Dead Kona EV!

[Toolworker]

This is what I would suggest:

1. Check the traction battery cell balance for anything over 0.04 V deviation, preferably at 80% SoC or higher. Significant cell balance deviations may be triggering excessive BMS attention.

​

2. Assuming you're parked in a home garage, leave all the windows fully down and let the car go to sleep, an hour is enough, leave car unlocked. Switch on your BM2 and monitor the battery voltage in real time (on the first screen). Sneak up on the Kona (without having the key fob) and wiggle the doors one at a time up and down, in and out, by grabbing the inside handle but don't open any. The hatch and hood as well. Watch the dash for signs of a door open or ajar light, and the BM2 for sudden voltage changes. Obviously we're looking for a flaky switch.

No sign of a flaky switch.

3. If practical use the factory portable EVSE for charging and post BM2 graphs over a few days. You could mark them up with periods driving, parked-not-plugged-in, and parked-plugged-in. Post them individually so they're readable on this forum.

Interesting. I plugged in the Hyundai cord, connected to 240v through a homemade adapter. And a few hours later, the battery died again. Although a few days ago it did the same thing when not even plugged in.

​

At the moment I'm not seeing any evidence to blame the 12V battery.

Blaming the battery would be blaming the victim. But after many 100% discharges, I ought to get it replaced before it's out of warranty. The only reason I mention the 12v battery's condition is that I'd like some idea of whether, when I drive an hour to the dealer and ask them to replace it, they tell me it's fine.

 

[KiwiME]

I would hope after seeing these crazy graphs that they would replace it, and perhaps the LDC! Or, they'll blame the BM2 for all of it ...

I don't understand the voltages seen at the latter half of the 28th ("left car on") and just after 1:00 am on the 29th, the 13.3 V approx.

The system voltage should be either at 14.6 (charging) 13.1 (float) when the LDC (meaning the car's systems) are "on", or when the car is "off" anything from 13 V and down. There's no normal state that I'm aware of where it would be regulated at 13.3, especially for 20 minute intervals. Since it looks like that is a repeatable situation ("jumpered, left car on") I'd suggest verifying the accuracy of the BM2 with another voltmeter at that time. You'll need to unplug the hood latch switch so that you can access the battery without disturbing the car. Fortunately that switch is 'open' when the hood is closed.

The 6 hours of sawtooth on the 28th is about the longest I've ever seen that happen on my own 2019. Usually it's 2-3 hours.

I can suggest other tests (like substituting a known-good 12V battery) but assuming the BM2 is not flakey I think that there is enough visibly wrong with either the 12V battery and/or the LDC (or whatever regulates system voltage) that the dealer should be able to justify carrying out some diagnostics.

 

[Toolworker]

I would hope after seeing these crazy graphs that they would replace it, and perhaps the LDC! Or, they'll blame the BM2 for all of it ...

I don't understand the voltages seen at the latter half of the 28th ("left car on") and just after 1:00 am on the 29th, the 13.3 V approx. The system voltage should be either at 14.6 (charging) 13.1 (float) when the LDC (meaning the car's systems) are "on", or when the car is "off" anything from 13 V and down. There's no normal state that I'm aware of where it would be regulated at 13.3, especially for 20 minute intervals. Since it looks like that is a repeatable situation ("jumpered, left car on") I'd suggest verifying the accuracy of the BM2 with another voltmeter at that time. You'll need to unplug the hood latch switch so that you can access the battery without disturbing the car. Fortunately that switch is 'open' when the hood is closed.

The 6 hours of sawtooth on the 28th is about the longest I've ever seen that happen on my own 2019. Usually it's 2-3 hours.

FWIW the 13.3v charging voltages after the car was turned off late on the 28th match the voltage while the car was on. When it returned to the four hour charging interval, it also returned to 14.6v.

I checked the BM2 against Torque Pro at several voltage levels; they always agree.

I can suggest other tests (like substituting a known-good 12V battery) but assuming the BM2 is not flakey I think that there is enough visibly wrong with either the 12V battery and/or the LDC (or whatever regulates system voltage) that the dealer should be able to justify carrying out some diagnostics.

