The Kona EV 12v battery issue

The problem

The 12V Auxillary / leisure battery voltage drops too low to keep the car ECU / systems alive.  The car is then essentially electronically dead.  The physical key needs to be turned in the driver's door, the bonnet needs to be released and a 12v supply connected to power up the systems / charge the 12v battery.

It happened to us

Soon after buying the car - when it was about 4yrs old - we experienced the 12v failue twice.  Hyundai did not replace it under warranty, but since a traditional car battery is cheap and easy to replace I did that and thought no further of it for over a year.

Then it happened again

It coincided with me querying the ECU via a Bluetooth ELM327.  The hardware I added used around 0.5w.  It could not have been the addition of the equipment since that had been in place for 6 months...  I'd only changed the software recently... the co-incidence was significant.

What I was doing...

I was querying the Battery Management System module (at header 7E4) specifically PID 220101 via an ELM327 adapter in the OBD-II port to get power use whilst driving.  After the car is turned off the BCM (7E4) stops responding so I presumed it went to a powersave mode and just kept trying as long as the ELM327 was connected (the queries started getting a response when the car was turned on again).

I removed all the equipment and added it back bit by bit as well as enabling each bit of software bit by bit.

Findings...

Along with querying the BCM for power consumed I logged the 12v supply voltage (ATRV) available from the ELM327.  

With my hardware installed, bluetooth connection active, but no CANBUS data all was good. 

Re-enable the BCM queries, even if there is no response - failure again.

Having logged the 12v and power consumed from the traction battery (via the BMS) during periodic 12v charging cycles some conclusions can be drawn.

The graph below shows the instance where the 12v battery failed (red) and 2 instances where it was OK (green & blue).

The rate of voltage drop (discharge rate) of the aux battery is clearly higher in the failed case.  Some basic maths (an excess of 45whrs every 4hrs) implies an excess draw of 10w.  I was unable / unwilling to place a ampmeter inline with the 12v battery so that is a deduction, not a measurement.

I used the presence of more than 13v at the ELM327 adapater to know whether the car was powered on (really if the DC DC converter was active) - and now only send OBD-II messages when the traction battery is the power source.  Once the voltage drops below 13v I stop.  Arguably what I should have done all along.

The 12v is queried with ATRV every 5 minutes so the data above will have lost (on average) 2.5mintes of data at each charge cycle, but it is representative.  Cycles are around 20min so a potential for a 10% under read, possibly sightly more since most the charge is not constant current - higher load first.

Conclusion

I presume that the problem experienced by other Kona owners is similarly caused...  with bluelink active something is querying the car's systems frequently enough to prevent them entering a low power state and consuming more power than the 12v battery can provide when it is only being charged once every 24hrs.  

In my case the 0.5w load is so low as to make no difference and the Aux battery sits at 12.5v for 11+ hrs a day when in the daily charge cycle.

 

Background

This page is based on hindsight - having had the opportunity to work through the 12v charging issue after (re)-creating it by accident myself.


There is a clear pattern to the automated charging of the 12v battery.  Once the car is turned off, it wakes after 4 hrs to charge (it seems the gap between charges is 4 hrs rather than a 4 hr cron job).  After 14 of these cycles or around 2.3 days / 56hrs it reduces to once per day.

I have also read that if the traction battery is below a certain level it only charges the 12v battery once per day.