Thanks Mickey. I'll do Cap2 after I take it down to low charge next.
I'm not convinced the battery indicator uses coulomb counting - at least not only coulomb counting.
You’re correct, the BMU uses both methods to determine capacity, the ‘fuel’ indicator merely dispays SoC. Problem is that the actual battery capacity value has to be correct for that to work as expected.
If it did, it shouldn't matter what the assumed fully charged capacity is. I had it at low charge. OBDZero reported 0.2kWh and 1.3% SoC - I presume these were estimated, after I had disconnected/reconnected the 12v battery, from voltages.
The BMU calculates the SoC after a 12V reset based on battery (cell) voltage, this is only accurate, if there was no (or very low) current flow for at least 30min.
SoC is a ratio of the current battery capacity, therefore the BMU’s (wrong) interpretation is that the capacity is 1.3% * 9.2kWh when in fact it’s more like 1.3% * 6kWh.
If the BMU capacity value was correct, SoC would show 100% after charging is completed.
Charging it fully took it to about 5.2kWh - which is what I would expect. That is, I'm pretty sure that's the present capacity of the battery. This showed, as expected, as about 9/16 bars (as the system view of total capacity is 9.1kWh) and 50km range.
During a charge the car stops twice (for about 30min) to measure voltage based SoC, but switches to Coulomb counting once current flows, therefore the car does measure the actual capacity, but doesn’t change the BMU value accordingly (from experience it only allows corrections of about +/- 2Ah per month)
The 5.2kWh measured are therefore correct and if you add the (little) bit before the charge, your actual capacity is around 6kWh.
After leaving it overnight, RR was 55km. Seemed odd.
It looks like the BMU did detect some capacity change overnight, could it be temperature related?
I drove 33km. I didn't record the battery indicator but I think it was about 3/16. After charging (just over 2hrs at max 8*360 ~ 3kWh) the battery was 16/16 and RR was 89km. If the battery indicator were counting coulombs in and out, it should have returned to a maximum of 5.2kWh ~9/16. Do you agree with my conclusion that this implies that it can't be just coulomb counting?
Once a cell reaches 4.1V and the car isn’t in ‘calibration mode’ (start with less than 2/16) the BMU assumes the car is fully charged and sets the SoC to 100%. The battery indicator displays SoC and the RR works off the 9.2kWh.
You are correct, the car uses a combination of both methods.
While I'm sorting out how to make the most of this little car (you'll have gathered that I'm pretty keen to do so...), I'm happy to use an app + OBDII port to do the coulomb counting.
Remember the guy with the MUT3? Once CAP2 gives you an accurate capacity reading in Ah (my money is on 19Ah), he can increase the ‘‘battery age” until the BMU value is correct and all will work as intended afterwards.
