P3D is just a software upgrade

[Knightshade]

I think the P is just a software unlock for launch control. AWD to me feels the same as a P from 60 to 90mph. I am very interested to see a test from 60 to 100 in a P vs a AWD.

I really think the P just dumps more energy from 0mph to get thr faster 0-60 and once around 60mph is reached the AWD and P is about the same.

Btw, AWD from 60mph+ is way faster than the RWD. RWD feels like a turtle now when I punch it.

We already have pretty solid evidence of that.

The 1/4 mile runs of the P vs the RWD and the AWD that are out there.

They show the 1 second difference at 60 mph is still about a 1 second difference at ~110 mph.... and that the trap speeds of the P and AWD are quite close to each other.

Suggesting nearly all the advantage on the P happens below 60 mph.

The RWD on the other hand the time gap gets even larger between it and the AWD than it already is at 60, and the trap speeds are much lower than either AWD or P.

 

[Canikony]

You can really tell the AWD is nerfed from 0-30. I've also noticed that if you sort of do a double tap, it feels like it accelerates harder with the AWD.

 

[mongo]

We already have pretty solid evidence of that.

The 1/4 mile runs of the P vs the RWD and the AWD that are out there.

They show the 1 second difference at 60 mph is still about a 1 second difference at ~110 mph.... and that the trap speeds of the P and AWD are quite close to each other.

Suggesting nearly all the advantage on the P happens below 60 mph.

The RWD on the other hand the time gap gets even larger between it and the AWD than it already is at 60, and the trap speeds are much lower than either AWD or P.

Perhaps above 60 MPH, the pack output is the limiting factor? Below that point, P could put out more torque/ power at the same RPM/ MPH based on its max current/ torque setting.

 

[voip-ninja]

The only way this question is ever going to get answered is when someone finds out what part #s the motors are in the P3D- cars vs. the AWD cars.

If they aren't different part #s then P3D- is nothing but a software uncorking.

It could be that P3D+, which will get track mode, has the bin sorted motors, but again, it would take having parts #s to verify it.

 

[nwdiver]

The only way this question is ever going to get answered is when someone finds out what part #s the motors are in the P3D- cars vs. the AWD cars.

If they aren't different part #s then P3D- is nothing but a software uncorking.

It could be that P3D+, which will get track mode, has the bin sorted motors, but again, it would take having parts #s to verify it.

I would expect they WOULD be different part numbers AND just software. A binned component is virtually identical but will still have a different part number.

 

[mongo]

The only way this question is ever going to get answered is when someone finds out what part #s the motors are in the P3D- cars vs. the AWD cars.

If they aren't different part #s then P3D- is nothing but a software uncorking.

It could be that P3D+, which will get track mode, has the bin sorted motors, but again, it would take having parts #s to verify it.

Disclaimer: I realize this would be a pain in the neck:

Tesla can look up vehicle type by calling up the configuration in their database with the full VIN as the key.
Tesla can look up motor performance data by calling up the test results in their database by using the drive unit serial number as the key.

So, technically, they do not need a separate part number. Service center orders a drive unit for VIN xxxx. Fremont has GF1 send appropriate type of unit, service center installs it. They need to track SN anyway for warranty, so the only annoying part is keeping the units sorted while on the shelf to avoid pulling the wrong type and having to put it back.

They could also encode performance in the serial number itself. Like how organic vegetable price lookup codes are just the normal code with a leading 9.

 

[David L]

It's also possible that they are binning and tracking motors after they get installed and that some non-P D cars have motors that are capable of the higher P-level output. Then when a P3D- order comes in, they just unlock one of the non-P D cars with a high binned motor. I imagine their practices will continue to be pure speculation unless an insider leaks evidence one day. Or if Elon goes on a Twitter rant and (accidentally) decides to share with us....

 

[nwdiver]

It's also possible that they are binning and tracking motors after they get installed and that some non-P D cars have motors that are capable of the higher P-level output. Then when a P3D- order comes in, they just unlock one of the non-P D cars with a high binned motor. I imagine their practices will continue to be pure speculation unless an insider leaks evidence one day. Or if Elon goes on a Twitter rant and (accidentally) decides to share with us....

Everyone keeps saying ‘motor’ but you mean inverter... right? I’ve never heard of binning motors.

 

[mongo]

Everyone keeps saying ‘motor’ but you mean inverter... right? I’ve never heard of binning motors.

Drive unit (combined motor and inverter). The rotors are balanced, and bearing load increases as the square of speed. Also winding/ magnet/ lamination variability.

 

[nwdiver]

Drive unit (combined motor and inverter). The rotors are balanced, and bearing load increases as the square of speed. Also winding/ magnet/ lamination variability.


View attachment 341936

Yeah... pretty sure it’s the inverter part that gets binned....

 

[mongo]

Yeah... pretty sure it’s the inverter part that gets binned....

What would they use for a load? Motor winding connections are bolted after the motor/ inverter mating.

 

[nwdiver]

What would they use for a load? Motor winding connections are bolted after the motor/ inverter mating.

They might be binned after assembly but the difference in performance is a result of the inverter not the motor. My understanding would be this is very similar to computer processors and solar panels. All solid state devices where small molecular differences can cause a measurable performance difference.

But the inverter could easily be binned prior to assembly in the same way processors are.

 

[mongo]

They might be binned after assembly but the difference in performance is a result of the inverter not the motor. My understanding would be this is very similar to computer processors and solar panels. All solid state devices where small molecular differences can cause a measurable performance difference.

But the inverter could easily be binned prior to assembly in the same way processors are.