I have little faith in a dealer fix after having gone through a long, futile process with my '20 Kona EV. Eventually I got them to show the graphs to a Hyundai techline agent, who said the car was fine because there were no error codes (!!), and the graphs were "normal operation"!!!

​

After the battery recall I wound up with a '21, and things were better for a while - until they got worse.

I'll see if I can find a savvy service advisor, and give it yet another try. But as I posted in another context, my expectations this time are:

​

 

[KiwiME]

It seems quite a coincidence that two Konas in a row have gone bonkers. What's in common? The BM2, the wall charger? 5G tower nearby?

Over on an Ioniq 5 group there is a Hyundai tech that insists that the BM2 could upset the car's assessment of the 12V batteries 'internal resistance'. I think he's wrong but I'm going to check it out if I can find another unobtrusive way of logging voltage.

 

[Toolworker]

It seems quite a coincidence that two Konas in a row have gone bonkers. What's in common? The BM2, the wall charger? 5G tower nearby?

Over on an Ioniq 5 group there is a Hyundai tech that insists that the BM2 could upset the car's assessment of the 12V batteries 'internal resistance'. I think he's wrong but I'm going to check it out if I can find another unobtrusive way of logging voltage.

I could remove the BM2 and give it a month. The way things are going, it should die before then. Even if not, it would give me a rest from checking the BM2 app.

Random web pages say lead acid batteries' internal resistances range from 0.005 to 0.030 ohms. My BM2 draws about 1.7mA, which comes to about 7,000 ohms at 12v. It's hard to imagine that affecting a resistance measurement in milliohms. Of course the tech may well observe BM2s on an unusual percentage of cars with 12v charging problems, but that's like calling hospitals dangerous because so many people die there.

It's possible there's something in the environment here. There was another problem with the ABS on the '20 occasionally just stopping for a day or three. That was observed with my 240v EVSE, and the included 120v EVSE plugged into both 120v legs of my 240v supply, and then in desperation into my neighbor's 120v.

Or it could be just me. When I was programming, I could break pretty near any system just by using it. But not all users notice when something's broken, fewer report it, and still fewer can document and duplicate it.

 

[Toolworker]

Updating with a couple more observations (including that I am an idiot).

ABS is running normally at the beginning. Then some driving, returning home, plugging in with the Hyundai charger at 240v., charge completes on TOU schedule ...

​

and then off the cliff once more.

​

Then today:

​

Again, ABS running normally with EVSE unplugged. Then drove, parked, drove back. Since things had been going normally, I thought to remove the BM2 for a while and see if problems stopped. After disconnecting it, I thought to check one more time, so tacked it back on and noticed (1) that its memory is volatile and (2) the voltage was 12.1v and dropping. After a few minutes it was down to 7v, so I jumped and restarted the car.

I am an idiot because (1) I should have downloaded data before disconnecting the BM2, and (2) I forgot to plug in the OBD2 and check the amperage load as the battery dove off the cliff.

I sure wish I had the data from 10:15 to 2:45!

 

[Toolworker]

BTW does anyone know the Torque Pro parameters that give the voltage and current for the 12v battery? There are several 12v parameters and sometimes they differ.

 

[KiwiME]

Because the LDC is normally operating whenever OBD data is live, what those PIDs tell you is the 12V system voltage and the LDC output current.
I’ll have to check tomorrow as which PIDs I’ve found are reliable.

 

[hobbit]

For aux battery current itself [measured at the shunt on the negative terminal] I've been using
7E2, 2102, (32700z((W*256)+V))/100 which you might have to translate to whatever your app wants.
For LDC current I have 7E2, 2102, ((Q*256+P)/100 which I think is simpler because it's always
a positive quantity. Come to think of it, 32700 in the first one might be a little off [32767??] but
I think I was trying to compensate for some small observed error. The derivation is someplace
in ooooold IEVF history...

I really should pull together the page on Kona OBD2 stuff I've been meaning to for like 2 years now
but never quite knew how to structure it.

