Phantom Drain

[Vach]

For the past couple days I will fully charge my car with my new 240 volt charger and when I go and sit in the car the charge time reads 10 to 20 minutes needed.

I will charge the car and then go down and unplug it before I sleep or do something else before work. The car usually sits in my cool garage for a few hours charged before I drive it.

Normally, as I was suggested because my commute is so short, I don't charge the car until I get under 10 feul bars. But I've just been checking out my new charger and then I noticed this.

has anyone else experienced this? Is the battery actually draining a tiny bit or is this just the computer system being stupid?

 

[arnolddeleon]

Don't know what you are seeing but all batteries have some amount of self-discharge. So it could be as simple as that.

I generally try to have charging end as closely as I can to my departure to minimize the amount of energy wasted in self-discharge (this was really big with NiMh batteries of the original RAV4-EV).

arnold

 

[edmc]

I notice no discharge... certainly not overnight. I reported elsewhere of parking my Rav4 for over a week outside and could not detect drain even then.

I have to ask: Why do you bother unplugging before you go to bed? If anything, I would program your car to charge after Midnight. That's what I do and it works great!

I have noticed three ways in which charge drops:
1) I leave the car "on"
2) I change my driving style (the car recalibrates over time)
3) I turn on A/C (or heat) (though that only effects the GOM on the left, not the right)

 

[Ampster]

Moderator screw up... Deleted, sorry!!!

Tony

 

[Vach]

I notice no discharge... certainly not overnight. I reported elsewhere of parking my Rav4 for over a week outside and could not detect drain even then.

I have to ask: Why do you bother unplugging before you go to bed? If anything, I would program your car to charge after Midnight. That's what I do and it works great!

I have such a short commute it isn't practical to charge every day, I'd be charging for 5 minutes on the 240v each morning (I work graveyard).

So normally, I wouldn't even plug it in, but I want to run my new EVSE through it's paces to make sure it has no problems and then I noticed this "issue." My concern was that I was noticing this when the car sits for less than 8 hours before I move to it. Normally on days I charge I'd just leave it plugged in - not charging.

I don't think it is detectable with the existing instrumentation. We don't have any way of measuring pack voltage or individual cell voltages on the RAV4EV. I am pretty sure the GOM number we see is calculated from the Amp hours that enter the battery when we are charging. A shift in the pack voltage is not going to affect that figure.

I am able to see that slight discharge on my VW conversion. Most lithium batteries reach a stable voltage within 24 hours of charging. For example, I charge the batteries in my VW conversion to 3.55 per cell and after 24 hours they settle in at 3.4 volts each. I also have a meter that counts Watthours in and out and that number does not change after the batteries have settled.

Is the same computer that measures the GOM giving me those charge times I see when I first sit in the car (on the orange display next to the center display)? If so, I can see where that value is coming from because it is just guessing.

But if I understand correctly, the battery isn't losing power it's just "settling" after a charge to come to a different voltage and the instrumentation interprets that incorrectly?

 

[miimura]

I notice no discharge... certainly not overnight. I reported elsewhere of parking my Rav4 for over a week outside and could not detect drain even then.

...........

I don't think it is detectable with the existing instrumentation. We don't have any way of measuring pack voltage or individual cell voltages on the RAV4EV. I am pretty sure the GOM number we see is calculated from the Amp hours that enter the battery when we are charging. A shift in the pack voltage is not going to affect that figure.

I am able to see that slight discharge on my VW conversion. Most lithium batteries reach a stable voltage within 24 hours of charging. For example, I charge the batteries in my VW conversion to 3.55 per cell and after 24 hours they settle in at 3.4 volts each. I also have a meter that counts Watthours in and out and that number does not change after the batteries have settled.

I think Tesla has characterized the cells so well that they know how much energy they contain by the millivolt reading from the pack at any given time - therefore, they would not use the energy that went into the pack or through the on-board charger for the GOM.

I did notice the 0:10 charge time for the first time today. Every previous time that I've noticed the value, it was 0:00. It may be related to the amount of time elapsed since unplugging or since charging since I did not drive my car until just after noon on Sunday and it was charged on schedule for 7am departure. So, it could be charge settling as mentioned above. I usually unplug my car shortly after the departure time when I walk my dog because the cord is a little bit in the way walking in and out of the garage.

 

[Ampster]

I think Tesla has characterized the cells so well that they know how much energy they contain by the millivolt reading from the pack at any given time - therefore, they would not use the energy that went into the pack or through the on-board charger for the GOM.

.......

