12V battery
- Thread starter electriceddy
- Start date
- Replies 506
- Views 131K
Woke up to a 4.6 V battery this morning. I should have taken appropriate action yesterday morning when I noticed the battery saver light was on at a weird time.
Slipped my old 12 V charger on the posts and after many relays clicking(5 minutes or so) I was able to start the car and remove the charger.
The tech is coming back from training (including 2 weeks self isolation) on Monday so it will be an interesting morning .
Perhaps the EMCU (EPCU) is bagged now as its maximum output is 13.6 V where it was 14.4 - 14.6 V before the incident
https://insideevsforum.com/community/index.php?threads/dead-kona-ev.6991/page-8#post-82199
Together with the tapping noise repairs procedures I mentioned are going to happen
https://insideevsforum.com/communit...g-vibration-type-sound.7262/page-6#post-95972
I can probably expect to be without my Kona EV hopefully not for too long
Slipped my old 12 V charger on the posts and after many relays clicking(5 minutes or so) I was able to start the car and remove the charger.
The tech is coming back from training (including 2 weeks self isolation) on Monday so it will be an interesting morning .
Perhaps the EMCU (EPCU) is bagged now as its maximum output is 13.6 V where it was 14.4 - 14.6 V before the incident
https://insideevsforum.com/community/index.php?threads/dead-kona-ev.6991/page-8#post-82199
Together with the tapping noise repairs procedures I mentioned are going to happen
https://insideevsforum.com/communit...g-vibration-type-sound.7262/page-6#post-95972
I can probably expect to be without my Kona EV hopefully not for too long
KiwiME
Well-Known Member
As I alluded to in an earlier post the lower voltage may be due to the LDC current limiting when charging from a very low SoC. Check the image in post #78. The aux battery could also have been damaged but my guess is that it will come right after a proper charge. The bigger question is why did it discharge so far down?
The lessor question that I have is why you thought the aux saver event was unexpected? I've never seen evidence of that function operating at any time other than on a 24 hour schedule after the completion of the last normal drive but would be happy to be proven wrong, especially if it was triggered by a low aux battery SoC. It may have just been the scheduled time as it's not restricted to nighttime.
I'll just add what I posted over at UK site SpeakEV where they also have such problems. This is what I understand happens to the aux battery so far:
While driving or charging past the "charge" period of 10-20 min, the system voltage will suddenly drop and dither just above 13 V for the duration of the trip (or charge) as the system (LDC) maintains a high-enough voltage as to draw effectively nothing from the aux battery. That system voltage is quite noisy so don't rely on random spot checks. 12.9 to 13.4 is perfectly normal for this period.
As I also mentioned, I've seen evidence that the car tests the aux battery, and I'll wildly speculate that it's trying to determine the health. It seems the i3 also tracks battery health and apparently there is a procedure required after a new battery is installed which owners refer to as 'registering the battery'.
The lessor question that I have is why you thought the aux saver event was unexpected? I've never seen evidence of that function operating at any time other than on a 24 hour schedule after the completion of the last normal drive but would be happy to be proven wrong, especially if it was triggered by a low aux battery SoC. It may have just been the scheduled time as it's not restricted to nighttime.
I'll just add what I posted over at UK site SpeakEV where they also have such problems. This is what I understand happens to the aux battery so far:
While charging it should be around 14.65 V. That may be affected with temperature compensation that I haven't encountered during our summer.
- It only charges for around 10-20 minutes at the start of each driving or charging event.
- Same with aux battery saver events which trigger daily once the first 24 h period has elapsed after the last proper drive, meaning over about 20 min. Shorter drives, moving the car, opening a door, or entering Run mode to check something doesn't reset the trigger or disrupt that schedule.
While driving or charging past the "charge" period of 10-20 min, the system voltage will suddenly drop and dither just above 13 V for the duration of the trip (or charge) as the system (LDC) maintains a high-enough voltage as to draw effectively nothing from the aux battery. That system voltage is quite noisy so don't rely on random spot checks. 12.9 to 13.4 is perfectly normal for this period.
