Repair issues and fixes

Yes I'm prepared to chance it, took the car on a run to the sons place 220km country miles each way, the GOM said 465 km range.
Ended up having to put in a charge on the way back, it would have only done 400km max, but the first and obviously last 120km were very hilly.
It is a great little car, but I will be annoyed if the WOF happens.

 

What Hyundai should be doing is extending the warranty on these gearbox/motor issues. Other manufacturers have done this sometimes just to preserve the confidence of their customers.

I had a 2018 Crosstrek, and in the past Subaru had some issues with their CVTs. While there were no issues with the newer cars, they increased their CVT warranty to 10 years, not only for the older ones, but also the newer ones which didn't have any problems. And when a valve spring issue surfaced (actually pretty rare) with some older engines already out of warranty they covered all those as well.

And they also had "hidden" warranties. My HU on the Crosstrek wouldn't properly accept a new phone that I had just bought, and when I brought it to the dealer, they informed me that it needed to be replaced. I was out of warranty by then and feared a big cost, as I know those things aren't cheap. But to my surprise, the dealer explained that this was a known issue, and so even out of warranty they replaced the HU at no cost to me.

Sure would be nice if Hyundai treated their owners and customer that way. Instead they seem to be avoiding and even trying to run out the warranty before fixing some problems.

My advice on this tick tick issue, is document it well with your reporting it to your dealer before the warranty runs out. Then you may have a chance of it being covered after the warranty period.

I sure hope with my Ioniq 6, Hyundai has learned enough with their earlier BEVs, so there will be no long term design type issues that will bite me later.

 

I flog my MINI Cooper SE frequently and nothing has broken yet. The 3-year warranty has expired, but I figure I can keep flogging it--if I have to eventually pay the piper for my folly, I'll still feel I got my money's worth. Babying a car while the warranty's still in effect doesn't make sense to me. Similarly, I don't baby my battery--I charge that 32.6-kWh power-pack fully every time I plug in.

 

The ticking started at around 90K for me. They replaced the motor mount brackets but that only resolved it for a few weeks. Next the reduction gear will be replaced, once the parts arrive, all under warranty.

The reduction gear has a longer warranty as it's part of the EV System. For my 2019 in Canada the warranty is:

EV System


HYUNDAI AUTO CANADA CORP. EV System Warranty covers the described components up to a total of 96 months from the warranty registration date, or 160,000 KM, whichever occurs first.

WHAT IS COVERED
The following original EV system components which are found to be defective in material and/or workmanship under normal use and maintenance will be covered during the "Warranty Period" described above:
Motor, Inverter unit, VCM, Reduction gear, DC/DC converter, Onboard charger, Onboard charger connector, Trickle charge cable, In Cable Control Box, High Voltage Battery.


Coverage
96 Months / 160,000 KM

 

At least one WoF report a week appearing now on FaceBook, all on 2019 models in these four examples. The second two are here in NZ and out of warranty, owners are hoping for goodwill repairs.

Kona EVs were a popular fleet car in 2019 and those are now re-entering the market. There are 10 up for sale at the moment, priced NZ$30k - 47k with 26,000 to 123,000 km. These are attractive prices for a long range EV and somewhat less than the new Chinese-designed offerings from GWM and BYD.

What's also attractive is that these have a 10-year unlimited kms battery warranty. I suspect buyers in this price range are ex-Leaf owners and focus mostly on the battery.








(... no)

 

KiwiMe,
You've spent a great deal of time on this issue over the years. Aside from the more deeply technical stuff (which I'll admit are a bit beyond me occasionally) I believe I've read most of your posts. As a 2019 owner who's had a full drive train replacement I wonder if you've ever rendered an opinion about whether the replacement parts should be expected to solve the issue long-term...now that we're actually getting real-world examples of "long term"? I know you're a believer in changing the oil and controlling stray metal bits as a prophylactic measure but do you feel there is any reason to believe that the replacement parts sufficiently buffer against a repeat failure?

 

That is a very good question. Up to now I've generally ignored model year because it adds complication. But last week it was suggested to me by a prospective buyer in the US that his search indicated only one noise report on a 2023 model.

But we're also seeing lately that the problem can appear at higher kms. Most of the early models (as noted above) are moving past 50,000 km or even 100,000 km. I had thought that the WoF would happen either early on or never, but clearly I'm wrong. Another complicating factor is that some of these noise incidents could be due to the motor 'tail bearing' failure, a potential problem that sounds the same but I believe has no connection to the reducer failure. We can't do anything to avoid that risk unfortunately and a new motor is far more expensive than the GRU.

If the GRU-related WoF noise is caused by the contaminated oil (as I believe it is) then it would only take a single oil change on any example (such as your own) to determine if that gear reducer had been appropriately revised. The oil should come out an opaque brown or gold colour after covering a considerable distance, say at least 15,000 km.