I agree that TESLA has characterized these cells well. However this is all speculation, because Tesla has kept this part of the IP proprietary. What I do know from my own testing and from conversations in the DIY electric car community is that the charge and discharge curves of these kind of cells is very flat for 80% of the cycle and voltage is a very poor indicator of State of Charge of these cells for that section of the charge discharge curve. In addition voltage will sag, depending on the amps being pulled out of the pack when the motor is running, and we do not see the GOM drop in range as the car rapidly accelerates. It is only until you get into the top and bottom knee of these charge/discharge curves that voltage changes rapidly and amperage drops quickly. These cells do not behave like lead acid batteries where voltage is a good indicator of SOC.

Much of the non proprietary equipment sold in the DiY community uses Amp hours as a predictor of SOC. I can only assume that Tesla uses an algorithm that includes amp hours. The fact that we see no decline in the GOM in the 24 hours after charging would support that assumption. Otherwise the GOM would decline as a result of the millivolt decline, that we know takes place in these cells after charging.

 

[edmc]

I don't think it is detectable with the existing instrumentation. We don't have any way of measuring pack voltage or individual cell voltages on the RAV4EV. I am pretty sure the GOM number we see is calculated from the Amp hours that enter the battery when we are charging. A shift in the pack voltage is not going to affect that figure.

Well I have no direct knowledge but I'm pretty sure there is a calibration of pack voltage (or, rather, a computed average across sections of the pack I suppose) that varies with charge... ever so slightly until "empty" (where circuitry shuts off battery drain to avoid damage). It might be asking a lot for this calibration to update itself through use (e.g. over months/years), but I believe (aggregate) voltage is the only input to the GOM other than when I turn on various accessories.

 

[Vach]

Much of the non proprietary equipment sold in the DiY community uses Amp hours as a predictor of SOC. I can only assume that Tesla uses an algorithm that includes amp hours. The fact that we see no decline in the GOM in the 24 hours after charging would support that assumption. Otherwise the GOM would decline as a result of the millivolt decline, that we know takes place in these cells after charging.

Argh! I'm getting confused here, trying to follow you two. So are you saying that Tesla uses a more advanced way of detecting SOC than Amp hours and therefore it can detect minor battery level changes... resulting in the popup of more charge time from previously "fully charged"?

I did notice the 0:10 charge time for the first time today. Every previous time that I've noticed the value, it was 0:00. It may be related to the amount of time elapsed since unplugging or since charging since I did not drive my car until just after noon on Sunday and it was charged on schedule for 7am departure. So, it could be charge settling as mentioned above. I usually unplug my car shortly after the departure time when I walk my dog because the cord is a little bit in the way walking in and out of the garage.

I've been keeping track, I noticed it a couple more times this last week as I again have been unplugging on purpose, but then there was one time it was unplugged for 12 hours+ and there was no mysterious 10 minutes.

One day, however, I went downstairs to do some laundry after the car has been fully charged (still plugged in) for over 5 hours, and as I was doing laundry I heard it start charging behind me. The lights and charging indicator on the EVSE all turned on (I have an Arrowvironment 240 unit). It stayed "charging" for over 5 minutes until I got too curious and opened the door, seeing that it said 00:00 even though I could see all said charging lights active. I wish I had too some readings, because that event made me a little more curious as to what was going on, HAH!

 

[Ampster]

Much of the non proprietary equipment sold in the DiY community uses Amp hours as a predictor of SOC. I can only assume that Tesla uses an algorithm that includes amp hours. The fact that we see no decline in the GOM in the 24 hours after charging would support that assumption. Otherwise the GOM would decline as a result of the millivolt decline, that we know takes place in these cells after charging.

Argh! I'm getting confused here, trying to follow you two. So are you saying that Tesla uses a more advanced way of detecting SOC than Amp hours and therefore it can detect minor battery level changes... resulting in the popup of more charge time from previously "fully charged"?

I did notice the 0:10 charge time for the first time today. Every previous time that I've noticed the value, it was 0:00. It may be related to the amount of time elapsed since unplugging or since charging since I did not drive my car until just after noon on Sunday and it was charged on schedule for 7am departure. So, it could be charge settling as mentioned above. I usually unplug my car shortly after the departure time when I walk my dog because the cord is a little bit in the way walking in and out of the garage.

I've been keeping track, I noticed it a couple more times this last week as I again have been unplugging on purpose, but then there was one time it was unplugged for 12 hours+ and there was no mysterious 10 minutes.

One day, however, I went downstairs to do some laundry after the car has been fully charged (still plugged in) for over 5 hours, and as I was doing laundry I heard it start charging behind me. The lights and charging indicator on the EVSE all turned on (I have an Arrowvironment 240 unit). It stayed "charging" for over 5 minutes until I got too curious and opened the door, seeing that it said 00:00 even though I could see all said charging lights active. I wish I had too some readings, because that event made me a little more curious as to what was going on, HAH!

