Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Wiki Sudden Loss Of Range With 2019.16.x Software

This site may earn commission on affiliate links.
Did stumble across this: (2018, so old news and broad. I don't recall reading the document though).

https://www.researchgate.net/public...ent_System_Hardware_Concepts_An_Overview#pf12

Extract:

The risk of fire can be lowered by placing temperature sensors inside the pack and reacting
appropriately to critical temperature values. In addition, sensorless temperature detection methods
using indirect sensing methods, like electrochemical impedance spectroscopy can be used
additionally. New methods to acquire temperature along wire harnesses within complex electrical
systems can also be utilized to add safety


I'll have to read up on

"electrochemical impedance spectroscopy"
 
I have the circuit diagram somewhere. 'L' signified inductor in my experience, and it might do nothing more than filter some ac component from the BMS circuit, (guess), but often inductors are a 'series' component, so if they fail open circuit they effectively disconnect part of the circuit.

Dead right about a few cents to fix 20k worth. A truly 'high integrity' sustainable business model would not allow this kind of fault to saddle the owner with a 20k repair bill. (although we don't know what Tesla would have charged to fix).

I'll have a look for the diagram.

[edit: found the incomplete diagram I was thinking of, and it doesn't include those components.]
EDITED: Upon further review of my comment, I have decided it would have gotten me higher on Tesla's $hit list than I already am so I have removed it. I'm already a fine hair away from getting permanently banned from ever being able to purchase from Tesla again.
 
EDITED: Upon further review of my comment, I have decided it would have gotten me higher on Tesla's $hit list than I already am so I have removed it. I'm already a fine hair away from getting permanently banned from ever being able to purchase from Tesla again.
You are probably better off that way. Tesla is sketchy company.
 
It seems to me that if this sort of situation is possible for a smattering of battery pack failures that this explains why some people have received a hybrid 350V replacement while others received a refurbished pack. This way, Tesla minimizes its warranty costs while replacing the old pack with a refurbished pack. The labor to remove and replace at the service center will be identical. Shipping costs to and from Fremont will be identical. Turn around time is faster, and Tesla can fix broken packs when convenient. These hybrid packs are used when there aren't sufficient refurbished 85 packs, or perhaps they are test mules for a future design plan.

It boggles the mind that Tesla would charge $20,000 for a refurbished pack that contains more used components than new. But then, I am not in tune with the cupidity that permeates the modern technological tidal wave.

It further boggles the mind that Tesla cannot train its technicians to repair these packs on site--it saves on shipping and likely the labor rate at service centers is less than the engineers and technicians in Fremont who would actually perform this sort of work. Then again, everything with Tesla is either a trade secret, secret formula, or other intellectual property that Tesla wants to guard to maintain a monopoly over all functions and features.

I wonder if any of this information has been disclosed to plaintiff's attorneys in the class action lawsuit. This could be a sticking point with the anticipated mediated settlement. One could see how $5-$10 of parts, plus ten hours of labor (just a WAG), plus an interminable wait for all these packs to be available, plus the strain on the understaffed service centers would affect a wide, prompt, and reasonable settlement.
 
Surely Tesla has regional battery recon, at least swapping BMS and other electronic parts at a module level.

Probably difficult to roll out network of trained / qualified personnel, especially internationally and given safety implications.
 
  • Helpful
Reactions: Droschke
I One could see how $5-$10 of parts, plus ten hours of labor (just a WAG), plus an interminable wait for all these packs to be available, plus the strain on the understaffed service centers would affect a wide, prompt, and reasonable settlement.

see below video; cost was 5k just to snip one cell off a module...interesting in this case that a SINGLE cell (shorted) took the whole car down (not typical I would assume).

 
see below video; cost was 5k just to snip one cell off a module...interesting in this case that a SINGLE cell (shorted) took the whole car down (not typical I would assume).


