Model 3 Motors on the Tesla Parts Catalog

[Knightshade]

My guess is that they binned motors early on in the older builds because it was necessary (consistent quality want there yet). Now that they have built so many of these motors and are on newer revisions, the need for binning is not necessary, as all 980s can now support the demands of the P configuration. They can replace any older 980 (binned/P or not) with a newer one knowing that it will take the power.

But the 980 never required a VIN to order a replacement.

So how did they know the "replacement" one was "special" if you were ordering it and didn't let them do the special handshake wink-wink that you had a P and needed a "special magical somehow hidden binning but no evidence ever of it actually happening" 980?

If all service parts were the higher performing motor then there would no need for a VIN or additional tracking outside of Fremont (beyond checking on upgrading D's to P's at the delivery center.

So not only would there need to be a special magic binning system that nobody has evidence of and isn't reflected anywhere in any catalog or service listing- they also need to insure that ONLY the "special" ones are ever used as replacement parts?

Also apparently, somehow, every person who got a delivered-as-AWD software flashed to a P because that's what they actually wanted/ordered when they showed up to get it, all just HAPPENED to have been lucky enough the car had one of these magic special nobody knows how to detect em 980s in it?

Oh and then after doing all of that for 18 months someone slapped their forehead and said "OH HEY MAYBE WE SHOULD USE A SECOND PN FOR THE INFERIOR-BINNED ONES"?

If their magic hidden system worked great- or even existed- why would they do that?


Occams razor man.

There was never any binning.

Elon said it before they'd ever actually done consumer production runs of Ps and discovered all 980s were "good enough" and it'd be a massive waste of time and money to "bin"... (which everyone else in MFG, who DOES bin, does by using DIFFERENT PNs to distinguish parts of different capabilities.


Every alternative to that is massively more complex, with 0 evidence, and usually falls apart when exposed to any critical thinking.


Same with the 990. It's a cheaper, less efficient, less max-power-capable motor.

If it wasn't cheaper it wouldn't exist, they'd just keep using 980s like that did for the first 18+ months in all cars.

If it wasn't less capable they'd use it in the P since it has to be cheaper to be worth existing.

And the EPA data tells us it's less efficient for a fact.

 

[Not Sure]

I'm really not sure

 

[t-rizzle]

What motor is used in the front of P cars? I have a LR AWD, 4/19 build, and I have the 990 rear motor and 960 front motor. I confirmed this two weeks ago when Mobile Service was here for tire rotation. Also, what motor are they putting in the S / X Raven for the front motor, the 980, 990, or 960?

 

[Knightshade]

What motor is used in the front of P cars?

The same as every other model 3 ever.

There's only 1 front motor, the 960.

Also, what motor are they putting in the S / X Raven for the front motor, the 980, 990, or 960?

None of the above.

It's listed as:


ASY, DRIVE UNIT, RAVEN, MS
1478000-00-D

and


ASY, DRIVE UNIT, RAVEN, MX
1478000-01-D

respectively.

 

[AlanSubie4Life]

(parts catalog only shows one inverter #, but that doesn't mean much)

No, it doesn't mean much, as you say. They have a history of hiding parts, like they did on & off with the remanufactured motors and other motors. I think there are probably two different inverters, but of course I'm just guessing, and being biased by my priors.

Still the most interesting to me is that there are clearly two different remanufactured powertrain units (of course you would expect there to be two) - and definitely they are not associated with the 970 since they are MOSFET units. But the interesting thing is the description seems to add information. And clearly the part description includes information on the inverters, not just the motor. So that makes me think the 980 is 840(A) and the 990 is the 630(A). I'm using artistic license with the (A). I brought this up before, though; it is not new information. Here are the RMN (remanufactured) parts.

 

[pdx_m3s]

But the 980 never required a VIN to order a replacement.

So how did they know the "replacement" one was "special" if you were ordering it and didn't let them do the special handshake wink-wink that you had a P and needed a "special magical somehow hidden binning but no evidence ever of it actually happening" 980??

Perhaps by the time replacements were needed, they were on a revision of the motor that didn’t require binning (i.e. all the replacement 980 motors were sufficient for P demands).

 

[mongo]

No, it doesn't mean much, as you say. They have a history of hiding parts, like they did on & off with the remanufactured motors and other motors. I think there are probably two different inverters, but of course I'm just guessing, and being biased by my priors.

