Real Or Imagined: Is The Tesla Model S Drivetrain Defective?

2 years ago by Jay Cole 90

Tesla Model S - Driven Like No Other Car?

Tesla Model S – Driven Like No Other Car?

Recently Edmunds sold their long-term Tesla Model S; but before they did so the car had four (yes, four) drivetrains replaced by Tesla Motors.  Motor Trend’s long term car is working on its 2nd.

This does not make us ask the question “is drivetrain failure a common occurence in the Model S?”

Because it  is.

Tesla Model S Drivetrain

Tesla Model S Drivetrain

If you check out the latest numbers on a Tesla Motors Club poll, you will find that 75 (supposed) Model S owners have had a least 1 drive unit replaced – 12 of which have had it replaced more than once.

(Fair disclaimer: There is always a couple jokers messing with the results of an uncontrolled poll like this. As a point of reference, TMC likely has somewhere around 5% of Model S owners as members)

To their credit, Tesla service has been doing a yeoman’s job taking care of its clients.  Despite the many failure occurrences we have yet to hear of Tesla giving a Model S owner a difficult time in solving the problem (ie-replacing the drive unit entirely).

The problem for customers will arise however when Tesla’s 4 year/50,000 mile warranty on the unit expires; that is when repairs reportedly greater than $10,000 a pop really start to hurt.

To ease the burden of that oncoming deadline, many owners have opted for the extended warranty – a great luxury to have; that is if you own your car in a location where the car is on sale and the warranty is available.

But is the Drivetrain defective?

We often get emails asking us to “look into the situation”, or to try and get a firm price from Tesla on what a particular service operation will cost outside of the warranty period.

Even more common, “Is Tesla working on a fix?”

To date the main inference has been that the unit is defective.   And why not?  The failure rate is already well above the norm for a normal car.  Therein lies the problem…it may not be defective, it may just be well used.

Is Everyday A "Track Day" For Many Tesla Model S Sedans?

Are Many Days A “Track Day” For Tesla Model S Sedans?

Now before anyone gets too excited and tells a story of an owner somewhere that had a failure through no fault of his/her own after a 1,000 miles … yes that drive unit could be defective, we are not advocating that all failures are the owner’s fault.  Sometimes things just break.  Heck, they could all be defective; but we seriously doubt it.

Electric cars are known for their instantaneous torque and silent performance.  They are known for making anyone with a driver’s license into “Johnny Professional” when it comes peeling away off the line, or passing other cars on the road.  And as EV drivers, we like to feel that silent performance  underfoot … a lot.  You find me a Tesla Model S that he/she doesn’t, and I will find you a liar.  It could be that desire of the common man that is ultimately causing the bulk of drive unit failures.

Drawing from a personal example, I have tested many a car in my day; but the closest thing I ever got to owning a true petrol sports car of my own was a BMW Z4 M back around 2008.  I can recall truly pinning that car on just two occasions to see “what she could do”. 

It was great fun!

Naturally, I first found myself a nice quiet, out of the way piece of open road where few people lived, and then let it rip!  There was spinning tires, a growling engine – the whole nine yards of unbridled driving..  And then I was done; as this was my car I didn’t fancy blowing her up and doing serious damage.

The Tesla Model S Has Performance Specs Roughly Akin To An Audi R8

The Tesla Model S Has Performance Specs Roughly Akin To An Audi R8

Now, when I bought my first electric car a few years ago, you know what I did the moment I was out of eye-shot from the dealership?  The exact same thing.  I put my foot to the floor and experienced the quiet awesomeness of all-electric acceleration.

Only this time, there wasn’t the spectacle of an internal combustion engine making all persons within a half-mile wonder “what adolescent borrowed daddy’s car and is now making a scene?”   I knew I didn’t need any special occasion or location to do so.

And what of that little voice inside me that once told me to “not blow-up” my BMW z4?  She was no where to be found – because the markers I had long associated with a car being pushed to its limits weren’t registering in my mind.

I probably pushed my new EV to the limit a half dozen times before it hit the family driveway for the first time.

And the first thing I did when I got home?  Took the wife for a spin to show here what it could do.  Then my friends.  Then later, complete strangers; something I never would have done in my Z4.  In total, my foot has hit the floor hundreds of times, maybe thousands of times in my plug-in vehicles.  I love it…it’s a cool sensation.  It is part of what owning an EV is all about.  That and hypermiling as far as one can past the stated range – it’s a dichotomy to be sure!

The Tesla Model S has roughly the same performance of a new Audi R8 V10 Coupe – a classic “supercar” –  400lb-ft+ of torque, and a 0-60mph time slip of about 4 seconds…except of course that the Model S can seat up to 7 and weighs a 1,000 lbs+ more.

Now if I were to say, “I just don’t know why my Audi blew up, I only drove the hell out of it a couple hundred times,” I think people would be less inclined to call any shop time for that car a result of a defect, or to say that Audi had a situation on their hands; they would call it an expected outcome.

Many times you hear statements from those owners affected by a Tesla drive unit failure like “I really didn’t push it that hard at all and it happened to me too!

Although in most cases their statement is quite likely misguided, I do believe they are being honest…because they aren’t hearing the screaming of a 10 cylinder engine, or getting the condescending looks that would normally be associated with what they are asking of their Tesla Model S sedan.

For some others, especially those more familiar with the rigors of maintaining a ultra-high performance automobile, I think they can appreciate the true sturdiness of the car as compared to its peers, and accept the costly consequences over time for the thrill of the drive.

Or then again, maybe it’s just defective.  Maybe there is a way to have a supercar have the longevity and reliability of a Honda Civic.

Tags: , , , ,

90 responses to "Real Or Imagined: Is The Tesla Model S Drivetrain Defective?"

  1. kdawg says:

    How is the drive train failing? Is it mechanical or electrical?

    1. Jay Cole says:

      It’s mechanical (for the most part), generally starts as a droning sound at speed then louder engine whine, degrades from there. Can also manifest itself as a “pull over safely” motor power loss until restart, etc.

