Separate names with a comma.
Discussion in 'Roadster' started by Bifff67, Sep 9, 2015.
I know I should know this, but what exact number do we want for CAC (Calculated Amp-hour Capacity)?
Higher. 160 is perfect. I don't think we know if the new, larger pack is going to change the scale.
thanks, was not clear, what I mean is what is the method of getting the desired number: I guess after normal mode charge I see 181 and range mode 206. So I claim 181. How do you get 2 more decimal points? I am missing something.
For those who have placed an order for a new 3.0 battery
CAC is the Calculated Amp-hour Capacity, not the range. It's available in the log file - you'll need to download the log to a USB stick and run it through a parser to retrieve the CAC.
Your 181 standard mode charge sounds high compared to your range mode charge. Did you allow your Roadster to sit for 1-2 hours before reading the std. mode charge? The battery needs to balance before you get the real range. Range mode you read it as soon as the charge has completed, so I guess you did that.
Also what would be nice from all this CAC data points is that what ideal miles do you get in Std. as well as Range mode. Just for the people in the side-lines really, I'd like to know how far the Roadster can go on a given CAC, especially the lower ones. Makes me gauge how much usefulness I have based upon how many miles I drive per day.
The calculated ideal range depends both on CAC and balancing. If one brick has a lower SOC than the rest, the range will be limited by that one brick. This formula seems pretty close from watching the CAC, SOCLIM and range for my own car, and it seems right intuitively based on nominal voltage, number of bricks and rated efficiency.
Ideal Range = CAC * 3.7V * 99 * SOC LIM / 230Wh/mi
A perfectly balanced pack would have a SOCLIM of 85% for a Standard charge and 95% for a Range charge.
Subtract 25 miles when driving in Standard mode.
If your pack is in balance then based on the formula above your CAC would be 152, which is pretty good. I assume the 206 miles is your Standard charge but driving in Range mode, not a Range charge.
Ideal mileage also varies based on how you charge. 240v/30A gives me 183 miles, while 110v/13A gives me 189. Quoting miles is not a consistent measure of battery health.
Ansd what is the ideal cooldown targer temperature ?
26C, but you typically have to shoot 1 to 2 degrees below that so 24-25C and after 5 mins it equalize back up to 26C. If ambient temp is cooler out, it will stay at 26 and no need to go down any further. Also if its hotter out, it will never hit 26, so you'll have to take what you get. And for me I usually hit 28-29C under that condition.
My CAC situation as calculated by the VMS Parser:
Approximate CAC: 143.40 Ah as of 11/10/2014 odometer: 11525 (shortly after I bought my Roadster)
Approximate CAC: 135.97 Ah as of 09/09/2015 odometer: 21645
My Standard charge range displays at 159 - 163. That is what is shows several hours after my nightly charge is complete.
I live in a mild climate (coastal hills on the SF Bay peninsula). I charge nightly at 56A from my Roadster HPC in my garage.
Is there a reason why you have to charge at night? I typically charge as late as I can, I schedule the Roadster to charge at 6AM, its done around 8AM and when I hop in for my daily commute at 10AM the Roadster is fully balanced. That way its not sitting at a higher SOC than it needs to be overnight. Over time those hours add up. The only time it sits at a higher SOC is when its balancing. And even then, I still need to get my OVMS up and working, I want to use OVMS to stop the charge at 174 ideal miles. I do this manually at times for now. That is just enough over the threshold so it can balance without going under the bottom floor where the balancing cuts off. There's no need for me to really go all the way up to a full std mode charge which is currently at 181. If I miss a day or two with balancing, then I will want to do go all the way up to 181.
I charge at night between 2400 and 0700 because that is when electricity is least expensive on my EV TOU plan ($0.11/kWh compared to $0.40 in the summertime).
I almost always do a Standard charge, I rarely Range charge so my battery does not sit at 100% SOC. I do not have a fixed daily commute, I start driving between 0500 or as late as 1200 (and some weekdays not at all), my distance varies and it can change on short notice. I need my car ready to go to up to 145 miles.
I appreciate that you are very careful about managing your SOC and timing your charge. That's not really feasible for me, nor do I want to spend time every day micro-managing my battery.
If there was a battery charging regimen I could follow for a few days every several months that could potentially raise my CAC by a significant amount I would consider doing that. What do you recommend?
Same for me. Also, I have solar and so I deliver energy to the grid during the day at that more pricey rate. I'd rather not consume it then by charging the car.
