Stop the Press! Tesla announces REAL HP numbers for P85D and P90L

[stopcrazypp]

Maybe it's semantics but isn't the car never torque limited (ie max motor torque which is not the same as max torque delivered by the motors limited by current), it's traction determines the max acceleration which is sustained until the battery limits on current draw kick in which determines when the acceleration drops?

Nope. The car is motor power / motor torque limited between ~13-25 mph.

Remember, the motor power includes the limits of motor controller and inverter (more accurately it is the drive unit limit). S85 has a 900A inverter, P85 (P85D rear) has a 1200A inverter and post-P90D rear ~1300A, 70D/85D front and rear (P90D front) ~650A. All of these limit the amount of torque the given drive unit can output independent of traction limits.

The fact that the increase in motor power from 728 hp of the P85D to 762 hp for the P90D launch raises previously measured acceleration from 1.05 G to 1.1 G shows that the tire traction limit was not reached yet under the P85D.

However, even ignoring that point about inverter limits, motors don't have infinite current or voltage carrying capabilities, nor can it physically handle infinite torque/power. I've seen claims that motors can output its rated power at all rpms, but this is false given the torque limits (and back EMF). If it was true that motors can handle infinite torque/power, there would be no need to keep a physically bigger motor for the rear for the P85D/P90D vs the 70D/85D. All Tesla would need is the smaller motor unit and then software limit it, but that obviously is not the case.

 

[rns-e]

The fact that the increase in motor power from 728 hp of the P85D to 762 hp for the P90D launch raises previously measured acceleration from 1.05 G to 1.1 G shows that the tire traction limit was not reached yet under the P85D.

The increase in motor power from 728 hp to 762 hp also includes the P85D but no performance improvement. So the limit here would be the fuse?

 

[lolachampcar]

Thank you for that - that's a point that wasn't coming out for me. Are you saying that you were expecting to be able to upgrade a P85D to a P90DL as a standard offering from Tesla? Or that you were disappointed that the P85D was released, then the P90D with L was released too soon and you would have chosen the latter if you had known about it?

Watching things unfold, I am of the opinion that the P85D was shipped with the higher speed current/power draw issue yet to be sorted. Elon indicated there would be an OTA update to provide performance at higher speeds not seen outside the factory. It would be released pending verification. That verification process drew out the weakening contractor spring and likely popped a fuse or two resulting in the need for a hardware change. Ludicrous was born and Tesla made lemonade out of lemons with a $10K Ludicrous upgrade to the P90D.

From my perspective, Tesla took three years to come out with a significant change in performance. It was not inexpensive to switch from an 18K mile P+ to the PD and my expectation was that I would be good for another three years. Circumstances beyond everyone's control has put a kink in that plan and more must be done if I am going to get to play with Tesla's top performer. It is a first world problem and "crap happens".

The irritation comes from how Tesla is handling all this. Tesla knows good and well exactly how the P85DL performs with respect to the P90DL. The only information they release is that it will be .2 seconds faster 0-60 and in the 1/4 then the current P85D. This does not explain a specification difference of 11.4 to 10.9 when comparing the 85DL to the 90DL. Battery capacity and small changes in chemistry should not make that much difference. Point five seconds is an eternity in the 1/4 and would be worth disposing of my current P85D and buying the P90DL. But wait. That 10.9 second 1/4 mile time may be a Tesla "price includes the time you spend pumping gas over the lifetime of your ICE" or "691 combined motor hp" shenanigan. The speculation that MotorTrend tested the P90DL at high density altitude and high ambient temperatures then applied ICE correction factors to take consistent 11.4 second runs and adjust them to consistent 10.9 second performance may be true. Owners are only getting 11.4 seconds out of their P90DLs and it has been too long since the release of the MotorTrend results to have cars in the field receive an OTA upgrade to bring them up to MotorTrend performance levels.

All the above has me thinking that Tesla knows the truth but chooses to present facts in a misleading way or simply choose the facts that allow them to craft a potential customer's understanding. That destroys trust which, like water under a bridge, is lost once it passes. All first world problems to be sure but a shame never the less. The cars and company have more than enough going for them then to have to rely on such tactics.

Sorry for such a long answer but a lot built up to the current state of affairs for me.




On the whole traction limited thing, we need to remember that most everyone posting here has no direct experience with high torque/hp AWD cars, myself included.

