-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Range dropping abruptly at end of 50+ mile commute
- Thread starter insaneoctane
- Start date
AlanSubie4Life
Efficiency Obsessed Member
I assume at this point you've tried every single vampire drain related "fix" concurrently (disable mobile access, remove USB key, remove any accessories, turn off wifi, turn off data sharing, no cabin overheat protection, etc.).
It's an interesting looking problem. It sort of does look like some sort of rebalancing, as was suggested earlier by someone. Does it lose about the same amount of miles regardless of the driving end point state of charge? Looks like a decaying exponential...rate of change proportional to the difference in two quantities...
scottf200
Well-Known Member
Back around October, I had similar experience with my Model 3, except it happened before my commutes. I have it set up to finish charging to 80% by 8am, and when I leave for work at 8:30am, the range was often ~10 miles less than what the push notification showed. (IOW, lost 10 miles between 8:00 and 8:30)
I have a hard-wired dashcam, so I've attributed the range loss to charging the 12V battery. The balancing theory also makes sense. It has stopped happening since the recent updates, but I haven't been monitoring it as closely.
I have a hard-wired dashcam, so I've attributed the range loss to charging the 12V battery. The balancing theory also makes sense. It has stopped happening since the recent updates, but I haven't been monitoring it as closely.
Last edited:
So armed with ScanMyTesla and OBDII/CANBUS data I really don't know much more! It's still happening and I don't see things like I'd expect (like cell voltage dropping). Here's the only things I've got which shows cell temp dropping:
AlanSubie4Life
Efficiency Obsessed Member
Cool. You should plot the cell imbalance, pack voltage, and other cell parameters (max and min cell voltages) like that to see whether anything correlates. It doesn't really appear to be temperature related and it's not like this is a very significant temperature shift anyway.
Last edited:
The other params you suggest are very very steady during this time (to my surprise). It's an hour of data with over 800K rows of data, so not easy to work with....but I can plot anything (including boring plots like the cell imbalance which is basically ~4.0 mV over the entire hour. My SMT imbalance data appears to be pretty coarse- it appears to be 2.0mV of resolution. I will plot it for completeness and share.
AlanSubie4Life
Efficiency Obsessed Member
Mysterious. Sounds like you need to pare that down a bit to a region that is changing (say 27 to 30 minutes) and then just plot everything.
I'm mostly out of ideas. I guess you should check the size of the buffer and nominal remaining and see what they are doing. Those would both affect SoC but not change full capacity. Presumably nominal remaining is just going down. But maybe the buffer size is going up! (Very unlikely to be changing much, as I believe it's supposed to just be 4.7% of the full pack capacity or so.) But one or both of them have to be changing of course.
Last edited:
omgwtfbyobbq
Active Member
It could just be that Tesla's SOC calculation is in part a function of temperature, especially over a specific interval, eg 27-29C.
Is the current draw from either the 12V or large battery changing during that same time period?
AlanSubie4Life
Efficiency Obsessed Member
Yeah, not much there. I suppose if it goes the opposite way in general for the reducing SoC times that might mean something. It's weird how your battery is sometimes magically recovering and other times magically decaying.
Sorry...the above may make more sense if you have time to fiddle...
If I had to guess it's going to be the really obvious (and not useful) observation that nominal remaining is increasing/decreasing as the SoC changes. And buffer size will remain unchanged.
Sorry...the above may make more sense if you have time to fiddle...
If I had to guess it's going to be the really obvious (and not useful) observation that nominal remaining is increasing/decreasing as the SoC changes. And buffer size will remain unchanged.
This one is nominal remaining, to charge complete and buffer vs SOC:
This one is min/max/mid cell voltage:
This one is min/max/mid cell voltage:
One more.
Here's the 12V current plot. Funny how much data there is for 12V, but there is zero data for big battery voltage or current for this whole hour that my car sits in my garage! Maybe because it's not in drive? Still sad......
Here's the 12V current plot. Funny how much data there is for 12V, but there is zero data for big battery voltage or current for this whole hour that my car sits in my garage! Maybe because it's not in drive? Still sad......
AlanSubie4Life
Efficiency Obsessed Member
I think your BMS is just defective in some way. It doesn't seem to be of much consequence though (it seems to be netting out to about zero?). It seems like it goes down if the state of charge is above some level, and up if it's below some level.
If you leave your car sitting, at say, an SoC of 50%, does it change very little? Seems like right around that point is what it is "trying to get to" (for unknown reasons).
If you leave your car sitting, at say, an SoC of 50%, does it change very little? Seems like right around that point is what it is "trying to get to" (for unknown reasons).
@AlanSubie4Life , can you elaborate on your understanding of the significance of the "nominal full pack" value with respect to SOC? Specifically, should it be possible for the GUI displayed SOC to change by 15 or 16 miles while the nominal full pack value holds the same? I was thinking about this and thought that SOC was just a constant X nominal full pack, no? I'm still trying to make sense out of my post-commute phenomena here....
AlanSubie4Life
Efficiency Obsessed Member
Of course. I don’t have any such tools, I’m just an armchair observer here, but the remaining energy is what is changing.
SoC% = (RemainingkWh - BufferkWh)/(NominalFullkWh - BufferkWh)
AlanSubie4Life
Efficiency Obsessed Member
Since Buffer kWh is 4.5 % of nominal full kWh, this can also be expressed as:
SoC % = (RemainingkWh - 0.045*NominalFullkWh) / (0.955*NominalFullkWh)
By the way, even if you modify this statement to " SoC just a constant x nominal remaining energy," it is still not a true statement. That is why the SoC and rated miles are not equivalent and not interchangeable. The SoC depends on both the remaining energy and the full energy, while the rated miles depend only on remaining energy.
SoC % = (RemainingkWh - 0.045*NominalFullkWh) / (0.955*NominalFullkWh)
By the way, even if you modify this statement to " SoC just a constant x nominal remaining energy," it is still not a true statement. That is why the SoC and rated miles are not equivalent and not interchangeable. The SoC depends on both the remaining energy and the full energy, while the rated miles depend only on remaining energy.
Yes, your replies have resulted in me thinking about this properly again (I think). For my unusual range loss phenomena, my nominal full energy is staying constant, but my remaining energy is what is decaying after my commute in and then growing back after my commute home. What's interesting is that when I plug my car in upon arriving at home, while I can verify no energy is being supplied from the wall (because I have CT taps on my EVSE), the API data shows the energy growth as "charging" under the energy_added attribute. So, it looks like crazy high vampire loss at work (15-16 mi per hour) and then magical charging from heaven at home (even faster!). Still mentally struggling to understand since the voltage imbalance reports things are balanced at around 4mV upon arriving at both work and home. I rarely see full recoup but close (lose 15-16 and gain 11-14), but getting most back makes me a little less hyper that my car is broke, but as an engineer this behavior is irritating me!
AlanSubie4Life
Efficiency Obsessed Member
Your situation is truly bizarre, and you have documented it very well. I have no idea what is happening.
Similar threads