Yes, you have to charge to fit your schedule of course and with your job you do have to be ready at a moments notice. I'm following this charge at the latest possible time from a tip that was provided from a Tesla engineer. The mail is buried somewhere here on TMC. And that practice is typically general for any lithium battery really where you don't leave the battery sit at a higher CAC then needed. But don't get too worried, I'm pretty anal when it comes to my ESS and taking care of it. Kind of a long term experiment. The high end of the std. mode charge SOC is still in the area where it does not stress the battery too much, I'm just trying to wring out as much life out of it as I can and see where it goes.
As for raising your CAC, its really experimental. Note that I have not range mode charged in a year and won't do it unless I really really need it. I know that 90-100% SOC and 0-15% SOC can be stress on the pack. I also try to keep my SOC sitting at around 55-65% over the weekends or when the Roadster is not in use. And I try not to punch the accelerator too hard when I'm below 50% SOC. Odd thing is that the Zero motorcycles really are set to abuse their batteries and they're holding up well. Constantly at 100% and 0-10% SOC, they have no thermal cooling so they heat up, they get hammered for amps below 50%, etc, etc. I'm not seeing anyone complain about a huge loss of capacity, so that's pretty wild. I do believe that the Roadster's cell chemistry is more delicate than Zero's current chemistry. And that's why I'm a little depressed that we can't upgrade our 1.5's to the 3.0 battery. I don't care about the extra range, what I want is a more durable battery chemistry.
So back on track... I saw my CAC dive as I mentioned in other posts, it dove from 153 and kept going down where it reached 146. My job location changed where I had access to the Chargepoint chargers. So when I moved to that location I started doing cool-downs whenever I arrived at work, [email protected] Sometimes I'll do 12amps on cool-downs, but feel that the 16amps covers the HVAC overhead better. And when I get home from work, I do a cool-down. Don't have any real data other than my own observation and personal charge/cool-down algorithm I created, but seeing promising results. Is it the trickle charge at lower amps? The cool-down in the morning and evening? Or both? I really don't know, but they appear to be a positive input for the ESS. Staying with the consistent cool-down protocol above, mostly cooling at work, I started seeing my CAC slowly creep up. Then after the creep, the climbs became bigger. I'm at 153.63 CAC now and its still going up, so we'll see where it goes.
Looking at my logs I discovered brick #75 & #90 were the weakest brick during the low CAC time. They'd rotate out with one another, so everything was normal and appeared to possibly be something weak going on in those cells. Also, the avg. CAC is really high for my pack except for those two bricks. So somehow I possibly managed to break a cell(s) degradation path for the time-being. All I know is that I found a behavior I'm doing to the Roadster that's in-turn providing a positive result to the pack. So I'll continue to do that behavior. Also seeing that the CAC dropped like a rock, I'll be more delicate to the pack. Such as not taking the Roadster on trips where there are really high temps. I know that the pavement on 280/92 where you live can feel like the surface of the sun. Whenever I drive in hot road conditions like that, I really want to for sure drive the battery temps down and not allow it to hot-bathe for too long of a duration.
Truth is every pack has its own character and quirks. You have to do something, watch your CAC and either stop or keep doing that thing depending on what direction your going. Its all about data really and what / how to process it.
