M3 Tear Down by Munro & Associates. Pictures By Motor Trend

[voip-ninja]

My takeaway from the Munro video (watch all the way to the end!) is that the Model 3’s drivetrain, suspension, battery pack, electronics, and autonomy chipsets are all two or three generations ahead of everyone else in the industry.

Simultaneously Tesla’s coachwork (stampings, body assembly, efficiency of materials allocation) is two generations behind everyone else. Not really a surprise here.

One of these is much easier to fix than the other and Tesla is known to redesign and improve parts from one week to the next on S and X. If the strengths were reversed and the tech was a kludge with stunning coachwork I’d be scared for the viability of the company.

Incidentally I rented a December 2017 build Model 3 in early February and it had a number of wonky panel fits and uneven gaps. Nothing I would have made a stink about, but it was assembled about as precisely as my 2012 Volt. Not fantastic but tolerable.

I took delivery of my April 2018 build Model 3 last Friday and it’s...perfect? At least very close to perfect. I’d put its build precision up against a BMW or Audi without shame, it’s really that good. The amount of improvement from December to April took me by surprise.

I agree with everything you wrote other than the suspension bit. While the suspension was praised I didn't see anything in the analysis that indicated they thought the suspension was even one generation ahead of what anyone else in the industry was doing.

I appreciate your response, it's great getting this kind of feedback from actual owners of the product.

 

[Doubletap67]

As someone with an industrial background I must tell you that you and all other "this is just an early production vehicle" sayers VASTLY overestimate the room for improvements a manufacturer with a running production has......While stuff like the gap dimensions or any modular part of the car can certainly be tweaked or replaced the heavy body frame and basically anything else that stems from design/engineering errors rather than material flaws will stay till the car gets a major overhaul......this is NOT some handmade low volume car!

........no

Totally disagree - I'm also in "industry", in my case the Medical Device Industry, and design changes affecting manufacturing happen at a rate of many a week - significant changes that replace major sub-systems happen many times a year for products early in their life-cycle. .

 

[mtndrew1]

I agree with everything you wrote other than the suspension bit. While the suspension was praised I didn't see anything in the analysis that indicated they thought the suspension was even one generation ahead of what anyone else in the industry was doing.

I appreciate your response, it's great getting this kind of feedback from actual owners of the product.

Thank you, that’s a fair criticism. Maybe I should have written that the “suspension dynamics are outstanding.” It seems every reviewer agrees on this point and Munro seemed impressed that they did double wishbone and five-link (this is expensive).

 

[ProphetM]

From their analysis, the problem with the model 3 is hubris.
The mechanical and manufacturing problems come from the fact the tesla thought they could ignore centuries of coachbuilding experience and do it their own way. They apparently were infected by the silicon valley mentality of "new = better!" and "high tech > low tech" ignoring that many times things have been done the same way for a long time for a reason, and more tech often creates more problems (cough cough PHONE KEY!)
Elon has admitted this to some degree with his confession that they were assuming robots would be better than they are.

This engineer's conclusion is that if they had built the drivetrain and suspension and electronics the way they are now, but used traditional means for the mechanics and manufacturing, they would have wiped the floor with even toyota. He was unsure if it was fixable at this point.

That might be true but I have to wonder, if they had used more of those traditional means, would the car be in production yet? Or would we still be waiting for it? And how would a longer timeline before production ramp have affected Tesla's viability?

 

[Brando]

I agree with everything you wrote other than the suspension bit. While the suspension was praised I didn't see anything in the analysis that indicated they thought the suspension was even one generation ahead of what anyone else in the industry was doing.

Dual wishbone front and Five link rear. I think that was what he said. I think this is "state of the art" for production cars.
Munro did say he thought the handling was what? great? fine? I don't remember any complaint about the suspension/handling. .

Oh, rear steering - that would be leading edge stuff. Model 3 doesn't have rear steering.
AND torque steering I think would require two motors on an axle, right? Model 3 doesn't have that either.
But what cars do? Sorry I don't know. Perhaps these are becoming common?

GM first with concept electrics (1960s two different Corvairs - I'm sure many others by GM, that just the one I know about)
GM first with Electric Vehicle EV1 (RAV4 and SMART electrics first generation about the same time)
GM first with +200 mile <$40,000 electric.
Tesla always the great follower, right?

Mostly I only think I know what I read - and what is printed, I know, isn't always accurate. Interesting times. reading on.

