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

Tesla's 85 kWh rating needs an asterisk (up to 81 kWh, with up to ~77 kWh usable)

This site may earn commission on affiliate links.
^ That's a whole different can of worms than the capacity issue. Suffice it to say the 90 packs are much more prone to SoC imbalance between cell groups. If one group has over about a 20 mV delta from the pack average, charge speed will taper. The higher the delta, the worse the charge speed reduction.

I swapped out the 90 pack on my S with a newer one because I had two cell groups at something like 40mV delta from the rest of the pack... my top supercharger speed was about 50 kW no matter where I was and it made charging time on my last trip with that pack take 3x as long. The new one is already starting to develop its own uncorrectable imbalance after just ~6000 miles of use.

Suffice it to say... **** the 90 packs.
That's exactly what I've seen get worse over time....

Follow-up question. I reported a problem to service where I would have observed battery percentage just disappear. 3 times I visually saw the pack at 17-19%, left the vehicle, came back 20 min later with no HVAC use in between, and the charge was at 12-15%. Could this be a possible side effect of a large uncorrectable cell imbalance?
 
Failing is the wrong word. More like random, rapid, and extensive degradation (increase in internal resistance) of individual cells causing an uncorrectable imbalance (where the ~100mA bleeder resistors don't have a chance of maintaining balance under normal use).
Is this a possible reason why they tried to put launch limits on the car, so as to slow down a scenario that could contribute to this problem?
 
Could it be from the damage in the car accident you got your salvaged pack from?

Definitely not a physical damage issue nor related to any kind of impact. The cells themselves really aren't affected by such things unless the pack itself, and thus the cells, were physically impacted and damaged in the accident. My testing would seem to confirm this, as well as data from non-crashed cars.

That sounds like a quality control issue in the cell manufacturing process as opposed to an inherent flaw in the general cell chemistry.

Possibly. I don't know enough about the underlying chemistry to make a real prediction on that. It's weird, because the issue seems to cause a high IR for the cell during charging only. During discharge the same cell can behave as expected. Again, I'm no expert on the chemistry itself, but my guess is a quality control problem wouldn't manifest quite like this. Could be wrong, but, don't know for sure.

That's exactly what I've seen get worse over time....

Follow-up question. I reported a problem to service where I would have observed battery percentage just disappear. 3 times I visually saw the pack at 17-19%, left the vehicle, came back 20 min later with no HVAC use in between, and the charge was at 12-15%. Could this be a possible side effect of a large uncorrectable cell imbalance?

Most likely you're seeing the range loss as a result of the pack cooling down while the car sits. This is more exaggerated at lower states of charge and is the reason for the ambiguous "There will be significantly less energy available from your battery if it gets colder" (or close to that) popup message. The "it" in that sentence refers to the pack itself, not the weather. There is a calculable amount of additional energy available to the user when the pack is warm, due to decreased IR, than when the pack cools down. The change in IR is more dramatic with temperature when the pack is at a lower SoC.

So, for example, you drive for a while, car is at 20% SoC. By then it's likely the pack has been allowed to warm to a higher working temperature, usually between 35-45C, allowing the best IR possible for discharging efficiently. Then you park the car, and it sits for a while, cools pretty quickly in Ohio winter weather, and when you get back the IR of the pack has gone up to the point where there is no way to pull the same capacity out of the pack without losing a chunk of that to IR and internal cell heating.

Hopefully that makes sense.

Is this a possible reason why they tried to put launch limits on the car, so as to slow down a scenario that could contribute to this problem?

For sure it's a contributor. They also put the launch limits in place because they seem to have found that wear and stress on the cell level fuses can be cumulative at high currents, eventually resulting in a failure of one or more fuses, and thus causing an uncorrectable pack imbalance.
 
Last edited:
Possibly. I don't know enough about the underlying chemistry to make a real prediction on that. It's weird, because the issue seems to cause a high IR for the cell during charging only. During discharge the same cell can behave as expected. Again, I'm no expert on the chemistry itself, but my guess is a quality control problem wouldn't manifest quite like this. Could be wrong, but, don't know for sure.

I was thinking QC issue if some cells in the same pack did not exhibit this behavior while others did. If the basic chemistry had an issue I'd expect all cells in a pack to have similar problems. Could even be a QC issue in the cell chemistry since small amounts of specific additives can have dramatic effects, but I'd think a pack would have cells with the same batch of chemistry inside. Have you tried cycling individual healthy cells and see if they too start to degrade or remain stable? Certainly a puzzle.
 
Most likely you're seeing the range loss as a result of the pack cooling down while the car sits. This is more exaggerated at lower states of charge and is the reason for the ambiguous "There will be significantly less energy available from your battery if it gets colder" (or close to that) popup message. The "it" in that sentence refers to the pack itself, not the weather. There is a calculable amount of additional energy available to the user when the pack is warm, due to decreased IR, than when the pack cools down. The change in IR is more dramatic with temperature when the pack is at a lower SoC.

So, for example, you drive for a while, car is at 20% SoC. By then it's likely the pack has been allowed to warm to a higher working temperature, usually between 35-45C, allowing the best IR possible for discharging efficiently. Then you park the car, and it sits for a while, cools pretty quickly in Ohio winter weather, and when you get back the IR of the pack has gone up to the point where there is no way to pull the same capacity out of the pack without losing a chunk of that to IR and internal cell heating.

Hopefully that makes sense.

Makes perfect sense, thank you. Now back to the supercharging weirdness....
 
