All LR motors are the same- MR one is NOT

[derotam]

Well now you are making assumptions about what the parts are for... Some parts in the catalog specifically state which version of vehicle they are for. Tell me where the DU's actually specify what vehicle they are for, whether we are right or wrong in making our assumptions about things in the catalog, it still remains fact that some things have more specific wording than others. If there is one part listed in the catalog, at this point, unless it specifically states it is for an LR, MR, SR, LR(AWD), Performance, whatever, then I say one can't necessarily definitively say one way or the other.

edit: removed statement about onboard charger because I looked at wrong section of part catalog.

 

[AlanSubie4Life]

Yes.

So it's not like that part said "DU for non P cars" and they just left out the P one from the catalog. There is no P one. Which was a major point of debate for a long time on here.

There's nothing that would prevent them from having tightened the specs on the SC MOSFETs over time (and that is what Elon sort of said - though it was misleading), so it's possible the early RWD vehicles (and possibly earlier AWD) have less well screened (SC MOSFET not capable of same output current) drive units. This would not require a separate part number. It would also avoid the issue of which drive unit would be used for a service replacement. You'd just get a better one than the original (they're probably only shipping & using more tightly screened units now). But the installed base of RWD (and maybe AWD) vehicles might be less capable than the newer vehicles in terms of max output (current) due to original screening limits.
This may not be what happened at all, I'm just trying to establish consistency with Elon's tweet, with the assumption of truthfulness.

In this way you're probably right that there's not a specific P drive units. But it does not necessarily follow that even early build LR RWD vehicles have equally capable drive units. Some might, some might not, depending on where they fell on the original distribution with the original screening limit.

All wild speculation of course, and we really don't know unless Tesla helps us out.

 

[Knightshade]

There's nothing that would prevent them from having tightened the specs on the SC MOSFETs over time (and that is what Elon sort of said - though it was misleading), so it's possible the early RWD vehicles (and possibly earlier AWD) have less well screened (SC MOSFET not capable of same output current) drive units. This would not require a separate part number.

Sure it would.

There had to be a replacement PN for the P, different from the RWD/AWD, during the time they had any units being made that didn't meet P specs.


Otherwise a P could end up with a replacement DU that didn't make P specs if it failed before they "all" got up to that spec.


Since that PN doesn't exist it tells us there's only ever been one rear DU in production.


(the catalog does in contrast list the springs/shocks that were original production on RWD that were replaced in December 2017- and also lists the springs/shocks that RWD used from 12/17 through 6/18, when they were replaced a second time.

In this way you're probably right that there's not a specific P drive units. But it does not necessarily follow that even early build LR RWD vehicles have equally capable drive units. Some might, some might not, depending on where they fell on the original distribution with the original screening limit

I suppose it's possible RWD cars could be covered by this- if the rear DU was up to P specs in all cases by the time the P was first produced.

But since P and AWD began production at the same time it's not possible for it to be the case on AWD cars. (otherwise you have the same issue I raised earlier on replacing a delivered-bad DU on a P in early days)

 

[Knightshade]

Well now you are making assumptions about what the parts are for... Some parts in the catalog specifically state which version of vehicle they are for.

I did, in the original post in the thread. Did you not read it?

Tesla has since removed that text, but tons of folks saw it at the time and I directly quoted what they said there.

It specifically said the one rear DU was for all cars with the 75KW LR battery, and the front was for all dual motor cars.

Which is how we know the "new" DU is for the MR/SR, RWD only, cars (and why the MR/SR cars have the same 0-60).

 

[AlanSubie4Life]

Sure it would.

There had to be a replacement PN for the P, different from the RWD/AWD, during the time they had any units being made that didn't meet P specs.

What I indicated is that they tightened the spec at a particular point in time and scrapped anything that didn't pass. During the transition to tighter screening they'd need to be careful with their inventory but it seems manageable.

(the catalog does in contrast list the springs/shocks that were original production on RWD that were replaced in December 2017- and also lists the springs/shocks that RWD used from 12/17 through 6/18, when they were replaced a second time.

Presumably that is because these particular springs/shocks/sway bars/etc. parts are a different design that changed with time. Screening drive units to a different level isn't necessarily a difference in design. Component part tolerances just may have improved.

I suppose it's possible RWD cars could be covered by this- if the rear DU was up to P specs in all cases by the time the P was first produced.

I don't really follow this. But I think you're saying what I was saying - their production process improved over time and they can screen to P specs with minimal yield loss.

I'm saying I *speculate* there's only one motor meeting one set of specs, that they've been making since the advent of the P. And it's an identical design to earlier motors, just different screening methodology. Which is kind of what Elon said.