 

[MP3Mike]

Anyone else that find it peculiar that the two side rows with fewer "bricks" of cells seem to be of equal lengh as the inner ones from a pack enclosure poiny of view? The pictures Electrek shared Tesla Model 3: Exclusive first look at Tesla’s new battery pack architecture showed some content here, but in the picture here it looks like a "spacer". If the 23 vs 25 bricks from Electrek is right, and there is space, why not fit 1 or possibly 2 more per side module? Downgrading to not overshoot Model S/X range?

People have speculated that the extra space in the front corners is there to help with the partial overlap impact crash test...

And interestingly Tesla did it on both sides. (Only the driver's side is tested.) Other manufacturers have put structure in place just on the driver's side to handle the test. Sure a partial overlap is most likely on the driver's side, but the passenger should get the protection as well. (Not to mention having the car be built the same for LH and RH drive countries.)

 

[nvx1977]

Incidentally I rented a December 2017 build Model 3 in early February and it had a number of wonky panel fits and uneven gaps. Nothing I would have made a stink about, but it was assembled about as precisely as my 2012 Volt. Not fantastic but tolerable.

I took delivery of my April 2018 build Model 3 last Friday and it’s...perfect? At least very close to perfect. I’d put its build precision up against a BMW or Audi without shame, it’s really that good. The amount of improvement from December to April took me by surprise.

I think I remember your review of that rental, and I responded in that thread that I had a february build that did not have a lot of the issues you had mentioned. Great to see some validation in your April build!

 

[TT97]

AND torque steering I think would require two motors on an axle, right? Model 3 doesn't have that either.

I thought torque steering is bad? Mostly occurs in FWD vehicles.

There are plenty of cars that distribute power to one side or the other and none have two motors on an axle - you just need to limit power on one wheel.

 

[mtndrew1]

I thought torque steering is bad? Mostly occurs in FWD vehicles.

There are plenty of cars that distribute power to one side or the other and none have two motors on an axle - you just need to limit power on one wheel.

I believe this person meant to use the term “torque vectoring” but they can correct me if I misinterpreted.

 

[voip-ninja]

Totally disagree - I'm also in "industry", in my case the Medical Device Industry, and design changes affecting manufacturing happen at a rate of many a week - significant changes that replace major sub-systems happen many times a year for products early in their life-cycle. .

Somehow I doubt changes on your products require sending them back to the government for crash testing.

 

[BioSehnsucht]

I don't think anyone sends their vehicles to the government for crash testing... it's all private facilities or internal facilities now. In fact you might be able to start selling them before the actual crash tests have been completed? As long as you're confident in your results and willing to take the risk ... computer modeling of crash testing has come a long way.

 

[voip-ninja]

I don't think anyone sends their vehicles to the government for crash testing... it's all private facilities or internal facilities now. In fact you might be able to start selling them before the actual crash tests have been completed? As long as you're confident in your results and willing to take the risk ... computer modeling of crash testing has come a long way.

Crash testing was a bit tongue in cheek.

I do believe it was pointed out they would potentially have to resubmit a car to continue selling it if substantial structural changes were made.

 

[mtndrew1]

I don't think anyone sends their vehicles to the government for crash testing... it's all private facilities or internal facilities now. In fact you might be able to start selling them before the actual crash tests have been completed? As long as you're confident in your results and willing to take the risk ... computer modeling of crash testing has come a long way.

I have bought several cars with no government crash rating.

I believe the vehicle only needs to comply with the current FMVSS to be sold. If a crash test has been performed and then structural changes are made perhaps the model can no longer be sold advertising the prior ratings. I don’t believe FMVSS has any specific crash performance metrics beyond bumpers.

 

[Brando]

I thought torque steering is bad? Mostly occurs in FWD vehicles.

There are plenty of cars that distribute power to one side or the other and none have two motors on an axle - you just need to limit power on one wheel.

@TT97

I was first made aware of "torque vectoring" from Rimac
Rimac All Wheel Torque Vectoring | Rimac Automobili
AND Elon mentioned that the Roadster 2020 had dual motors in the rear (one in the front) and would do torque vectoring.

Sorry for my miss-statement and confusion. My Saab turbo has plenty of torque steering - yeah, the bad kind.

PS - I don't think vectoring can be done with brakes. Traction control apparently can be done with brakes.
- SEMI with four motors on four wheels will use torque vectoring to prevent Jack-Knifing - which some are sceptical - and want to see it before they believe it. good to be sceptical

thanks @mtndrew1

 

[omgwtfbyobbq]

Crash testing was a bit tongue in cheek.