If the GRU-related WoF problem is due to something else and information to suggest that it's fixed does not surface, then it will take years for the 2022-2023 relevant statistics to shake out such that we can reach a conclusion, just as what's happening with the 2019 models now.

The last comment I'll make is regarding Hyundai/Kia's powertrain engineering, the branch known as Mobis. The design choices that triggered the dirty oil problem and motor bearing failures have been essentially eliminated in the 2023+ Niro and presumably the 2024+ Kona. The Niro has been out perhaps a year and I've not heard even one report of a drivetrain issue. I fully expect the new Kona will be as equally reliable and as-such I believe that Mobis must have figured what went wrong. Frankly, this is just about good engineering design practices rather than what's required to land on the moon.

Just for the record, here are the technical issues I've seen:

Powertrain design defects found on 2019- 2023(TBD)
a) particle magnet not secured rigidly leading to nearly non-existent sequestration performance.
b) potentially-risky bearing geometric-overconstraint exists between motor and GRU ...
c) ... along with motor-end GRU input bearing outer raceway left as a slip-fit, which in combination can allow intermittent outer race spin due to lack of consistent radial loading.
d) motor tail bearing rusts and loses grease, possibly due to moisture ingress or insufficient axial thrust spring preload leading to outer race spin.

What I can see appears corrected on the 2023+ Niro and expectedly on the 2024+ Kona
a) 3.5 times more oil in circulation plus addition of screen-type oil filter with pump to catch large particles. The same magnet appears to be present.
b) one bearing deleted near motor and GRU input spline to eliminate the over-constraint situation. Nearly every other EV already has this, the Leaf excepted.
c) a 'wet' motor design which will preclude rusting and keep the motor bearings well-lubricated. Toyota did this 20 years ago in the Prius (to reduce frictional losses) and it's a common feature in EVs now. Tesla moved to this starting with the Model 3 after coolant leakage issues on the Model S.

 

As I posted a page or two back, I had the GRU replaced under warranty due to WoF noise.

Not lucky, I'm afraid
Occasionally I can hear a slight moaning noise at speed which was present before the GRU was replaced. It was not there when the car was new and is too faint and hard to identify to make it worth reporting to the dealer. They would just test drive and report "Nothing found", I'm sure.

I changed the oil at 1000km. Black as the ace of spades with the same colloidal metal in suspension as per my post#225 on page 12. It's on its way to failure again, just a matter of time.

I don't think I want to keep the car any more. Will see what the dealer says, and may have to wait for more failures/attempts at rectification before commencing consumer action. Or sell it, with remaining warranty, suck up the $17000 loss it will cost me over my 2.5 years of ownership and get a 2023+ Niro. My son can lease one of those for me at very affordable fortnightly rates. Or I might get a mid-range EV6 for a few more $$.

With this drive train, plus the towbar issue, this car is not the right one for me.

The pic shows the colloidal metal in the oil:

 

We don't have too many oil change reports at very low kms and your photo shows exactly what I feared, that the internal magnet isn't doing what it was intended to do, nevermind that Hyundai still don't realize this after 5 years. It's incomprehensible to me that ball bearings could tolerate that level of contamination without incurring life-threatening damage.

For the uninitiated reading this, if left alone those particles would disappear by around 20,000 km because the bearings and gears will chew them up into a black dust. But the bearing damage would have already been done and their lifespan reduced. If the oil was changed several more times at similar intervals, some or much of that damage could be avoided.

The other noise could very well be the motor tail bearing going south. I read that Auto Repair Videos reported that the hollow motor shaft rusts inside and those particles exiting the tail end had gotten inside the bearing. Noise reports of late from owners lead me to believe that the motor noise can be present at higher speeds and not only during regen.

I spend much of the day exploring the motor bearing application and sure enough it has a glaring deficiency, see below.

It's totally understandable to move on.

 

Today I took two oil samples to my service rep. One was taken at 5000km before the GRU was replaced. The other was from my oil change posted above, 1000km after GRU replacement.
He started off a bit hostile - "It's a sealed unit, you're not supposed to touch that, you'll void your warranty", etc
We batted back and forth a bit and then he asked what prompted me to do an oil change at 1000km. I explained that given the GRU failure and the condition of the oil 5000km before replacement, I thought a run-in protocol was a reasonable precaution. We discussed the condition of the oil, including the colloidal contamination. He had to tow the company line and tell me I shouldn't be in there, and I shouldn't be using magnetic plugs as they are not Hyundai parts, etc, etc, but I could see his attitude softening as the evidence of something wrong in the GRU is irrefutable, and I said that.

I asked him "Where to from here?". He said he cannot approach Hyundai unless the vehicle physically exhibits a problem, such as an abnormal noise. Hyundai won't do anything based on an oil sample and if presented with one may deny a future warranty claim because of unauthorised mechanical work. To his credit, I suppose, the advisor said he hadn't seen the samples and we hadn't had this discussion. So I guess the warranty is intact, the problem will rear its ugly head in about 2 years (actually mine was evident at about 6 months/5000km from new, so it may be less than two years) and it will be fixed under warranty. But it won't be under my ownership.