I asked for input from such owners and one posted that her/his AWD 9xx variant would loose traction on all four wheels at significant speed when the traction control was turned off. The 9xx is significantly lighter than Model S and with much wider/stickier tires. If it can blow off the tires at speed, it is reasonable to assume Model S can as well. The reality of things may be that Model S is traction limited well above the speeds we are talking. It would take examination of Model S' traction control data to tell and we simply do not have access to that data. As for parsing what Elon says, I now have enough experience listening to him to take what he says with a grain or two of salt.

 

[WarpedOne]

The cars and company have more than enough going for them then to have to rely on such tactics.

Hanlon's razor to the rescue.

 

[lolachampcar]

I posted earlier that, intent aside, the results are the same. It does not matter if they intentionally or un-intentionally provide insufficient or misleading information; I'm still mislead.

 

[Bill D]

...But wait. That 10.9 second 1/4 mile time may be a Tesla "price includes the time you spend pumping gas over the lifetime of your ICE" or "691 combined motor hp" shenanigan. The speculation that MotorTrend tested the P90DL at high density altitude and high ambient temperatures then applied ICE correction factors to take consistent 11.4 second runs and adjust them to consistent 10.9 second performance may be true. Owners are only getting 11.4 seconds out of their P90DLs and it has been too long since the release of the MotorTrend results to have cars in the field receive an OTA upgrade to bring them up to MotorTrend performance levels...

I'm sure many folks, like me, bought a P90DL because they rightly believed "This one goes under eleven".

So, your new comments above on 10.9 are troubling. Previously your position on 10.9 was "sit tight", which made sense because Tesla still unequivocally specs 10.9 sec in the Design studio and they haven't added an asterisk.

Not delivering 10.9 based on obscure technicalities would cause a major loss of trust in Tesla, far beyond the 691 HP controversy where they have some "wiggle room" as evidenced by this thread.

 

[kennybobby]

.. The speculation that MotorTrend tested the P90DL at high density altitude and high ambient temperatures then applied ICE correction factors to take consistent 11.4 second runs and adjust them to consistent 10.9 second performance may be true. Owners are only getting 11.4 seconds out of their P90DLs and it has been too long since the release of the MotorTrend results to have cars in the field receive an OTA upgrade to bring them up to MotorTrend performance levels.

TM seems to put a lot of stock in test data from MotorTrend, but when you start peeling the onion it looks like mistakes, errors and blunders.

The link from TM to MT test procedures, Testing, Testing - The Motor Trend Way - Motor Trend, indicates that they test and make weather corrections according to SAE J1349:

"In an attempt to ensure fair comparisons between cars tested in the high desert heat of summer and the dense cold of a Michigan winter, we record the weather conditions using a Computech RaceAir system and then use the SAE’s J1349 procedure as a guide to correct all test results to standard operating conditions of 77 degrees F, 29.32 inches mercury barometric pressure, and zero-percent humidity."

This is just a red herring by MT to appear to have a technical basis. The problem here is that when you go read SAE J1349, it refers to making corrections to ICE engine performance when testing the engine on a dynamometer, or corrections to an ICE engine performance when instrumented for a vehicle acceleration run.

"Instrument the vehicle with any suitable data acquisition system to record all engine controller parameters
that will be required to replicate the engine’s time-based variable controls on the dynamometer net power
test. This may include, but is not limited to: RPM, MAP (or Mass Airflow), Spark Advance, A/F Ratio, Cam
Timing, fuel injector timing and pulse width, EGR valve and throttle position, coolant and oil temperature"

J1349 has nothing to do or say about making corrections to 1/4 mile run times. Also there is a mistake in pressure since sea-level is at 29.92.

Testing at high desert would be at a lower atmospheric pressure than sea level, although the weather channel would report a barometer or altimeter pressure reading already adjusted for sea level. If any correction to 1/4 mile times were done it should have increased the sea level time since the air is denser at SL.

My guess is that MoronTrend used the reported barometric pressure when testing in the desert (already adjusted for sea level), then reduced the times to their "29.32" pressure correction value which is less dense at about 600' above sea level. Or they just blundered by reducing the 1/4 mile time instead of increasing it for the SL correction. Either way it's a joke that TM uses MT as their reference for performance metrics.