As I also mentioned, I've seen evidence that the car tests the aux battery, and I'll wildly speculate that it's trying to determine the health. It seems the i3 also tracks battery health and apparently there is a procedure required after a new battery is installed which owners refer to as 'registering the battery'.
Last edited:
Desulfator waveforms can include pulses or other shapes from micro seconds to milliseconds. I wonder if Hyundai is doing 12V aux battery desulfation at times?
There may also be aliasing from other switched or pulsed waveforms that might be influencing the data.
There may also be aliasing from other switched or pulsed waveforms that might be influencing the data.
Last edited:
hobbit
Well-Known Member
When I pulled the little connector off the "intelligent" battery sensor at the
negative, *with* the car powered up if I remember right, the LDC output
went to 13.6 or 13.8 and stayed put -- over several more power cycles!
I didn't want to make any assumptions because the weather [temps] was
rather changeable at the time, so ambient may have been affecting the
LDC algorithms too ... but it's clear that the LDC has a fallback mode if
it can't get the data, and it may want to do a lot of sanity-checking or
decisions over a time domain before going back to its usual.
One could try just running around with the sensor disconnected, and
watch battery health. Of course you wouldn't be able to monitor the
battery current via OBD2, but eyeballing the voltage and knowing what
loads you're using, and checking in both "on" and "ready" modes when
the LDC is running or not, could bring some insight and possibly a
simple workaround for this whole thing. I'll try to do some playing
around with this, and make sure I've got a booster battery handy.
If we're ultimately allowed to drive around at all, that is ... us car-owners
are very lucky, when you think about it, having our own little isolation
bubbles on wheels. As long as we don't bring anything nasty inside...
_H*
negative, *with* the car powered up if I remember right, the LDC output
went to 13.6 or 13.8 and stayed put -- over several more power cycles!
I didn't want to make any assumptions because the weather [temps] was
rather changeable at the time, so ambient may have been affecting the
LDC algorithms too ... but it's clear that the LDC has a fallback mode if
it can't get the data, and it may want to do a lot of sanity-checking or
decisions over a time domain before going back to its usual.
One could try just running around with the sensor disconnected, and
watch battery health. Of course you wouldn't be able to monitor the
battery current via OBD2, but eyeballing the voltage and knowing what
loads you're using, and checking in both "on" and "ready" modes when
the LDC is running or not, could bring some insight and possibly a
simple workaround for this whole thing. I'll try to do some playing
around with this, and make sure I've got a booster battery handy.
If we're ultimately allowed to drive around at all, that is ... us car-owners
are very lucky, when you think about it, having our own little isolation
bubbles on wheels. As long as we don't bring anything nasty inside...
_H*
Toolworker
Well-Known Member
@KiwiME, @mikeselectricstuff - A meta-question: Can you post links to other threads that discuss the dead 12v battery problem? Is there one good place to focus on it?
I started this one and posted a graph and video documenting my EVSE and the Kona in a loop that drains the 12v battery. I'd be happy to combine it with another one specifically on the topic.
@apu, @electriceddy - Were you plugged into your EVSE when you found the battery dead? What model EVSE?
The car was not plugged in or charging
If it followed the 24 hour cycle , it should not have come on until late Friday afternoon as I drove the car on Thursday (the day previous) and parked it about 3pm. It had to have been triggered by low SOC as you mentioned.
KiwiME
Well-Known Member
There is a bit more technical expertise here, best just keep it here in this thread. I barely get a 'like' over there, and never a 'thank you' for providing info.
R P
Well-Known Member
Just want to say, I for one, am grateful to have you and others on this forum provide such technical and useful insight to some of the problems/issues with our cars. I may not always respond or like (I should do more of the latter), but am certainly following. Keep it up...
apu
Well-Known Member
Your presence here together with the other kind contributors is very appreciated,thank you.
Thank you for your efforts, most interesting reading. You inspired me to look more closely at my 2019 Ioniq electric 12V system. Ioniq electric owners are having the same, or similar 12V problems. Same as you, I am curious, but have had no 12V problems yet myself.