This is speculation, as I said. Waste of time, mostly.

 

[Knightshade]

What I indicated is that they tightened the spec at a particular point in time and scrapped anything that didn't pass. During the transition to tighter screening they'd need to be careful with their inventory but it seems manageable.

It wouldn't be unless by the time the P and AWD were in production, all motors were P-grade.

I don't really follow this. But I think you're saying what I was saying - their production process improved over time and they can screen to P specs with minimal yield loss.

I'm saying I *speculate* there's only one motor meeting one set of specs, that they've been making since the advent of the P. And it's an identical design to earlier motors, just different screening methodology. .

But you suggested maybe early RWD and AWD cars got a below-P-level tested motor.

That's not possible with 1 part number for the AWD. It could be for the RWD I suppose.

If they stuck with a single part, they had to have every example of that part in inventory up to P spec by the time the very first production P shipped (which was the same time or slightly earlier than the very first production AWD shipped)

Otherwise you'd risk a P getting a not-up-to-P replacement part.

Hope that clears it up?

So while it's possible very early RWDs got a below-P spec rear motor, it's not possible any AWD cars did (and likewise not possible any AWDs got a below-P spec front motor) based on only 1 PN existing for front DU and only 1 Pn existing for rear DU.

 

[AlanSubie4Life]

I
Hope that clears it up?

So while it's possible very early RWDs got a below-P spec rear motor, it's not possible any AWD cars did (and likewise not possible any AWDs got a below-P spec front motor) based on only 1 PN existing for front DU and only 1 Pn existing for rear DU.

Yes, that's what I thought you might be saying. Yes, it does seem that most AWD cars would have P-spec (rear) DU.

But, it's still speculation - you assume:
1) They actually didn't build *any* AWD vehicles before they built P vehicles. I think it's hard to guarantee this didn't happen. Maybe there is info out there to prove one way or the other, but I have no idea. I think we know P vehicles were delivered first? I don't even know that, but I have heard it.
2) They don't have a way to track which motors are tested to P-spec and which ones aren't. Assuming each DU has a serial number (I have no idea how/if they're marked), it doesn't seem inconceivable that they have an inventory program/tool that identifies which test program a DU was tested with when it is scanned prior to installation. Maybe they have no such thing, I have no idea.
3) They have very strict rules for when they have to have a new part number. You're assuming a different test program with zero design change means a different part number for Tesla. I've been in places producing millions of units, where this is not the case. Just have to be careful...

All of the above may be true, but I'm speculating, and I think you're speculating as well. Which is fine!

I think this is the last time I'll talk about this, as there's no way to know right now!

 

[Knightshade]

Yes, that's what I thought you might be saying. Yes, it does seem that most AWD cars would have P-spec (rear) DU.

But, it's still speculation - you assume:
1) They actually didn't build *any* AWD vehicles before they built P vehicles. I think it's hard to guarantee this didn't happen. Maybe there is info out there to prove one way or the other, but I have no idea. I think we know P vehicles were delivered first?

Yup- they did highest profit cars first. Lots of reviews and deliveries of P models weeks before the first AWD deliveries. Seems unlikely they'd build AWDs and let them sit. Or that this mystical window of "non P" DUs would just happen to have wrapped up RIGHT in that tiny window as well.

2) They don't have a way to track which motors are tested to P-spec and which ones aren't.

Right- that's why the fact there's only 1 PN tells us they are all P spec capable.

Assuming each DU has a serial number (I have no idea how/if they're marked), it doesn't seem inconceivable that they have an inventory program/tool that identifies which test program a DU was tested with when it is scanned prior to installation. Maybe they have no such thing, I have no idea.

Can you present any reason doing that would make any sense from a repair, parts ordering, or supply chain perspective- rather than using a different PN to denote a higher spec DU?

3) They have very strict rules for when they have to have a new part number.

Who is they? I've worked with quite a few major electronics OEMs and ODMs and it's not that hard at all to get another PN if there's a business reason to do so.

You're assuming a different test program with zero design change means a different part number for Tesla.

No, I'm not doing that at all.

I'm stating as a fact that replacement parts need to be correct, and that having a different PN for different spec parts is the industry standard way to do that, and nobody has suggested an alternative that isn't measurably inferior to it AND harder/more complicated to no apparent benefit.

I've been in places producing millions of units, where this is not the case. Just have to be careful...

I can't think of any factory or supply chain or repair ordering system I've ever worked with where they would ever make it possible to order a PN and get a part that doesn't meet required spec.

Which would be possible in the AWD scenario you described where they put non-P motors in AWDs after production Ps had already been built and both cars used the same DU part number.