MT is not an engineering or motor testing organization--they are a bunch of journalists selling advertising for a magazine. They buy a techno gadget to use for testing, and like 99% of folks, plug it in and start using it without reading the manual. You know it's true--how many people read their owner's manual before driving the car...?

 

[Krugerrand]

I posted earlier that, intent aside, the results are the same. It does not matter if they intentionally or un-intentionally provide insufficient or misleading information; I'm still mislead.

Well, that's interesting. It certainly matters to me when dealing with a person/company what their intentions were all along. It matters a great deal. That's as if you want to live in a black or white world? :wink:

 

[chickensevil]

The only information they release is that it will be .2 seconds faster 0-60 and in the 1/4 then the current P85D. This does not explain a specification difference of 11.4 to 10.9 when comparing the 85DL to the 90DL. Battery capacity and small changes in chemistry should not make that much difference. Point five seconds is an eternity in the 1/4 and would be worth disposing of my current P85D and buying the P90DL. But wait. That 10.9 second 1/4 mile time may be a Tesla "price includes the time you spend pumping gas over the lifetime of your ICE" or "691 combined motor hp" shenanigan. The speculation that MotorTrend tested the P90DL at high density altitude and high ambient temperatures then applied ICE correction factors to take consistent 11.4 second runs and adjust them to consistent 10.9 second performance may be true. Owners are only getting 11.4 seconds out of their P90DLs and it has been too long since the release of the MotorTrend results to have cars in the field receive an OTA upgrade to bring them up to MotorTrend performance levels.

So this has me a bit confused, because if you look at the site they advertise the quarter mile time and 0-60 time irrespective of 85 vs 90 when you order it new. So why can they put that on their site, but then tell you that if you do the retrofit that it won't give you the full performance?

Also the MotorTrend numbers said 2.6 instead of 2.8 0-60 (both assuming 1ft rollout), which is also totally un-achievable that I have seen outside of that test.

 

[Kalllagtunet]

I have the answer to this, MT have tested the car that TM have rated at 780 hp without the battery but using the same external power source connected with a quarter mile long extention cord....

 

[Andyw2100]

I posted earlier that, intent aside, the results are the same. It does not matter if they intentionally or un-intentionally provide insufficient or misleading information; I'm still mislead.

Well, that's interesting. It certainly matters to me when dealing with a person/company what their intentions were all along. It matters a great deal. That's as if you want to live in a black or white world? :wink:

Intentions matter to me too, Krugerrand. And I expect they matter to lola as well. He never said they didn't. All he said was that the result (winding up mislead) was the same, without respect to the intentions.

For me one good way to analyze intentions after the fact is to take a look at how things are being handled. In this case I think it's pretty clear that Tesla now realizes that intentionally or not, they wound up misleading many of their customers. Whether or not they do anything to rectify the situation for those mislead may not be a perfect litmus test for whether or not there was intent to deceive (as an example of why the test fails--not saying this is the case--just using it as an example--there could have been intent to deceive, but the hope to completely get away with it, and failing that, the recognition of the need to make reparations), but I think it goes a long way towards letting us infer something about whether or not there was intent. More importantly, it tells us how much (or how little) Tesla values our relationship.

I'm still reserving judgment, as I don't believe this is over yet, but like lola and many others, my feelings towards Tesla have been moving in the wrong direction for some time now. Tesla needs to act to correct that.

 

[vgrinshpun]

Ok, I think I understand that.

What that means is that Elon really should have said the car begins being power limited at ~30 mph, not "traction-limited until 30 mph", no?

This is what I understand from what you all have done to educate me in the last few posts:

* The car is traction-limited for the first fraction of a second (to roughly 10 mph),
* The car is then motor torque limited for the next 20 mph or so, until ~30 mph,
* The car is then power limited beyond that.

Do I understand it?

If not, I'll shut up and let the technical discussion continue without my interruption in the interest of understanding how the variables play, but reiterate my concern for any additional parsing of what Tesla says or doesn't say.

I do not believe that Elon should've said anything different. What I am trying to explain is that I do not think that what Elon said in the context of answering the specific question and what I posted discussing the power/torque curve was contradictory.