Here's a link showing how people have observed the Roadster's CAC going up and down:
Tesla Motors Club - Enthusiasts & Owners Forum
Below is a capture from my VMSparser -b call where it was @ 146 and showing the climb. Also you can see that the avg CAC of the bricks is 157.22 and lowest brick is 153.29. CAC in diags always reports a little higher for some reason, where it says 153.63:
timestamp, brickahmin, brickahave, bricknumber
05/20/2015 00:49:27, 146.79, 150.61, 75
05/21/2015 00:49:27, 146.91, 150.50, 75
05/22/2015 00:49:27, 146.85, 150.38, 75
05/23/2015 00:49:28, 146.51, 150.16, 75
05/24/2015 00:49:27, 146.51, 150.16, 75
05/25/2015 00:49:28, 146.51, 150.16, 75
05/26/2015 00:49:27, 146.51, 150.16, 75
05/27/2015 00:49:27, 146.39, 149.98, 75
05/28/2015 00:49:28, 146.28, 150.04, 75
05/29/2015 00:49:27, 146.51, 150.16, 75
05/30/2015 00:49:28, 146.51, 150.16, 75
05/31/2015 00:49:28, 146.51, 150.16, 75
06/01/2015 00:49:28, 146.51, 150.16, 75
06/02/2015 00:49:28, 146.68, 150.27, 75
06/03/2015 00:49:28, 146.91, 150.38, 75
06/04/2015 00:49:27, 147.19, 150.50, 75
06/05/2015 00:49:28, 147.36, 150.67, 75
06/06/2015 00:49:29, 147.59, 150.73, 75
06/07/2015 00:49:29, 147.59, 150.73, 75
06/08/2015 00:49:28, 147.59, 150.73, 75
06/09/2015 00:49:29, 147.76, 150.90, 75
06/10/2015 00:49:29, 148.27, 151.35, 75
06/11/2015 00:49:30, 148.62, 151.70, 75
06/12/2015 00:49:30, 148.62, 151.75, 75
06/13/2015 00:49:30, 148.79, 151.87, 75
06/14/2015 00:49:30, 148.79, 151.87, 75
06/15/2015 00:49:30, 148.79, 151.87, 75
06/16/2015 00:49:30, 148.50, 151.52, 75
06/17/2015 00:49:30, 148.39, 151.35, 75
06/18/2015 12:11:55, 148.45, 151.18, 75
06/19/2015 12:11:55, 148.67, 151.41, 75
06/20/2015 12:11:55, 148.73, 151.52, 75
06/21/2015 12:11:56, 148.73, 151.52, 75
06/22/2015 12:11:56, 148.73, 151.47, 75
06/23/2015 12:11:56, 148.67, 151.52, 75
06/24/2015 12:29:23, 148.27, 151.13, 75
06/25/2015 12:29:23, 149.07, 151.81, 75
06/26/2015 12:29:23, 149.70, 152.32, 75
06/27/2015 12:29:22, 149.47, 152.09, 75
06/28/2015 12:29:23, 149.47, 152.09, 75
06/29/2015 12:41:57, 149.47, 152.04, 75
06/30/2015 12:41:58, 150.27, 152.78, 75
07/01/2015 12:47:15, 150.38, 152.89, 75
07/02/2015 12:47:16, 150.38, 153.06, 75
07/03/2015 12:47:16, 150.33, 153.06, 75
07/04/2015 12:47:16, 150.33, 153.06, 75
07/05/2015 12:47:16, 150.33, 153.06, 75
07/06/2015 12:47:16, 150.33, 152.95, 75
07/07/2015 12:47:16, 150.73, 153.35, 75
07/08/2015 12:47:17, 150.27, 153.12, 75
07/09/2015 12:47:17, 150.10, 153.01, 75
07/10/2015 12:47:17, 149.70, 152.61, 75
07/11/2015 12:47:18, 149.70, 152.61, 75
07/12/2015 12:47:18, 149.70, 152.61, 75
07/13/2015 12:47:18, 149.70, 152.61, 75
07/14/2015 12:47:19, 149.70, 152.61, 75
07/15/2015 13:32:04, 149.53, 152.44, 75
07/16/2015 13:32:04, 149.93, 152.84, 75
07/17/2015 13:32:03, 150.27, 153.23, 75
07/18/2015 13:32:04, 150.16, 153.23, 75
07/19/2015 13:32:03, 150.16, 153.23, 75
07/20/2015 13:32:04, 149.93, 153.01, 75
07/21/2015 13:32:05, 150.21, 153.35, 75
07/22/2015 13:32:05, 149.87, 153.12, 75
07/23/2015 13:32:05, 149.76, 153.18, 90
07/24/2015 13:32:05, 149.64, 152.72, 75
07/25/2015 13:32:05, 149.76, 152.84, 75
07/26/2015 13:32:04, 149.76, 152.84, 75
07/27/2015 13:32:05, 150.10, 153.18, 75
07/28/2015 13:32:05, 150.50, 153.52, 75
07/29/2015 13:32:06, 150.73, 153.98, 90
07/30/2015 13:32:05, 151.01, 154.26, 90
07/31/2015 13:32:06, 150.95, 154.26, 90
08/01/2015 13:32:07, 150.67, 153.92, 90
08/02/2015 13:32:06, 150.67, 153.92, 90
08/03/2015 13:32:07, 150.73, 153.75, 75
08/04/2015 13:32:08, 150.27, 153.63, 90
08/05/2015 13:32:08, 150.10, 153.35, 75
08/06/2015 13:32:07, 150.10, 153.52, 90
08/07/2015 13:32:08, 150.73, 154.03, 90
08/08/2015 13:32:09, 150.84, 153.98, 90
08/09/2015 13:32:08, 150.84, 153.98, 90
08/10/2015 13:32:09, 151.24, 154.32, 90
