-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Kraken
Member
I think we might be viewing this all wrong....
i think they made a change to the algorithm sometime around 4.5 or 5.x that was having un foreseen consequences. I think the end result was that it was mis-estimating the range more and more over time due to not using the full battery. The unaccounted for range was increasingly being pushed essentially below zero. Because the car was guessing low, the amount it was missing by was going below zero.
So, what they did was correct it (or revert back to the old algorithm), so now it is more accurate, however there is less below zero than there was before because those miles were put back up at the top where they were intended to be.
so yes they did decrease the range below zero, but only because it wasn't supposed to be there in the first place.
alternatively, perhaps tesla is just moving numbers around into different combos to see which is customers prefer.... A bigger number with no buffer or a smaller number with more buffer. Initially they had the first and we complained. Then they moved to the latter, and others complained. Now they have moved back to their original because there is always a complaint.
i randomly gained back 9 miles today before installing the update. I got home with almost no miles and I didn't want to update with a low battery, so I hit the super charger. Got back to the house with 103. Then I remembered I had forgot to start the update, so I went back to start it and it said 112. No setting changes or anything. I hadn't done the update yet. I was going to see what the post update charge was... But the update reconfigured my charging settings and started charging instantly (during peak hours too)... So I was at 196 when I checked on it.... Or maybe I gained 84 miles!
i think they made a change to the algorithm sometime around 4.5 or 5.x that was having un foreseen consequences. I think the end result was that it was mis-estimating the range more and more over time due to not using the full battery. The unaccounted for range was increasingly being pushed essentially below zero. Because the car was guessing low, the amount it was missing by was going below zero.
So, what they did was correct it (or revert back to the old algorithm), so now it is more accurate, however there is less below zero than there was before because those miles were put back up at the top where they were intended to be.
so yes they did decrease the range below zero, but only because it wasn't supposed to be there in the first place.
alternatively, perhaps tesla is just moving numbers around into different combos to see which is customers prefer.... A bigger number with no buffer or a smaller number with more buffer. Initially they had the first and we complained. Then they moved to the latter, and others complained. Now they have moved back to their original because there is always a complaint.
i randomly gained back 9 miles today before installing the update. I got home with almost no miles and I didn't want to update with a low battery, so I hit the super charger. Got back to the house with 103. Then I remembered I had forgot to start the update, so I went back to start it and it said 112. No setting changes or anything. I hadn't done the update yet. I was going to see what the post update charge was... But the update reconfigured my charging settings and started charging instantly (during peak hours too)... So I was at 196 when I checked on it.... Or maybe I gained 84 miles!
Yeah, maybe, but what I would've really liked to see them do is incorporate a balance cycle at the end of each charge (or at least make it an option). Eliminating the below 0 mile buffer is just a quick fix for degradation concerns, even if it was placed there by accident, which I doubt.
You can pretty much guarantee that Tesla is already doing it's best to keep the pack as balanced as possible. Kraken's theory does make a lot of sense.
Kraken
Member
What if there wasn't any serious balancing issues....? Maybe in most or almost all cases it was the continuation of the algorithm slowly causing more and more range to be hidden below zero, and we interpreted it as a lack of balancing occurring over time.
even our rebalancing methods might only have been having much success because those same techniques help teach the algorithm the batteries true potential (except for the rare cases where someone was truly far out of balance).
perhaps we we won't be talking about rebalancing much anymore because our packs will rarely "seem out of balance" because that wasn't the issue...
or perhaps it was a combo of mainly algorithm with a bit of out of balance and some minimal degradation. Maybe the balancing fix is more thorough and in V6
but for now maybe they are just putting back what wasn't supposed to be slowly sliding below zero. Say they originally intended for 15 miles below zero. Also assume that 2 months ago I was getting (90% charge) 225 rated with 20 below zero, but a month ago I was getting 220 with 25 miles below zero. now I'm getting 215 rated with 30 miles below zero. If they fix the algorithm so I'm back to getting 230 with 15 miles below zero which is hypothetically where I started, is that a bad thing for them to do? Maybe then I am 7 miles worth of out of balance and 5 miles degraded... Then the car naturally rebalances someday or they do improve that too in v6.0, and I'm able to get 237 with 15 miles below zero and 5 of degradation... That wouldn't be too bad. Maybe the start is just getting the algorithm to work the way it was intended to. (Note that all numbers and everything are hypothetical).
