-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
wiztecy
Active Member
Well like I mentioned, Tesla has the brick's Vmin when the ESS is getting hammered, that alone is a strong data-point Tesla grabs and uses for the CAC data that's fed to the algo. That is not dependent on a shallow charge behavior. Its the same, the weak brick no-matter what SOC will dive like a rock compared to others.
And where do you think the "logs" go??? Into the CAC algo. Have you even realized Tesla keeps all the Temp data as well as the charge behavior in their logs? Why do you think they're collecting that? Its more data fed to the algo. Have you even analyzed what's on the stick/logs? Understand what it means? Everyone passed up and didn't know what the ahr.log file was, but I was looking though all this data and understood the relation and knew this was a key part of the data on the ESS, and identified it to the forum. Honestly I'm sure others have understood what this file was, but nobody pointed it out to open a discussion about it. And now that people who understand what this file is, they now can use it as a tool to understand what's going on with their ESS better than without it.
And again, a pass on the challenge. When someone says its "pointless" I say that's a cop-out. I'm trying to provide more information with real data, not arguing back in forth which gets people no where.
So after a range mode charge and low discharge, and even after a year where if my CAC is higher than 146, you'll still be a disbeliever. But like I mentioned before, I'm not trying to impress anyone nor tell people what to do. I'm providing data points and my best guess on how this stuff works to contribute in a positive fashion to the Roadster community.
Your personal CAC may vary depending on your attitude with life.
Last edited:
supersnoop
Tesla Roadster #334
Now it seems we have a comprehension problem. CAC is the single most accurate measure of a batteries health. Or it's the most accurate single measure. Would a full analysis of the complete history of the vehicle provide more information? Yes. Would it provide better information? Maybe, maybe not.
My CAC increased from 141 to 154. Why? Because the battery was replaced. But you won't see that in the logs. And it takes a long, long time for the CAC to recover from that event; eight months in my case. So, would I pay more for a car with a higher CAC? As I qualifi d in my initial response, maybe, maybe not. It depends on why. I never said I would pay more for wiztecy's car, so stop basing the success of your argument on that falsehood.
Voltage only tells a small part of the story. This is why Range predictions vary so wildly between different manufacturers. If you have driven a Leaf, you will know what I mean.
wiztecy
Active Member
But that's what Tesla got right and the others didn't, that's what's so proprietary about their algorithms, computations that give them the competitive edge. I *never* had a false range reading on my Roadster with a properly balanced pack. Ever. And I'm sure part or all of that same team worked on the CAC algorithm.
That's my point. If taking only Voltage into account was most of the picture, then there wouldn't be such a wide variance in range estimates.
wiztecy
Active Member
Battery Voltage Information – Battery University
"The voltage behavior under a load and charge is governed by the current flow and the internal battery resistance. A low resistance produces low fluctuation under load or charge; a high resistance causes the voltage to swing excessively. Charging and discharging agitates the battery and full voltage stabilization takes up to 24 hours. Temperature also plays a role; a cold temperature lowers the voltage and heat raises it."
"Voltage" does help us understand and it directly represents what's happening in terms of resistance is of the battery, and that's very important as to what's going on and how its doing overall unless I am not interpreting the above information correctly. A battery that has degraded chemically will have higher resistance holding other variables like temp constant. And there are other tests that can be done, as I mentioned to get a clearer picture and overall health of the pack.
Last edited:
smorgasbord
Active Member
You know, supersnoop, when your argument is based parsing individual words, you should watch out.
When your argument is based on reordering individual words differently, you're in trouble.
And when your argument is based on reordering individual words that weren't actually said, there's no hope.
witztecy said: The CAC is our and Tesla's best gauge for overall health of the ESS. There was no use of the word "single" in that statement. So, um, who's got the comprehension problem?
I'm screaming inside my head again. That's an argument against my saying using logs is better than just using CAC?
Hey, I gave you fair warning in advance that it was a trick question. I'm sorry it succeeded in tricking you, but maybe it's taught others not to put too much faith into CAC values that differ by only about 5%.
P
I really don't know what point you are trying to make. If it's as easy as you say, maybe you need to forward that link to Nissan's engineers, obviously they must have missed it...
smorgasbord
Active Member
That doesn't mean the CAC fully replaces the logs when evaluating battery health.
Have you even realized that I recently posted my battery temp data in this very thread? Post #54 to be exact.
Only because you're ignorant of the reasons, and apparently don't need to hear the reasons to attack them.