Still the most interesting to me is that there are clearly two different remanufactured powertrain units (of course you would expect there to be two) - and definitely they are not associated with the 970 since they are MOSFET units. But the interesting thing is the description seems to add information. And clearly the part description includes information on the inverters, not just the motor. So that makes me think the 980 is 840(A) and the 990 is the 630(A). I'm using artistic license with the (A). I brought this up before, though; it is not new information. Here are the RMN (remanufactured) parts.
View attachment 496532

That would sure line up with each MOSFET section bring rated to 210 A.

 

[CMoZ]

So, I’ve only skimmed this thread thus far and to preface my current position

• I don’t believe the existing info we have definitively proves the 990 and 980 are different
• My guess at this point is that the 990 is the same or better ( cheaper, more reliable and/or higher performance ceiling )
  
• The only way to settle this is bench testing a 980 and 990 in isolation

So the question I have is given that the actual acceleration curve (hp/torque) is a function of software design ( otherwise it would be something similar to an exponential curve in all cases and pretty unpleasant ) if the 990 and 980 have different capabilities how is the software managing that there’s not 1 expected motor in the LR AWD? This could be needlessly complex depending on earlier design decisions

Eg

• How does the software/firmware tell what motor is in the car
• What happens if a motor has to be replaced and it’s replaced with the wrong one

 

[Knightshade]

The VIN tells them every single part number that went into the car at the factory- so that's a complete non issue.

The computer knows what motor is in the car... (just like it knows if your car is a P+ or P- or a non-P....and just like it knows if you have FSD or not, or AP or not, and every other hardware and software PN associated with the bill of materials for any specific VIN)


It also eliminates the chance of Tesla putting the 'wrong' motor in at repair time.


The 980 and 990 being the same makes zero sense for any number of reasons.

If it was just an overall improvement tweak for the existing 980 there wouldn't be a 990. There'd be a 980-H (current rev is 980-G)

The reason it's a different PN is to indicate it's a different part

Exactly to avoid someone getting the wrong replacement part.

(This is also the same reason why the oft suggested idea there's a "special binned" 980 but having the exact same PN as the "normal" one makes zero sense and runs counter to the whole reason you HAVE part numbers and BOMs when manufacturing things)

 

[CMoZ]

The VIN tells them every single part number that went into the car at the factory- so that's a complete non issue.

It also eliminates the chance of Tesla putting the 'wrong' motor in at repair time.

this is laughable for a couple reasons. So background I spent the start of my working life in an auto shop while I was working my way through school and I've been a Software Engineer for 15 years so I have personal experience in both worlds Tesla has a foot in.

Wrong parts especially when varying from year to year get put in all the time for all sorts of reasons, some are mistakes some are supply issues but it happens A LOT.

The 980 and 990 being the same makes zero sense for any number of reasons.

That's an opinion and I disagree with it. and "Same" is a loaded word here. I've been very pointedly saying at least functionally the same

The motors should not be the "Same" if they have different part numbers they have a difference. It could be something as small as a different paint code used on a painted part to something as big as a casing redesign or like you contend a different performance profile.

Sense doesn't even come into it. Reason does and there could be many reasons for Tesla to slap a different part number on it

• Perception - Because a lot of people don't seem to understand the differences between EVs and ICE cars a large segment of people seem to feel ripped off if they have a car that is physically every bit as capable as the P but they can't access some of that performance. This would be a really good marketing reason to slap a different number plate on motors going into the LR AWD even if there was no difference and while I don't think it's literally identical this is still a good theoretical reason for Tesla as a business
• Identification - It's Exactly as performant as the 980, cheaper to produce but they don't have confidence in it's reliable power band or real world failure rate yet. You can put this in the back of the LR AWD's to see how many are coming back for repairs or issues without stranding them or making your flagship 3P seem less reliable to your highest paying customers in the segment. If you put it in the RWD cars and it had a high failure rate customers in those cars would be completely stranded and press would be terrible. I'd suspect a lag of 2 or more years (maybe a % over 100K miles ) before this made its way out to the rest of the fleet.

If it was just an overall improvement tweak for the existing 980 there wouldn't be a 990. There'd be a 980-H (current rev is 980-G)

Sorry didn't realize you worked for Tesla part number assignment. This is complete speculation none of us know exactly what the Tesla strategy behind part numbering is it's all speculation at this point.