Last edited:
(LMB spouse)
LMB's car is a US S85 with a max rated range of about 252 miles using 5.8.x firmware after about 12,000 miles. Sometime last fall, we range charged the car, then drove about 80 miles and visited the local service center. They checked the battery and said that everything looked normal and also that the battery was well balanced.
We typically charge the car somewhere between 60 and 75% i.e. we try to "baby" the battery. This seems to be typical of those reporting similar max ranges. This suggests that the issue is algorithmic rather than balance related, assuming that lots of batteries are not experiencing substantial losses in the first year,
LMB's car is a US S85 with a max rated range of about 252 miles using 5.8.x firmware after about 12,000 miles. Sometime last fall, we range charged the car, then drove about 80 miles and visited the local service center. They checked the battery and said that everything looked normal and also that the battery was well balanced.
We typically charge the car somewhere between 60 and 75% i.e. we try to "baby" the battery. This seems to be typical of those reporting similar max ranges. This suggests that the issue is algorithmic rather than balance related, assuming that lots of batteries are not experiencing substantial losses in the first year,
Geez, don't spoil this. Isn't it much more fun to just read all of the emotional responses and wild assed guesses?
Seriously, though… the "science" of this forum has advanced to the point where - given a couple of weeks for people to measure - we should see some pretty useful reporting. I'd discount anything reported within 14 days of 5.9 release unless it's from a beta tester who's had it a while. Good analysis requires more than 1 or 2 hasty data points.
Kraken
Member
Yeah... Give it a couple days. Sometimes bound to wind up hitting 0 a few feet short of their driveway... They'll let us know when they get stranded.
You and your fancy new D pack. :tongue:
For those that have 5.9: When does your 30 mile average W/mi = rated consumption? Is it back up to 308 W/mi or is it the ~280 some such number reported in cinergi's thread?
If the range was being pushed below 0, I would expect that to show (to me) as an increasing pack voltage when 0 range was shown on the dash. Unfortunately, my 0 rated range pack voltage seems to be very consistent at 320V though there seems to have been a change from 4.X to 5.X where my 0 rated range pack voltage changed from 316V (Consistent since new) to 320V. Those 4V of potential are only worth a few miles at that point in the V-Wh curve.
Once I get 5.9 I'll recheck this and let you know.
Peter
Once I get 5.9 I'll recheck this and let you know.
Peter
I brought it down to 8 miles. SOC was 3%. Hope this answers the others' questions.
- - - Updated - - -
I should add that previously it was over 10% for the same number of miles. I am wondering if "B" packs will see the same symptom, or if this is a fix for "A" packs to look more like "B" packs.
Well now, that is interesting.
Assuming you could drive it those remaining 26.7 miles to 0 miles range, that would imply a 0% SoC as well, which is bricked battery territory.
So the possibilities as I see them are that the new firmware:
- Eliminates the below-zero buffer, as well as the anti-brick buffer, allowing you to drive to zero and make a $40,000 error
. - Eliminates the below-zero buffer, and suppresses reporting of the anti-brick buffer, allowing you to drive to zero and no farther, in essence "hiding" the bottom of the pack charge from things like Visible Tesla
. - Calibrated range such that there are no more hidden buffers, but you can no longer drive to 0 range, thus preserving the anti-brick buffer, but making the estimated range goofy on the bottom end.
Quite frankly, none of these scenarios seem likely as intentional to me... but #2 would seem the most plausible.
I too am really interested in finding that out, as not only did my max rated (and ideal) range drop drastically, but based on the point at which projected miles matched rated miles, which dropped from 305 to 280, it became much more difficult to attain even that diminished range. If it has gone back to 305, I will be very happy.
Last edited:
FWIW a Leaf owner in Chicago is reporting zero degradation on his Leaf in 12 months and 16k miles. This measurement is using a CAN bus meter that captures data directly from the battery management system.
On the other hand there are many others from the same area who have reported anywhere between 5 and 10% loss.
On the other hand there are many others from the same area who have reported anywhere between 5 and 10% loss.
Kraken
Member
I'm going to go with number two, which is sort of what I was leaving towards. I also wouldn't be surprised if they are masking the amount below zero that you can drive, thus making the reporting of 0% SOC to things like visible Tesla not inclusive of the hidden driving range or the brick protection amount. I don't they really wanted us knowing the actual numbers. It's better to have people playing out safe. I know people want the info, but from their perspective, some things are better if they aren't public knowledge, even if some people might make good use of it out prefer to have it.
My Wh/mi --> consumption has not changed with 5.9. It's around 288 to match rated.
Similar threads
- Question