And we now know that the drive units are cryptographicly linked to the car. So you wouldn't be able to just add a motor, you would have to bring both motors over from the Performance car along with the ECU, unless you had the software access to update the configurations in the car.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Model 3 Motors on the Tesla Parts Catalog
- Thread starter EVMike3
- Start date
And we now know that the drive units are cryptographicly linked to the car. So you wouldn't be able to just add a motor, you would have to bring both motors over from the Performance car along with the ECU, unless you had the software access to update the configurations in the car.
Knightshade
Well-Known Member
Such software access doesn't seem that hard though?
I've seen threads on folks doing this, and obviously there's a few "famous" folks who do this stuff for a living on Teslas...
From a thread on replacing a DU on an S for example (without moving the ECU with it)-
"You will need to reflash firmware after the install. We confirmed that today. The car would not run until the firmware on the new drive unit was sync'd up to the gateway."
The Model S doesn't have cryptographiclly linked DUs. Tesla started that with the Model X. As I recall even @wk057 couldn't replace a drive unit in a Model X.
Manufacturing practices only count for original build and you're extrapolating that to the whole service and repair industry down the line. Tesla and non Tesla service centres will be working on Model 3's down the line and I'm telling you from experience they don't always do vin checks they do Year model and part number checks so it's 100% likely a 990 will end up in a car that originally had a 980 or vice versa. This has nothing to do with manufacturing practices.
I didn't say they're physical the same I said I suspect the 990 is as capable if not more capable than the 980. This has 0 to do with them being physically the same
I never said with certainty they have "spare" 980's lying around that's a possibility but my suspicion is that production is transitioning from 980 to 990
That's irrelevant a vocal passionate minority can be just as damaging. In fact you're arguing my case the people not on the forums and understand less about how EV's work who hear "Your car is physically identical to the one $8000 cheaper, look it even uses the same motors" are the ones that this would placate
The tread you've pointed to even points it out it totally depends on the way it fails. My point still stands it's "LESS" risky to test it our in the AWD than a RWD
Not they wouldn't you protect your high paying customers you don't do risky things at the high end when you're a mass market automaker. Again my example of the Honda Prelude is a perfect example of this.
Yes, it is. You're extrapolating from incomplete information. You assume you know what's been enough to trigger a rev letter but you until now have no basis to know with certainty the smallest change that necessitates a PN change to them.
You overlook a very basic example and while I don't think this is the change it very well could be. A change to the casting of the casing for the motor would be considered a Class 1 change and even if none of the other parts in the motor changed even if the casting was only changed slightly to add strength or to reduce the thickness of a wall to save weight or cost a Class 1 change if you're following best practices requires a full part number change not just a revision in traditional manufacturing if both parts are still being produced at the time of revision. There's other changes that would be similar
But Tesla will not always be doing it and that make it an issue
Sigh you're not understanding this part the issue isn't downloading the code it having multiple code paths for the same car and the complexity that introduces to the code itself
You're not listening to what I'm saying here
SKU's before
- 960
- 980
SKU's now
-960
-980
-990
2 SKU's to 3. why?
2 production lines to 3 why?
2 inventory buckets to 3. Why?
All of these add significant costs
And for the last damn time I'm not saying they're sitting on an inventory of hundreds of thousands of 980's that would be idiotic and terrible inventory management but I also don't think they're going to expand to multiple production lines when it's been working with fewer for years already. There's no cost savings or sense in that. The simplest answer here is the 990 is being proven out and eventually the 980 will be retired in favour of the 990. Then they're back to 2 SKU's. 2 Part Numbers, 2 Production Lines, 2 Inventory buckets
I'm including that in the 'Misc HV cabling'. It does require penthouse disassembly, but it's unclear to me how much work that is. Maybe you can just swap some penthouse circuit boards and the penthouse housing. It might even be a connector hidden behind a blanking panel.
Originally they did have a blanking panel in the battery case for the front motor HV cable. But then Elon said that was stupid they were wasting time and money putting the hole in the pack, and then more time and money sealing the hole back up with a blanking plate when it would never be used. So they switched to two different battery bases depending on if you were getting a RWD or AWD Model 3.
So it might be easier on a very early RWD Model 3, but I doubt we will ever see anyone actually do it, short of swapping the whole running gear and electronics of a collision totaled AWD car into a flood totaled RWD car. (Like @Btr_ftw did with his original Model S build.)
Knightshade
Well-Known Member
I'm not "extrapolating" I'm telling you how a bill of materials works.
It's used AFTER manufacturing not just for things like repairs and replacement parts but it's also how they know WHICH VINs they need to include in a recall.
They know all of Part # XYZ they put into cars of a given range need to be recalled.
So they can query the database to find every BOM where that specific part was used- and then narrow that down to just the range in which the "bad" part was installed. Voila- list of VINs to recall.
Again this is very basic stuff in manufacturing.
