P3D horsepower?

[SammichLover]

Yes, I see that. As I clarified above - yes, that is flat for the AWD, but not related to electric motors - it's due to the AWD software limitation.

*facepalm* "Continued reading fail alert".

 

[AlanSubie4Life]

*facepalm* "Continued reading fail alert".

The original quote from @MoreAgain made it clear there was a software limitation to contrast between P and AWD. Which is correct. All I was pointing out (since it got a bit confused in the follow-up post) is that the flat area between 50mph and 100mph wasn't due to electric motors (which you could argue was implied in that subsequent post). But I understand why you're posting "reading fail alert" - that's why I clarified my original post. It's cool - I did not read the original post very carefully at all and was focused more on the content of your follow-up post. Which perhaps I'm still mis-reading. Like I said, it's fine.

 

[AlanSubie4Life]

I'd be entirely unsurprised if the instrumentation didn't operate down there in an any meaningful way, and even if it did the software probably isn't set up to react and treat the info in that small window as relevant and likely throws out that data.

Exactly. That's why it is probably a better plan to just use accelerometer data in conjunction with the speedometer readout - which seems terribly primitive, but works surprisingly well and produces results that appear to be much more reasonable than the dyno measurements - of course, since they are based on a real world run. Or just use a VBOX (which is annoying since getting the data in usable format off the device takes work).

Of course, those results don't take into account various environmental factors (hills, winds, density of the air), but it appears they are still more predictive. Comparisons vs. prior software versions are also a lot less fraught with error sources - dynos lose calibration, they need all these factors, etc.

Qualitatively, of course, the dyno measurements are interesting (as long as you exclude the data below 3300rpm).

 

[MichaelP90DL]

For what it's worth, there are HP calculators out there.

Engine Horsepower Calculator

Determine your car's weight and then look up the best quarter-mile drag strip run for that same car, enter the data and it will give you a very close estimate.

 

[Knightshade]

Oh come on. Look, nobody is denying that it's possible Tesla is using rollout for some models and not for others. But you said Tesla EXPLICITLY said they are.

I mean- you've been shown Tesla.com saying exactly that

Then also shown both Tesla.coms own change from 3.5 to 3.3 without any actual change to the car or software when they switch the P3 from honest reporting to the same "rollout for one but not the other" system they've been using on the S for years.

Then also shown calibrated testing data of the 0-60 times on the P3D and the LR AWD proving exactly that is what's happening.


So other than trolling there's no explanation left for your doubt about this.

I think you're using the word dishonest incorrectly. Just one opinion.

They're reporting the same performance metric on 2 different versions of the same model, and using two different ways of measuring it to do so.

Not sure how you can call that anything but dishonest.

Especially when no other car company in the world does this.

It would, as mentioned earlier, be like a company advertising crank HP for the most expensive version of a model but wheel HP for the cheaper ones, and not telling you they were doing it.

How's that not dishonest?

 

[AlanSubie4Life]

For what it's worth, there are HP calculators out there.

Engine Horsepower Calculator

Determine your car's weight and then look up the best quarter-mile drag strip run for that same car, enter the data and it will give you a very close estimate.

The problem with these calculators is that they were originally derived (empirically I think, but not sure) to fit to ICE vehicles (as far as I can tell). It appears to me they do take into account shift points, etc., and make some basic assumptions regarding about how close the vehicle is kept to the HP peak during the run (your link says "To estimate peak horsepower, maximum work output should be applied from start to finish."...whatever that means).

Auto Math Handbook

One issue is that Teslas generate very small amounts of power relative to a similar peak power ICE vehicle, between say 0 and 20mph. At 20mph the Model 3 Performance is only putting out ~240HP at most (though it is producing maximum torque)! An ICE vehicle with 480HP peak (to the wheels) will likely be putting out close to 480HP at that speed, so will be accelerating faster (if it is not traction-limited, which it may well be at 20mph).

On the other hand, Teslas don't have to shift, and you don't have to worry about getting a good solid launch without wheel spin (since currently with the Model 3 Performance there is no wheel spin, due to insufficient torque to the wheels). And they don't have to spin up nearly as much motor and drivetrain and gearbox, etc., etc. And as alluded to above, there is no point in generating more power at a given speed (torque) than the tires on the vehicle can handle (and the Model 3 Performance is only 10-30% or so off of this torque limit (depending on the tires in use), all the way up to 40mph - so a considerable amount of that extra power "available" on the ICE can't be used to destroy the Tesla at very low speeds unless the ICE has a LOT of grip). All the myriad reasons Teslas do so well relative to similarly powered (and higher powered) ICE vehicles in a sprint to 60mph...

That being said, these factors appear to balance out in the formulas and give you a calculation which is "in the ballpark" - which of course it has to be, due to physics (average power to get to a given speed for a given weight is easily calculated and will of course be quite well correlated with peak power). But it's just in the ballpark, so won't really resolve any of the discussion here.