I am about to take some data with a DC split core current transformer (10 us, 50A, 4-20 mA loop) and Keithley 6500 digitizer, no idea yet if anything interesting or meaningful will show. May I post some ioniq electric results here, or would you prefer that I start a new 12V thread in the Ioniq electric forum?
Toolworker
Well-Known Member
Please post here, and can you give us a link to the Ioniq forum discussion(s)? Maybe we can get this all in one thread, so all of us with the same problem are talking with each other.
KiwiME
Well-Known Member
I'd sure like to see any results either here in this thread or as a link to an Ioniq thread. I have no easy access to such useful kit. Formulating a test plan that will identify the fault(s) is the challenge. Do you have BlueLink app capability on your Ioniq?
Thank you, I think a collaboration would be good. No links as such to the other forum, at the Ioniq electric forum (https://www.ioniqforum.com/forums/hyundai-ioniq-electric-ev.234/) there are any number of 12V battery posts, some take voltage readings on occasion.
I need to verify my setup as we work, I will try to be careful, but in the past, I've found mistakes, so caution work-in-progress. For the moment, I am only looking at current on the positive 12V battery cable, which appears to be a single loom, no idea how many conductors in there yet. It almost does not matter, as long as the DC current transducer is over all of them. I need to look again to verify only the one cable(s). The 12V negative terminal has the additional sensor wires, but I should probably move the CT over there to be sure the results are about the same.
For a first quick test, running 10 second blocks of 100 microsecond sample rate (10 kHz). Using Blue link, I tap the round repeat symbol on the first page on my iPhone. I get something like these graphs every time. Scale is on the order of, near 0 mA is about -10A (charging direction) and full scale should be 20 mA, or 50A (discharge). So, the first odd thing is that I'm seeing readings approaching 30 mA (4-20 mA loop, with some under range to -10A), which assuming the sensor is working properly means some over scale reading above 50A. (4 mA is zero A, 20 mA is 50A) I have another 200A sensor, but it is not fast enough (100 mS rise time) to see these 300 uS high current pulses (two every time): Is that just charging some capacitors somewhere, starting the LDC?, or are there two very short (300 microseconds), short circuits?
I need to verify my setup as we work, I will try to be careful, but in the past, I've found mistakes, so caution work-in-progress. For the moment, I am only looking at current on the positive 12V battery cable, which appears to be a single loom, no idea how many conductors in there yet. It almost does not matter, as long as the DC current transducer is over all of them. I need to look again to verify only the one cable(s). The 12V negative terminal has the additional sensor wires, but I should probably move the CT over there to be sure the results are about the same.
For a first quick test, running 10 second blocks of 100 microsecond sample rate (10 kHz). Using Blue link, I tap the round repeat symbol on the first page on my iPhone. I get something like these graphs every time. Scale is on the order of, near 0 mA is about -10A (charging direction) and full scale should be 20 mA, or 50A (discharge). So, the first odd thing is that I'm seeing readings approaching 30 mA (4-20 mA loop, with some under range to -10A), which assuming the sensor is working properly means some over scale reading above 50A. (4 mA is zero A, 20 mA is 50A) I have another 200A sensor, but it is not fast enough (100 mS rise time) to see these 300 uS high current pulses (two every time): Is that just charging some capacitors somewhere, starting the LDC?, or are there two very short (300 microseconds), short circuits?
Last edited:
hobbit
Well-Known Member
I think the LDC output cable is a single wire, but with a fat plastic loom and tape around
it for some distance. But I haven't sliced it open. Another place to pick up the current
might be where it comes under and up into the fuse block, next to the battery wire.
The LDC lug should have a splash of red paint on it; here's a visual reference.
It's pretty tight in there, though; a large clamp probably won't fit.
Today I pulled the little connector out of the current-sensor, fired up the car, and went
out on a small errand. The voltage quickly rose up to 14.7 and basically just stayed
there. LDC output was the normal 9 or 10 amps, and interestingly, the "battery current"
which I now couldn't get data for was locked at reading +32A (discharge). Ambient
temp was about 45F.
I left it that way for most of the trip, thinking it might eventually sink down. It
didn't. Finally I stopped and with everything still powered up, reattached the
sensor. Voltage dipped briefly to 13.6, and then rose slowly back to 14.6 or
14.7. Battery current was basically nil, so clearly it had been maintained in
the LDC "fallback" mode.
