-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
tomsax
Member
The survey just collects the data that's available to the driver. If the differences in apparent battery capacity are due to charging habits, that should show up on the survey because I am collecting data on both of those things. If we find a better measure of battery pack capacity, I'll add that to the survey. If someone has a theory about how to improve the accuracy of the readings, or ways to improve battery longevity, the survey data should be able to help validate (or disprove) that theory.
In which thread? I looked at post 28 in this thread and in the survey thread. Neither seems relevant. You can get a link to a specific post by right-clicking (Windows) or control-clicking (Mac) on the post number.
If you have thoughts on how the survey could be improved, I'm always open to suggestions. However, this probably isn't the best thread for that discussion. Please PM me or post on the Model S survey thread.
Kalud
Active Member
Just saw the pics of different battery packs labels, decided to go check mine, delivered late November. Vin P25610.
Part number: 1014114-00-D
Part number: 1014114-00-D
chickensevil
Active Member
I think the good take away from all this (always looking for a positive takeaway when I can!) is going to be that they have revised their batteries at LEAST 4 times now in the span of about 1 year's time. Most likely this consists of either A: a better performing battery (as far as longevity and such) or B: a lower cost battery (maybe it's both?).
If cost is a factor, this is great news on the "lowering the cost of the battery front" as far as getting to Gen3, and also increasing their profit margins since they get a cheaper-to-make battery but haven't changed the price of the car.
If cost is a factor, this is great news on the "lowering the cost of the battery front" as far as getting to Gen3, and also increasing their profit margins since they get a cheaper-to-make battery but haven't changed the price of the car.
Post 28 in the thread page I linked.
That's the thing, there is no way to measure how/if a pack is out of balance, unless you have the service password. Rated range will not tell you if your pack is seeing degradation, unless you know how well your pack is balanced. Inaccurate information is useless.
Mods, can you please move these four posts to the thread I linked above, or better yet, to the thread titled "decreasing rated range".
wraithnot
Model 3 VIN #2942 Model S VIN #5785
Bluetinc shared a really nice dataset where he range charged starting from 1 rated mile (8% SOC as gauged from by the REST battery_level parameter) on the 120 KW supercharging thread: Finally 120KW Supercharging! - Page 23
Since he has an "A" battery, and I have a "B" battery (battery SN 6864) I thought I would compare this dataset to the the data I got in July when I was 90 kW limited by the v4.5 software and to the data I got in December when I was 120 kW capable (v5.8). The first plot compares Bluetinc's charging power vs. SOC (state of charge from the REST battery_level- values are integers) in Delaware on Jan. 3rd to the same type of data I gathered at various California 90 kW superchargers between July 19th and July 27th.
The second plot shows bluetinc's SOC vs. time and my SOC vs. time for my most complete 120 kW charging sessions (Tejon Ranch on Dec. 20th).
Since bluetinc started charging at a much lower SOC, I used the portion of his data that overlapped with my data and adjusted his time values accordingly. At least for these two charging sessions, it took bluetinc's 90 kW capable "A" battery about 5 minutes longer to go from 14% SOC (16 rated miles on my car) to 89% than it took for my 120 kW capable "B". It's risky to draw any general conclusions from this since these weren't side-by-side charging sessions and Tesla probably tweaks the supercharger software all the time without telling anyone. But I found the comparisons interesting so I decided to share them. And yes, I'm a total data nerd. But I don't think I'm alone in that respect . . .
- - - Updated - - -
I'm always a fan of having more data out there rather than less data. I actually find the survey data quite encouraging as some of the cars with the most mileage are reporting very minimal degradation: Plug In America (perhaps because people with the highest mileage have done the most road trips that required deep charge/discharge cycles). If people figure out a robust pack balancing procedure then perhaps the last time an owner balanced their pack can be added to the survey.
Since he has an "A" battery, and I have a "B" battery (battery SN 6864) I thought I would compare this dataset to the the data I got in July when I was 90 kW limited by the v4.5 software and to the data I got in December when I was 120 kW capable (v5.8). The first plot compares Bluetinc's charging power vs. SOC (state of charge from the REST battery_level- values are integers) in Delaware on Jan. 3rd to the same type of data I gathered at various California 90 kW superchargers between July 19th and July 27th.
The second plot shows bluetinc's SOC vs. time and my SOC vs. time for my most complete 120 kW charging sessions (Tejon Ranch on Dec. 20th).
Since bluetinc started charging at a much lower SOC, I used the portion of his data that overlapped with my data and adjusted his time values accordingly. At least for these two charging sessions, it took bluetinc's 90 kW capable "A" battery about 5 minutes longer to go from 14% SOC (16 rated miles on my car) to 89% than it took for my 120 kW capable "B". It's risky to draw any general conclusions from this since these weren't side-by-side charging sessions and Tesla probably tweaks the supercharger software all the time without telling anyone. But I found the comparisons interesting so I decided to share them. And yes, I'm a total data nerd. But I don't think I'm alone in that respect . . .