All of the above may be true, but I'm speculating, and I think you're speculating as well. Which is fine!

I think this is the last time I'll talk about this, as there's no way to know right now!

I think the difference is you've offered no reason they'd go a much more dangerous and complex route, with no apparent benefit of any kind, versus just using another PN if there was ever a substantive performance difference.

Your explanation requires some kind of incredibly ideal timing where they magically got the rear DU perfect between RWD and P/AWD, and/or some wacky "secret second part number" system, using serial #s, because.... REASONS!

Mine simply requires MFG tolerances good enough that all motors are capable of P output. Which is hardly a stretch given how much praise every engineering tear-down of the Model 3 motor has seen.

Occam's razor. The simplest explanation is usually the right one.

There's only 1 rear DU for all LR model 3s, they're all P capable.

There's only 1 front DU for all model 3s period, they're al P capable.

No special timing needed, no special secret second PN using serial #s to encode testing, no nothing but good engineering.[/QUOTE]

 

[AlanSubie4Life]

Mine simply requires MFG tolerances good enough that all motors are capable of P output. Which is hardly a stretch given how much praise every engineering tear-down of the Model 3 motor has seen.

I agree with this! And it's consistent with what I've been saying. The AWD thing certainly you may be right - they could all be P-spec DUs.

No special timing needed, no special secret second PN using serial #s to encode testing, no nothing but good engineering.

Entirely possible, even likely. The only thing left to explain is Elon's tweet! As I said a couple messages back, that's the only reason for my speculation & getting away from the "Occam's Razor" explanation.

I'm trying to find a way to make his tweet consistent with what we think must be the case (in other words, trying to understand how this tweet could have been accurate):
"Performance drive units are lot sorted for highest sigma output & get double the burn-in."
(One question I have: "double the burn-in" implies a comparison. Double relative to what, exactly?)
Of course it's entirely possible that the tweet isn't accurate. Wouldn't be the first time. I understand that. It's just a thought experiment. The question: Let's *assume* that it was accurate - how do you explain it? That's all I'm doing here. I guess because I'm really bored.

 

[Knightshade]

Double burn in just means they run the test cycle twice as long.

It doesn't help with anything, but it looks good on paper.

You see this a surprising amount in QC- tests that don't really ever find a problem but they make it seem like they're doing something.

Highest sigma just means the "best" units go into Ps, but the difference is too small to really matter so they don't bother marking replacement units special.

It might reduce warranty costs 1% or something but it wouldn't stop them from unlocking an AWD to a P with software if they wished.

 

[AlanSubie4Life]

Double burn in just means they run the test cycle twice as long.

Sure, burn-in is very common terminology. I wasn't asking about burn-in, I was asking about the comparative nature of the statement:
...twice as long...as...they do for the AWD/RWD rear DU? Is that the comparison? I think that's what you're saying.

Highest sigma just means the "best" units go into Ps, but the difference is too small to really matter so they don't bother marking replacement units special.

Sure, there's not any confusion about sigma in regards to distributions/test limits here or what any of that means; that's basic stuff.

You're saying your hypothesis is that the better units go in the P's, but it's not a huge difference? I guess it's semantics, but I feel like that sort of doesn't mesh with the thread title "All LR motors are the same". I mean, technically one set of units (in the P) is from the better part of the distribution, and the ones in the AWD are from the lower part of the distribution (based on whatever metric they happen to choose - "output"). At least from the factory. (If that's what you're saying, which I think it is.)
To me this isn't that different than buying a processor from Intel that is marketed as being overclockable - rather than multiplier locked - Intel is able to have a supply of these parts because they select the fastest parts from the distribution for such unlocking to increase the chance that they work decently at extreme F/V/T (this is likely an oversimplification). Anyway, from Intel, they happen to have a different part number and are fused to enable/disable overclocking (again, oversimplified likely). Otherwise they're 100% identical silicon masks & designs to an equivalent "consumer-grade" CPU (might be oversimplified). Just from a better lot or place on the wafer & then fused to enable/disable overclocking. But a consumer wouldn't call them the same.

Back to DUs: whether they're labeled differently or not, I don't think if you've lot sorted them you can really call them "the same". It's definitely semantics though.

I guess to summarize: you're saying that initially the Performance units are better to lower warranty exposure, but if you have to replace a drive unit, you'll likely just get whatever they're shipping from the factory on that day (won't be guaranteed to be double burn-in or a high sigma unit)? Because all LR drive units are "the same".

I'd agree that your scenario is a good simple way to explain things. Just Elon exaggerating the necessity/benefit of the enhanced screening in his tweet, mostly. But it does mean that Performance customers got something for their money! The hardware isn't the same, initially. It's screened.

