-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Remember, this only happened when I was intentionally trying to get the battery to low SOC. A little while earlier it was Vmin 3.57, Vmax 3.65 (and still at very low SOC). I think maybe one sheet just had a tiny bit less charge than the others and was falling off the cliff first.
Here's an update with new data for three cars: #277 (EU), #255 and #670 (mine). There's most of a year's new data for the first two, which is always interesting.
#277 (EU) is now the third highest mileage in the study, after me and #181 (Dave Denhart). Interestingly, all three cars showed a flat spot in the CAC decline around 11K miles, though that may just be an accident related to when we all range charged. I'll be very curious to see what happens with 435 and 95 (EU) (both of which haven't updated in > 6 months, and so probably have thousands of more miles driven than are on the graphs).
255 showed a big jump, that immediately went back down. Three consecutive days show 192.5, 193.1, 191.4. The car wasn't driven between days 1 & 2, but had 205 miles between 2 & 3. I suspect that this was a range charge (which bumped the CAC up a lot), followed by running the battery down most of the way (which took it back down). This is pretty consistent with what I've seen in other cars.
My car continues to lose a little CAC after my running the battery to very low SOC. I hope the decline will stabilize, but only time will tell.
I'd love to get updates from some of the higher mileage cars that haven't provided data in a long while, like 181, 425, 95 (EU) and 537. You all know how to contact me.
#277 (EU) is now the third highest mileage in the study, after me and #181 (Dave Denhart). Interestingly, all three cars showed a flat spot in the CAC decline around 11K miles, though that may just be an accident related to when we all range charged. I'll be very curious to see what happens with 435 and 95 (EU) (both of which haven't updated in > 6 months, and so probably have thousands of more miles driven than are on the graphs).
255 showed a big jump, that immediately went back down. Three consecutive days show 192.5, 193.1, 191.4. The car wasn't driven between days 1 & 2, but had 205 miles between 2 & 3. I suspect that this was a range charge (which bumped the CAC up a lot), followed by running the battery down most of the way (which took it back down). This is pretty consistent with what I've seen in other cars.
My car continues to lose a little CAC after my running the battery to very low SOC. I hope the decline will stabilize, but only time will tell.
I'd love to get updates from some of the higher mileage cars that haven't provided data in a long while, like 181, 425, 95 (EU) and 537. You all know how to contact me.
backmost
Member
dhrivnak
Active Member
Yes. Make a USB with a directory called "VehicleLogs" in the top level (capitalization matters). Plug it in the car, and it will download the logs onto it (this is painfully slow, I always go away for a while and come back). Once it's done, a few levels below VehicleLogs will be a file who's name starts with the date, then some more numbers then ".tar". That's what I need.
Once you have it, DM me and I'll tell you how to get it to me.
VehicleLogs (note capitalization), at the top level. The stick needs to be 4 gb or less, formatted FAT32. The file is about 8 mb in size, so be sure you have enough free space. Other contents on the stick are fine; they're not bothered with.
The car can be plugged in while doing this, but not actively charging and not "on".
Excellent!
If you'd like to take a look at what's inside (it can be quite eyeopening), take a look at VMS Log Parser for Tesla Roadster for programs that can decypher and display the contents. I recommend VMSParser.exe for status and event (error) messages, and TeslaGLoP for graphical views of temperatures, power/current, speed, etc. over time, e.g. the course of a drive. Both are Windows programs, but work under Wine on Linux (and I presume Mac) if they are your platform of choice.