The reason it's a different PN is to indicate it's a different part

Exactly to avoid someone getting the wrong replacement part.

Yup but not only or always for just that reason. Plenty of parts in the auto world that are completely equivalent but different.

(This is also the same reason why the oft suggested idea there's a "special binned" 980 but having the exact same PN as the "normal" one makes zero sense and runs counter to the whole reason you HAVE part numbers and BOMs when manufacturing things)

Agree binning makes zero sense every industry where binning takes place assigns part numbers or identifiers after the binning occurs to identify them

there no vin requirement on the 980 or 990 so there is a very real possibility the “wrong” one could end up in the wrong car and thus the wrong motor control profile would be applied.

The software component seems to be super overlooked in this thread in particular and the importance of it can't be understated. There needs to be specific code paths that match up for any differences in the motors

So originally we would have had

• AWD P
• AWD LR (980)
• MR
• SR

Now we've branched out to

• AWD P
• AWD LR (980)
• AWD LR (990)
• AWD LR (990) with Boost
• AWD LR (980) with Boost
• MR
• SR

As well now we have increased manufacturing complexity now now we need production lines for 3 motors instead of 2 the cost of a new production line would likely be more expensive than any saving of creating a new underpowered motor

Manufacturing 101 limit sku's and inventory

 

[dmurphy]

If it was just an overall improvement tweak for the existing 980 there wouldn't be a 990. There'd be a 980-H (current rev is 980-G)

Yes and no ...

Just a tweak, as in, running improvements or, say, supplier part changes etc - absolutely. That'd be a rev.

But if it were, say, a different design - i.e. "Hey, we could do this with 3 MOSFETs and not 4" - the performance characteristics might be the same as a 980, but it gets a different part # because servicing it would be different.

We just don't know what the characteristics and capabilities are of each. That said, it's fairly unlikely we ever will. It would take some serious software changes to treat a 990 like a 980, and that may or may not make sense.

It's cool to speculate, but I don't think in a practical sense, we'll ever know much of a difference between the two ...

 

[CMoZ]

Yes and no ...

Just a tweak, as in, running improvements or, say, supplier part changes etc - absolutely. That'd be a rev.

But if it were, say, a different design - i.e. "Hey, we could do this with 3 MOSFETs and not 4" - the performance characteristics might be the same as a 980, but it gets a different part # because servicing it would be different.

We just don't know what the characteristics and capabilities are of each. That said, it's fairly unlikely we ever will. It would take some serious software changes to treat a 990 like a 980, and that may or may not make sense.

It's cool to speculate, but I don't think in a practical sense, we'll ever know much of a difference between the two ...

If someone wants to send me one of each I'll bench test them it's pretty easy to tell that way. No promises on timeline though pretty busy with work and an ICE restoration project ( Don't worry EV conversion to come on that one)

 

[dmurphy]

If someone wants to send me one of each I'll bench test them it's pretty easy to tell that way. No promises on timeline though pretty busy with work and an ICE restoration project to eventually convert to EV eventually

I suspect that's exactly what it would take.

Even better, I'd love to see someone swap a 990 into a P3D and see what happens.

With their car, of course, not mine. ;-)

 

[AlanSubie4Life]

AFAIK nobody has ever found a single model 3 of any trim with that 970 motor actually installed in it.

@Knightshade, just FYI on this. Technically you are right; the motor was not installed on a car. . I don’t remember seeing any other reports and I think even the LR RWDs I have seen which were pretty early VINs still had 980, so I can’t explain this claim.
Picture of a 970 drive unit

 

[MP3Mike]

Even better, I'd love to see someone swap a 990 into a P3D and see what happens.

With their car, of course, not mine. ;-)

The drive units are cryptographicly linked to the car so you can't even swap in one of the same model.

 

[dmurphy]

The drive units are cryptographicly linked to the car so you can't even swap in one of the same model.

Yup... which is why I said above, we'd need a ton of software changes, most of which we're very unlikely to ever see.