 

[Knightshade]

Sure, burn-in is very common terminology. I wasn't asking about burn-in, I was asking about the comparative nature of the statement:
...twice as long...as...they do for the AWD/RWD rear DU? Is that the comparison? I think that's what you're saying.

Right. They do it so it looks like they're doing "something" for your $11,000 even though it makes literally no difference.

Sure, there's not any confusion about sigma in regards to distributions/test limits here or what any of that means; that's basic stuff.

You're saying your hypothesis is that the better units go in the P's, but it's not a huge difference? I guess it's semantics, but I feel like that sort of doesn't mesh with the thread title "All LR motors are the same". I mean, technically one set of units (in the P) is from the better part of the distribution, and the ones in the AWD are from the lower part of the distribution (based on whatever metric they happen to choose - "output"). At least from the factory. (If that's what you're saying, which I think it is.)

Yeah, but since all motors (being there's only a single PN) already score better than needed to go into a P, so the "better" ones are likely a very tiny difference in score- and make literally zero actual performance difference- again it's mainly a marketing thing.

It might potentially reduce warranty cost on the P by some tiny fraction, and great if it does- but given the claimed 1 million mile life on the motors I doubt that'll matter much either.

To me this isn't that different than buying a processor from Intel that is marketed as being overclockable - rather than multiplier locked - Intel is able to have a supply of these parts because they select the fastest parts from the distribution for such unlocking to increase the chance that they work decently at extreme F/V/T (this is likely an oversimplification). Anyway, from Intel, they happen to have a different part number and are fused to enable/disable overclocking (again, oversimplified likely). Otherwise they're 100% identical silicon masks & designs to an equivalent "consumer-grade" CPU (might be oversimplified). Just from a better lot or place on the wafer & then fused to enable/disable overclocking. But a consumer wouldn't call them the same.

Yeah- because they're not the same.

They have a significant difference in ultimate performance and capabilities.

It's why Intel gives them their own part number.

The fact Tesla uses the same PN for all DUs is exactly the point of why that analogy doesn't stand up here.

There's 0 practical difference in the DUs in a P and a RWD and an AWD- which is why if the DU in any of them dies they get the same part to replace it.

I guess to summarize: you're saying that initially the Performance units are better to lower warranty exposure, but if you have to replace a drive unit, you'll likely just get whatever they're shipping from the factory on that day (won't be guaranteed to be double burn-in or a high sigma unit)? Because all LR drive units are "the same".

I'd agree that your scenario is a good simple way to explain things. Just Elon exaggerating the necessity/benefit of the enhanced screening in his tweet, mostly.

Pretty much.

The only alternative would be any time a service center orders a replacement P DU they somehow specify that, and then wait around for the original assembly plant to build a "new" batch and hope there's an extra that "bins" higher.

Since that's insane, what must really happen is a DU off the shelf in a warehouse gets shipped to the service center- and it's the same DU an AWD or RWD owner would get as a replacement.

And honestly they might not be binning at all anymore if their MFG tolerances have gotten that good- eliminating that process might be part of what Elon meant when he talked about reducing MFG time and costs on the Q3 call.

But it does mean that Performance customers got something for their money! The hardware isn't the same, initially. It's screened.

Except, not really.

Because it's not like the "replacement" not screened part will slow the car down.

It's functionally the same part.

So I don't see how you'd ascribe any real value to the customer when it makes no functional difference in the car they bought.

 

[AlanSubie4Life]

I think we agree on the substance of what happened. Only real difference vs. my original thinking is the width of the distributions.

Personally, if this is what is going on, I would prefer to have the P DUs, as they may have slightly higher margin to failure and less infant mortality due to extra burn-in. Assuming of course you don’t ever use the extra initial torque of the P! Anyway, it has value to me - more real value than the silly 0-60 time. So I ascribe value to it. Assuming it’s true.

Of course, I don’t have the car until this weekend, so maybe my value system will change after I get the car.

 

[brandonee916]

I went to the EPC.teslamotors.com and I can't find any differences in any of the RWD motors... it's just a single part number. On the HV Battery system side it's a different part number for MR and LR....

 

[Knightshade]

I went to the EPC.teslamotors.com and I can't find any differences in any of the RWD motors... it's just a single part number. On the HV Battery system side it's a different part number for MR and LR....

When originally listed it stated "Only for VINs with 75 kWh (E3) battery pack, Option Code BT37. Comes pre-filled with oil"

And there was, briefly, a second IGBT rear motor that is for the SR/MR cars listed- but that seems to keep dropping in/out of the catalog.

 

[brandonee916]

Maybe it's related to the system outage they're having globally... hehehe