Upgraded performance Metrics Summary

[SucreTease]

See the note on the rollout in the summary tab. You can add a quarter second to remove the rollout. Basically since I don't know when folks are pressing the accelerator I am simply starting with the first distance measurement that is greater than one foot, like the one foot rollout. Some folks have big power increases before zero calculated that way, even though my understanding is cars can't without launch mode can't simultaneously brake while accelerating the motor, so I have started graphing the second before my calculated zero. I know about the hate for rollout, but it seemed the best way to normalize data.

Coming up with 3.75s 0-60 mph time from a reported 4.5 sec 0-100 kph time seems quite fictional to me. I don't think you should be reporting a performance metric that is estimated from guessed numbers. That belies the very meaning of "metric" (i.e. something measured). If it isn't directly measured, it shouldn't be in a document you title "Upgraded Performance Metrics Summary".

 

[Krash]

...estimated from guessed numbers...

Nobody is guessing any numbers. I have a rule that lops off the rolling time prior to one foot distance travelled to simulate rollout rules. I am not going to rehash the rollout argument. I only use it to normalize data.

You are welcome to contact all the forum members and try get get the powerlog summaries if you think that is an easy task and is more accurate. It would be great to have you maintain a second set of metrics. I'll put a link to them at the top. But you'll notice a lot of runlogs have a rolling start. So one run log will take an extra three quarters of a second to get started. When people ask why that metric is off you can point them back here. Be sure to tell the moderator why your metrics are more metered and they can retitle my thread and let you use this thread's title.

You could start with your own runlog.

 

[chrisschorsch]

I guess that I understand somehow the way you subtract the rollout time. What I still don't get (there are no formulas in your Excel file) is how you calculate the distance? You must get it somehow out of the Latitude and the Longitude, but I don't really know what those numbers are and in which unit they are in the runlogs. In my opinion we should try to stick to runlogs that have SoCs higher than 80%. We know very well by now that the SoC has a big impact on the max Power and therefore on the 0-60 times. Same holds true for the elevation. I would say max 1m up or down.

Anyhow, what is independent of the rollout is the max Power. If we want to understand the secret behind the corked/uncorked my feeling is that this parameter is the most relevant/unbiased.

We still don't know for sure if there are only corked/uncorked cars out there or if there are cars in intermediate stages, i.e. if both hardware and software are required for a full uncorked car then one can imagine that there are some cars out there that benefitted only from the software change.

If SoCs below 80% and other obvious deviations are ignored from you metrics than we should be able to come out with a simple metrics showing the max Power for corked cars (75, 75D, 90D, 100D). Any significant deviation above that max Powers can be seen as a hint for a kind of uncorking.

Many thanks for your great effort collecting all these data!

 

[Krash]

For distance I calculate =ACOS(COS(RADIANS(90-$G$2))*COS(RADIANS(90-$G3))+SIN(RADIANS(90-$G$2))*SIN(RADIANS(90-$G3))*COS(RADIANS($H$2-$H3)))*5280*3958.756 for cell Q3. Where G2 and H2 have initial GPS location data. I overwrite it when I remove the GPS data before posting.

I think having lower state of charge (SoC) runlogs is interesting to see the SoC effect on actual torque, power and current, but yes to figure out where the limits are set we would need to either just use high SoC or know how to adjust accordingly.

 

[Krash]

@Krash just out of curiosity, how do I know which battery version I have on my S 90D? Do you have the part #s?

Take a photo of the battery sticker and look for the part number. The sticker is in the lower right wheel well. Turn the wheels all the way to the left and reach in with a phone camera at an angle as the sticker is sometimes covered at the top where the most useful information is.

 

[Krash]

Updated. See the file attached to the first post in the thread for individual vehicle runlogs and graphs.

Don't read too much into my swag on estimated torque. Someone much better than me is doing the real math. Hopefully we'll have their results soon.

 

[Speedr117]

@Krash, Thank You for creating this! Your efforts are much appreciated by the vast majority of us. I finally signed up to say that.

From your chart, it looks like the new Drive Unit is actually less powerful and results in cars being slightly slower than the old DU. Am I drawing the right conclusion from your chart?