      1. kdawg says:

        Sounds like more torque than it can handle. They’ll have to beef it up or reduce the torque (not so good for the 0-60 #’s though).

      2. JakeY says:

        I don’t know how you say that with such confidence. If you look at Roadster replacements from the Plug-in America survey, the inverters/electronics (PEM) failed multiple times more than the motors. In the Model S, the inverters are replaced as a single unit with the motor (one of the two cylinders is the motor, the other is the PEM).

        It’s curious that the PEM+motor failure rate was surveyed as higher for the Roadster than the Model S and there was never any reports claiming the Roadster drive-train was “defective”. AFAIK this topic came up for the Model S mainly with pushing by some of the TSLA shorters.

        This is discussed in this thread:

        And if you read the thread, a lot of the replacements are for noise that the technicians detected. There are also members who posted in the poll thread that they voted just to mess with the numbers.

        For what it’s worth Consumer Reports gave the Model S a top pick status for 2014 and they don’t do that unless it was a car that got better than average reliability. According to one of the members with a subscription, the 2012 and 2013 Model S got a “best” rating for drivetrain reliability. So you have to be very careful about drawing conclusions (esp. from online poll numbers that are not vetted and even have people that said they purposefully voted to mess with the results).

        1. JakeY says:

          Sorry, some terminology correction for my post. When I say “failure rate” I really should be saying “replacement rate”. Given some were replaced for noise that only the technician detected (and not even owners themselves) it’s not clear replacement = failure.

        2. SeattleTeslaGuy says:

          I think this is an important point. A lot of the replacements weren’t for outright failures (ie people didn’t get stranded). Replacement rate does not equal failure rate.

          1. pjwood says:

            Exactly. A “whine” @60mph is the symptom, not “dead on road”.

            1. Mint says:

              What’s more is that Tesla is very interested in getting many samples as they can of these working but possibly wearing motors.

              They’re about to produce a model in mass numbers for the first time. Any defect there will be brutal.

    2. Mike I says:

      There are also electrical failures. The inverter is part of the same assembly and if there is any problem in the whole drive unit, they just pull it and put in another. Eventually, when a significant number are out of warranty, Tesla will have to sell refurbished drive units with a reasonable core charge. If they don’t, a third party definitely will.

  2. Anon says:

    High torque eats away at any point of friction, even in an electric drivetrain. Improved bearings might help… But wear and entropy are as certain as death and taxes.

  3. Josh says:

    Amen, Jay.

    I used to AutoX my S2K, and I won’t mention what I did with it before I got that healthy habit. But the cold air intake, and VTEC, would clearly remind we what I was doing to it at 9,000 RPM (bouncing of the rev limiter). I got stopped in that car once because a cop told me I must have been speeding based on the sound he heard!

    I traded the S2K for my LEAF, and enjoyed dusting German nameplates 0-30 regularly, minus any official attention. Every time I gave a test drive or let someone test drive, pedal to the floor was a must to help explain why the electric motor is so awesome. The explanation usually followed with, this is 80 kW and the Model S is 310 kW.

    Many people that never had a track car, or never pushed an ICE to (or past) the limit, may never have dealt with real engine trouble. Engine overhauls on high end sports cars is a regular occurrence. I still believe Tesla is aiming for a higher standard. The fact that Tesla can do an entire drivetrain swap in one day is a three week improvement on the German comparisons.

    More direct to the topic. The first grumblings I hear of the “milling sound” were late 2012 or early 2013, and only on the Performance models. In fact my first test drive (Oct 2012) was a P85 that had the humming, or “milling sound”. The most reasonable explanation I read for it was vibration in the motor coil windings. The sound I heard in the test drive scaled with torque (~current), which would confirm that diagnosis. The noise clearly didn’t inhibit the Model S from hitting its acclaimed acceleration.

    Tesla is replacing these drivetrains based on the noise, not based on a failure in the motor. I have not read any evidence to show that the sound is evidence of an imminent failure. I don’t know of any other OEM that gives you an entire new drivetrain because there is a sound you don’t like. Tesla may be inflating the “failure” numbers by replacing drivetrains to keep the car quiet and the driver happy.

    Once the Model S is out of warranty, there will be many more Model Ss “milling” around. Overall, I don’t think we know the full story on this yet, but I have no doubt that Tesla has a full team based on fixing it. They build their own motors and have proven to address all known issues in the past. I can’t imagine them moving to dual motor AWD (confirmed for Model X) without having a handle on the issue.

    All of that said, if I owned a Model S and was headed back for my second drivetrain swap, I would be concerned and hoping I had an extended warranty.

    1. Jason says:

      It’s not just a noise….It’s like driving down the street with a Vita-Mix on high. You can hear it over the radio, and everyone outside can hear it too. With the windows down at 20mph, you can hear the noise reverberating off of the houses you drive by. I’m on my second drivetrain. A 3rd may be on the way.

  4. Josh says:

    I just checked that poll, unfortunately they didn’t have an option for “0 replacements”, so we really can’t get a handle on what the rate is. It is showing 76 total replacements right now. Tesla has delivered 25k+ Model S and many with issues go to TMC for answers, so I wouldn’t say it is an issue that is hitting every Model S.

    I look forward to hearing something more official from Tesla.

  5. See Through says:

    I don’t quite believe in the internet polls or even other polls, as these aren’t always unbiased, and can be grossly wrong. But both Edmunds’ and Motortrend’s Model S have gone through drive unit replacements. Edmunds had 3 replacements in 30K miles. Motortrend probably got one replacement, but their road test doesn’t seem very thorough.
    This tells me, the problem is quite common. Motortrend did mentioned that the service center seemed to take drivetrain replacment as an everyday event.

    Read Motor trend’s update 3 here:
    “Oh it got a new motor, a new motor..” – As if it’s a good thing.

  6. 76 instances from ~30,000 Model S’s now in the hands of owners in the U.S. This appears to be a 0.25% failure rate … no idea how this compares to transmittion, or differential failures in similar 400+ HP sports cars?

