HEV 12V Battery investigations

[PhilHornby]

I connected a few instruments, turned the car on (without pressing the brake) and applied as big an electrical load as I could

The attached graph shows the resulting battery voltage. The voltages I logged are in keeping with a 4-cell Lithium Iron Phosphate battery, charged to about 80%, if this web site is correct.

26 minutes in, the car turned itself off and (surprisingly?) the voltage at the fuse box terminal disappeared as well. I had to press the "12V battery Reset Switch" to start the car. Unsurprisingly, the Climate Control settings etc were lost.

I need to repeat the test and then WAIT before pressing the Reset Switch...

Looking at the graph, the 12V battery is disconnected at approx. 11.2V. The load on the car is clearly varying slightly and is affecting the voltage, resulting in the earlier dip to 11.5V. I suspect that the 12V disconnect isn't necessarily a 'latching' condition - in other words, if the voltage hits the 11.2V threshold, it will trigger - but should the voltage rise (if the battery 'recovers'), the 12V battery may be reconnected . I panicked somewhat in this test and wanted to make sure I'd not immobilised the car

Should this theory prove correct, it would explain the loss of settings, without the Battery Reset Condition being present. The alternative (that modules are being turned off pre-emptively to save power) makes little sense (apart from the Bluelink) - since 'sleeping' Microcontrollers usually have tiny quiescent currents.

[old man]

To me and I suspect many others, this is gobbledygook, so can you explain your findings in laymens terms and what they mean in regard to why our HEV cars lose settings and sometimes the car needs to be unlocked manually and the reset button is required ?
For info, and I've no idea if it makes any difference to your research, the findings in the link you attached refer to LiFeP04 lithium batteries, which ours are not. Ours is a lithium polymer battery which requires different treatment as I found out in my quest buy a trickle charger. i.e. chargers made by NOCO must NOT be used on our batteries and Ctek chargers are designed for LiFeP04 batteries and must NOT be used on our batteries.

[terryp]

I thought I'd try Hyundai with this to see what response I'd get. Here's the reply

"Thank you for contacting the Hyundai Customer Contact Centre.
If your local retailer is unable to diagnose the fault they will need to take the vehicle back in and open a technical case to find a solution.
If there are any other problems feel free to contact us back."

Hardly inspiring and it's as if they'd never heard of it before.

[PhilHornby]

To me and I suspect many others, this is gobbledygook, so can you explain your findings in layman's terms and what they mean in regard to why our HEV cars lose settings and sometimes the car needs to be unlocked manually and the reset button is required ?

The graph is a representation of the 12V battery voltage vs time.

I annotated the points at which I switched on various items and the overall shape of the discharge curve is very much as expected. However, it is very jittery, which leads me to suspect that the vehicle load was varying. I hoped I might spot some bursts of pre-emptive charging, but I didn't, so that theory seems dead in the water. But, there is that sudden drop to 11.54V, from which the battery quickly recovered. I don't know what induced it, but what if it had made the voltage drop to 11.2V? The Battery Reset module would have spotted it and cut power to everything (losing all the settings).

So: What if the Battery Reset module automatically resumes operation if or when the voltage recovers⃰ The settings would still be lost, but normal operation of the door locks etc would then be restored. This might be the scenario we see, when seemingly just the settings are affected. (As an example, perhaps - as suggested in another thread - Bluelink drains the battery a lot trying to 'phone home' in a bad signal area. When the power is cut & restored, it might stop doing it for a while; though this is pure conjecture at this stage!)

By 'recovers⃰', I mean the phenomenon you see when your phone battery dies. If you wait a little while, you can normally switch it back on again, for at least a short period. I assumed the Battery Reset module works like the phone and cuts the power until there is a user intervention. But - just maybe - it resets itself, if the voltage rises a little, for whatever reason? I will try and test that scenario...

Ours is a lithium polymer battery which requires different treatment as I found out in my quest buy a trickle charger.

There is a marked lack of documentation about the detail of our vehicles and a lot of confusing/conflicting documentation about Lithium-ion batteries. There are many variants of Lithium battery, though the only pertinent difference for this investigation, is the voltage range at which they operate. Most of them are very similar, but Lithium Iron Phosphate (LiFePO4) has a significantly lower voltage than the other types. I was trying to estimate the state-of-charge of the battery, by guessing at its chemistry.

There is some useful info. at: https://batteryuniversity.com/article/b ... ithium-ion, which details the different types of battery, by chemistry. Lithium Polymer (LiPO), on the other hand, describes the physical construction of the cell. See: https://batteryuniversity.com/article/b ... ce-or-hype (I acknowledge there are other websites with conflicting information on this subject)

Li-polymer can be built on many systems, the likes of Li-cobalt, NMC, Li-phosphate and Li-manganese, and is not considered a unique battery chemistry

So our batteries may be both Lithium Polymer and Lithium Iron Phosphate

UPDATED 1/10/23

I notice that this web page says:

Hyundai Motor Group's latest hybrid vehicles use lithium iron phosphate batteries instead of conventional 12V lead-acid batteries for low-voltage batteries. A small and light lithium iron phosphate battery is integrated with a high-voltage battery and applied under the rear seat.