Yah, it could. They could also have STMicro bin the SiC parts before PCB assembly.

I would say CPUs are a little different since they are stand alone parts. The drive unit as a whole is the item whose performance matters. (Parallel would be CPU+heatsink + motherboard + memory + power supply and testing the overclocking of the entire system).
My feel is that they have an end of line test for the drive units anyway, so they use that as the final efficiency/ burn in check. That would be somewhat redundant to an inverter only burn in. I do expect they have a basic functional inverter check post PCB assembly.

 

[Nocturnal]

@JeffK that's really weird - I have a P3D- and my monroney sticker (which was emailed to me as a PDF and never received in print form) is almost exactly the same as yours, except at the top left where yours says Performance in bold, mine says "Long Range AWD". However, in the right-hand column where it lists charges, it lists Performance Dual Motor All Wheel Drive. Weird.

Sidenote, I can't understand why no one can imagine that it's quite possible that Tesla IS binning and using the top-performing parts for performance cars, AND occasionally installs the wrong (AWD) software on these performance models. Just because they can flip a switch to uncork those cars doesn't mean those cars didn't also have superior hardware to begin with.

Last note, why is this such a big deal? If you bought the performance, your car is faster, whether it's software or not. Who gives a *sugar* otherwise? Ugh what a waste of time, can't believe I typed this much even

That was my thought. Correct hardware, just that some were flashed as AWD rather than P.

 

[voip-ninja]

Disclaimer: I realize this would be a pain in the neck:

Tesla can look up vehicle type by calling up the configuration in their database with the full VIN as the key.
Tesla can look up motor performance data by calling up the test results in their database by using the drive unit serial number as the key.

So, technically, they do not need a separate part number. Service center orders a drive unit for VIN xxxx. Fremont has GF1 send appropriate type of unit, service center installs it. They need to track SN anyway for warranty, so the only annoying part is keeping the units sorted while on the shelf to avoid pulling the wrong type and having to put it back.

They could also encode performance in the serial number itself. Like how organic vegetable price lookup codes are just the normal code with a leading 9.

That's a heck of a lot of assumptions.

It also means that you have great faith in Tesla replacing a part on a P3D- car and the technicians request a part and Tesla back-end systems are smart enough to fulfill the replacement request with an appropriately binned part.

It also means that if an error is made in gathering a 'performance' version motor, that has the same part # that the owner can suffer premature failure of the part. The motors are only warrantied for eight years, which really isn't that long in the grand scheme of things.

The smart way to do this is to bin the motors and put different part #s on them so that there is no confusion or room for error when the car is being produced or a part is being replaced.

If the part #s are the same I would have to assume that Tesla are in fact not binning, or that Musk was sincere that they would be binning the parts but it turned out all motors that went into production were able to pass the stricter burn in tests so they ultimately just used the same parts for all cars and determine P vs non-P by software.

 

[Msjulie]

I want my binned motor <- actually I don't care but if Tesla offers a software unlock for AWD a bunch of folks who wished they could have ordered their P with 19" OEM wheels will be bumming.. artificial silliness those options of aero or 5k for rims you may not want..

I just want my car eligible for any additional software-induced speed improvements (Tesla software upgrades are a gateway drug)

 

[mongo]

That's a heck of a lot of assumptions.

It also means that you have great faith in Tesla replacing a part on a P3D- car and the technicians request a part and Tesla back-end systems are smart enough to fulfill the replacement request with an appropriately binned part.

It also means that if an error is made in gathering a 'performance' version motor, that has the same part # that the owner can suffer premature failure of the part. The motors are only warrantied for eight years, which really isn't that long in the grand scheme of things.

The smart way to do this is to bin the motors and put different part #s on them so that there is no confusion or room for error when the car is being produced or a part is being replaced.

If the part #s are the same I would have to assume that Tesla are in fact not binning, or that Musk was sincere that they would be binning the parts but it turned out all motors that went into production were able to pass the stricter burn in tests so they ultimately just used the same parts for all cars and determine P vs non-P by software.

If the drive unit has a calibration for max current (set after burn in testing along with the serial number), you would not be able to over drive it.

Like I said, it would be a pain in some ways, but basing the bin/not-binned on the assumption of a different part number is itself an assumption and does not allow for the use of a distinct digit in the serial number to provide clear differentiation.

 

[Nocturnal]

I think Tesla should give a free 30 day trial Performance software unlock to AWD NonP to entice them to upgrade to P. I’d sure enjoy that trial period

Sure, at long as I get 30 days of ludicrous mode as a trial.

 

[sperkin]

We already have pretty solid evidence of that.

The 1/4 mile runs of the P vs the RWD and the AWD that are out there.

They show the 1 second difference at 60 mph is still about a 1 second difference at ~110 mph.... and that the trap speeds of the P and AWD are quite close to each other.

Suggesting nearly all the advantage on the P happens below 60 mph.

The RWD on the other hand the time gap gets even larger between it and the AWD than it already is at 60, and the trap speeds are much lower than either AWD or P.

I saw that as well. Just wasn’t sure if it’s just me feeling it wrong or others felt the same. My Butt Dyno says AWD is the same as P at highway speeds -hella fast

 

[Nocturnal]

Nobody complains about paying 5k for EAP, and that's pretty obviously all software.

It's pretty rare for any car to get to add the amount of performance either RWD->AWD or AWD->P get you for the amount they charge in the model 3, regardless of how much of either is HW vs SW.

That wouldn't be materially different than a car company using software to only allow a stock turbo to produce X lbs of boost or detuning the same engine in a down market variant.