    Unknown is how many drivers the Edmunds Model S had and how much time it spent on the track, or taking friends and family for a ‘ride’?

    Something that would be interesting to know is when the ~76 drivetrains were manufactured, but expect forum details will be sketchy. It may be a case they were from earlier production, or had gears, bearings, or other components from a common supplier.

    We just don’t know … so pretty much speculation unless more specific details show a common pattern.

    BTW: Nice writeup.

    1. Lensman says:

      According to reports on the Tesla Motors Club forum, there is no consistency regarding when the “milling noise” first appears. New cars, older cars, cars bought early in the production run, cars bought more recently. Seems to be surprisingly random.

      But no indication it’s a -frequent- problem, despite what some what us to believe.

      1. pjwood says:

        The probability of “tries” suggests to me that it happens more when cars are placed in the hands that “explore” their performance, like car mags.

        Nonetheless, someday: A car guy, like me, shows up at a service center with two barrels, and says, “here’s your core, can I get the thousand dollar rebuild”? Will I get, “No, no, you have to bring your car in, for the $5,000 job.”

        It’s encouraging the roadster owners have conditioned Tesla to a savvy bunch of enthusiasts, but I understand it is also within their right, as “non-dealers”, to play hardball. I know most will shell out for warranty, but what will access be like on the other side for others with the Model S?

        1. Lensman says:

          pjwood asked:

          “It’s encouraging the roadster owners have conditioned Tesla to a savvy bunch of enthusiasts, but I understand it is also within their right, as “non-dealers”, to play hardball. I know most will shell out for warranty, but what will access be like on the other side for others with the Model S?”

          Two points, one of which has been made upstream already:

          1. This isn’t a “failure” in the drivetrain. It’s a noise. If someone wants to pay for out-of-warranty replacement of a “drive unit” for mere noise instead of failure, a noise that in many cases the customer didn’t even notice, that’s their business.

          2. This isn’t a case of “We have to replace the drive unit, so we’re going to charge you for a new drive unit.” This is a case of “Well, we could wait for it to be shipped back to the factory, disassembled and refurbished, with only the parts that need it being replaced, and returned here, but you’d have to wait three or four weeks for that. Instead, we’re going to replace the unit with a a refurbished unit that has no more wear than yours.”

          In other words, they’re not replacing the entire “drive unit”– that’s motor, gearbox, and inverter– because all three need replacement. They’re replacing them because they can be slipped out and replaced as a unit in just one day’s work at the service center.

          So again, claiming that a service center doing an out-of-warranty replacement will charge the customer the full rake for a new set of all three modules is downright silly. At worst, the customer could rent a car for that 3-4 weeks, and just wait for the original unit to be refurbished and returned. (On another forum, someone cited a 3-4 week waiting period for a drive train rebuild on some German high-performance gas guzzler. I’m guessing it would take no longer for the Tesla Motors factory to do that. Perhaps it would be somewhat faster, given the much simpler nature of the electric drive.)

          1. TomArt says:

            Good points.

    2. Taser54 says:

      76 instances in a forum that represents 5% of Tesla Model S owners. At this point, extrapolating to the entire Model S ownership base would suggest a huge problem.

      Again take the poll with a grain of salt, but Tesla will not publish numbers so, it’s all we’ve got for now.

  7. Robster says:

    slightly off topic, but funny you took a picture of a Dutch Model S as an American website. still wondering where the picture was taken

    1. Jay Cole says:

      Hey Robster, good eye. We like to mix it up with unique/scenic Model S pictures whenever we can. That shot was actually taken in the UK…just as FYI.

  8. Jouni Valkonen says:

    As it was pointed in in Tesla Motors Forum, it seems that Tesla’s drivetrain failure rates are getting lower when sample sizes are becoming statistically significant.

  9. DaveMart says:

    The rationale given in the article makes no sense to me.

    If the problem is caused by flooring it, why are there no reports of similar problems in the Leaf?

    Sure, it is not as fast as the Tesla, but perhaps that is all the more reason to floor it a lot.

    If it were a cause of repeated failure, then the design is flawed just as much as if the cause were anything else, as presumably the performance could and should be limited a little if it can’t handle full power repeatedly.

    I don’t know if this is a major problem with a lot of cars or not, but most cars develop more faults as they age, and this seems to be occurring in more or less new cars.

    Hopefully it won’t, but this could be the start of a spectacular failure rate as the fleet of Model S cars age.

    Car engineering is tough, and Tesla can have problems like anyone else.

    Lets hope they can fix it.

    1. Lensman says:

      The Leaf doesn’t develop anywhere near the torque of the Model S. The Model S is much heavier and it has a much faster 0-to-60 time.

      If the problem is in the gearbox– and one report says a service person said the “milling noise” was coming from the gearbox– then it’s likely a result of the very powerful torque the motor puts out.

      Tesla Motors is also having service personnel replace the motor mounts, which might suggest (guessing here) that there might be a problem with vibration in the motor, which may be contributing to developing noise in the gearbox.

      Again, I don’t think we have any indication the “milling noise” is generally indicative of an imminent failure. It may be just a noise which is noticeable because the Tesla Model S is normally so quiet.

      1. DaveMart says:

        I’m aware that the Tesla has much greater torque.
        That is why all the components have, or should have, much higher specifications to suit.
        That does not alter the fact that the performance envelope has to be designed to cope with the power available.

        I gave no opinion on whether this is a common or serious issue, we will find that out in due course.

        My comment referred to the exculpation offered in the article, which makes no sense to me if it does prove to be a major or common problem then it is an engineering mistake if caused by the car being pushed to its allowed limits repeatedly as if from any other cause.

        1. Stimpacker says:

          Fully agree with DaveMart and fully disagree with Jay. I am a Leaf owner and I drive it heavily sometimes. Plenty of other Leaf owners I know do the same. Oh, come on you say. It has less torque and is nowhere near the supercar the Tesla Model S is. Well, if the Leaf is designed to handle what is engine/motor can do, then so should the Model S, right? Basic engineering…..