_H*

 

[KiwiME]

I'll note that Car Scanner Pro has a plausible display of Aux Battery Current. Despite being a highly configurable app, it (unsurprisingly) doesn't reveal the PID as best as I can see.

 

[Toolworker]

No need to wonder whether the 12v battery is bad. It's gone. So any voltage recordings the last few days don't mean much.

If I'm quick, I can turn the car off and lock the doors before it goes totally dead. I open it with the manual key and keep the jumper pack in the center console.

Is there any reason not to drive with it until the dealer can put a new one in?

 

[KiwiME]

Not that I can see. You could always switch to Utility mode instead of off at home and just leave it until the car is next needed.

 

[Toolworker]

Not that I can see. You could always switch to Utility mode instead of off at home and just leave it until the car is next needed.

I think that would draw a couple of kWh per day. (Maybe it's dumping energy into the dead battery.) Or I could just jump it.

Where it would be really handy would be running errands - if the car could be locked in Utility Mode. But it can't.

I could carry nothing valuable in the car and leave it unlocked. I'd have to take the jumper pack with me; can't leave that home and can't afford to have it stolen.

 

[Keith Smith]

Well, I don't carry a voltage meter on me so that's easier said than done!

Just what I want from a 7 month old $62 000 car! Not impressed. Especially when I have it set up to use the battery saver function. Shouldn't happen.

You are right, it shouldn't happen. Unfortunately batteries are batteries, and the 12v's are a weak link. Sometimes they just "Die Hard" (heheheheh).

 

[Keith Smith]

I assume you must be dealing with salt air? It's funny, I've owned 35 years worth of cars and have bought new batteries maybe twice in that span. And that's in Ontario where we deal with huge temperature swings - and salty winter roads.

In North Carolina and Virginia an overpriced battery lasted seemingly forever. In Phoenix, Arizona, after 2 years it'***** or miss. And when they go, they generally just completely die. No dome light nuthin'.

 

[Keith Smith]

The DC to DC converter runs whenever your car is turned on and likely in motion, the aux battery saver is just software that engages the converter when it detects low voltage from the lead acid battery during periods when the car is turned off. I suspect the utility mode is useful when the car is turned on but not in motion as the converter would not otherwise engage.

I am going to speculate automatic mode engages whenever it meets low voltage criteria, while cycle is just that, a schedules duration and frequency of charge not dependent on actual battery voltage.

One might note, it is highly likely that the relays that turn on the DC-DC converter are 12v, so if you battery goes to toast, there is a high probability you are dead in the water.

 

[Keith Smith]

Yea, in some regards, BlueLink is pretty useless.

I'd go more with pathetic. I find it stunning just how bad BlueLink is.

 

[Zem]

Where it would be really handy would be running errands - if the car could be locked in Utility Mode. But it can't.

Hello. You can lock the last door with the metallic key.

 

[Toolworker]

In the last episode of this telenovela, we left our hero frantically trying to get a human being at the Hyundai dealer on the phone to see if Parts had the battery. No point taking the car in otherwise.

Wow! They did have it, and could replace it right away. (organ music)

After several hours with the car, Pradep, the tech, explained that he had to charge the battery overnight and put instruments on it to document its demise for HMA. They need the specific printouts before authorizing replacement under warranty. (sad organ music)

So I left it for a couple of days and picked it up today with a new (flooded) battery. Pradep had left the instrumentation on the car, but nothing unusual happened.

I asked what I should do when in five or six weeks I find the battery dead again. He said don't touch it, just call him. No BM2, no checking error codes - let him have the first shot at diagnosing it. For the first time, I feel like there's a competent tech and if there's a way to find the problem, he's my best shot. (fade up inspirational organ music; cut to commercial)

FWIW Pradep said he has seen cars with a dead 12v battery traced to a module under the dash that's supposed to shut all the systems down three minutes after you turn it off and leave. Sometimes the module will intermittently leave it on and kill the battery. Replacing it always fixes that problem.

 

[KiwiME]

Sounds like a plan. I know my Kona takes 9 minutes to drop to the normal standby current after being powered off. Perhaps you should loan them your wall charger for the full experience?