What I was stating was consistent with the convention regarding power/torque curve, i.e. any values that are directly on the curve and below it are achievable, while values above the curve are not. Since this particular curve includes two distinct regions - constant torque and constant power - I used terms torque limited and power limited to refer to the areas on and under the curve. Note that torque limited in this context does not mean that torque corresponding to the values on the curve is not achievable, it means only that values above the curve are not achievable.

When EM was answering the question about engineering steps that lead to the improvement in performance for Ludicrous, he essentially was following the same curves, but instead of talking about torque and power he was talking in terms of corresponding physical limitations of the car. Instead of power limited he used term battery pack limited, because battery was the component that limited power. Similarly, rather than discuss torque limited portion of the curve, he addressed physical limitation that was dictating the design of the drive unit with specific maximum torque - maximum practical limit of the size of the wheel/tire combo. So when he mentioned traction limited in that context he just implied that increasing torque in constant torque portion of power/torque curve would have had no bearing on performance improvement because current wheel/tire setup can't handle such an improvement while increasing the size was not a practical solution. His use of traction limited does not mean that P85D is not able to transfer all available torque anywhere during the whole duration of the run (0 to about 30+mph), just that an increasing this maximum torque in practical terms could not have lead to improvement in acceleration, and therefore was a design limiting factor.

So with all the above in mind, I agree with all three items (asterisks) in your summary, however, I do not believe that anything EM said during the call contradicts this summary. I do not believe that he implied that P85D is not able to put full available maximum torque to the pavement anywhere between 0 and 30mph ("Dodge Hellcat syndrome"). This would not make much sense from the design point of view (designing components for a torque that car can't utilize), nor it is consistent with the Consumer Report chart.

 

[rns-e]

TM seems to put a lot of stock in test data from MotorTrend, but when you start peeling the onion it looks like mistakes, errors and blunders.

The link from TM to MT test procedures, Testing, Testing - The Motor Trend Way - Motor Trend, indicates that they test and make weather corrections according to SAE J1349:

"In an attempt to ensure fair comparisons between cars tested in the high desert heat of summer and the dense cold of a Michigan winter, we record the weather conditions using a Computech RaceAir system and then use the SAE’s J1349 procedure as a guide to correct all test results to standard operating conditions of 77 degrees F, 29.32 inches mercury barometric pressure, and zero-percent humidity."

This is just a red herring by MT to appear to have a technical basis. The problem here is that when you go read SAE J1349, it refers to making corrections to ICE engine performance when testing the engine on a dynamometer, or corrections to an ICE engine performance when instrumented for a vehicle acceleration run.

"Instrument the vehicle with any suitable data acquisition system to record all engine controller parameters
that will be required to replicate the engine’s time-based variable controls on the dynamometer net power
test. This may include, but is not limited to: RPM, MAP (or Mass Airflow), Spark Advance, A/F Ratio, Cam
Timing, fuel injector timing and pulse width, EGR valve and throttle position, coolant and oil temperature"

J1349 has nothing to do or say about making corrections to 1/4 mile run times. Also there is a mistake in pressure since sea-level is at 29.92.

Testing at high desert would be at a lower atmospheric pressure than sea level, although the weather channel would report a barometer or altimeter pressure reading already adjusted for sea level. If any correction to 1/4 mile times were done it should have increased the sea level time since the air is denser at SL.

My guess is that MoronTrend used the reported barometric pressure when testing in the desert (already adjusted for sea level), then reduced the times to their "29.32" pressure correction value which is less dense at about 600' above sea level. Or they just blundered by reducing the 1/4 mile time instead of increasing it for the SL correction. Either way it's a joke that TM uses MT as their reference for performance metrics.

MT is not an engineering or motor testing organization--they are a bunch of journalists selling advertising for a magazine. They buy a techno gadget to use for testing, and like 99% of folks, plug it in and start using it without reading the manual. You know it's true--how many people read their owner's manual before driving the car...?

Agree 100%

 

[lolachampcar]

Well, that's interesting. It certainly matters to me when dealing with a person/company what their intentions were all along. It matters a great deal. That's as if you want to live in a black or white world? :wink:

Krug,
Andy kinda nailed it. I have my own opinions about intent but really do not think they belong here. What seems on point is what is the information coming out of Tesla and what impression does it leave with the buyer. It would seem we could discuss this without inflammatory talk of intent.

 

[brianman]

What seems on point is what is the information coming out of Tesla and what impression does it leave with the buyer. It would seem we could discuss this without inflammatory talk of intent.