Can you give any examples of a Tesla service center putting the wrong motor in a car... ever?
Even one?
Certainly a 3rd party isn't going to have internal access to Tesla build databases- they might WELL screw up depending on their level of incompetence.
I'm not sure how that matters to the discussion at all though.
Especially since we've established a third party might not be able to replace a drive unit without software access on newer Teslas.
Then it'd be weird they've gone a year putting a just-as-capable but cheaper unit in ONLY the AWD.
Very weird in fact.
nonsensically weird.
You don't transition hundreds of thousands of parts for an entire year....after which you've still installed zero of them in any other trim of the car.
That's not a transition. That's either gross incompetence of the worst kind, or they're not ACTUALLY comparable enough to be interchangeable and no transition is happening at all.
But...it's not.
Because the rear DU failing in either car is equally disabling to the car.
As that very thread mentions.
Like cut the price thousands of dollars without warning?
LOL as the kids say.
Again this is simply not true.
Rev PNs replace the previous one. If you order a 980 today you get a G. Doesn't matter what rev you had before because it's a newer rev of the SAME PART. It drops right in and there's no reason to offer the "old" version to anyone.
You only bother with a NEW PN if there's a difference that matters for things like both initial build AND/OR replacing the part later.
Hence we know that's the case.
A 980 will be replaced by the current-rev 980. It won't be replaced with a 990. Which is a different part.
If it were something like that they'd be using the 990 in all cars. Not exclusively the AWD non-P.
So again- no.
There has to be a REASON it's only in the LR AWD. And has continued to ONLY be there for over a year now.
The simplest reason, and the only one that really fits the info we have, is it's less capable of max output than the 980.
You've offered no other actual explanation that fits the facts.
If the drive unit is mated to the VIN in Teslas system as it has been suggested is the case for the 3 (and newer Xes at least) it sounds like it WILL always be Tesla doing it unless someone hacks around that.
No, I understand it fine- you've just failed to make a case for it being an issue.
There's ALREADY a bunch of paths.
They just added TWO MORE with acceleration boost (980 vs 990). T
hey created new profiles for EVERY trim....TWICE... with the two different 5% power upgrades too (which obviously were different for each trim of the car based on testing done by owners).
One more doesn't appear to be much burden on Tesla.
It's weird you';re asking- since YOU keep insisting the 990 isn't really "different"
I'm the one who pointed out there'd be no reason to HAVE a 990 unless it is different and in particular cheaper.
Because otherwise it wouldn't be worth the cost.
Your argument is that it's somehow NOT different but STILL worth the cost of having a different part.
It's yet another on the long list of reasons it fails occams razor.
So your premise is the 990 is just as good as the 980....(why does it exist at all in that case again?)
Or, no, wait, the 990 is "better"...somehow....but they're gonna spend YEARS running three DU lines until they're really super extra double sure of that.... because apparently Tesla has no idea how to actually test anything to tell if it's good enough without running it for multiple years in production...or something....? But not actually testing it in production on ANY trim other than the LR AWD.... because REASONS!
Your story isn't making any more sense, but it keeps getting funnier.
Depends on when they swapped it then, I suppose, no? Even then, how hard do you think it is to get your hands on JUST the penthouse housing? I've got a June 2018 build, so if it happened after that... that's ~50k Model 3's driving around with blanking plates just waiting to be upgraded
Depends on when they swapped it then, I suppose, no? Even then, how hard do you think it is to get your hands on JUST the penthouse housing? I've got a June 2018 build, so if it happened after that... that's ~50k Model 3's driving around with blanking plates just waiting to be upgraded
I don't think the difference is the penthouse housing... It is the front of the main pack.
There's no connections on the front of the main pack for the motor. Those are for the AC only I believe.
See this post:
Model 3 Motors on the Tesla Parts Catalog
There's 2 channels on all battery packs that are meant to hold the HV wires that run from back of pack to front of car for the front motor. On the LR RWD (and I assume SR/SR+) they are just empty. I can see them if I jack up my car and poke my head under.
EDIT: Actually, re-watching the videos, even the AC lines are run from back of the car. The penthouse is all the connections for the car it seems.
Yes I don't disagree with that you're trying to redirect and I've never disputed what a BOM is or how is is/can be used.
That will fly in the face of Right to Repair legislation that's coming down the pipe in NA and elsewhere. So I don't see this being the case long term
Why is that so weird? How long is a successful test bed. Off the top of my head I'd start around a certain % hitting somewhere near 60000Miles that's not going to happen in a year.
Why not, see above. Or they're comparable enough to be interchangeable full stop.
This is just plain false
That's apples to oranges. This is about the long term reliability and functionality of a car someones already sunk a significant amount of money into. Whinge all you want about price changes that happens and it's completely unrelated to this.
It is not false. You're half right but your on the wrong side of an all x's are y's but not all y's are x's type situation here.