I was thinking QC issue if some cells in the same pack did not exhibit this behavior while others did. If the basic chemistry had an issue I'd expect all cells in a pack to have similar problems. Could even be a QC issue in the cell chemistry since small amounts of specific additives can have dramatic effects, but I'd think a pack would have cells with the same batch of chemistry inside. Have you tried cycling individual healthy cells and see if they too start to degrade or remain stable? Certainly a puzzle.
That would mean that having a newer repalcement pack would hopefully not contain cells with the same defects. Or at least less likely. But the only way to answer that is to look at @wk057 's data. Surely he has the 90kwh packs broken down by revision, but his comments earlier lead me to believe that all 90kwh packs are affected. Interesting idea.
 
Oh, I'd like to hear this can-of-worms......

@islandbayy What supercharging issues are you seeing? I'm seeing weird rapid tapers within 60 seconds of starting supercharging, tapering down to 90kwh and random capping in the 60kwh range even with less than 30% SOC.
Supercharging issues, wildly varying charge rates, and a maximum rate cap of 94kW, if I'm lucky.

I'm also seeing one module of cells, reach a higher temperature much quicker then others. As usual, Tesla says this is fine and dandy.

That one module, can be as high as 10*F hotter then the other 15 modules.

And I can assure you it's not a faulty sensor. At complete rest, such as sitting overnight, all modules are within 1-3 degrees of each other. This is mostly seen while supercharging or CHAdeMO Fast charging.
 
Last edited:
See attached for example.
 

Attachments

  • image.png
    image.png
    81.9 KB · Views: 116
  • image.png
    image.png
    83.5 KB · Views: 111
I also see slightly higher temps in the middle of the TM-Spy graph on my car (I can upload screenshots tomorrow). Do we know which module in the pack layout this is? I was thinking this perhaps could be the top one of the stacked modules in the front of the pack resulting in slightly higher temps?
 
  • Informative
Reactions: supratachophobia
That would mean that having a newer repalcement pack would hopefully not contain cells with the same defects. Or at least less likely. But the only way to answer that is to look at @wk057 's data. Surely he has the 90kwh packs broken down by revision, but his comments earlier lead me to believe that all 90kwh packs are affected. Interesting idea.

Good point. The newest variation of the 90 packs has a lot less data than the rest in my database, but the trend seems to be the same as the rest. My guess was the revision isn't cell chem related and instead just other hardware.

Supercharging issues, wildly varying charge rates, and a maximum rate cap of 94kW, if I'm lucky.

I'm also seeing one module of cells, reach a higher temperature much quicker then others. As usual, Tesla says this is fine and dandy.

That one module, can be as high as 10*F hotter then the other 15 modules.

And I can assure you it's not a faulty sensor. At complete rest, such as sitting overnight, all modules are within 1-3 degrees of each other. This is mostly seen while supercharging or CHAdeMO Fast charging.

@islandbayy - seems your observation is consistent with what @wk057 has indicated to be the case with these 90 kWh packs. The IR of a cell group in one of your modules is significantly higher than the others resulting in this module heating up more quickly.

See attached for example.

I also see slightly higher temps in the middle of the TM-Spy graph on my car (I can upload screenshots tomorrow). Do we know which module in the pack layout this is? I was thinking this perhaps could be the top one of the stacked modules in the front of the pack resulting in slightly higher temps?

The thermal loop is supposedly setup so that the modules are circulated in parallel. In practice, though, it's not completely perfect and some modules get less cooling or heating than others. This is usually fine, but when something like supercharging causes the cooling system to run at full power with no real chance of actually stopping the thermal rise in the cells completely, some deltas do happen. The modules at the front of the pack end up getting more cooling than those at the back, so it's usually an obvious gradient across the pack when looking at the data in larger sets.

I can't say for sure what @islandbayy or @bjornb are seeing with the single module temperature variation can be attributed to the high IR of the 90-type cells during charging, but it's possible. As I said, this only seems to manifest during charging... and I supposed if the cells this is happening to arent evenly distributed and there are more of them in one module, then sure it would cause a temperature spike during charging. You might be able to somewhat confirm this by checking that temperature delta during high discharge... like, run the car pretty hard for a little bit while the temperature is even and see if that same module ends up hotter. Probably not worth the effort, though.
 
  • Informative
Reactions: MP3Mike
I admit that I haven't read this whole thread, but thought this tidbit useful anyway...at Centralia, WA a few days ago (2014 S85, 40K miles, fair amount but not huge of supercharching) I was only getting 50kw on charger rated 135kw, called them, they checked out all the cars charging, and said that the electric utility was stiffing them and only delivering much less than contracted for....so that is a complication to factor in. I don't know if this is a common issue or not - it certainly has often been the case at Centralia.
 
I am also curious on Cabby’s monitoring. Any recommendations on getting started with canbus monitoring?
So probably not the thread for this *exactly* and I know @wk057 has a lot more sophisticated/intrusive access to the cars than a normal person would have, but..... here is what I found:

There is a thing called a ODBlink MX (also an LX model available, not sure if it matters for our cars. I bought the more expensive one, because you know, if it costs more, it has to be better, right?). That particular brand also seems to be able to take signals at a faster rate than the cheaper units, which does actually matter for our cars.
Then you'll need a custom cable which you can make yourself, or they can be found on Panjo. Keep in mind there are two types that are roughly split by model year; 2015 and older is one type, 2016 and newer is another. There might even be a color difference. Some have blue connectors.
Lastly I found two apps for Android (because why would anyone use anything else....?), TM-Spy and ScanMyTesla. Both are under $10 and I'm not sure if one is better than another. So basically that's it unless someone corrects me.
 
Agree, old 85s were 81. I had 71KWh usable after 1 year. Degradation has been excellent (low). I'm at about 6% after 3.6yrs and 78,000 miles. It wouldn't be so bad if I didn't have to make 200 mile trips without superchargers (none in ND yet, but I'm working on that), and those trips are impossible at highway speeds after year one and anything but perfect weather.
 
  • Informative
Reactions: FlatSix911