Concur.

 

[stopcrazypp]

The increase in motor power from 728 hp to 762 hp also includes the P85D but no performance improvement. So the limit here would be the fuse?

We do not know if it includes the P85D (whether cars from factory before or after they announced the change in motor power) or even the P90D (without Ludicrous). We don't know if is from physical changes or software changes.

However, what we do know is:
1) Tesla advertised 691 hp motor power and 1G of acceleration at P85D launch
2) Manual was changed to reflect a number of 728 hp motor power around time of the 6.2 update and ~1.05 G has been measured (see sorka's graphs)
3) Tesla advertised a number that reflected 762 hp motor power and 1.1G of acceleration at P90D Ludicrous launch.

You can do the math, it scales perfectly. You can throw in the P85 at 470 hp motor power and 0.7 Gs of peak acceleration. The main point is that for the P85D specifically, it was not traction limited yet or it would be impossible to raise the peak acceleration, no matter how much you raise the motor power.

P85 graph:
http://www.teslamotorsclub.com/attachment.php?attachmentid=93832&d=1442012006

- - - Updated - - -

TM seems to put a lot of stock in test data from MotorTrend, but when you start peeling the onion it looks like mistakes, errors and blunders.

The link from TM to MT test procedures, Testing, Testing - The Motor Trend Way - Motor Trend, indicates that they test and make weather corrections according to SAE J1349:
...

I would suggest you read that article more carefully. They say explicitly that they reduce the correction factor drastically for turbo and hybrid based cars. If you see their WRX numbers, they made no adjustment to ET and only adjusted the trap speed by 0.1mph. So assuming they treat EVs similarly, the number for the P90D Ludicrous should have little to no adjustment.

 

[jpet]

So this has me a bit confused, because if you look at the site they advertise the quarter mile time and 0-60 time irrespective of 85 vs 90 when you order it new. So why can they put that on their site, but then tell you that if you do the retrofit that it won't give you the full performance?

I have an answer from Tesla on this topic:

Will P85D have the same performance as P90D? Not quite. Upgrading to Ludicrous Speed will provide for a 0.2 second faster 0-60 time over the customer's current vehicle performance. It will not be an exact equivalent performance spec as a P90D. (if customers press on for absolute times, please reiterate the relative improvement over the current performance and not the absolute time. As the vehicle ages, the performance level is dependent on how people use it, which is hard to generalize across the population. But for a brand new P85D, the ludicrous retrofit with the 0.2 second improvement over the stated 3.1 seconds 0-60 should provide for a 2.9 seconds 0-60 time).

 

[vgrinshpun]

However, what we do know is:
1) Tesla advertised 691 hp motor power and 1G of acceleration at P85D launch
2) Manual was changed to reflect a number of 728 hp motor power around time of the 6.2 update and ~1.05 G has been measured (see sorka's graphs)
3) Tesla advertised a number that reflected 762 hp motor power and 1.1G of acceleration at P90D Ludicrous launch.

You can do the math, it scales perfectly. You can throw in the P85 at 470 hp motor power and 0.7 Gs of peak acceleration. The main point is that for the P85D specifically, it was not traction limited yet or it would be impossible to raise the peak acceleration, no matter how much you raise the motor power.

The scaling part clearly demonstrates that motor horsepower is not a meaningless rating, not a "fake" hp, by a long shot. It is not a very good spec, though, for comparing EV performance to what would be expected from an ICE car of the same hp rating and weight, even when EV output is not battery limited.

 

[LetsGoFast]

The scaling part clearly demonstrates that motor horsepower is not a meaningless rating, not a "fake" hp, by a long shot. It is not a very good spec, though, for comparing EV performance to what would be expected from an ICE car of the same hp rating and weight, even when EV output is not battery limited.

To my eye, the chart does no such thing. It demonstrates the value of increased torque. "Motor horsepower" doesn't even appear on either axis.

 

[dsm363]

Krug,
Andy kinda nailed it. I have my own opinions about intent but really do not think they belong here. What seems on point is what is the information coming out of Tesla and what impression does it leave with the buyer. It would seem we could discuss this without inflammatory talk of intent.

Agree. There have been multiple posts saying Tesla intentionally mislead or lied which we have no way of knowing at this point.