In the end, it'll be interesting to see @SleeperService run his Stealth again once the new update is rolled out. That's my gold standard! Though I'm a bit afraid that this boost is going to break free those MXM4s...it's getting pretty close...you need about 0.96g (derived from v^2/(2*d) for the stopping distance test with new MXM4s) - but if the tires are worn/heat cycled, surface isn't good...it is going to be close, at around 0.91g after the boost (assuming they do boost the torque with this update, and not just a lame HP boost ).

 

[MagnusMako]

I mean- you've been shown Tesla.com saying exactly that

Then also shown both Tesla.coms own change from 3.5 to 3.3 without any actual change to the car or software when they switch the P3 from honest reporting to the same "rollout for one but not the other" system they've been using on the S for years.

Then also shown calibrated testing data of the 0-60 times on the P3D and the LR AWD proving exactly that is what's happening.

Yep. And I actually remember when it was 3.5 and then one day a video came out on Youtube that did 0-60 testing with a VBOX and reported their time consistently as 3.2 with rollout. I think it was DAEriK or Model 3 Owners Club. Coincidentally by some act of randomness, Tesla changed their sites time to 3.2 less than a week later. Hrmmmmm I wonder why.

 

[gcmak]

Here's 3 runs above 85% SOC. The gains are there and I was able to do more runs (than those attached) at under 3.2 seconds (if you count roll-out).

 

[AlanSubie4Life]

Here's 3 runs above 85% SOC. The gains are there and I was able to do more runs (than those attached) at under 3.2 seconds (if you count roll-out).

Do you happen to have any data from the same vehicle on the same course at the same weight at high SOC before the update? Could look at the splits and do some simple math to see if the peak torque increased. It kind of looks to me like the torque is simply sustained to a higher speed, and the results sort of imply that as well (it's not that much faster...but that's why I was curious about comparative data from your car, since that relatively slow time could just be due to junk in the trunk )

10-40mph looks like about 1.61-1.64 seconds here. Wonder what it was before for your car... (on my car with the prior update, I got 1.58 seconds 10-40mph at sea level)

But, your 40-60mph is about 1.26-1.28s...whereas with the prior update my 40-60mph was 1.32s. So that is a LOT faster than mine (though mine was kind of a dog in that region from what I recall possibly due to battery temperature).

So if that data pattern holds, this looks more like peak power increase than a peak torque increase. You'd have to have a few runs with the prior update in the same conditions to really know. So many variables.

 

[SammichLover]

The original quote from @MoreAgain made it clear there was a software limitation to contrast between P and AWD. Which is correct. All I was pointing out (since it got a bit confused in the follow-up post) is that the flat area between 50mph and 100mph wasn't due to electric motors (which you could argue was implied in that subsequent post).

Wrong.

It is an entirely normal thing to see in electric motors with fixed power input (a choice in operating, not a "software limitation"), and an entirely abnormal thing to see in an ICE engine. Because of the nature of electric motors, which have a very wide RPM band of near-optimal operating in contrast to ICE motors.

 

[AlanSubie4Life]

Wrong.

It is an entirely normal thing to see in electric motors with fixed power input (a choice in operating, not a "software limitation"), and an entirely abnormal thing to see in an ICE engine. Because of the nature of electric motors, which have a very wide RPM band of near-optimal operating in contrast to ICE motors.

Sure, with a fixed power input, the output power of the electric motor will be roughly fixed, assuming the motor output power limit has not been reached for the given excitation and motor speed.

All I was saying that the blue curve here (which I have used some artistic license on to better fit to reality) is not very flat, though it is broad! While the dashed red curve is artificially limited (as the motor and battery are the same in both cases) and looks "flat" (at least above 5k RPM). That "flatness" is a software limitation (which limits the input power!).

That's all I was saying. (Again with the caveat that the original comment was about the flatness of the AWD curve - which I agree: yes, it's pretty flat. It was saying the same thing I'm repeating here, in fact.)

 

[AlanSubie4Life]

Look, I "fixed" the dyno plots! So proud of myself. All very approximate....

 

[gcmak]

Do you happen to have any data from the same vehicle on the same course at the same weight at high SOC before the update? Could look at the splits and do some simple math to see if the peak torque increased. It kind of looks to me like the torque is simply sustained to a higher speed, and the results sort of imply that as well (it's not that much faster...but that's why I was curious about comparative data from your car, since that relatively slow time could just be due to junk in the trunk )

10-40mph looks like about 1.61-1.64 seconds here. Wonder what it was before for your car... (on my car with the prior update, I got 1.58 seconds 10-40mph at sea level)

But, your 40-60mph is about 1.26-1.28s...whereas with the prior update my 40-60mph was 1.32s. So that is a LOT faster than mine (though mine was kind of a dog in that region from what I recall possibly due to battery temperature).