 

[Krash]

...it looks like the new Drive Unit is actually less powerful and results in cars being slightly slower than the old DU...

I envision SX engineering team getting presented with a new, more efficient drive unit and inverter from 3 engineering team and management says "figure out what settings for torque and power make the new unit as fast as the higher model (100) and what settings make it as fast as the previous model (75). When complete, management says, oh, now add a new setting for the old unit to make it at least as fast as the new faster setting for the new unit."

Did that happen? I don't know how to interpret the data yet.

 

[Speedr117]

Gotcha. It sure seems that way, doesn't it. And they figured within a tenth of a second was close enough. Have you had any data from a 100D with new new DU? Also, for the BTX8 batteries, were they all over in Europe?

 

[Krash]

...Have you had any data from a 100D with new new DU? Also, for the BTX8 batteries, were they all over in Europe?

No S100Ds with the new drive unit. None of the following either: 40, 60, 60D (original or limited 75), 70, 85+ or 85D.
All of the BTX8 I have are from Europe: UK, Norway, Spain. There have been some sold in the middle east.

 

[Speedr117]

You should put a link to this thread in the Model S forum with an ask for the logs. I only found this thread from deep within another thread within the Model S forum. I bet that's why you haven't seen a lot of logs yet.

 

[Krash]

You should put a link to this thread in the Model S forum with an ask for the logs. I only found this thread from deep within another thread within the Model S forum. I bet that's why you haven't seen a lot of logs yet.

I was actually thinking of asking for it to be moved to the technical forum now that we have X data but yeah, if we wanted more data I could request it in the S and X.

 

[Zero CO2]

Question concerning the performance metrics summary , I have seen many folks posting on other threads that the Model S 75 cannot be uncorked, maybe i am not reading the metric correctly but it looks like there are a few Model S 75 RWD that have been uncorked?
Krash... Can you advise ?

 

[Krash]

Yes, you are reading ot correctly. S75 RWDs get\come uncorked. It likely requires hardware. Maybe a fuse. Unfortunately I don't have a lot of RWD samples and have no matched pre and post uncorking pairs of RWDs yet.

 

[Zero CO2]

Yes, you are reading ot correctly. S75 RWDs get\come uncorked. It likely requires hardware. Maybe a fuse. Unfortunately I don't have a lot of RWD samples and have no matched pre and post uncorking pairs of RWDs yet.

Thank you sir your diligence on this topic is appreciated

 

[PhaseWhite]

Take a photo of the battery sticker and look for the part number. The sticker is in the lower right wheel well. Turn the wheels all the way to the left and reach in with a phone camera at an angle as the sticker is sometimes covered at the top where the most useful information is.

@Krash looks like my S 90D March 2017 production date has battery part #: 1088790-00-A I was just curious where this fits in your chart?

Interestingly it looks like this battery has been used in P90D configurations as well:
See:
Battery table
2016 Model S P90D - $92k

 

[Krash]

@Krash looks like my S 90D March 2017 production date has battery part #: 1088790-00-A I was just curious where this fits in your chart?

According to the 90 crowd and battery experts that is a v4 battery.

 

[Speedr117]

@Krash, just put a deposit on an inventory S 100D. I know you can't know the future, but looking at the data you have, do you think if there is a significant acceleration improvement in the 100D, that the old drive (non DU00) would be able to match the new DU00 times like the 75D does?

 

[Krash]

@Krash...do you think if there is a significant acceleration improvement in the 100D, that the old drive (non DU00) would be able to match the new DU00 times like the 75D does?

Yep. But my opinion has no engineering basis. I think the old and new drive units could do 3.3 sec 0-60 without a significant increase in the failure rate. Many folks smarter than me think this is impossible.

I plan to test my theory - just as soon as my warranty expires in seven and a half years. Grin.

 

[Speedr117]

@Krash, the 100D was added to my Tesla account and I did the right click on the "View Spec". Within the options URL was "DU00", so even though it's a mid-June build, it appears to have the new drive unit, correct? Once I get it (3 weeks if I'm lucky), I'll send you the logs.The only 100Ds (both S and X) you have so far are the older DUs, correct?