 

I've written a short description with photos of how the gear oil condition deteriorates over kms driven. I'm focusing more on the fact that the manufacturer has failed to sequester particles adequately to ensure a reasonable product lifetime rather than be distracted by the exact mechanism leading to this.

 

Just completed a gear oil change on our 2019 Kona EV at 89446 km , 20235 km on the oil since last change, magnets added to drain plug only at last change. Results here for KiwiME to add to the high end of the graph. Lower levels this time round but last sample had three times the km so not unexpected. Thanks for all the info sharing!

 

Thanks for that. Results look similar to other 2nd changes, low iron presumably because of the magnets and highish aluminium possibly due to unabated internal magnet rubbing the interior of its pocket. Both numbers of no concern though.

Full hi-res image here.

 

Can you put the link of Amazon for magnets?

 

Regarding the GRU oil contamination problem, I've been wanting to 'prove' that the factory internal magnet moves while driving due to oil turbulence. I had an idea for a sensor a few months ago but had delayed putting that into action due to the effort required. But recently I thought I would do the software component of that as an exercise, only because I needed a simple display application to help develop my hobbyist-level Arduino programing skills.

That turned out to be easier than expected and since I already had the remaining hardware, I really couldn't find an excuse to avoid finishing the job.

Apparatus
The basics are that I have two linear Hall sensors that provide X and Y data for a graph. Both sensors have a steel screw placed behind them to focus the field lines though the sensor, which significantly increases the sensitivity. They are sensitive enough to detect the change in a magnet's field when it translates or tilts, including through the aluminium housing. However they can't detect a pure rotation (spinning) component because that doesn't cause the axisymmetric field to change on a round or toroidal magnet.

The Hall sensors respond to static north and south magnetic strength (gauss) as deviation (positive or negative, respectively) either side of a nominal 2.5V. Passed on to (2) analog inputs on an ESP32S3 microcontroller development board, those analog signals are read every 25ms, autoscaled and mapped linearly to the display pixels as a large dot. There is a tracer line that tracks the last 1.25 seconds of data to help understand the recent X-Y pattern. Note that the pattern on the graph doesn't necessarily translate to the magnet's actual spatial movement, it's just a visual indication that there is movement.

The first video shows what the GRU magnet looks like when influenced by an external magnet with no oil present. It has limited room to move in all directions.

The second video shows testing the apparatus using a drill press to simulate a wobbling magnet. It worked exactly as I had hoped, automatically scaling the signal's range to fit within the graph area. The numbers on the graph are meaningless, that was just part of the programming lesson.

Test Procedure
The board holding the sensors was attached to the GRU underside with Blue Tack. While installing the sensors I was able to verify that the magnetic field was present and readily detected by both sensors. I'll note that the autoscaling is one-way, it will fit the largest signal found during the session and not return to a finer scale unless a button is pushed to reset it.

The third video shows the first road test in the Kona with the sensors oriented to detect magnet motion side to side in the retention pocket.

I did a second run with the sensors placed at 90°, to detect the magnet entering and exiting the pocket. The outcome was much the same as the first, so I didn't record the second run.

Results
The result in both runs is that there was no evidence that the internal magnet moves while driving. All motion on the display could be attributed to electrical noise or extremely small movements. If there were larger movements the scaling would zoom out automatically to accommodate them, making subsequent steps smaller in movement. The large jumps indicate that the signal changes are very small.
There only remains the possibility that the magnet is spinning but I doubt that it could do that so perfectly that there were no associated and detectable translation or tilting artifacts.

Conclusion
We now have far less plausibility to the theory that the magnet catches, grinds down and releases wear particles once they lose their magnetic attraction. And we don't have a good reason why the internal factory magnet can stay relatively clean over tens of thousands of kilometers or miles. It should catch most of the particles soon after they are generated and hold onto them for the life of the car, looking like a steel hedgehog.

This doesn't change the reasons to change the oil very early (500 km/300 miles and 1,500 km/900 miles) and at scheduled intervals far less that the factory recommendations. We know that the GRU suffers damage due to inadequate particle retention but we don't yet understand the role the magnet plays in that.

 

There is a problem with Kona, it stops after this error P1B2500 High Voltage Path fault, P1BBA00 BMS Owercharge Protection Device Active.
After restarting it goes, but not for log time. How to solve this problem?

 

Not much choice here. If your car’s HV battery (including all high voltage components) are under warranty take it to the dealer. If not, still take it to the dealer. I would not fiddle around high voltage components.

 

I'm getting random, brief Turtle Mode warnings at 100kph in smart cruise control. No loss of speed but accelerator doesn't respond while it's activated.