So if that data pattern holds, this looks more like peak power increase than a peak torque increase. You'd have to have a few runs with the prior update in the same conditions to really know. So many variables.

Too many variables, honestly. I don't have a strip that I might reliably test on and even then there's a lot of other variables.
Attached is an image from when I had V9. I basically could get these figures repeatably so I kept this as a record.

 

[mact3333]

Man why are you always on here arguing with people?

I dare you to ask him about brakes and stopping power...a 2 inch by 2 inch rotor/pads will stop a car as fast as 14 inch brembo's if they had the same tire size hehehe.

 

[Daniel in SD]

I dare you to ask him about brakes and stopping power...a 2 inch by 2 inch rotor/pads will stop a car as fast as 14 inch brembo's if they had the same tire size hehehe.

Haha. @Knightshade is right about brakes and right about 1 foot rollout and I admire his tenacity in correcting everyone who spreads misinformation about the subjects.
I feel like he should have his own Knightshade Manifesto sticky.

IMHO If you’re using 1 foot rollout for 0-60 then you’re not actually measuring 0-60. At least with the 1/4 mile times the car is actually traveling 1/4 of a mile in that time (just not from a stop).

 

[AlanSubie4Life]

Too many variables, honestly. I don't have a strip that I might reliably test on and even then there's a lot of other variables.
Attached is an image from when I had V9. I basically could get these figures repeatably so I kept this as a record.

Well that is 1.62s 10-40mph (vs. 1.61-1.64 now)

And 1.35s 40-60mph (vs. 1.26-1.28s now)

So looks to me like all this improvement is up top (which buys a significant amount - about 0.07 seconds).

But this means no change in peak torque, and only maintenance of that peak torque to higher speed (meaning higher power).

You can kind of see this in the plots.

It looks like Dragy poorly captures the rollout time (not too surprisingly) so all the 0-x times should be ignored, but can use deltas and just subtract 0-10 time so everything is 10-x. And the “1ft” time is ok too of course because it gets rid of that rollout time.

 

[hydro 481]

I'm not the one arguing.

I've simply stated the same documented facts a bunch of folks before me mentioned about how Tesla misrepresents the P versus non-P performance differences.

It's other folks who seem to want to "argue" about if these facts are facts (they are- as cited by screenshots from tesla themselves, and then confirmed by numerous calibrated measurements from both owners and car magazines)... and then when that argument fell apart they now want to argue if this dishonest is "important" or not.

It seems like you are always engaged in some sort of "debate" on here. You must enjoy confrontations

For the record, I do agree with your stand on most things. Including the testing inconsistencies for Performance vs Non-Performance. But I'll be honest, the way you talk sometimes comes across as a bit pompous and condescending. I'm sure your intentions are good but it might be rubbing people the wrong way. But hey, I'm not here to tell you how to act. You do you bud.

 

[Knightshade]

It seems like you are always engaged in some sort of "debate" on here. You must enjoy confrontations

No, I just dislike incorrect information

If I'm ever wrong about my facts I'd absolutely want someone else to correct me on em as well.

I
For the record, I do agree with your stand on most things. Including the testing inconsistencies for Performance vs Non-Performance. But I'll be honest, the way you talk sometimes comes across as a bit pompous and condescending. I'm sure your intentions are good but it might be rubbing people the wrong way. But hey, I'm not here to tell you how to act. You do you bud.

Honestly I think I'm fairly polite the first couple of times I correct someone on something- it's when it's like the 5th time I'm pointing the same thing out, generally with well sourced facts, to the same guy that I possibly get a little less polite.

(see the dude in this very thread who kept insisting he still isn't actually convinced Tesla uses rollout on the published P numbers despite a mountain of evidence they do including from Teslas own mouth)

 

[SleeperService]

@AlanSubie4Life Here's the data you are looking for! I've plotted three runs: before the first bump (2019.5.15), after the first bump (2019.8.4), and after the second bump (2019.36.1). I did a run yesterday morning as well on the latest version prior to the second bump (2019.32.12.2) to be certain nothing had changed since 2019.8.4, and it had not. It looks like there was a very modest increase in the peak torque, but a more substantial bump in the 40-100 mph region this time. It's definitely not as big of a bump as last time, but it's still there and measurable. 0-60 (1 ft rollout) was 3.081s. I'm betting it could be a bit better if my battery was warmed up, as I had just pulled out of my garage and still had a dotted line on regen (but not power). Anyway, the Chart:

Edited to add: max G's were 0.901. Didn't break loose the MXM4s.

 

[WilliamG]

(see the dude in this very thread who kept insisting he still isn't actually convinced Tesla uses rollout on the published P numbers despite a mountain of evidence they do including from Teslas own mouth)

I have to admit, there’s only so much BS I can put up with. Tesla is currently not admitting to any such thing (whether they do it or don’t is a different argument). Your skewed narrative does nobody any favors, and just spreads misinformation, as if there isn’t enough of that out there already...