          My next BEV WAS going to be the Tesla Model X (for me) and GenIII (for wife) but this sort of reliability issues does not bode well. I am re-considering and might wait a couple more years instead to see if reliability improves.

          1. DaveMart says:

            Just as there was not a problem with electric cars in the heat per se, but a problem with Leaf engineering there is not a problem with electric cars as such, even powerful ones, but questions about the specific engineering that Tesla have done.
            Even should the back or my sofa yield unexpected riches I would certainly not rush out to buy a Tesla S at the moment, until there is a better handle of this.
            The overall rate of failure may be low, so far, although we don’t really know that from the haphazard data to date, but the same car going back several times? Come on!

            1. DaveMart says:

              I pressed the wrong key.
              I meant to add that repeated failure in the same car means that either there is an underlying cause which the engineers don’t understand, or that the rate of failure is REALLY high.
              We have seen the same thing umpteen times in components from different manufacturers.
              Dismissing them never worked, and normally a complete re-jig of some kind is needed.

          2. James says:

            There must be point where premature ware on the gears is unavoidable due the limitations of the materials available. Even if the LEAF reduction gear is half as beefy as the Model S’s, it develops 1/4 the power and weighs quite a bit less. Personally I think that if you push the limits of acceleration many times on every journey you do in ANY car, you are going to have problems alot sooner than otherwise. That must be particularly true for a car that is at the limits of performance anyway, like a Model S.

            Also, just because the drive train responds immediately, doesn’t mean that you should shock (sudden changes of acceleration and decceleration) it every time you want to do anything. I would have thought that would put a lot of additional stress on the gear as well, going from lots of tension in one direction to suddenly having lots of tension in the other direction and then back again.

            Actually I’m now starting to be quite impressed that there isn’t a much greater replacement rate!

            1. ELROY says:

              My 2012 LEAF developes over 200lb of torque, and I floor and abuse it all the time with no reliability issues. The Model S is only a little over twice the torque, but the engineering should have made allowances for it. The modded 2014 BMW M5 I was driving made over 670 rwhp, and 650lbs of torque with many many drag strip type runs. Differential, driveshaft, trans, engine, etc…withstand this upgraded HP. Many 300hp 335i BMWs making over 500hp at the engine with no problems. For factory stock HP levels, the platform should be dead reliable with thoroughly over engineered components.

    2. DaveMart says:

      I should add that derating so that the performance limit is not pushed to the screaming limit is perfectly standard engineering practise, so for instance tyres are always rated at way above the top speed the car is capable of, and all the other engineering components also have, or should have, margins built in.

      Unless you are talking about an Italian exotic from the 60’s cars are normally designed with adequate margins.

      Not doing that is just as much an engineering failure as any other cause of the problem would be.

      1. Lindsay Patten says:

        Engineering usually involves trade-offs, reliability needs to be balanced with performance and cost. I read the article to be arguing that ICE sports cars aren’t engineered for 100% reliability no matter how hard you push them either.

        Without knowing the underlying cause of the problems that have occurred we really aren’t in a position to come to any meaningful conclusions.

  10. Lensman says:

    Sorry to see this “stock shorter” FUD repeated in an InsideEVs article.

    How in the world can anyone rationally conclude this is a “common problem” when the only data we have are 75 replacements out of more than 28,000 Model S’s sold? That’s only 0.27% of the total. Sure, the actual percentage is almost certainly a bit higher. But if this was a -frequent- problem, how could it possibly have gone on for two entire years before it rose to the level of anyone noticing?

    This is at best not a rational complaint, at worst it seems some are deliberately trying to smear Tesla Motors’ hard-earned reputation for what some reviewers have called “The best car in the world”; a car which has earned a higher rating from Consumer Reports than -any- other car… including gas guzzlers, not just EVs.

    Here are some facts:

    1. Many or most of the replacements are due to a Tesla Motors serviceman hearing what is described as a “milling noise” in the drivetrain, probably from the gearbox. In other cases, the customer hears it and calls the service department. The number of actual failures is quite low compared to the replacements.

    2. Numbers like $10,000 or $15,000 for a repair are being thrown around rather casually. This is the cost quoted for replacement of the entire “drive unit”, which includes the motor, the gearbox, and the inverter. If the problem is just a mechanical one in the gearbox, then it seems rather silly to claim it will cost $10k or $15k to fix.

    3. TM is pulling the “drive unit” and shipping it back to the factory for detailed analysis and refurbishing, rather than having the unit repaired in the shop. This is being done for the convenience of the customer, because it’s faster to do that than to have the unit torn down, diagnosed, and rebuilt.

    4. TM has the repair shop replace a drive unit it pulls out with a refurbished unit, at least according to reports on the Tesla Motors club forum. Again, the idea that a customer would have to pay for a brand new drive unit, including a motor and inverter which needed no repair, is nothing but FUD.

    The article also mentions some problems with the inverter. Well, of course, every model of car has various problems with various things. But rarely does anyone try to claim cars occasionally needing servicing to be some sort of serious issue, unless it happens so frequently that a recall is justified.

    And nobody seems to be taking notice that most of those replacements– including three of the four replacements for the car– are a result of Tesla Motors taking extra care of their customers, and dealing with a potential issue before it becomes a real problem.

    It should also be noted that the various reports at indicates the team of testers is very happy with the car, despite the servicing it has received.

    Seems odd to me that this short-seller FUD is being spread around so far.

    1. Mark H says:

      In the early days of Inside EVs, at times I thought the chief editor was the grim reaper in disguise until I sat back and realized his most uncanny hand to cover all stories good and bad. I have watched InsideEVs restrain from printing stories good or bad until they felt like they had the data.

      My read on this one is that the story is out there and the writer did a marvelous job of calming the mob with pitch forks.

