-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Do P85Ds and S85Ds have inaccurate odometers that overstate distance traveled?
- Thread starter Andyw2100
- Start date
-
- Tags
- Model S User Interface
New trip, new data point: When the car clicked over 100.0 exactly, the GPS Odometer read 99.19. Even my mental math can see that's right on the 0.8%.
I did several other intermediate calcs, and they were all in the range of 0.79 to 0.81 error with the car always showing the larger (longer distance traveled) miles.
Danal
The fact that the error is so consistent in my mind really detracts from the argument that the error is within the acceptable margin of error for odometers. I mean if some cars were overstating mileage and others were understating it, and there was a lot of randomness involved and everyone was saying, "Hey, in general the odometers are close enough", that would be one thing. But in this situation, it would seem that the odometers may all be working extremely precisely, and possibly with very little "error", if they really are all consistently off by almost exactly .8% in the same direction. So I'm hoping it won't be a big deal for Tesla to make whatever change is required to bring these odometers that are probably very precise into line with reality.
I'm looking at this basically as if we all have really well made watches that keep perfect time. The problem is that all the watches were just set to the wrong time at the factory, and we can't reset them ourselves. Once Tesla resets the time on our watches to the correct time, our watches will work perfectly. Once Tesla figures out what calculation they've got wrong that is resulting in this .8% error and corrects it, we'll all have very accurate odometers.
Does that make sense?
billarnett
Member
Sure, that is exactly what I was thinking. I suspect the Tesla firmware just has a constant somewhere that is the revs per mile used for the odometers calculations. I would be surprised if the service centers can't tweak it.
If Tesla wants to fix the issue, I would think the easiest thing to do would be to just correct it in a coming firmware update.
Danal
electricmotorglider.com
Phrased another way: 0.8% consistent mis-calc by all the NEW "D" cars is not near as concerning when you consider that every one of them will go through a 2.3% change as the tires wear to end of life. Assuming that Tesla and Michelin exactly nailed the center of that 2.3% range, we should EXPECT new cars to be 1.15% off in one direction, and worn tires cars to be 1.15% off the other way. Only mid-wear cars would be "perfect".
And there may be some very good reason they picked a 75% point, or a 2/3 point, or whatever, that works out to .8 off when new, and 1.5 off when worn. No matter what they do, the car cannot "keep perfect time" as the tread wears. It is going to change (for those particular tires) by a little over 2%.
Phrased yet another way: I'd expect all the new P85Ds to have matching errors, because they have matching tread depth. Because they've only been shipping for about 3 months. If it is made perfect now, it will be way off when the tires wear.
Having said all of that... the only thing concerning to me, is that the error is in the OPPOSITE direction I'd expect for new tires. But, it is really easy to get all of this backwards, and I may have it backwards... so I'd like somebody else to logic it out independently. I have confidence in the 2.3%; that's the simplest math in the world. In fact, you can derive the change by simply (TreadDepth * 2) / Diameter. Comes out to the same number. Cross check it numerous ways, it is always (for those tires) landing on that same amount of change.
billarnett
Member
Sure it can, but it's probably not worth the bother: instead of a constant revs per mile make it a variable which is periodically re calibrated against the GPS. Not that hard to do but if I were in charge of Tesla's software it would be pretty low priority
So owners are sending emails to Tesla over a 0.8% difference between the miles traveled as reported by the car compared to a GPS tool?
Wow, what a waste of Tesla's time. Have you also thought about the accuracy of the speedometer, too? Maybe we should send Tesla emails because our speedometers may read slightly differently than those street-side radar displays. And of course, the testing that we lay people do must be more accurate than that of Tesla, with 10,000 employees. Yah... :tongue:
Wow, what a waste of Tesla's time. Have you also thought about the accuracy of the speedometer, too? Maybe we should send Tesla emails because our speedometers may read slightly differently than those street-side radar displays. And of course, the testing that we lay people do must be more accurate than that of Tesla, with 10,000 employees. Yah... :tongue:
EarlyAdopter
Active Member
You guys are also overlooking the fact that the P85D comes equipped standard with several different make and models of tires, each of which vary slightly in diameter despite being the same size designation.
While the car reads slightly over with 21" Michelin's, perhaps it reads slightly under with 21" Continentals.
Point being that a variance of 0.8% is in the noise and possibly unavoidable to accommodate more than one tire supplier.
While the car reads slightly over with 21" Michelin's, perhaps it reads slightly under with 21" Continentals.
Point being that a variance of 0.8% is in the noise and possibly unavoidable to accommodate more than one tire supplier.
Why are you so often so condescending? It's really becoming tiresome.
Perhaps this isn't an important issue to you. No one asked you for your input. No one asked you to write to Tesla.
Before I "wasted Tesla's time" writing to them about this, I started this thread, to make sure I --wouldn't-- be wasting their time. I confirmed that others are seeing the same issue that I am. Many of these others have had Model S vehicles before, and they did not experience this inaccuracy with those vehicles. I also made sure of that before writing to Tesla.
