 produce the same outcome as method 2, especially after resting. Need to test again tonight.

BTW: All data shown in my little video on B - levels and coasting in B0 is obtained by grabbing existing data from CANBUS, not via Q & A on OBD. Much, much faster than the Dog could ever be (https://www.youtube.com/watch?v=IQeMVTG77OQ&t=165s).

 

[anko]

So, did another registration. Directly upon arriving at home:

SoC calculated by the Dog: 7.4 Ah out of 26.2 Ah = 28.2%
SoC reported by the BMU: 28,13%

Pretty close . After about 2 hours:

SoC calculated by the Dog: 7.5 Ah out of 26.2 Ah = 28.6%
SoC reported by the BMU: 27,80%

So, Ah has gone up (slightly, but still) while SoC has come down a bit, according to BMU.

After sitting in my car for just a few minutes, Dog calculated value went back to 7.4 Ah out of 26.2 Ah = 28.2%. By that time, BMU reported value was down to 27.60%.

Would this mean anything?

 

[elm70]

So ... apparently the DOG and your tool are making the same result about SOC in % and capacity left in Ah

As well ... since by driving down to 100% to 28.2% ... the SOC does not change after rest .. it means you car has really 26.2Ah capacity left from 100% SOC ... at least accordingly to the "error" that the current flow in your car may introduce.

So ... I would expect that the BMU of your car will not reduce anytime soon the SOH .. so from 26.2Ah down to 26.1Ah may take a bit longer then normal .. at least .. if my assumption about the BMU implemented logic is correct

 

[anko]

We'll have to see. Last 0.1 Ah drop was between end of July and 15th August. Last 0.5 Ah drop was between 15th and 18th of August. All within about 3000 km. Since then I have driven just little over 1000 km. It would be rather annoying if Ah value got stuck on 26.2 :evil:

 

[elm70]

We'll have to see. Last 0.1 Ah drop was between end of July and 15th August. Last 0.5 Ah drop was between 15th and 18th of August. All within about 3000 km. Since then I have driven just little over 1000 km. It would be rather annoying if Ah value got stuck on 26.2 :evil:

I can understand ...

Probably there could be some trick ,or usage pattern that can cause the BMU to see less capacity and downgrade the SOH

But ... the battery smoothing procedure will equalize it ... actually the smoothing procedure is looking to update the BMU with an optimistic high SOH

PS: What is your EV range now ? I guess you can still do over 30 or even 35km in EV mode ... right ?

PPS: My only 0.5Ah lost in a single BMU update , was after my vacation which I did 3000km without any charging ... I would have expected zero battery degradation while using the car in "ICE" mode .. but possibly using often the "CHARGE" functionality for gain 10/15km of EV range ... it could have upset the BMU, and possibly also "aged" faster the battery too ...

 

[anko]

We'll have to see. Last 0.1 Ah drop was between end of July and 15th August. Last 0.5 Ah drop was between 15th and 18th of August. All within about 3000 km. Since then I have driven just little over 1000 km. It would be rather annoying if Ah value got stuck on 26.2 :evil:

I can understand ...

Probably there could be some trick ,or usage pattern that can cause the BMU to see less capacity and downgrade the SOH

But ... the battery smoothing procedure will equalize it ... actually the smoothing procedure is looking to update the BMU with an optimistic high SOH

PS: What is your EV range now ? I guess you can still do over 30 or even 35km in EV mode ... right ?

PPS: My only 0.5Ah lost in a single BMU update , was after my vacation which I did 3000km without any charging ... I would have expected zero battery degradation while using the car in "ICE" mode .. but possibly using often the "CHARGE" functionality for gain 10/15km of EV range ... it could have upset the BMU, and possibly also "aged" faster the battery too ...

Last night, the car reported 36 km EV range. But that was several hours after charging. But I think 32 is more realistic.

The last three drops (2 x 0.1 and 1 x 0.5) were indeed during my vacation when I had, compared to the amount of KMs driven, relatively few charge sessions. But still, at least once a day.

Think it would 'help' to charge to 95% rather than 100%?

 

[elm70]

Think it would 'help' to charge to 95% rather than 100%?