Sorry for hijacking this thread. We may need a separate thread about the GOM and how it works. In the meantime I will try to answer your questions. What I am saying is the battery management system in the Tesla the RAV4EV use Amphours to make the calculation on our dash about how many miles of capacity is in the pack at any one time (SOC). My understanding of what others are saying that Tesla calculates the capacity (SOC) by measuring Volts. None of us know for sure since Tesla or Toyota have not disclosed how it works. I stand by my assumption simply because Amphours and KWatthours are a measure of capacity in the world of electricity. Volts and Amps are measures of flow and pressure at any one time.

As to the second question raised by you and in the beginning of this thread I think there is another explanation. In the case of your last example when the charger started up again. My theory, and it is only a theory, is that the battery management system was in balancing mode or in temperature management mode. My theory is that in those modes additional capacity is probably not added or subtracted from the pack but capacity is bled off of the cells that are above the designated voltage and put into cells that are lower. At least that is how the battery management system on my VW Conversion works. I have no doubt that the Tesla system is way more sophisticated and includes heating and cooling the pack. I have not seen a decline in the pack capacity (phantom drain) that you reported earlier.

 

[Vach]

^^^

That makes a ton of sense; I would assume Tesla would put in considerable measures to make sure the battery is properly temperature controlled and that the cells are moderated evenly.

Based simply on experience I am going to have to agree with you because it would be insane that my battery would be slightly draining after only 3000 miles. Much more likely the battery is settling. I am still recording evidence but I'll post some screenshots and/or data when I get it, but I would think the car is just taking advantage of the connected power to charge the batteries back up after draining them slightly to cool or condition the battery itself.

That said, I thought the TMS in the RAV4 EV did not turn on when the car was inactive? Although that doesn't make much sense to me either.

 

[Ampster]

^^^

That makes a ton of sense; I would assume Tesla would put in considerable measures to make sure the battery is properly temperature controlled and that the cells are moderated evenly.

Based simply on experience I am going to have to agree with you because it would be insane that my battery would be slightly draining after only 3000 miles. Much more likely the battery is settling. I am still recording evidence but I'll post some screenshots and/or data when I get it, but I would think the car is just taking advantage of the connected power to charge the batteries back up after draining them slightly to cool or condition the battery itself.

That said, I thought the TMS in the RAV4 EV did not turn on when the car was inactive? Although that doesn't make much sense to me either.

One of the compelling reasons for me to buy the RAV4EV was the Tesla technology in the drivetrain (including a very sophisticated battery management system) and the Toyota reputation for low maintenance costs. I know several Tesla engineers and recently heard a talk by a Spacex engineer. None of these people have disclosed any proprietary information but they have given me a goodsense for the philosophy that Elon Musk has articulated in public.

As to your second question about the TMS, that has been discussed in another thread and the answer is not clear. Certainly the TMS is engaged when the car is plugged in or operating. Some have speculated that it is also engaged when the pre climate controls are turned on.

I live in a little beach town in Southern California and the temperate climate here has not given me much opportunity to observe the TMS in action.

 

[n3ckf]

i notice that i sometimes have a couple miles less on the GOM when i start the car than when i turn it off when parking overnight. For instance today, it went from 40 when i parked last night to 37 this morning. I've seen it the last couple of times i've started the car in the morning (i keep a paper log of the GOM and Odometer reading every time i turn the car *off* so i notice these things). I wish i could do this electronically but paper seems the only way. Mainly i'm doing this to compare ground truth distance vs. miles of discharge.

 

[TonyWilliams]

Well, as many of you might expect, I won't engage in trying to guess what algorithm might make up a GOM. I prefer the rated range of the Tesla Model S. Simple, always the same. I will know if I'm operating above or below the Rated Range. Or they could offer an indicator for that.

Or an calculated range based on preprogrammed route data, including elevation, temp, etc, just like we do with an airplane.

The LEAF uses AmpHour for calculating capacity, Hall effect device to trigger Low Battery warnings, lowest cell pair voltage for shut-down (contactor open) and highest cell pair voltage for stop the charge. SOC% times amphour equals the fuel bars.

 

[Vach]

i notice that i sometimes have a couple miles less on the GOM when i start the car than when i turn it off when parking overnight. For instance today, it went from 40 when i parked last night to 37 this morning. I've seen it the last couple of times i've started the car in the morning (i keep a paper log of the GOM and Odometer reading every time i turn the car *off* so i notice these things). I wish i could do this electronically but paper seems the only way. Mainly i'm doing this to compare ground truth distance vs. miles of discharge.