I do believe it was pointed out they would potentially have to resubmit a car to continue selling it if substantial structural changes were made.

Did you read my link on page 1? NHTSA buys cars it's going to crash test, not the other way around. If the 3 doesn't meet whatever minimum standards, Tesla has to fix whatever problems there are, but it's up to NHTSA to purchase and test the updated car.

 

[slipnslider]

That might be true but I have to wonder, if they had used more of those traditional means, would the car be in production yet? Or would we still be waiting for it? And how would a longer timeline before production ramp have affected Tesla's viability?

What's more important, getting a crappier car done faster, or making a brilliant car that puts all other cars to shame in both technology and reliability?

 

[chibi_kurochan]

Dual wishbone front and Five link rear. I think that was what he said. I think this is "state of the art" for production cars.
Munro did say he thought the handling was what? great? fine? I don't remember any complaint about the suspension/handling. .

Oh, rear steering - that would be leading edge stuff. Model 3 doesn't have rear steering.
AND torque steering I think would require two motors on an axle, right? Model 3 doesn't have that either.
But what cars do? Sorry I don't know. Perhaps these are becoming common?

GM first with concept electrics (1960s two different Corvairs - I'm sure many others by GM, that just the one I know about)
GM first with Electric Vehicle EV1 (RAV4 and SMART electrics first generation about the same time)
GM first with +200 mile <$40,000 electric.
Tesla always the great follower, right?

Mostly I only think I know what I read - and what is printed, I know, isn't always accurate. Interesting times. reading on.

Agreed on all accounts,
more like: Tesla "first" with "desirable" +200 mile

My takeaway from the Munro video (watch all the way to the end!) is that the Model 3’s drivetrain, suspension, battery pack, electronics, and autonomy chipsets are all two or three generations ahead of everyone else in the industry.

Simultaneously Tesla’s coachwork (stampings, body assembly, efficiency of materials allocation) is two generations behind everyone else. Not really a surprise here.

One of these is much easier to fix than the other and Tesla is known to redesign and improve parts from one week to the next on S and X. If the strengths were reversed and the tech was a kludge with stunning coachwork I’d be scared for the viability of the company.

Incidentally I rented a December 2017 build Model 3 in early February and it had a number of wonky panel fits and uneven gaps. Nothing I would have made a stink about, but it was assembled about as precisely as my 2012 Volt. Not fantastic but tolerable.

I took delivery of my April 2018 build Model 3 last Friday and it’s...perfect? At least very close to perfect. I’d put its build precision up against a BMW or Audi without shame, it’s really that good. The amount of improvement from December to April took me by surprise.

i might agree with you. Our December/January car with 003xxx vin has those panel gaps and stamping issue. But I just saw a 020xxx VIN at the parking lot and this car has MUCH better fit and finish compared to mine.

 

[ProphetM]

What's more important, getting a crappier car done faster, or making a brilliant car that puts all other cars to shame in both technology and reliability?

If the company will go under before they can accomplish the second one, then I choose the first one.

 

[TT97]

@TT97

I was first made aware of "torque vectoring" from Rimac
Rimac All Wheel Torque Vectoring | Rimac Automobili
AND Elon mentioned that the Roadster 2020 had dual motors in the rear (one in the front) and would do torque vectoring.

Sorry for my miss-statement and confusion. My Saab turbo has plenty of torque steering - yeah, the bad kind.

PS - I don't think vectoring can be done with brakes. Traction control apparently can be done with brakes.
- SEMI with four motors on four wheels will use torque vectoring to prevent Jack-Knifing - which some are sceptical - and want to see it before they believe it. good to be sceptical

thanks @mtndrew1

I knew what you meant and tried to make a joke - apparently didn’t convey well. I wasn’t meaning to insult.

I used to have an Acura that had SH-AWD - it would send more power to the outside wheel in turns, mostly done by a differential in the rear axle (Similar to a limited slip differentials). There are plenty of other systems very similar although they all seem to be on AWD and not RWD but no reason couldn’t be RWD. Others common are MB 4matic, Audi Quattro, BMW xDrive, Subaru Symmetrical AWD.

 

[alseTrick]

I suspect the cost has increased since the early builds.

The number of vehicles currently produced per week is 20 times greater than early builds. There's literally no chance the cost has increased.