 

[scubastevo80]

So, I’ve only skimmed this thread thus far and to preface my current position

• I don’t believe the existing info we have definitively proves the 990 and 980 are different
• My guess at this point is that the 990 is the same or better ( cheaper, more reliable and/or higher performance ceiling )
  
• The only way to settle this is bench testing a 980 and 990 in isolation

I think most of us are of the opinion that the 980 motor has a higher performance ceiling given it is still used in the performance models and the 990 is used in the regular dual motor models. It is speculation that the 990 is more reliable or cheaper, although I suspect it's cheaper given it isn't used in the performance model and would arguably use compenents that don't need to handle the higher current that the P does.

 

[Knightshade]

this is laughable for a couple reasons. So background I spent the start of my working life in an auto shop while I was working my way through school and I've been a Software Engineer for 15 years so I have personal experience in both worlds Tesla has a foot in.

Wrong parts especially when varying from year to year get put in all the time for all sorts of reasons, some are mistakes some are supply issues but it happens A LOT.

I can't speak to your experience. But I can speak to my own which includes dealing with a lot of manufacturing for several of the largest electronics companies in the world among others.

Every complete unit built has a detailed BOM. It lists every single individual PN that went into the thing and that's the bible the MFG line uses to build the thing.

So the "wrong" PN going in at the factory is impossible short of multiple individuals AND automatic systems all screwing up in exactly the same way (since several check each other)

As to repair, again, it'd require the person who takes the service call to pull up the finished products serial # and somehow select the wrong part, AND it'd still have to be a part that came in the BOM or it won't let them order it for warranty purposes (without a manual intentonal override)- AND require the actual repair tech not catching it either despite the fact they're ALSO supposed to check. So again it'd require multiple separate mistakes by different people/systems.



I don't "know" if Tesla has the same # of checks and balances on replacement parts for warranty repairs- but I have no doubt at all they use a BOM system for original manufacturing.


So I dunno what crap companies with no checks or balances or industry-standard MFG practices you've worked for... but they are the ones who seem pretty laughable.

That's an opinion and I disagree with it. and "Same" is a loaded word here. I've been very pointedly saying at least functionally the same

The motors should not be the "Same" if they have different part numbers they have a difference. It could be something as small as a different paint code used on a painted part to something as big as a casing redesign or like you contend a different performance profile.

If the part is physically and functionally the same, but they say put in a slightly "better" something- as in it was intended to replace the old part without changing the fit or reducing capability- they would slap a letter change on it. Not a new PN.

Because as you noted, running 2 PNs costs money (and time and complexity).

And if that "new" part is BETTER, they stop using the OLD one when they run out of it.

The idea they had SO MANY spare 980s they could put them in ALL Ps, SRs, and even the RWDs they sold for a bit, for an entire year, is, again, a lot more "laughable" than anything "the parts are actually different" side has suggested.


Tesla sold ~300,000 Model 3s in 2019.

Even if 50% were LR AWD that would suggest Tesla has 150,000 spare drive units sitting around taking up space and money at the start of 2019.


Laughable indeed.

Perception - Because a lot of people don't seem to understand the differences between EVs and ICE cars a large segment of people seem to feel ripped off if they have a car that is physically every bit as capable as the P but they can't access some of that performance. This would be a really good marketing reason to slap a different number plate on motors going into the LR AWD even if there was no difference and while I don't think it's literally identical this is still a good theoretical reason for Tesla as a business

Outside of a tiny # of people on the forums I doubt you could find any significant # of owners who know what a 980 is. Or a 990. Or have any idea WTF we're talking about here.

So again your explanation seems the laughable one here.

Indeed the more mass market the car gets the lower the already tiny % of people who know or care about this stuff gets.

Identification - It's Exactly as performant as the 980, cheaper to produce but they don't have confidence in it's reliable power band or real world failure rate yet. You can put this in the back of the LR AWD's to see how many are coming back for repairs or issues without stranding them or making your flagship 3P seem less reliable to your highest paying customers in the segment. If you put it in the RWD cars and it had a high failure rate customers in those cars would be completely stranded and press would be terrible. I'd suspect a lag of 2 or more years (maybe a % over 100K miles ) before this made its way out to the rest of the fleet.

And even more laughable.

For one- a failed rear motor leaves you stranded in a Model 3. Even AWD.

One of the many threads discussing it including someone who actually had it happen-

Myth: AWD able to drive if one motor fails?


Failing front motor may still drive- not so rear.


And of course if the motor had problems they'd prefer to put it in the car they sell far fewer of- which is the P, not the LR AWD.