If everything in this video is as stated, this could be the end of a journey. The cell being 'resistive' (self shorting) is evidently not drawing enough current to blow its own fuse, but it is somehow able to render the pack unusable. He says that as soon as the cell is disconnected, its voltage drops rapidly. That means either that cell would get hot (wouldn't make sense) or is storing little energy. So the bad cell is no longer capable of storing energy but is capable of effecting the battery (by discharging it). This will discharge one brick only, and more than can be corrected by the balancing circuits.

A 'resistive cell' is a resistor effectively, turning energy into heat. Here is the point where things could go wrong.

This is my guess, but makes complete sense to me.

All the resistive cell has to do is create an imbalance that can't be rebalanced by those rows of 4 tiny 158 ohm resistors on the BMS board. A 'resistive cell' only needs to dispipate a couple of Watts to throw things off.

I guess it would be interesting to examine those resistive cells to find out exactly what the resistance is caused by (if not Dendrites or 'pre-dendrites') and understand how much such a cell can dissipate under worst circumstances. Edit: in order to know how much heat could be created and potential for cascading effect to surrounding cells.
 
Last edited:
Someone is reporting that Supercharging speeds have been uncapped for them on Twitter - maybe after the latest update (2020.48.26 ?)

https://twitter.com/GeorgKonjovic/status/1343239483920150530

Anyone else notice any difference?
I did see a small improvement, but not much.
V3 does make a difference indeed. When plugged in I saw 125Kw, which is even more than the 'before capping' 117Kw I was getting. Still though the charging curve is aggressively dropping and it still charges slower than before. So from 50 minutes for a 10-90% charge, V3 now gives roughly 1 hour and 5 minutes. V2 still sucks though (around 1 hour and 20 minutes).
In both cases, the estimated time remains tampered and shows pre-capping estimates
 
  • Informative
Reactions: Matias
V3 now gives roughly 1 hour and 5 minutes. V2 still sucks though (around 1 hour and 20 minutes)

Not doubting your observations. It's hard to see how a Tesla Supercharger (designed and managed by Tesla and with full access to car / battery data - V2 & V3 could both push earlier MS batteries to the battery's max) can somehow change a battery's ability to absorb energy (safely) .

Your V3 rate is not far off the pre-nerfed v2. If Chargegate wasn't a thing, you might guess they were reducing stress on v2 chargers.
 
Last edited:
V3 now gives roughly 1 hour and 5 minutes. V2 still sucks though (around 1 hour and 20 minutes).
In both cases, the estimated time remains tampered and shows pre-capping estimates

Yes, the estimates are so off when choosing anything from 90% or up. Albeit, when setting max charge to e.g. 80% or below the calculation seems to be better, similar to the estimated available range in the energy app. The newest for my 2015 S85 so far was 2020.36.11 and it seemed the charging times when routing through several SuCs was more accurate, if the charge goal was below 80% than with earlier versions. But then again, since I live and drive in Germany, the routing were never able to reflect the 150kmph I used to drive when allowed, thus I am running the abrp/classic web routing in parallel to setting the next SuC in the Tesla routing.

Did anyone notice if the routing still does the recalculation of charging times?
With earlier versions, I noticed when routing a longer distance the SuC stops shows a relatively short charging time, but after a couple of seconds the charging times are significantly increased.Thus the SuC stops seems to be optimized for newer cars whom are capable of higher charge rates per soc. With e.g. abrp/classic I am able to limit the max charge, add average consumption and max velocity plus degradation, ambient factors as temperature and loaded weight etc.

I'm wondering, did the non-linear portion of the soc in percentage changed in the newest f/w?
You wonder what I'm referring to with non-linear? Let me explain, when it comes to a gas tank, I expect a certain linearity, filling it up to e.g. 60l I expect to have 30l at 50% and so on. This seems not the behavior in my S85, I've noticed this back with v8 or v9: When looking at the estimated range in e.g. the energy app at high charge levels, the kWh equivalent is higher than the actually used energy.
It will become more clear when you test it the other way around:
Next time you are at a SuC and to reach your destination or next SuC you need more than 15% soc to add. The navigation shows the approximate charging time and you are able to choose "Weiterfahren" (drive on?). The navigation algorithm will calculate a negative soc value for your next stop, if it is too negative, you will get some placeholder. But somewhere in the minus 20%ish it starts showing negative values. When it reaches 5% soc at next destination, it stops increasing albeit the soc is still increasing, this goes on for ~6% and then it starts increasing again in parallel to the soc.