You're clearly arguing about something I never said. Earlier you said: "some believe it is, that a dive is purely degradation of chemistry and the CAC no-way can jump back up, that this number is useless." This is wrong on many counts:
1) No-one said that a CAC dive is "purely" due to degradation of chemistry.
2) No-one said the CAC can't "jump" back up.
3) No-one said that CAC is "useless."
What some, like me, have said is:
1) CAC values are not very accurate (no more than 5% overall), and Tesla reporting them in hundredths of an amp-hour is silly.
2) CAC values can be manipulated by driving and charging in certain ways.
3) CAC values aren't as good as looking at the car's entire log file history.
4) One should not spend more money on cars solely on the basis of reported CAC values.
Last edited:
wiztecy
Active Member
Ohh really? I easily could make the same statement. Its part of a discussion and educational argument. And how am I "attacking"? I'm disagreeing and that is a totally different thing.
What I'm seeing here is eristic.
You're stating (a) and I'm stating (b) based upon observations, habits, and characteristics of a battery. I am trying to validate my side in a concrete and educational manner to come up with a real true result. Not this back & forth thing. But you don't want to validate your side. If you have nothing to lose I don't see why you don't line them up to see where the numbers end.
Last edited:
Take a breath guys....
My CAC is at 149ish. It has been there for about a year now, having fallen gradually and steadily over the 4 1/2 years and 40,000km (25,000miles) I've driven. I've no doubt that my pack degradation has been impacted by the brutal summers we have here (where even with cooldown cycles it is hard to get the pack <31C for six months of the year). I've range mode charged perhaps once or twice in my 4 1/2 years of ownership. I standard mode charge to 90% most days in the summer, having driven it down to perhaps 80% on my 40km/day commute. In the winter, I worry less about pack temperature, don't need cooldown, so charge perhaps once or twice a week. I don't see any appreciable difference in CAC between the winter deep-cycles and summer shallow-cycles.
I guess the background is that I understand how a low CAC can suggest a pack is worse than it really is (unbalanced, bad estimate, whatever), but can a CAC ever be higher than the pack condition really is?
So, my question is: how do I bring my CAC back up? Tell me, and I'll try it for the next few months to see if we can see any difference.
My CAC is at 149ish. It has been there for about a year now, having fallen gradually and steadily over the 4 1/2 years and 40,000km (25,000miles) I've driven. I've no doubt that my pack degradation has been impacted by the brutal summers we have here (where even with cooldown cycles it is hard to get the pack <31C for six months of the year). I've range mode charged perhaps once or twice in my 4 1/2 years of ownership. I standard mode charge to 90% most days in the summer, having driven it down to perhaps 80% on my 40km/day commute. In the winter, I worry less about pack temperature, don't need cooldown, so charge perhaps once or twice a week. I don't see any appreciable difference in CAC between the winter deep-cycles and summer shallow-cycles.
I guess the background is that I understand how a low CAC can suggest a pack is worse than it really is (unbalanced, bad estimate, whatever), but can a CAC ever be higher than the pack condition really is?
So, my question is: how do I bring my CAC back up? Tell me, and I'll try it for the next few months to see if we can see any difference.
smorgasbord
Active Member
You said I was copping out and you said it without knowing my reasons.
Because the test you propose isn't worthwhile. But, you don't really want to know why, you just want to attack me.
I would like to find the truth. But, what you propose has too many variables, and with CAC's that vary by about 3%, you're going to find that the variables outweigh the CAC differences.
wiztecy
Active Member
djp
Model 3 Performance
There's another test that any Roadster can run to verify CAC without the variables of driving:
1. Charge to Standard
2. Turn on heat and fan to max
3. Watch the #kWh on the SOC diagnostic screen
4. Run the battery down to the "Cannot estimate range" level and record #kWh
Actual CAC Ah = Wh / 3.7V / 99 / 85%
For 153 CAC the test should take just under 16 hours and produce 47.6kWh. 147 CAC would be about 15 hours and 45.7kWh.
1. Charge to Standard
2. Turn on heat and fan to max
3. Watch the #kWh on the SOC diagnostic screen
4. Run the battery down to the "Cannot estimate range" level and record #kWh
Actual CAC Ah = Wh / 3.7V / 99 / 85%
For 153 CAC the test should take just under 16 hours and produce 47.6kWh. 147 CAC would be about 15 hours and 45.7kWh.
supersnoop
Tesla Roadster #334
You're splitting hairs. My comment is obviously discussing the one best measure of battery health. You're claiming that some in-depth, detailed analysis of the vehicle logs (which may or may not actually be available) is more accurate? You cannot definitively prove that. Your personal analysis of the logs will never be the measuring stick. Now cool it on the personal attacks. That BS is out of line.