I have been doing the same thing except digitally with my phone and so far I have only experienced the range go down without driving the car once but it was only a couple miles. As Tony has said, the GOM is relatively stupid, so I try not to pay attention and use the trip meters and the after trip miles to determine real world range.

Based on what Ampster has said and based on how the car is functioning I think this minor charge loss Is due to the battery conditioning itself in temperatures and other reasons it would need to do so. Aka drain power.

 

[cwerdna]

^^^
I can't speak to how the Rav4 EV determines its state of charge and all, but when the Leaf's Chief Vehicle Engineer (Kadota-san) came to the Bay Area in December 2011, he sorta explained the challenges of determining state of charge along w/the Leaf's GOM challenges.

http://www.mynissanleaf.com/viewtopic.php?p=155787#p155787 and http://www.mynissanleaf.com/viewtopic.php?p=155618#p155618 I think sums it up. Once you put a load on it, it's just a guess.

Maybe a year ago, I ran by basically that by a friend of mine who has a computer science and engineering degree (basically CS + EE) and also flies battery powered RC planes. And yes, he basically confirmed the same issue. (My degree was just in CS, no EE.)

See below and some posts around there for some info/discussion like http://www.mynissanleaf.com/viewtopic.php?p=155842#p155842.

So, given the guessing that has to be done and only knowing when at startup (before putting on a load) and when the car's off, I wouldn't be surprised if there's "drift"/deviation, besides any parasitic losses and natural self-discharge inherent to batteries. And, there may need to the need to rest and possibly balance, after charging.

I met Tony Williams there, as he flew up to attend. I unfortunately sat in the back corner and the CVE stepped away from the mic during some of his explanation, making it even harder to hear, besides him having an accent.

That was an incredible meeting, BTW.

 

[Vach]

I was thinking it might be the TMS running in some kind of low power state to adjust battery temperature in the low or high extremes. This makes sense, because the days I have seen it, it has been very hot in my garage (it's getting into summer temperatures out here) and seems to occur more often when I charge in afternoons instead of mornings.

This also coincides with Ampster's lack of experiencing the same thing, because the temperature where he lives has been relatively mild making the TMS not as important.

 

[Vach]

i notice that i sometimes have a couple miles less on the GOM when i start the car than when i turn it off when parking overnight. For instance today, it went from 40 when i parked last night to 37 this morning. I've seen it the last couple of times i've started the car in the morning (i keep a paper log of the GOM and Odometer reading every time i turn the car *off* so i notice these things). I wish i could do this electronically but paper seems the only way. Mainly i'm doing this to compare ground truth distance vs. miles of discharge.

As an update I did experience this recently, I had 70 miles when I parked the car and when I got back in it dropped down to 65. When I drove it was suspicious though, it stated 65 and then up to 66 and then finally 67 after I had driven it for 5 miles.

If it is related to the GOM then we can assume those numbers are based on estimates, which means there probably isn't anything wrong with anyone's battery... not that I was thinking there was this early, but better safe than sorry!

EDIT: Fixed grammar stupidness.

 

[TonyWilliams]

... If it is related to the GOM then we really can't base those numbers off anything and there probably isn't anything wrong with anyone's battery... Not that I thought there might be... but curiosity killed the cat!

Wouldn't ya think that revelation would kill these kinds of threads? It won't.

 

[Vach]

Well, barring apparent immaturity and lack of evidence, I feel this thread contains much more discussion than submission. If you don't want to be a part of it you are free to believe whatever stoop you want.

I for one, am legitimately curious as to where certain piece of information are coming from and what we can base them on, considering I'm planning on driving the car until it is driven into the ground, or for like 300,000 miles, especially if it proves ICE enthusiasts differently. Or at least helps promote electric vehicles in general.

Things I think would be worth knowing: what is being measured in the charge read? How does the system measure it? Does the thermal control system turn on when the car is off? Is there some sort of residual drain as the car finishes settling - this one I think would be cool to discuss because there could be things going on that the car does when it shuts down that I'd like to know. As Ampster brought up, I'd also like to know what sophistication Tesla is taking to measure battery that Toyota may or may not be using.

Worst case scenario, it is some kind of bad system that is draining the battery a lot faster than it should and and as a community of all or mostly all drivers that own it we could help discover that and therefore prolong the car's life. I mean, obviously there is some kind of issue with the meshing of technology as we see in the screen of death thread.

And you of all people should understand the importance of discovering warranty issues now so that people don't have their car sitting in Toyota dealerships instead of their garages. Especially people out of state.

EDIT: more grammar stupidness