I'll repeat the experiment in warmer temps, but I'm already thinking just leave
it all hooked up, let it happily do its complex little management game, and see
how long the battery lasts. Maybe its assumption about AGM and ambient
temps isn't so wacked after all.
_H*
it for some distance. But I haven't sliced it open. Another place to pick up the current
might be where it comes under and up into the fuse block, next to the battery wire.
The LDC lug should have a splash of red paint on it; here's a visual reference.
It's pretty tight in there, though; a large clamp probably won't fit.
Today I pulled the little connector out of the current-sensor, fired up the car, and went
out on a small errand. The voltage quickly rose up to 14.7 and basically just stayed
there. LDC output was the normal 9 or 10 amps, and interestingly, the "battery current"
which I now couldn't get data for was locked at reading +32A (discharge). Ambient
temp was about 45F.
I left it that way for most of the trip, thinking it might eventually sink down. It
didn't. Finally I stopped and with everything still powered up, reattached the
sensor. Voltage dipped briefly to 13.6, and then rose slowly back to 14.6 or
14.7. Battery current was basically nil, so clearly it had been maintained in
the LDC "fallback" mode.
I'll repeat the experiment in warmer temps, but I'm already thinking just leave
it all hooked up, let it happily do its complex little management game, and see
how long the battery lasts. Maybe its assumption about AGM and ambient
temps isn't so wacked after all.
_H*
The Ioniq electric cabling looks a little different, the charge strategy is probably similar.
In Ioniq, the 12V battery cable (red line) (probably one wire) goes to the under the fuse box.
The LDC cable (yellow line) runs from all the way over on the left side of the compartment under the electronics boxes over to likely the same junction point under the fuse box.
The thin grey cable running over the battery is a two pair twisted shielded cable for the DCCT power and the 4_20 mA loop which runs out a convenient hole in the bottom of the car and then across the home to my lab. The blue donut battery upper left is the DC CT. (there is a connector on the CT, the grey cable comes out for driving)
In Ioniq, the 12V battery cable (red line) (probably one wire) goes to the under the fuse box.
The LDC cable (yellow line) runs from all the way over on the left side of the compartment under the electronics boxes over to likely the same junction point under the fuse box.
The thin grey cable running over the battery is a two pair twisted shielded cable for the DCCT power and the 4_20 mA loop which runs out a convenient hole in the bottom of the car and then across the home to my lab. The blue donut battery upper left is the DC CT. (there is a connector on the CT, the grey cable comes out for driving)
Last edited:
KiwiME
Well-Known Member
I'm unsure that the LDC would wake up just to support a request from BlueLink. Checking LDC output would clearly show if it does, as possibly would the accompanying battery terminal voltage. I think though the takeaway is that the current draw due to a BlueLink request is short, under 1 second at about 22 amps, 79 mWh if my math is correct. Could the 300uS pulses coincide with receiving the request?
I would imagine that's all it can do without knowledge of aux battery current and terminal voltage. It seems the LDC has just two "on" modes, charging at that fixed voltage, or dynamically adjusting system voltage to ensure that the battery sees no net discharge current, effectively a very slight charge current.
One of the interesting comments from John Kelly in his Prius LDC-equivalent video is that if driving mode (Run) proceeds for more than 20 hours the aux battery is put through another charge cycle. Just an operational corner case but it would cover the situation of the car left accidentally in that mode, or perhaps being used for the Cannonball Run.
Just as you expected, it came back to full life on its own - after jumping it and leaving the car charging on an L2 for about 4 hours. Dealer Tech said another "ghost" and I will have to monitor for any more issues.
Sitting at 12.65 V resting and LDC at 14.8 V again when in "run" mode as usual
Car has been sitting and not driven for a week now.
I attribute the original problem to a "hung" Bluelink command as this was the only thing I did differently than the norm a couple of days previous to the occurrence.
I now have a set of jumper cables in the back as a just in case measure and will avoid Bluelink to see if that makes any difference.