(if it wasn't it'd make your "they had 100,000+ spare 980s in a box" idea even more hilarious)

Sorry didn't realize you worked for Tesla part number assignment. This is complete speculation none of us know exactly what the Tesla strategy behind part numbering is it's all speculation at this point.

No, it's not.

It's having noticed they've already revved the 980 motor enough times to get up to the letter G.

They didn't switch to a new PN any of those times they revved it. They wouldn't suddenly add cost and complexity for something as simple as part rev. A DIFFERENT PN means a DIFFERENT part.

Yup but not only or always for just that reason. Plenty of parts in the auto world that are completely equivalent but different.

Most commonly that happens when they're going on different model cars (especially different brands made by the same company)- for example a Toyota and a Lexus might use the "same" oil filter- but there's a PN for "this oil filter shipped in a toyota box" and "this oil filter shipped in a Lexus box"

Because the PHYSICAL COMPLETE UNIT (which in this case is Filter+Box, maybe even docs if they include any and they're different for lex/toy or language/country specific) IS DIFFERENT.

No such thing here-- they're all Tesla parts, and all going into a Model 3.

And even in the above case the parts system will usually show the sub-units that make up the BOM of the parts- and you'd find the actual filter part is the same PN and the box # is different

there no vin requirement on the 980 or 990 so there is a very real possibility the “wrong” one could end up in the wrong car and thus the wrong motor control profile would be applied.

Not if Tesla is doing it, no.

No VIN is required because if you're ordering a 980 you'll get whatever the latest 980 is. Doesn't matter what car the original 980 came from. So "dude off the street" says he wants to order a 980- no problem, they order him a 980. No VIN needed, they don't care why rando dude wants one.


But when you go TO order a replacement rear (as Tesla for a repair) you will enter the VIN because the repair itself will be associated with the VIN.

That VIN will then pull the BOM parts to show you the correct motor to order.

And the system shouldn't (without manual override of some kind) even allow you to order a part that didn't come on the vehicle as a REPAIR part.

The software component seems to be super overlooked in this thread in particular and the importance of it can't be understated. There needs to be specific code paths that match up for any differences in the motors

So originally we would have had

• AWD P
• AWD LR (980)
• MR
• SR

Now we've branched out to

• AWD P
• AWD LR (980)
• AWD LR (990)
• AWD LR (990) with Boost
• AWD LR (980) with Boost
• MR
• SR

Wrong AGAIN.. the 980 code could have profiles for all cars with a 980, and just pick which to use by VIN since they know what motor came in it.

In fact we can be pretty SURE they do it that way because the Boost update happens instantly in the car. They just flip a flag to add the software PN to your VINs BOM.... it doesn't "download a new profile"

Ditto the 990.

Also the P3D+ and P3D- have different code (the warning when turning on track mode is different for example, so the software "knows" which car it's being run on).

Manufacturing 101 limit sku's and inventory

So... you're saying you want to limit SKUs--- but somehow it makes sense they ADDED a SKU for the same part....oh, AND to limit inventory....but they somehow had an entire year of well over 100,000 spare 980s sitting around start of 2019?


I agree there's some laughable stuff being posted in this thread....but not on whom it's being posted by.

 

[t-rizzle]

If someone wants to send me one of each I'll bench test them it's pretty easy to tell that way. No promises on timeline though pretty busy with work and an ICE restoration project ( Don't worry EV conversion to come on that one)

I bet if everyone on this thread donated a couple bucks we could buy some motors from salvaged cars

 

[dennis_d]

So, who's going to be the first person to do a RWD -> P-AWD DIY mod?

Sounds to me that it's pretty easily doable (assuming you can figure out the software side of things). I guess, worst case scenario you can rip the ECUs out of a wrecked P-AWD and add the front drive unit/inverter/etc. Either way, pretty intensive. I bet it'll be done though.

As far as I can tell, the main differences are (from LR RWD -> P-AWD):
* Added front drive unit and inverter stack
* Misc HV cabling to handle all the FRU
* New front axles to replace non-driven ones
* Software?

Honestly, further down the road I bet it wouldn't be that bad of a mod. All you need is a P-AWD that is totaled in a back end collision. Should keep the main bits you need in tact.

RWD units have completely different battery pack (see part number) , it missing one connector from front DU