      The article does not declare the problem is torque related but certainly considers it with some high potential. I owned a International Six Day Trial bike designed by John Penton in the 70’s (showing my age). It was powered by a magnificent KTM motor that had a power band out of this world. So powerful that it allowed me to literally twist the crankshaft into two halves. Never underestimate torque under high performance conditions, especially as the writer points out, when it is applied over and over….

    2. sven says:

      “Sorry to see this ‘stock shorter’ FUD repeated in an InsideEVs article.”

      What’s your stock position on Tesla? Are you “long” on Tesla (hoping the stock rises in value) or are you just a Tesla apologist?

      1. pjwood says:

        He just enumerated what these “replacements” are doing to Tesla. Their sourcing completely new drive units, for say a total recall, versus refurbishing several dozen, are two very different outcomes. That said, I agree the issue isn’t FUD.

      2. JakeY says:

        I have $0 in TSLA and I have seen two similar articles written in Seeking Alpha by TSLA shorters as thinly veiled FUD articles.

        I’m sure Jay does not intend this article to be FUD, but AFAIK this issue was mainly brought up in the media as a FUD point.

        And looking at the article posted by acevolt: the *best* failure rate for an ICE is 0.29% (Honda)! So 0.25-0.27 is nothing to be alarmed about.

        This is before accounting for the fact that these replacements do not mean failure (as there’s plenty of evidence of units replaced for noise), that it includes the electronics too, and that there are members that purposely voted in that poll to mess with the numbers (not hard to do and keep in mind that 95% of TMC members are not Tesla owners).

        1. Boris says:

          Come people, I love Tesla just like the next guy here, but after thinking of buying Model 3 when it comes out, I thought the wait would be too long, so in 2 years I’d like to buy a used Model S (depending on pricing of course) and when I read that at Edmunds they had to get the motor replaced for 4 times, that raises a gigantic red flag for me. Come on now, that’s not once or twice but FOUR times. I am sure Elon will get some questions regarding this one very soon, I just hope he’ll have an answer that will satisfy me.

          1. JakeY says:

            Edmunds is an auto publication, but in terms of data for this issue, it’s just a data point and anecdotal evidence.

            Plus apparently 2/3 of the replacements (it’s 4 units, 3 replacements) is for noise. Only the first unit failed (had faults along with the battery).

            I think it’s mainly the way Tesla designed this unit to be serviced (having the entire PEM and motor in a single unit). The Roadster and all other EVs have it separate. Also Tesla doesn’t do any repairs at the shop, they just ship the whole unit to the factory. This makes things seem more serious than for example with an ICE car (where it is rare the whole engine is swapped out).

        2. sven says:

          “And looking at the article posted by acevolt: the *best* failure rate for an ICE is 0.29% (Honda)! So 0.25-0.27 is nothing to be alarmed about.”

          That article is about the failure rate of old ICE cars after their powertrain warranty has expireed, not ICE cars still covered under their OEM powertrain warranty. See my response to AceVolt below.

      3. Lensman says:

        “[Lensman said] Sorry to see this ‘stock shorter’ FUD repeated in an InsideEVs article.”

        What’s your stock position on Tesla? Are you “long” on Tesla (hoping the stock rises in value) or are you just a Tesla apologist?

        Gee, and when did you stop beating your wife?

        I own no TSLA stock, nor have any financial interest in Tesla Motors. And no one needs to be an apologist for a car which has earned more “car of the year” ratings than any other car; a car which has earned the highest rating (99 out of 100) that Consumer Reports has ever given a car… gas guzzlers included.

        I -am- an EV enthusiast, and I’m more than a little irked that Tesla Motors going out of its way to deal with a noise, as extra effort in customer service, is being touted by bashers as a “failure” of the drivetrain.

        I’m also more than a bit irked that Jay asserted in his article:

        “This does not make us ask the question “is drivetrain failure a common occurence [sic] in the Model S?”

        Because it is.”

        Unless Jay has some source of info he didn’t mention, then he’s stating unsupported opinion as fact.

        And if it’s true that -none- of the 600 Tesla Model S owners surveyed by Consumer Reports reported a drive train problem, then it’s clearly not that common a problem.

        Now, sven, if you want to be an apologist for those giving undeserved criticism to a car just because a couple of high-profile car review companies happened to have a rare problem crop up in the car they were using… well, it’s a free country.

        And even in this country, rare coincidences do happen, despite what Jay wrote.

        1. sven says:

          My wife died from cancer two years ago. She was the most loving and caring person I’ve ever met, and I am truly blessed to to have spent 23 years of my time on this earth together with her.

          You on the other hand are a lonely, pathetic Tesla fan boy who spends his Friday nights whacking off to a pic of Elon Musk. I feel sorry for you. 🙁

    3. See Through says:

      It is time you change your lenses, or at least rinse them with soap water, so you can SEE things that are lying on the table in front of you.
      When one car (Edmunds) goes through 3 drive unit replacements in 30K miles, at exactly 10K miles gap, no argument exculpating the quality issue works. This is really a piece of crap car. And if you read their road trip from LA to NY and back, you will see that those poor fellows were driving at 52 mph to reduce the number of charging sessions, or to reach their next charger whiel other were whizzing by at 80 mph. It can hardly be called rash driving. Initialy, they probably did the 0-60 test, but I’m sure they do it with all cars to test it out.

      Then, consider the Motortrend car, that also had a drive unit replaced. Most likely, Tesla is suppressing the other failures through their chain of service centers. Consumers should thank Edmunds, that they wasted over $100K to test this car out independently ( as much as they could stay independent) without the watchful eyes of Mr. Mask.

      The ‘milling noise’ is basically a beckon of death, not to mention the craziness it causes in drivers due to the continuous humming noise as they drive. There are examples. where when the drive unit was not replaced, it eventually died completely.

      I think it is time for Tesla to do the right thing. Recall all cars, and refund people the money. InsideEVs should do its job of protecting the consumers, not protecting Tesla shareholders or employees.

      1. Boris says:

        What brings you here See Through? You seem to have so much anger inside of you. Even if Tesla has to recall all of theri Model S cars and even if the motor replacement costs them half of their margin, the stock will not come down. You don’t get the bigger picture. TEsla is changing the world and one possibly faulty motor is a non-issue in the big picture.