If the part "CANNOT" replace the original part 100% yes it must have a different part number. but you can and companies often do iterate the part number while maintaining it's ability to replace a previous part number.
this is 100% assumption.
Testing
Or when right to repair laws come into effect
No you're really failing to understand here. It's not that simple.
How is the firmware identifying what VIN = what motor. We've already seen that VIN's aren't sequential.
So what it has a constantly growing list of all the VINs and what motors they have in the firmware? That's not efficient.
nooooo I've repeated this several times. The part is different. Our argument is what constitutes "Different" I contend that they're not going to spin up a new production line to put out a new less capable motor. That's nonsensical. To put the engineering time, production line, and all the logistics in place to have a new part is a better indication the motor is capable of more not less than the 980
No The motor is clearly good enough for production and certainly good enough for the LR AWD. Even if you look back at Elon comments about the 980 it took them quite some time to determine that it 100% did not require any binning to separate out the P destined motors. You're seeing the same thing here live instead of 2 years before production starts.
"Because reasons?" I've not been that vague. But lets be clear here are some of the reasons testing a new motor on the LR AWD make sense
- LR AWD is a subset of the P - you can run the motor for extended miles while it feeds metrics back to Tesla over the long haul while not risking premium customers where more abuse is potentially put on the motor
- Testing in the LR AWD over the SR you hedge your bets and see at least an ability to tolerate some failure or degradation without stranding the driver ( you contest this but sorry your wrong there's so many ways a motor can fail that don't make the AWD inoperable, yes there still are some that do )
I wouldn't be surprised if we start seeing AWD P's with the 990 by the summer
I'm not saying I might not be completely wrong. But I'm saying is we don't know either way at this point. And until we tear one down we won't that's what Iv'e said since the beginning
jebinc
Well-Known Member
Knightshade
Well-Known Member
The 'test bed' of the 980 was not only MUCH less than a year, it was in FAR fewer cars... (the relatively few LR RWD cars Tesla sold before the P and AWD was launched)
So it's nonsensical they'd need a MUCH larger AND longer live test of the 990 if it's somehow more capable AND virtually the same part/compatible.
<citation needed>
I cited a source saying it's exactly true. With multiple owners confirming.
If you disagree, provide a genuine source (ie an actual owner with a failed rear DU in a 3 who was able to still drive on the front unit- not the generic statement from Elon prior to actual production of the AWD)
Utter, utter, utter nonsense.
YOU said they might've changed the PN (adding significant cost to the company) JUST to stop AWD owners from being "upset" about an obscure thing very few are even aware of.... from a company who does things like large unannounced price changes that "upset" far more owners and are far more public. It's clear that as long as they're selling more than they can produce they're really NOT concerned about this kinda stuff and the idea they'd add PNs to avoid it is outright laughable as an excuse.
In which case they STOP USING THE OLD PART NUMBER.
Which a year later Tesla not only still isn't doing- they're still using the old one in the majority of their trims.
Which is nonsensical if the "new" part can fully replace the old one.
Because as you pointed out- continuing to do that adds cost and complexity to many many parts of their entire chain.
No... math and basic logic.
If the 990 is cheaper and "just as good or better" they'd use it in ALL cars- because it's CHEAPER AND JUST AS GOOD OR BETTER. Plus save the cost and trouble of having an additional DU that's more expensive and no better.
For double the time and nearly 10x the miles they tested the 980 in production? But apparently not testing ANY in ANY other trim but the LR AWD?
Again your grasping for increasingly nonsensical and complex explanations over the simple and obvious ones keeps getting funnier.
Why in the world would they need to be?
Why? It'd essentially be a very very simple, and (data wise) tiny database.
In fact you can make it even MORE efficient by just having a single database per model year since you know the model year...FROM THE VIN.
There's a gateway config file on each Tesla that tells it exactly what options the car has available. It changes as options change. The car knows if you have FSD or not. The car knows what wheels you have (well, what wheels the CFG file says you have), the car knows if it's a P+, P- or non P, and on and on. Why would it ALSO knowing "980 or 990" somehow NOT easily be done the exact same way?
Not if it's significantly cheaper, no it's not nonsensical at all.
Especially if the difference is simply LEAVING OUT some parts.
Specifically it's been suggested fewer MOSFET gets you a less capable, cheaper, rear DU that adds relatively LITTLE cost from a manufacutring perspective since everything ELSE IS THE SAME.
It just costs less, and can handle less power.
Which would mean you'd only use it in... the LR AWD.
Which is exactly what is happening
It really, really, does not.
If it did the 980 wouldn't still be in use on all cars requiring more rear DU power than the LR AWD over a year after it came into use.
Uh...Elons binning quote that was disproven in real world testing was a couple months before production.... not a couple years.
So again the idea they're still "testing" a year later is hilarious nonsense.
How large a wager would you care to make on this, specifically?