08/11/2015 13:32:10, 151.58, 154.66, 90
08/12/2015 13:32:09, 151.18, 154.60, 90
08/13/2015 13:32:09, 151.24, 155.00, 90
08/14/2015 13:32:09, 151.35, 155.29, 90
08/15/2015 13:32:09, 151.35, 155.29, 90
08/16/2015 13:32:09, 151.35, 155.29, 90
08/17/2015 13:32:10, 151.75, 155.57, 90
08/18/2015 13:32:10, 151.92, 155.80, 90
08/19/2015 13:32:09, 152.21, 155.63, 90
08/20/2015 13:32:10, 151.92, 155.12, 90
08/21/2015 13:32:11, 152.21, 155.34, 90
08/23/2015 11:57:58, 152.15, 155.40, 90
08/24/2015 12:17:10, 151.64, 155.34, 90
08/25/2015 12:17:09, 151.70, 155.57, 90
08/26/2015 12:17:10, 151.58, 155.57, 45
08/27/2015 12:19:18, 151.64, 155.69, 45
08/28/2015 12:51:25, 151.92, 156.03, 45
08/29/2015 12:51:26, 152.09, 156.14, 45
08/30/2015 12:51:25, 152.09, 156.14, 45
08/31/2015 12:51:26, 152.15, 156.31, 45
09/01/2015 12:51:25, 151.64, 156.37, 45
09/02/2015 12:51:26, 152.04, 156.83, 45
09/03/2015 12:51:27, 152.32, 156.43, 45
09/04/2015 12:51:26, 152.55, 156.60, 45
09/05/2015 12:51:27, 152.27, 156.48, 90
09/06/2015 12:51:27, 152.27, 156.48, 90
09/07/2015 12:51:26, 152.27, 156.48, 90
09/08/2015 12:51:27, 152.38, 156.37, 90
09/09/2015 12:51:28, 152.61, 156.54, 90
09/10/2015 12:51:28, 153.01, 156.94, 90
09/11/2015 12:51:28, 153.12, 157.00, 90
09/12/2015 12:51:29, 153.29, 157.22, 90
I'd like to get real here:
1) CAC is a number from an equation that is somehow related to how much energy (supposedly in AmpHours) the BMS (Battery Management System) thinks the battery can store.
2) CAC is NOT an indicator of the health of the battery.
3) That many Roadster owners have seen their CAC go down the then up (146 to over 153 in wiztecy's case) shows that the number is not a true reflection, except in the most crudest terms (which makes the calculation to hundredth of an AmpHour silly), of the battery's health.
4) Whatever chemical degradation has occurred in the battery is not reversible simply by charging and driving.
5) The range numbers supplied by the car are apparently calculated from the same information used to calculate the CAC, as the two move in the same direction. We do know the relationship between the two is not just a simple multiplier. My guess there is that how you charge, and from what initial charge, has more of an effect on reported Ideal Range than on CAC.
Therefore, all the games we play tracking our CAC and trying to boost our CAC number are just that - games. Whether it's cooldown cycles or not keeping the battery "full" or whatever, the actual chemistry of the battery is NOT being improved - at best it's merely being preserved.
When we see CAC numbers go up, do not be fooled into thinking that your battery is better now than it was. Your battery can only be worse than it was before, never better. What you're seeing is the BMS calculations getting a different handle on what's going on inside.
I recommend re-reading Martin Eherhard's post from 2006 on Roadster batteries. Yeah, it's still live. Some tidbits from there:
A) Charge limiting to 4.15 volts/cell (rated 4.2 volts), or 95%. Eberhard says: We advise and encourage a full (4.15V/cell) charge only when it is needed. Meaning that is for the Range Charge and that the Standard Charge charges to 4.1 volts/cell (~90%). Talking about the 4.15V Range Change: Unfortunately, further reduction of charge has a much smaller benefit on cycle life.
B) Temperature: Our cooling system engages to try and keep the temperature of the cells below 35° C at all times and the lifetime average temperature at or below 25° C. On the recent NorCal Roadster drive, a number of cars were pushing 39° C when we stopped for lunch (little shade and no chargers to cool down, btw). So, perhaps Tesla has changed their algorithm to limit at 40 instead of 35.