- - - Updated - - -
I'm always a fan of having more data out there rather than less data. I actually find the survey data quite encouraging as some of the cars with the most mileage are reporting very minimal degradation: Plug In America (perhaps because people with the highest mileage have done the most road trips that required deep charge/discharge cycles). If people figure out a robust pack balancing procedure then perhaps the last time an owner balanced their pack can be added to the survey.
I have already suggested directly to Tesla using the A packs in "real" loaner cars that do not get sold out from under the service centers (in addition to continuing to provide sexy P85/P85+sales cars to induce existing customers to upgrade). This solution does not cost the company anywhere near the 120M fix that has been suggested. Plus, the "real loaner" cars can be sold in a year and will provide many people who can't afford a new Tesla with a Model S at the mythical 50K price point (I may buy one as an everyday driver instead of a Model E-----p.s. after going on a few road trips to wine country 2 cases latter I like the bigger car and want to upgrade to the fastest SC possible). Also, I would like to clarify that I am NOT asking for a FREE replacement pack. I am advocating for a prorated battery swap upgrade with a Sig tax credit of 3-5K and a $1,000 credit for 2012-13 production A pack cars. They can give us a certificate now for the battery swap effective as much as 2 years from now to allow the company more cash flow growth. IMO this is a fair solution to "A-packers" and Tesla.
Thanks for pulling that together, I had wanted to do this also, but hadn't had a chance. One other fact that should be noted with the datasets overlaid like this, is that at the point noted (89%), Wraithnot has ~230 miles, while I have ~214. This about matches the difference in our battery capacities.
Peter
I'd just like to point out that it's possible that the "improved taper" Elon Musk promised may not have been implemented yet, in which case the current data is irrelevant on that point. Even if it hasn't been implemented yet, Tesla has allegedly stated that it *won't* be implemented for the "A" packs. Hopefully, this is subject to change.
In that vein, would you mind glancing at post #1267 and explain why my taper from 90 kW begins at < 40% while both you and bluetinc begin the same taper at 45% (judging from your graph)?
Where does Tesla state this?
I saw that too, and I can't explain it. I can tell you that I had just driven a couple of hours to the supercharger, it was cold that day (~20 degrees), and I was the only one charging while plugged into 2B in Delaware, and I was running 5.8.4
. I also think that was the fastest supercharge I've ever seen, but I don't have any data to back that up. I'll do another charge on a warmer day to see if it changes then.Peter
Very inaccurate data doesn't do anyone any good. More inaccurate data is definitely not better, in fact it makes things worse as people freak out for nothing.
That's my point, there is no easy balancing procedure. Your average user is not going to figure any of this out.
A neanderthal sure can!
Ok, ok. I understand the nuance - but what observable variation does it introduce?
My 90% SOC to 100% SOC is 220 -> 250 vs. a newer car that might be 235-265. You'd have to shift the graph 15 rated miles to account for that. If you look at any recent graph and imagine shifting it 15 miles, it makes a HUGE difference.
wraithnot
Model 3 VIN #2942 Model S VIN #5785
It looks like you have a more complicated power/SOC curve than I observed in the data from my car and bluetinc's. Our data seems pretty consistent with a constant 90 kW or 120 kW, an exponential decay, and then a linear drop at really high SOCs. The power curve in your car seems to behave differently- and your two replicates look really close to each other. For the datasets I plotted, the superchargers were nearly empty and the temp was below 100F- I exlcuded datasets where I was the second to plug into a supercharger stack or when things got really hot at Harris Ranch and the power never got close to 90 kW. Was your thermal management system making a lot of noise when you were recording your two charging sessions? While this is complete and total speculation on my part, perhaps different battery types have different sensitivities to temperature and are programmed to start limiting power when the battery pack reaches different target temperatures. Once again, total speculation on my part. But this would be consistent with the available data.
wraithnot
Model 3 VIN #2942 Model S VIN #5785
If after 10,000 or 20,000 miles a Model S displays much less range than it did when it was new then people are going to start freaking out and posting about it no matter what. Some nefarious TSLA short won't have too much trouble finding such posts. If it's easy to access a lot of reported range data from different users and that data includes reports from high mileage cars that still have lots of reported range then hopefully non-enthusiast owners will start to look into things and learn about balancing vs. real degradation. Of course if that collection of information included a prominent statement about the possible explanations for a Model S showing less than expected range, that would certainly help.
@bluetinc, @wraithnot - Thanks for your feedback. I'm still scratching my head over it. No, I don't think thermal management played a role here. Both sessions were in the 60-70 F temp range. I also didn't hear much noise coming from the car itself (most noise was from the supercharger tower).
If that were the case would it not finish FASTER with less potential to store??
No, because the tapers should all be based on SOC, not rated miles. The rated miles is just to make it easier for the end user to understand.
Similar threads