Is this still the same with the most recent f/w or even worse?

BR! Oaito.
 
Just sharing here as it seems to be another potential data point -- apologies if this instance is already known, it's hard to keep on top of such an active thread:

https://www.washingtonpost.com/technology/2020/12/28/tesla-battery-fire/?utm_source=reddit.com

"
Seconds after Usmaan Ahmad heard metallic bangs in his Tesla Model S last month and pulled off a suburban Dallas thoroughfare, flames started shooting out of his five-year-old car.


The sound was like “if you were to drop an axle of a normal car” on the ground, Ahmad, 41, said. Only the car was intact, he recalled. Suddenly, as he stood on the side of the road, the car ignited in flames, concentrated around the front passenger-side wheel.

The National Highway Traffic Safety Administration (NHTSA) is evaluating the fire of Ahmad’s vehicle in Frisco, Tex., and has contacted Tesla over the matter, NHTSA spokesman Sean Rushton said this month. The agency opened an investigation last year into alleged battery defects that could cause fires in older Tesla sedans and SUVs.
"
 
Just sharing here as it seems to be another potential data point -- apologies if this instance is already known, it's hard to keep on top of such an active thread:

https://www.washingtonpost.com/technology/2020/12/28/tesla-battery-fire/?utm_source=reddit.com

"
Seconds after Usmaan Ahmad heard metallic bangs in his Tesla Model S last month and pulled off a suburban Dallas thoroughfare, flames started shooting out of his five-year-old car.


The sound was like “if you were to drop an axle of a normal car” on the ground, Ahmad, 41, said. Only the car was intact, he recalled. Suddenly, as he stood on the side of the road, the car ignited in flames, concentrated around the front passenger-side wheel.

The National Highway Traffic Safety Administration (NHTSA) is evaluating the fire of Ahmad’s vehicle in Frisco, Tex., and has contacted Tesla over the matter, NHTSA spokesman Sean Rushton said this month. The agency opened an investigation last year into alleged battery defects that could cause fires in older Tesla sedans and SUVs.
"
This case is disturbing in a couple of ways..
1: it happened suddenly at normal speeds. No indication of an issue from BMS. Just Boom.
2: His battery was at 60%. So much for the nerf to 'protect the longevity the battery'.
3: Tesla seems to be going back on the promise made by Elon, that they would replace your Tesla and its contents if the battery caused the fire.

Why do I get the feeling that I am about to suddenly lose another 25% of my capacity "in an abundance of caution"?
This is getting plain out assinine.
 
Just sharing here as it seems to be another potential data point -- apologies if this instance is already known, it's hard to keep on top of such an active thread:

https://www.washingtonpost.com/technology/2020/12/28/tesla-battery-fire/?utm_source=reddit.com

"
Seconds after Usmaan Ahmad heard metallic bangs in his Tesla Model S last month and pulled off a suburban Dallas thoroughfare, flames started shooting out of his five-year-old car.


The sound was like “if you were to drop an axle of a normal car” on the ground, Ahmad, 41, said. Only the car was intact, he recalled. Suddenly, as he stood on the side of the road, the car ignited in flames, concentrated around the front passenger-side wheel.

The National Highway Traffic Safety Administration (NHTSA) is evaluating the fire of Ahmad’s vehicle in Frisco, Tex., and has contacted Tesla over the matter, NHTSA spokesman Sean Rushton said this month. The agency opened an investigation last year into alleged battery defects that could cause fires in older Tesla sedans and SUVs.
"

Yes, this incidence was previously reported, but the recent WP article, referred to as a "hit piece" below, is already "pissing off" some:
Washington Post Tesla Fire hit piece

On Edit: Fixed the link.
 
Last edited:
  • Informative
Reactions: lightningltd