• Not doing a full Standard Charge is not worth the effort unless you're not going to be using your Roadster for months (plural) and/or is being stored in very high ambient temperatures.
• Cool-down of the battery is a good thing, especially if the battery temp is running over 35° C. Cooling down below 30° C may not be worth the energy you're consuming. I personally use OVMS to cooldown to 28° C (BTW, 28° C = 82° F, which is an easy to remember data point).
• Do not believe that CAC is a true measurement of the health of your battery. It's merely an indication of what the BMS thinks the capacity is, and the value can change by up to 10 AmpHours without your battery chemistry being actually better or worse. And especially, if your CAC goes up, do not believe that your battery is better than it was. It's worse - it's always worse the next day.
• Very low states of charge (less than 2%) will hurt battery life.
• Keeping the battery cool will help preserve your battery. Tesla says best life is from average lifetime temperature of 25° C, and to avoid exceeding 35° C as much as possible (but we know that the Roadster itself doesn't kick in cooling while driving until 40° C).
Finally, in 2006, Eberhard expected 100K "aggressive" miles and 5 years to result in a battery capacity of only 70% of new. That would mean CAC values of 112. We have Roadster that have exceeded both that mileage and time and are doing better than he/they expected.
I have to disagree, CAC is the best indicator we have for the health of the battery pack. Amp hour capacity is how much the battery can hold in terms of energy, as the battery degrades its less and as its healthier its more. So the 7 extra standard ideal miles I'm getting on a charge is just smoke?
Also depth of discharge is a HUGE factor on how many cycles you can get out of your pack, so 2% is really low. Look at the table in this link and compare the number of cycles when a pack is discharged to 25% vs. 10% over time:
How to Prolong Lithium-based Batteries - Battery University
The battery state of health (SOH) is a measure of the
battery’s ability to store and deliver electrical energy.
Typical SOH methods characterize either the battery power
or energy. In this paper, new SOH estimation methods are
investigated based on the battery energy represented by the
Ampere-hour throughput (Ah). The methods utilize
characteristics of the Ah to estimate the battery capacity or
the useable energy for state of health estimation:
Look, a while ago your CAC was 146.
Today it's over 153.
Is your battery healthier today than it was some months ago? No, it has suffered from use degradation and calendar degradation. All that has happened is that your BMS has a different view of what the true state of the battery's capacity is.
Yes. Have you ever recovered a bad/dead cell before? Its possible. If you trickle charge it you can break it out of its weak pattern it learned. Typically it won't go as high in terms of Ah capacity, but it will be usable. So in that case, its always dead and that recovery is just an illusion? Again, where did these magical 7 miles come from? Wouldn't you say gaining more miles is better?
In my case I've identified two bricks that were going south, and from what I can take out of this the low amps and cooler temps have snapped them out of their downward cycle. Question is, will this be a long-term fix or just a short-term.
One last thing I need to do for this experiment is to do a higher SOC charge and bring it down to 30% on the one key drive to see where things settle in at. The pack needs to do close to a full charge and discharge to find their "real" true capacity, but what we have is just an estimation really(I'll do that step when observe the CAC has hit its ceiling). If the CAC drops back down to 146 after that, then I'll agree with you. But since I'm seeing that my amp hour capacity of the cells are rising, I'm happy. If I get more miles out of my pack per charge, I'm happy.
I still have to analyze my ahr.log and the individual voltages, need to import that data into a spreadsheet.
For NMH, yes, but not LiIon, which doesn't exhibit the memory effect.
They are AmpHours, not miles. They came from different imperfect data being supplied to an imperfect equation.
I believe all you have done is convince the BMS that the bricks are actually better than it thought they were. You haven't shown a chemical change inside the cells.
it's just a number from an equation based on very non-perfect data. The actual chemical state of the battery isn't better.
Using reported estimated miles out of the VDS is silly, as chances are that's closely based on the same imperfect data on which the CAC calculation is based.
To show us that you're actually getting more miles out of your pack per charge, you'd have to actually drive a certain number of miles to reach a known, indisputable battery state, and then show that your efforts to cool the battery and not full charge tricks equate to being able to drive more actual miles before that known, indisputable battery state is reached. The problem here is that all the battery states we know of are from equations based on very imperfect data. I think you'd have to do a chemical analysis of the cells, which is very cost-prohibitive.
EDIT: Note that I'm not saying that cooling isn't helpful. I believe it slows degradation down. I don't believe it improves the chemical state of the battery, however. There is a difference.