If you'd like to take a look at what's inside (it can be quite eyeopening), take a look at VMS Log Parser for Tesla Roadster for programs that can decypher and display the contents. I recommend VMSParser.exe for status and event (error) messages, and TeslaGLoP for graphical views of temperatures, power/current, speed, etc. over time, e.g. the course of a drive. Both are Windows programs, but work under Wine on Linux (and I presume Mac) if they are your platform of choice.
backmost
Member
I use a Macbook as my daily but can use my wife's Windows laptop if needed. Definitely curious to see what's going on behind the scenes
Updates for #573 and #209 (first from its new owner). 209 is smack in the middle of the I-never-range-charge curve with a bunch of other cars at the same mileage. 573 had been doing well on the by-days graph, probably because it doesn't have a lot of mileage on it, but it's been driven some over the summer and has moved back into a more normal part of the curve. Thanks for the data!
I had a not-so-great experience with my car Thursday. I did a standard charge and wound up needing to drive it a little more than I'd expected. After maybe 115 miles, the battery was down to about 60 miles range (after starting just over 200). OK, that's a little more ideal miles/mile than usual, but nothing big. I stopped to get some food, and when I came back out it had dropped to 40 ideal miles. I drove it home and when I got there I had driven 133 miles and had 13 ideal miles left (in standard mode, plus whatever the bottom 10% is worth). The max sheet was about 3% more charged than the average and min, so maybe it's a little out of balance. The next day, the CAC dropped a couple of amp hours. This is what I'd expect to see if the algorithm had overestimated the battery capacity. I really worry that I did some damage by getting the charge state so low a few weeks ago.
I had a not-so-great experience with my car Thursday. I did a standard charge and wound up needing to drive it a little more than I'd expected. After maybe 115 miles, the battery was down to about 60 miles range (after starting just over 200). OK, that's a little more ideal miles/mile than usual, but nothing big. I stopped to get some food, and when I came back out it had dropped to 40 ideal miles. I drove it home and when I got there I had driven 133 miles and had 13 ideal miles left (in standard mode, plus whatever the bottom 10% is worth). The max sheet was about 3% more charged than the average and min, so maybe it's a little out of balance. The next day, the CAC dropped a couple of amp hours. This is what I'd expect to see if the algorithm had overestimated the battery capacity. I really worry that I did some damage by getting the charge state so low a few weeks ago.
backmost
Member
#2 is a straight line because it's never driven, and the CAC only updates after some number of miles are put on. It's got 132 miles on the new battery and 4170 on the car overall. I wouldn't read too much into it.
The overall summary is this: the 3.0 batteries degrade pretty quickly, both based on age and on miles driven, so letting the car sit around unused doesn't help much. There's some indication that there is a knee in the curve after maybe 12% degradation, but that's based almost entirely on one car (mine). It's also clear that the CAC algorithm incorrectly estimates the battery capacity, and that it doesn't seem to get much information unless you charge fully or run the battery down to very low SOC. Full range charges tend to increase the CAC (repeated range charges as an experiment produced the big bump in #670 around 12K miles, for instance). Running the battery to low SOC reduces the CAC. Most likely, neither of these are actually doing much to the battery capacity, but rather are giving information to the estimation algorithm.
We still need more data to see what's really going to happen, but at least it's not as bad as I thought it might be when I first started this thread. If that had happened, my car would be around 145Ah by now and I would have made a warranty claim. Instead, it's in the 185-190 range, which isn't what I'd like but isn't terrible, either, depending on what it does over the next few years.
backmost
Member
I'm hopeful that the degradation levels off at some point, but if not it's sad to think these batteries will eventually match the capacity of the original batteries.
Not cool when you're spending $30k on an upgrade for a decade old car!
Not cool when you're spending $30k on an upgrade for a decade old car!
Kerios
Member
Agreed, and if the CAC drop off stablises that would be okay, if not great, but what bothers me and I'm sure it's true of everyone who has experienced the sudden drop of range from something useful e.g. 50miles to either 10 or 20, or in somes case to 0miles. This is a really good way to accelerate the onset of heart failure for the owner, and detracts a great deal of confidence in the overall capable range of the car. If the battery CAC starts to be consistent I'd be okay I guess with that, but I'd want the algorythm fixed so that one doesn't see this sudden death syndrome that appears prevalent.
Similar threads