      2. Lensman says:

        See Through said:

        It is time you change your lenses, or at least rinse them with soap water, so you can SEE things that are lying on the table in front of you.
        When one car (Edmunds) goes through 3 drive unit replacements in 30K miles, at exactly 10K miles gap, no argument exculpating the quality issue works. This is really a piece of crap car.

        I presume you are a short-seller posting FUD in an attempt to cause panic selling among TSLA investors. Either that, or you really REALLY got up on the wrong side of the bed this morning.

        Anyone who wants to know the -truth- should read all the road tests/reviews of the car. Overall, the team loved the car despite the servicing it needed.

        The article also clearly states it received more frequent checkups by Tesla Motors service people than the average car. If there was a persistent problem with a drivetrain noise, then it’s hardly a surprise this particular car had more “drive unit” replacements than just about any other Model S.

        * * * * *

        If anyone reading these comments really thinks there is a serious problem here, just ask yourself this: How many gas guzzler dealerships would take a car, brought in for a scheduled checkup, on a test drive to listen for an odd noise in the drivetrain? And having noticed one, how many would then pull out, diagnose, and refurbish under warranty the engine and transmission, even though the customer hadn’t noticed anything wrong?

        Tesla Motors service people going out of their way to make sure Model S customers don’t have any problem with their cars isn’t a “problem”. It’s exemplary customer service.

    4. Nix says:

      Sounds to me like Tesla is pulling every unit they can get their hands on that might even faintly sound of failure in order to get as many sample sets as they can for their engineers to examine.

      It sounds more like an aggressive quality control measure, trying to get as large a sample size to study, than anything else.

  11. 75 failures out of 28,000 sold is an R/1000 of 2.7. That’s BAD (ask any automotive engineer). Gasoline engine failure rates are expected to be much lower than that in mass production–anything over 1 repair/1000 typically draws out the quality ninjas.

    1. acevolt says:

      This Auto Express article shows failure rates of ICE vehicles are much worse. The best is Honda with a failure rate of 1 failure for every 344 engines,Audi is 1 in 27:

      1. sven says:

        Wrong. You’re comoaring apples to pears. Those are the failure rates of ICE cars after their powertrain warranty has expired, not ICE cars still covered by an OEM powertrain warranty. The Auto Express article is based on failure rates reported by Warranty Direct, a company which sells extended warranties. Also, Warrany Direct didn’t define what qualifies as an engine failure. It probably includes broken fan belts or a bad sensor, since it’s in their interest to report higher numbers of engine failures. Thus, a engine failure as reported by Warrany Direct doesn’t mean a catestrophic failure that requires an engine replacement.

        Did you even read the story you linked to?

  12. Anton Wahlman says:

    Tom Saxton’s survey of 256 Model S owners yielded 13% drivetrain replacements.

    1. Spec9 says:

      That is an unacceptably high level. Tesla had better fix this issue.

    2. James says:

      What’s your source for this statement. Not saying you’re not correct, I’m just interested in reading anything related to this first hand.

      1. Anton Wahlman says:

        You can download the survey data from this web page:

        The survey of Tesla owners was conducted by Tom Saxton of

        I can’t vouch for either the accuracy or the statistical significance of the survey. It was 256 as of this Saturday. Hopefully his sample size will increase. You have to provide VIN and so forth, in order to participate.

  13. GeorgeS says:

    Looks to me like Tesla has an engineering problem to deal with plain and simple.

    People are going to be slamming the peddle on this car all the time. I know I would. I slam the peddle in my Volt 8 times out of 10.

    I slam the throttle on my fjr 1300 Yamaha motorcycle all the time also. I also hit the rev limiter on many ocassions.

    It’s the difference between an old Bull Taco and a Yamaha that makes the difference. If you are going to offer the buyer all that power you had better design it to take the beating.

    Looks like Tesla was a little short sighted in that department.

    1. PJS says:

      “Bull Taco” — now that’s a hilarious spellcheck error!

      1. Mark H says:

        Nah, just a kick back from the day when the European bikes started showing up. Good on ya George!

  14. Omar Sultan says:


    Huge fan of this site, but I am gong to call you out on this “story”.

    1) Not a huge fan of basing a story on a a web poll – it lacks any kind of statistical validity

    2) You need to define “fail” – most “reporting” I have seen makes it seem like the drive units are dying and the cars left immobile by the side of the road. Unlike an ICE, the drive unit is not-field repairable, if anything is wrong (bad connector, logic board failure, mechanical issue) they will swap the entire drive unit out and replace it. If you want to compare failure rates, compare the Tesla drive unit rate to ANY kind of issue with an ICE (bad water pump, balky transmission, clogged fuel injector), the difference is that ICE are designed to be repaired in place, so it makes no news. I my case I had a drive unit replaced at ~10K miles because it “droned” at ~70 mph. It was perfectly drivable, no impact to performance and I actually drove it around for a couple of thousand miles before I took it into to the shop. The repair was painless, they picked the car up, swapped the unit, then returned it to me.

    3) No one outside of Tesla knows what the policy going forward will be for drive unit replacements–again its dangerous to take an anecdotal web posting and extrapolate company policy from that. With a couple of exceptions, most Model S are under warranty, so its a non-issue, but I will guess they will handle non-warranty repairs in a different manner than an ICE dealer because the service model is different.

    So, the one actual data point you do have access to is the cost of warranty repairs, which has to be disclosed as part of SEC filings. If you look at the quarterly warranty costs compared to the number of cars actually in service, warranty costs are down slightly Y/Y.


  15. taser54 says:

    My question: what does Tesla do with all these drive units they’ve replaced?

    1. Anon says:


      1. Taser54 says:

        I thought warranty parts were required to be new parts. Perhaps Tesla’s use of refurbished drive units is part of the problem.

        1. TeV says:

          Perhaps you simply have no clue what you’re talking about.