In case of my car, it would help to trick the BMU ... or better

When I charge my car to 100% .. and I drive it down to 75% .. after rest it believe I gain capacity (~1Ah win)

Then after having a fake gain, and I drive other 25% of battery ... then after a rest it believe it lost over 2Ah

So .. if I would charge my car to 80/90% .. and discharge down to 50% ... it will be regularly see 2Ah lost while resting.

Anyhow .. this will only trick the BMU ... so a smoothing procedure will equalize it ...

Possibly .. if the aim is to age faster the battery ... then ... don't charge the car ... and use the "CHARGE" mode for make win 50% SOC, and then go down to 0km EV ... this should be more or less forcing a fast charge in the car ... also ... if the battery is over 30deg using CHARGE can extra age the battery too

But ... it is quite an expensive way to age the battery ... in my case charge from the network is 4 time cheaper then use fuel

 

[Trex]

Ok catching up with where I am with my drive battery.

The dealer did a Auto capacity measured procedure on the Phev which took it back above 80% health. This procedure uses the electric heater to empty the battery then measures the amps and time when charging the battery back to 100% to work out capacity or health. If you do not have a electric heater it will use the air conditioner which will take longer. Best to take the phev in with as low SOC as possible to take the shortest time to to do test.

They would not do a BMU reset on my Phev while it is under warranty.

Regards Trex.

 

[anko]

So ... I would expect that the BMU of your car will not reduce anytime soon the SOH .. so from 26.2Ah down to 26.1Ah may take a bit longer then normal .. at least .. if my assumption about the BMU implemented logic is correct

Just happened.

 

[elm70]

So ... I would expect that the BMU of your car will not reduce anytime soon the SOH .. so from 26.2Ah down to 26.1Ah may take a bit longer then normal .. at least .. if my assumption about the BMU implemented logic is correct

Just happened.

Very strange indeed ...

So .. possibly it is looking like that my "error" don't count that much ... so .. maybe it is more related to the charging current

PS: About rest time .. I find out that if the car is not used for 1h .. and then used again ... it is looking the battery re-calculation happen without waiting the 2h for rest and fully bounce back the voltage ... I did seen strange result with very pessimistic SOC due to multiple usage without having enough time for the battery to rest and pop up in voltage

 

[elm70]

I just got my car from service (last time was almost 2y ago, I did jump one since I was driving 95% in EV, so for only few thousands km done in ICE, I did not believe was needed to make a new service ...

Anyhow ... as I mention in a different thread ... this time at the service they did mess up with the firmware

Not only I lost the remote climate control and the keyfob hack .. but I find out something "strange" too in the battery area

The first full charge I did, for the first time ... the DOG did report me 102.1% , 34.7Ah (SOH is 34.0Ah) .. at 4.081v per cell :shock: :shock: :shock: ... not only my car now get a full charge 0.01v less then normal/before, but as well, it believes the capacity is much more than the assumed real one

As result my morning trip 10km from 34.7Ah down to 27.2Ah ... after 8h parking, the car started with 24.9Ah .. so 2.3Ah lost .. when normally I was winning 1Ah in the previous run.

Now .. every time I park and leave my car unused/uncharged .. I came back for discover always a relevant SOC loss.

The firmware before was not good .. but the last one is looking terrible !!!

I need to consolidate the data, similar to some pages before which I show 6 trips without charging, and the lost in SOC while parking (total was 1.5Ah) ... but now I'm afraid this level is above 3Ah .. so 10% less capacity .. so 31Ah usable ... so ... my real SOH is not far from 80%

PS: I will not fully charge my car till Friday ... so I need to wait if also next time after a full charge my SOC is above 100% with less then 4.09v per cell

PPS: I'm surprise that with so many and huge deficit of SOC while parking, the car did not reduce already the SOH .. which is steady at 34.0Ah since 31st August ... next SOH change, I'm afraid it will be a 0.5Ah change

EDIT : Possibly this implies that SOH is updated only after "charging" ... can it be that the PHEV does only count the charged Ah, and ignore the consumed Ah while driving ?

Note ... something strange too .. after a short trip, with battery down from 19.7Ah to 17.1Ah (Dog report 3.5Ah consumed but 0.9Ah regen ... so total 2.6Ah net usage) ... after rest ... I lost further 0.6Ah .. so down to 16.5Ah ... I'm wondering if the car count wrongly the regen 0.9Ah in 4km short trip maybe is wrong ... so maybe only 0.3Ah was going back in the battery .. :?: