Does anyone else feel like the first 0-10mph are limited in the M3P?

[AlanSubie4Life]

with the car "cruising" at like 20-30 MPH and then you floor the accelerator up to 60? Asking because when i do this in my car, it feels more violent than the 0-60 run but not sure if it's just my imagination.

Do you mind measuring the g-forces starting at 20, 30 and 40? 30-40 feels the tingliest to me after plenty of "tests".

I think this is possible. You just trigger on the 20mph threshold.

However, I am about 99% certain the results will be no different than what we have already.

The tests you are suggesting imply that the car has “memory” and “catches its breath” by coasting at a fixed speed and starting acceleration at a certain speed. So somehow the idea is that the acceleration will be higher in this situation?

I believe we will see exactly the same acceleration (0.8-0.85g - calculate it from the slopes in the pictures) with the new test.

Regarding acceleration at 30-40, the plots show it is about the same as every point earlier. The acceleration starts to taper around 45, where the straight line deviates from the plot of speed vs. time.

With the raw data, if I ever download it, we should be able to reconstruct torque and horsepower curves.

Anyway, I will add these tests to my last couple tests when I get around to it. But ir is likely they won’t be interesting.


Regarding “jerk”, jerk is the derivative of acceleration. So a ramp to full acceleration that happens in a set time will give higher jerk for a higher “full acceleration” value.

But I don’t think jerk has anything to do with your perceptions, because jerk is very near zero from 5mph all the way to 60mph (start to get some negative jerk at around 45). I think the perception has something to do with the combined perception of speed + acceleration. I think we have been trained that acceleration will drop with increasing speed in an ICE car as it comes off its torque peak (before shifting). When this doesn’t happen (below 45mph) in a Tesla, it feels fast. But I don’t know. To me it still feels fast at any speed but I get used to it with time...

 

[Drizzle]

I think this is possible. You just trigger on the 20mph threshold.

However, I am about 99% certain the results will be no different than what we have already.

The tests you are suggesting imply that the car has “memory” and “catches its breath” by coasting at a fixed speed and starting acceleration at a certain speed. So somehow the idea is that the acceleration will be higher in this situation?

I believe we will see exactly the same acceleration (0.8-0.85g - calculate it from the slopes in the pictures) with the new test.

Regarding acceleration at 30-40, the plots show it is about the same as every point earlier. The acceleration starts to taper around 45, where the straight line deviates from the plot of speed vs. time.

With the raw data, if I ever download it, we should be able to reconstruct torque and horsepower curves.

Anyway, I will add these tests to my last couple tests when I get around to it. But ir is likely they won’t be interesting.


Regarding “jerk”, jerk is the derivative of acceleration. So a ramp to full acceleration that happens in a set time will give higher jerk for a higher “full acceleration” value.

But I don’t think jerk has anything to do with your perceptions, because jerk is very near zero from 5mph all the way to 60mph (start to get some negative jerk at around 45). I think the perception has something to do with the combined perception of speed + acceleration. I think we have been trained that acceleration will drop with increasing speed in an ICE car as it comes off its torque peak (before shifting). When this doesn’t happen (below 45mph) in a Tesla, it feels fast. But I don’t know. To me it still feels fast at any speed but I get used to it with time...

I would love to see the results. I just the car for a spin and I definitely get that tingly feeling between 20-40... but I'm starting to think that it might be because I launch it after I'm slightly slowing down for a few seconds. I'll have to test this more tomorrow.

 

[SigNC]

I'd love to see the same measurements on a P85D with "just" insane mode. To me it feels way more violent from 0. Also, i think at 57 degrees you might not be getting optimal power unless you've been driving around a while.

 

[AlanSubie4Life]

I'd love to see the same measurements on a P85D with "just" insane mode. To me it feels way more violent from 0. Also, i think at 57 degrees you might not be getting optimal power unless you've been driving around a while.

Well...I did get basically the rated acceleration. There were no power-limiting dots (just regen but maybe due to state of charge). If I were power-limited there would presumably be power-limiting dots, though I do not know for sure.

Next time I will go for no dots!

 

[kbecks13]

Seems like Engineering Explained got a Model 3 (awesome!) and interestingly enough, he has some similar feelings about a bit of a "delay" on the Model 3 acceleration punch. It's very weird to hear so many people reporting such a similar feeling but not being able to see it in the data. I think maybe we need to get more granular on the initial acceleration and jerk once you mash the accelerator.

Here's the part where he talks about the slight delay and even compares Model 3 to a Nissan Leaf (Go to 11:10, link is being weird):

 

[MountainPass]

It is indeed limiting your torque momentarily to preserve the drivetrain. Even the Nissan GTR has to apply a small amount of torque to take up gear lash before it unleashes in launch mode. The Model 3 does this at all speeds, but most noticeably at low speed. If you were to replace the inverter as we have done on other EV's you can request the torque uninhibited. It sounds and feels like a sledgehammer to the drivetrain though.

 

[Drizzle]

It is indeed limiting your torque momentarily to preserve the drivetrain. Even the Nissan GTR has to apply a small amount of torque to take up gear lash before it unleashes in launch mode. The Model 3 does this at all speeds, but most noticeably at low speed. If you were to replace the inverter as we have done on other EV's you can request the torque uninhibited. It sounds and feels like a sledgehammer to the drivetrain though.

You guys rock! Do you feel that it is over limited at all? And do you know how it compares to the model S?

And when you say request the torque uninhibited, do you ask tesla for that?

Lastly what can normal owners do to I create torque or quickness themselves if money wasn't an issue?

 

[MountainPass]

You guys rock! Do you feel that it is over limited at all? And do you know how it compares to the model S?

And when you say request the torque uninhibited, do you ask tesla for that?

Lastly what can normal owners do to I create torque or quickness themselves if money wasn't an issue?

The money part is the entirety of the problem. Reduce weight, unsprung then sprung! If you have $30k burning a hole in your pocket and are ok with a pure race car, we can do it!

 

[Drizzle]

The money part is the entirety of the problem. Reduce weight, unsprung then sprung! If you have $30k burning a hole in your pocket and are ok with a pure race car, we can do it!

That makes sense, but what about a bigger motor or unleashing more power within the current motor? Obviously there aren't superchargers or turbos for these, but is there anything else that can easily add power?

 

[MountainPass]

That makes sense, but what about a bigger motor or unleashing more power within the current motor? Obviously there aren't superchargers or turbos for these, but is there anything else that can easily add power?

We could recharacterize the OEM motor with an aftermarket inverter, the problem is the OEM inverter is very much integrated with the rest of the car. If it was a project like our Lotus, it would be fine. But for a nice, under-warranty street car, it is simply not a viable option. So the short answer is - no!

 

[Drizzle]

We could recharacterize the OEM motor with an aftermarket inverter, the problem is the OEM inverter is very much integrated with the rest of the car. If it was a project like our Lotus, it would be fine. But for a nice, under-warranty street car, it is simply not a viable option. So the short answer is - no!

Sounds like you guys need some investors to pay for a few experimental cars. Become the first speed enhancing tesla company! Wonder what tesla would think.

Anyways, I'll grumble until I can afford something like a model s ludicrous.

 

[kbecks13]

Sounds like you guys need some investors to pay for a few experimental cars. Become the first speed enhancing tesla company! Wonder what tesla would think.

It gets pretty complicated, you would likely need to root the software and get fairly invasive into the car. At that point you might as well start from scratch and not use a Tesla - but it's still very cool and should be/will be done

 

[HeatproofGenie]

I can't wait for the day when I can spend somewhere around $1500 or less and get a "tune" for the car just like I could with a turbo ICE vehicle.

 

[AlanSubie4Life]

This is a MR Model 3. I imagine it feels a bit “delayed” compared to the Performance, based on my experience comparing the P3D to the standard LR and the AWD.

Seems like Engineering Explained got a Model 3 (awesome!) and interestingly enough, he has some similar feelings about a bit of a "delay" on the Model 3 acceleration

Here's the part where he talks about the slight delay and even compares Model 3 to a Nissan Leaf (Go to 11:10, link is being weird):

 

[Drizzle]

I can't wait for the day when I can spend somewhere around $1500 or less and get a "tune" for the car just like I could with a turbo ICE vehicle.

Just drop an additional battery in the back and crank up the amps!

 

[AlanSubie4Life]

It is indeed limiting your torque momentarily to preserve the drivetrain. Even the Nissan GTR has to apply a small amount of torque to take up gear lash before it unleashes in launch mode. The Model 3 does this at all speeds, but most noticeably at low speed. If you were to replace the inverter as we have done on other EV's you can request the torque uninhibited. It sounds and feels like a sledgehammer to the drivetrain though.

This is definitely a good thing!

This likely makes a huge difference to the jerk. A step function has infinite/undefined slope (Dirac delta)!

It appears this preloading doesn’t affect 0-60 times (though it might impact the time from pedal push to start of motion). If it does affect 0-60 significantly, it probably affects the first inch or two of motion (at most). After all, not much sense in “preloading” if you start actually accelerating the gears! As can be seen from the VBOX plot (unless it is altering the data somehow which seems unlikely), at 1ft the car is moving at ~5mph with acceleration of ~0.8g.

In any case, the ramp up of acceleration (which definitely exists; it has to, obviously!) appears to happen over a distance which cannot be detected by the VBOX.

Video above and other things make me want to add a chill mode run to the requested datasets...

EDIT: if you zoom way in on the plot near 0mph, you can see there is some instability in the slope, and it starts out considerably lower between the first two time points. Additionally, right around 5mph there's likely an absolute maximum of acceleration between two points. It's possible this is related to the preloading, but it's difficult to distinguish from GPS position error/noise. The first two time points would likely happen in the first inch or less; I haven't bothered to calculate it.

 

[Drizzle]

Thanks again for your input. From the other side, I just want to feel more g's. I think the reason it personally bugs me so much is because if you slow down into 30mph and then slam the throttle before it hits a consistent 0 or positive acceleration, it feels soooo good. Like a true theme park magnetic launch ride. I'll have to see if Dragy picks up the g's in that scenario.

 

[AlanSubie4Life]

Thanks again for your input. From the other side, I just want to feel more g's. I think the reason it personally bugs me so much is because if you slow down into 30mph and then slam the throttle before it hits a consistent 0 or positive acceleration, it feels soooo good. Like a true theme park magnetic launch ride. I'll have to see if Dragy picks up the g's in that scenario.

We'll see. I'll try to test a situation where I maintain load on the drivetrain until somewhere below 20mph, then hit it, and provide results. Maximum acceleration is determined by torque though, and I don't anticipate that peak torque is going to be any different in this situation. There is no evidence in the VBOX data or the dyno data, or posted Dragy data, that there is any way to drive the car to get more than the current (software-limited) value.

EDIT: For the first two datapoints (let's assume it's real), if we estimate about 0.4G of acceleration (looks like about half the slope initially, though this is just an estimate):

x= 1/2 * a* t^2 = 1/2 * 0.4*9.81m/s^2 * (0.05sec)^2 = 5mm

So this effect drops off after about half a centimeter. Could take as long as an inch or two before you've really got great acceleration!

It actually seems pretty likely this would be noticeable. Body is presumably pretty sensitive to acceleration - we are able to stand up with our eyes closed, after all.

EDIT: very sensitive! Thresholds for perception of direction of linear acceleration as a possible evaluation of the otolith function

 

[Drizzle]

We'll see. I'll try to test a situation where I maintain load on the drivetrain until somewhere below 20mph, then hit it, and provide results. Maximum acceleration is determined by torque though, and I don't anticipate that peak torque is going to be any different in this situation. There is no evidence in the VBOX data or the dyno data, or posted Dragy data, that there is any way to drive the car to get more than the current (software-limited) value.

I'll be seriously confused if not. If not, then I'll have to start measuring stuff like my seat angle, how much I ate that day, if I had coffee and how much sleep I got next...

Semi interesting but also annoying just to determine when I'll feel tingles or not.

 

[AlanSubie4Life]

This was in my previous picture/post with the speed vs. time plot, but it is worth noting:

It take 0.27 seconds or so to go the first FOOT in the zero to 60 run. This is nearly 8% of the TIME, but it's less than 1% of the distance. So it might seem slow if visual perception is a cue to perceived acceleration.

I don't recommend this AT ALL unless you are on a closed course with others monitoring, but one thing you could do is to close your eyes and then stomp on it, to test this hypothesis. Maybe it's a visual effect. Seriously, don't try it. Probably not even on a closed course.

As a passenger, though, you could try it. See whether the perception is different. For a controlled experiment, you'd need to experience a run as a passenger with your eyes open, and with your eyes closed. Probably a few alternated runs, to try to eliminate any recency effects.