          1. taser54 says:

            Leave your personal attacks at home.

        2. See Through says:

          You are right. That’s why Tesla avoids the dealer network. They will NEVER put in a refurb part as warranty replacement. But Tesla’s own SC? They will do whatever the Boss says.

  16. ffbj says:

    Refurb Madness.

    1. Mark H says:

      good one!

  17. Ocean Railroader says:

    As long as Tesla investigates it and fixes it I will still respect them.

  18. Mike says:

    Here’s one defect in your INVESTING strategy.
    Index Funds are Heavily Weighted in the LOSER Carbon companies, and these are going to take serious LOSES as Solar and Wind become cheaper and cheaper.

    Wind contracts now down to 5 cent per KWh.
    Coal is Dead.
    But, your Index Fund is going to hold onto them until they’re in the grave.

  19. Aaron says:

    We need to hear from Tesla as for what the nature of these replacements is. Bearing issues? Gears that are shearing off? A breakdown or contamination of the reduction gear fluids? Motor issues?

  20. Nick M says:

    I’m going to call Betteridge’s law on this one. On the one hand you have an anonymous internet poll and some anecdotal evidence. On the other hand you have the Consumer Reports reliability survey which found that out of 600+ owners “no owner in our survey has reported problems with the electric drivetrain”.

    Since there’s no real statistical evidence, InsideEVs appears to have fallen back on complete speculation. Unless I’m missing something, the idea that “drivetrains are failing left and right because people are pushing the cars too hard” appears to completely fabricated, with no hard evidence to back it up whatsoever.

    I’m disappointed. I expect real news from this site. But I guess manufacturing a controversy draws more attention than simply reporting the facts.

    1. Jay Cole says:

      Clearly you have strong feelings on this Nick. We can appreciate that.

      I do suggest that you have a look at the source poll and the people reporting the replacements in the thread. These are established persons/owners with hundreds or even thousands of posts behind them raising their hands. And then there is also Edmunds and Motor Trend cars. And I have also personally been contacted by Model S owners with the issue. I really don’t believe their is a question of “if” something is really happening here.

      Again, this is not a new issue, and it is not a source of some joy to write a piece on it…however, if we did not acknowledge it at all here at InsideEVs and rather pretended it was not happening, we would be doing the reader a greater disservice overall IMO.

      We love plug-ins, but are committed to reporting ALL the newsworthy stories on them that are out there – good or bad. I don’t think this story offers up/flames a particular controversy (especially in relation to how the situation is presented elsewhere), it doesn’t even come to any solid conclusions when it comes to any kind of fault.

      Anywhoo, that’s the spirit it was written in, perhaps it did not read that way to yourself – sorry to disappoint

      1. TomArt says:

        Hey, Jay – have you been able to reach Tesla on this issue? Do you know if this question has been asked during a quarterly earnings call? Has any conclusive data, monetary or otherwise, appeared in their SEC filings?

        Are there really no facts available, or are they just indirect and hard to find? The only evidence we can count on, and that has been reported, is the following:

        1) Edmunds (4th drive unit)
        2) Motor Trend (2nd drive unit)
        3) Consumer Reports poll of actual Model S owners (no problems?).

        What gives? We can’t have it both ways. It’s either a “problem” or it is “not”. Maybe we all need to hound Tesla Motors…e-mail them…post in the forums…grill their reps in the stores/showrooms…put pressure to get an answer we can reasonably believe.

        At the very least, is it true that the drive unit cannot be serviced for essentially anything in situ, no matter how trivial – policy is to replace and then deal with the original drive unit in the shop? The answer to this question makes a material difference in the meaning of “replacements.”

        IMHO, statistics that owners, potential owners and/or stockholders should have include:

        1) % of all replacements initiated by a vehicle failure;

        2) % of vehicles with at least one drive unit replacement, broken out by model (S60, S85, P85 and P85+) and model year;

        3) % of vehicles from #1 with multiple replacements;

        4) % of all replacements initiated by mechanical reasons, broken down by issue (bearings? gearbox? motor imbalance?);

        5) % of all replacements initiated by electrical reasons (pretty much just power electronics, AFAIK);

        6) % of all replacements that were a minor repair (replacing a sensor or something of that nature);

        7) % of all replacements that were a major repair (gears in gearbox, bearings or bearing races, etc.).

        Those are all I can think of at the moment – this got pretty long, anyway, and I have other things to do…

        Would you guys from be able to tease any of this info out from a reliable source (Tesla, Consumer Reports, etc.)?

        1. Jay Cole says:

          Honestly, we haven’t tied.

          Although Tesla’s CEO is more than forthcoming ‘in the moment’ to give out details/make statements, the rest of the company is pretty guarded when it comes to stuff like this (as one would expect all things considered).

          There has been a warranty allowance/reserve disclosure of $72 million, but this is about it.

          The quarterly report/call is where the opportunity lies to address it. And I can’t see how this situation won’t be touched on in some way during the next call on Thursday. If so, we’ll be on the call and will pass it along.

          1. TomArt says:


      2. Nick M says:

        Jay, thanks for the followup. Perhaps I wasn’t clear enough in my posting; I’m not arguing that the issue isn’t happening at all, I’m arguing that there is no clear indication that it’s a *common* issue. Let’s review the evidence:

        The Tesla Motors Club poll shows 76 respondents indicating they had drive units replaced. Now, even if all 76 responses represent real individuals who have all had units replaced, that’s out of tens of thousands of Model S’s on the road today. On the one hand, not every one of the owners who’s had trouble is likely to respond to a poll on the TMC forums. On the other hand, people who experience problems are more likely to search online to try and find out if others have had the same problem and/or to vent their frustrations.

        It is notable that Edmunds and Motor Trend both had drive unit failures, but two anecdotes does not make for a statistically significant dataset. It could very well be bad luck that two problematic cars ended up in the hands of reviewers.