[PhilHornby]

I didn't annotate the graph as such, but the section to the right (from 00:26:02 onwards), shows what happens at the 12V terminal, once the Reset Button has been pressed and the car "started".

The fact that the voltage immediately climbs to the sort of values you see in a 'normal' car, leads me to believe that the 12V battery isn't normally used at all to supply the car - it couldn't possibly have charged that fast .

It also explains how 'jump starting' works, as in, how you can connect seemingly any old charger (or another car) without damaging the 12V Lithium battery (which requires a current-limited charging source). If that is the case, then you can never actually charge the 12V battery from an external source - only provide power to make the car "ready" and let it charge the battery.

I reckon the 12V battery is only used when the vehicle is "Off/Not ready" and is otherwise kept on-charge at that ~80% level when the car is "Ready to drive" - with the car "12V" being supplied instead (only) from the 270V battery via the "Low voltage DC converter" aka "LDC".

I found a little more technical information here: https://www.hyundai-kefico.com/en/busin ... content.do (you have to hunt around!). One of the snippets, is that the 270V Traction Battery is comprised of 72 cells. Does that equate to a LiFeP04 battery pack? (I'm not sure it does); you would think the 12V and 270V batteries would be built the same way...

[roadster]

Well done for finding out a bit more about how the 12 volt supply is integrated with the main batteries and how it is protected from excessive discharge. Not gobbledygook to me at least. I'm about to post a bit more about the PHEV lead-acid system based on similar but less exhaustive measurements.

[PhilHornby]

I need to repeat the test and then WAIT before pressing the Reset Switch...

I tried it again, but didn't manage to produce the scenario where the battery recovers enough to reset the Battery Saver

This time I tried the highest load I could think of - so seat heaters, steering wheel heater, air con on maximum, fog lamps etc etc. I figured this should result in the fastest possible descent to 11.2V and leave the most 'recoverable' energy in the battery. It didn't work, so I might need to think of another way

I waited an hour or so before hitting the Reset switch - I would have thought this was long enough. One thing I tried, was connecting an external 12V source. This was partly to see if I could simulate a 'recovery' and also to see what sort of power is required to do this. What I found was that my 3A Bench PSU was not man enough for the job - I just got a lot of clicking relays and about 5V across the terminals

12minutes in the life of a Lithium battery :-

[PhilHornby]

I was curious, as to how long the 12V battery would run the Parking Lights...

On my N Line S, which I assume is all LED, the answer is approximately 2½ hours.

The test results were weirdly different from the ones that I've done previously. The Battery Saver triggered at a voltage of 12.41V (much higher than the 11.2 seen in the earlier tests). Yet, this was far from the lowest voltage seen during the test. (There had been many sporadic excursions to lower values. There is clearly quite a lot of background activity, even when the car is supposedly switched OFF.)

The Battery Saver definitely triggered - I had to press the button to restore power.

[Deleted User 1130]

There is clearly quite a lot of background activity, even when the car is supposedly switched OFF.)

Interesting Phil ......

I've noticed that since the latest Bluelink update, I'm getting more notifications at decreased time intervals. Eg: left doors unlocked and received 3 notifications within a 10-15 minute period. I hadn't had that sort of response previously. I think it's been mentioned previously, but could this 'background activity' be related to Bluelink?

Would it be worthwhile doing some further tests with Bluelink on/off and/or simple Bluelink app connections to interrogate the car. Just a thought ....

[Philr]

It's interesting the large number of dips at irregular intervals and then a steady decline with no dips between 1:34 and 1:52.
I wonder if the dips are scheduled events maybe checks on any status changes e.g. has a door opened or whether events like that are treated as interrupts and not checked regularly only when they happen.
There may also be timers again for example if the radio hasn't been used for x minutes stop receiving until an event occurs like door opening or changing the radio channel, turning the car on etc etc
Scheduled events could then have a result so say after 3 checks if the door hasn't opened drop into sleep mode and don't check again until a door actually does open. This would go some way to explaining the irregularity of the dips.
There would almost certainly be a mixture of firmware embedded on chips and software that runs on them. Bear in mind that there will be multiple cpu's all doing their own thing - one or more say the lights, a different one or more for maybe wipers and so on and individually they would not necessarily all do their functions with the same time intervals.
I find all this stuff fascinating - I just wish we could get some real facts from Hyundai on some (any?) of this.