In my example, the CAC is more accurate than the logs. It's one argument. I never said the logs were useless, but, unless you're a Tesla engineer with a comprehensive background in battery management, the CAC is the better measure, and the logs may assist in explaining any anomalies.
No, you didn't trick me. I can't even tell if you actually read my answer. My answer was, "it depends." But I set the hook for you anyway. And now, while you claim others are building a straw man, you're building one of your own. If you have to resort this tactic to make your point, then you have no point to make.
- - - Updated - - -
I'm pretty sure we already got there.
dhrivnak
Active Member
I will try to weigh in here as well hopefully as a bit of middle ground. I do agree that CAC is directly related to range an is Tesla's best estimate of a simplistic number representing the health of the pack. I believe they calculate it based on 3 major variables, the true health of the pack, the state of balancing and a recalibration of the sensors/algorithm.
We can really affect only 1, the balancing. Balancing only occurs when the pack is above about 82% hence the reason the standard charge charges to 83%. There is a slight balancing that occurs and if you perform a full standard charge 2 or more times a week your pack will remain balanced and so there is little one can do to improve this. Now if you get a new pack like Wiztec has done or you get a pack from a car that spent a long time in storage mode it is VERY likely the pack is out of ballance and here is where a range charge can help significantly. As it will bring you pack in balance much faster. This is mostly due to more time above 83%. But if the pack is already balanced there is little to gain from this. The second issue is calibration of the software and sensors. Like many EV'S Tesla is counting amps going in and out of the pack. These instruments are very bad for drifting. Time and temperature affect them so the can easily be off. The only way to recalibrate is to fully charge your pack and drive to near empty. If you go from full range charge to empty on one key turn that is best but you can get most of the benefit going from 90 to 10% without any charging inbetween. Here you are towing the dice and CAC most likely will change but it can change up or down. For me it goes up in summer and down in winter but differences will vary.
And we have pack age and possible abuse. This can only hurt. Mileage, hot packs and frequent range charges and performance modes will all hurt the pack and all in one direction.
Finally as with any number there is variation. My weight can vary 4 lbs in a day. That does not necessarily mean in am losing or gaining weight as much an if I am fully hydrated or not. So one should not be overly excited or concerned with a change of a few CAC.
Over 5.5 years mine has drifted down with a cycle between summer and winter. I have posted graphs on the battery health thread.
We can really affect only 1, the balancing. Balancing only occurs when the pack is above about 82% hence the reason the standard charge charges to 83%. There is a slight balancing that occurs and if you perform a full standard charge 2 or more times a week your pack will remain balanced and so there is little one can do to improve this. Now if you get a new pack like Wiztec has done or you get a pack from a car that spent a long time in storage mode it is VERY likely the pack is out of ballance and here is where a range charge can help significantly. As it will bring you pack in balance much faster. This is mostly due to more time above 83%. But if the pack is already balanced there is little to gain from this. The second issue is calibration of the software and sensors. Like many EV'S Tesla is counting amps going in and out of the pack. These instruments are very bad for drifting. Time and temperature affect them so the can easily be off. The only way to recalibrate is to fully charge your pack and drive to near empty. If you go from full range charge to empty on one key turn that is best but you can get most of the benefit going from 90 to 10% without any charging inbetween. Here you are towing the dice and CAC most likely will change but it can change up or down. For me it goes up in summer and down in winter but differences will vary.
And we have pack age and possible abuse. This can only hurt. Mileage, hot packs and frequent range charges and performance modes will all hurt the pack and all in one direction.
Finally as with any number there is variation. My weight can vary 4 lbs in a day. That does not necessarily mean in am losing or gaining weight as much an if I am fully hydrated or not. So one should not be overly excited or concerned with a change of a few CAC.
Over 5.5 years mine has drifted down with a cycle between summer and winter. I have posted graphs on the battery health thread.
Last edited:
wiztecy
Active Member
Try not to get confused between the CAC doing its recalculation and the balancing of the brick/cells. The CAC will re-calculate multiple times taking account new dynamic information it receives. And as I mentioned I can see this occur after a (and some) charge no matter what CAC, after a drive (even a 35 mile one). Balancing happens only after the SOC reaches a minimum state ( I found that to be 82% in my case and after the charge has stopped by the car or request) and you can verify that in the diags screen in the grid. After the key is turned all balancing stops and will not continue unless a charge again has been initiated and the SOC is greater than or equal to 82%.