        The tone of the piece is set right from the beginning with your statement in bold type that “it is [a common problem]”. This conclusion doesn’t seem to be adequately supported with enough evidence to warrant your stating it as a 100% certain fact. I’m not upset that InsideEVs would decide to cover a problem like this, but I feel that some of the claims you made are not adequately supported by evidence. It certainly makes for a more exciting, attention-grabbing piece to start the article off with a boldface proclamation of fault, but I feel you went too far in this case.

        1. Lensman says:

          Bravo. Well said.

  21. James says:

    Because there are so few parts, any defect seems major. Think of how many parts can, and do, fail on an ICE engine vs. the tiny lump of a motor on a Tesla. Pretty much need to replace it every time. We have 21k miles on our the Tesla drivetrain in our RAV4, no problems. Brilliant driving.

  22. jmac says:

    The problem can be fixed. Fix it, and move on,

    Every car company has recalls.

    1. TomArt says:

      Yeah, but normally automakers, Tesla in particular, will come out and say it. This drivetrain issue has been almost exclusively from anecdotal sources – I am not aware of any statement from Tesla regarding the drive unit replacement rate.

      1. Lensman says:

        No, normally auto makers do -not- come out and admit that they have a problem with any part on their car. Not unless they have to do a recall. In fact, auto makers often go to great lengths to avoid admitting there is a problem, even when the problem is a serious one. Remember Toyota and the sticking accelerator problem? Or the Pinto gas tank leaks and fires? Denial, denial, denial in both cases, and many others you can name.

        It would be nice to hope that Tesla Motors doesn’t try as hard to hide any problems, but frankly, nothing I’ve seen leads me to believe they’re being any more transparent about problems than the average auto maker.

        1. TomArt says:

          Actually, Tesla has – one of their assembly line workers discovered a weld fault – Tesla voluntarily recalled those vehicles to fix the weld. They have voluntarily fixed a number of issues as they find them. I would hope that they wouldn’t change their behavior on this issue.

  23. @Jay,

    Is the Edmunds Model S mentioned (up top, needing 4 drivetrains) the same vehicle that set the 3332 mile cross-country trip record in 56 hours?

    1. Jay Cole says:


  24. Murrysville EV says:

    Don’t forget that the Edmunds’ Model S – during one of its drivetrain replacements – ALSO received an entirely new battery pack. This happened after it died on the freeway, without so much as the ability to turn on the hazards at night when it happened.

    AND – A commenter on the Edmunds forum noted that his Toyota RAV4 EV drivetrain was replaced, which is a Tesla unit. It wouldn’t have near the performance of the Model S, yet he experienced the same symptoms and it died.

    Like one of the commenters above, a Model 3 has been the leading candidate after my Leaf lease expires, but I’m not so sure now.

    As for the Edmunds long-term Model S: I was not aware they had sold it, as their blog hasn’t reported this. If this is true, they are once again covering up an important story. Their credibility fell when their long-term Dodge Dart mysteriously disappeared after some sort of catastrophic engine failure that left them to limp home. No explanation was provided.

    Numerous commenters on the Edmunds site have begged Edmunds to keep their Model S for a much longer time, to see how it goes. If they’ve ditched this car, it actually looks worse for Tesla, and very bad for Edmunds’ journalistic integrity.

  25. Jeff D says:

    I am not going to throw Tesla to the wolves, but I am not going to let them off the hook either. There does not seem to be large amount of data just yet and it seems that most owners to this point have not dealt with this issue yet. In the small sampling of data we do have though seems to indicate that there is some problem that tends to reoccur. If I was running Tesla, I would be very interested in getting to the bottom of the issue before the current small amount of data becomes huge.

  26. Bill Howland says:

    Hi guys…

    Like it or not, “I’m Back” because Jay Cole needs some support and also thanks from this longtime EV driver (about 55,000 EV miles under my belt currently – in 2 EV’s).

    Jay is to be congratulated for broaching this issue, even though it could conceivably harm future advertising revenue.

    The best information I’ve seen is with Tom Saxton’s survey. Whether people like the results or not, and whether it is ‘scientific’ or not (whatever that means); it does seem to give a ‘slice of life’ feel to what is happening. I took part in the Roadster survey and it is to these eyes a truly objective attempt to find out what is really going on with these cars in day to day driving.

    A simple helical reduction gear set should not wear, or get increasingly noisy, in a car with so few miles on it. To these eyes, as mentioned previously “13 % is unacceptable”.

    I was bitterly attacked for saying the 14-50P male plug adapter included with the “S” was designed by an incompetant. This design fails Nema standards of current density, but has the advantage of being easily fixed by changing the plug, or else putting it on a short, homemade extension cord to get it away from a combustible wall surface. Tesla also mitigated the issue themselves by lowering the charge current and putting in a fusible link.

    A gearbox is a much more serious matter, seeing as the average do-it-yourselfer, *Cannot* go to the local hardware store and order their own drop in reduction gearing.

    To these eyes it looks to be just another place where Tesla has scrimped in the wrong place. Just as GM years ago had the ability to have better ignition switches for only a $1 a piece more.

    I’m in the market for my 3rd EV, and I’m looking for either a stripped model “S”, Cadillac ELR, or else waiting until the new Volt comes out, with as the Chevy dealer just yesterday told me it will have a 40 mile range, with an extra cost 50 mile range option (the one option I’m certain to get), and a 3 passenger bench seat in the rear. Funny that I haven’t heard anything about it on Insideevs yet.

    In view of these gearbox problems, its making it look increasingly like my 3rd ev will be a GM product instead of a Tesla. (I currently own a Tesla and a Volt).

    1. sven says:

      “Like it or not, ‘I’m Back’ because Jay Cole needs some support and also thanks. . .”

      Like a phoenix rising from the ashes. 😀

      I agree. Thanks Jay for providing fair and balanced coverage of all the EV news!

      1. Jay Cole says:

        Hey Bill, nice to hear from you, it has been too long, (=

        ..and thanks for the kind words (you too sven)

  27. Anon says:

    The quickest Model S is NOWHERE NEAR as quick as a R8 V10 period.