And yes, the speedometers are off too, because the odometers are off.
As for our testing as lay people being less accurate than Tesla's, with the implication being that we must be wrong and Tesla must be right, that's just patently ridiculous. It's not a complicated procedure to figure out whether or not the car's odometer is accurate. We've had a bunch of people determine in several different ways that the odometer is not accurate. You'd have us keep this information to ourselves, rather than alert Tesla to it, for fear of wasting their time?
Finally, no one, as far as I know, is up-in-arms, demanding a fix for this. I simply let Tesla know that this problem exists. It's entirely possible that they had no idea the odometers were off. Perhaps their testing showed everything fine, and then somehow an older, incorrect formula made it's way back into the firmware, and has stayed there because no one informed Tesla there was an issue. It could be any number of things. If it's a simple fix, I hope we'll get it.
- - - Updated - - -
We didn't overlook that. I stated that I was using the Pirelli Sottozero Series II winter tires, from the Tesla winter wheel package. Others have stated the tires that they are using. What you are overlooking, however, is that this hasn't been a problem with Model S models in the past. Those models also had different tire options, yet somehow the odometers were accurate. Why should we believe Tesla suddenly doesn't know how to do something that they've been doing just fine in the past? It seems a lot more likely that this was just some sort of glitch or mistake that may be easily corrected. I really don't see the harm in bringing it to Tesla's attention, and allowing them to act on the information as they see fit.
Last edited:
Danal
electricmotorglider.com
Sottozero Series II now added to the sheet, and tabs arrange so you can click back and forth and watch things change.
This indicates the difference in calibration between these two tires is .04%. This is one twentieth of 0.8% measured difference. Comparing these two tires yields only one data point... so it is a bit early to conclude that tire model is/isn't a contributor... at the same time, that one data point tends to not support the idea that tire differences is causing the observed 0.8%.
Danal
Google Sheet of Tire / Revs calculations
This indicates the difference in calibration between these two tires is .04%. This is one twentieth of 0.8% measured difference. Comparing these two tires yields only one data point... so it is a bit early to conclude that tire model is/isn't a contributor... at the same time, that one data point tends to not support the idea that tire differences is causing the observed 0.8%.
Danal
Google Sheet of Tire / Revs calculations
Last edited:
EarlyAdopter
Active Member
I'm not suggesting it isn't worth mentioning to Tesla - it clearly is if many people have noticed and are concerned. I'm just noting that before people begin reaching for conclusions more data is necessary. Specially, on the 21" Continentals, which are conspicuously absent from this thread and complaints.
Tesla recently switched from staggered Michelins to non-staggered Continentals on the P85D, so they're out there. Also, the Continentals were standard equipment on P85's for many years, so the odometer may have originally been calibrated to them.
Anyone with Continentals on their P85D willing to do a distance check?
I think it was this part of the message you posted earlier that caused me to respond as if you thought we shouldn't be bringing this to Tesla's attention:
If I inferred incorrectly from that post that you felt my writing to Tesla was uncalled for, I apologize.
This is an interesting point. I hope someone with Continentals on a D chimes in.
Danal
electricmotorglider.com
Maybe this is part of Elon's announcement. To fix range anxiety they decided to redefine the mile.
:biggrin::biggrin::biggrin::biggrin::biggrin::biggrin:
Danal
electricmotorglider.com
When I hadn't received a response to my inquiry by a couple of weeks after I wrote, I wrote again. That was about ten days ago. I heard back immediately, from Jerome, saying eh'd check into it. I heard from him again just now. This time the answer was essentially that the error is within the SAE standard, and that all cars have some inaccuracy in their odometers.
Needless to say, I am disappointed. I had really hoped that together we were helping Tesla find and correct what was going to be a really simple problem for them to fix. The answer really doesn't sit well with me, for the reasons mentioned in my response, which I'll copy below. But as you will see from that response, I'm not pushing things any further. I think the point has been made, and the information is here, in black and white, if Tesla wants it. Clearly Tesla did something differently with respect to the odometers in the D models as compared to previous models. Perhaps it is something that can't be easily corrected.
This was my reply:
--
Thanks for the response.
We were just hoping that perhaps since all previous Tesla Models had much more accurate odometers (based on what those who have owned those vehicles have been saying), and since all the newer cars seemed to be off in the same direction, and by about the same amount, that there may have been some formula that simply needed to be adjusted.
Thanks again.
A) You have to understand what a lawyer is. If they admit a fault here, they could and would be sued for the mileage discrepancy with respect to tires, warranties, resale, etc. That's the reality of this situation.
B) For all you know the next update will solve this issue. Tesla famously doesn't tell you what bugs will be fixed and when.
C) It's the truth. The discrepancy is within the acceptable range as many have noted in this thread. That said, see #2. I think it'll get set closer to true at some point.
B) For all you know the next update will solve this issue. Tesla famously doesn't tell you what bugs will be fixed and when.
C) It's the truth. The discrepancy is within the acceptable range as many have noted in this thread. That said, see #2. I think it'll get set closer to true at some point.
