First service including a charging system update

[Dsinned]

While I was getting a 5000 miles service checkup today, my local dealer did a charging system update to my 2012 RAV4 EV, which I purchased back in November.

The update is known as:

ECU Software Update: Charging System Improvement

Applicability: 2012 RAV4 EV

Introduction: The EV Gateway Control ECU logic has been updated to improve vehicle operation and charging timer operation. Use the following procedure to reprogram the ECU.
.
.
.
Calibration Information: Previously ID: 1.3.37 New ID: 1.3.57

The technician's comments on the work order said, "Checked and found firmware update on Tesla ECU side. Performed Tesla ECU update and cleared codes."

Tonight, I will do a scheduled (normal) charge for completion by 7am tomorrow morning on Sunday (Mother's Day). I will park the car for the night and plug in around 7pm this evening in preparation for the scheduled charge. Then, I will be monitoring my Leviton 32A L2 EVSE charging station from that point on throughout the night to see what happens. Anyone also interested in monitoring this charge can click on the following link to "see what I see" on my PC home energy monitor, called "Enerati", which updates the graph at one minute intervals.

https://www.enerati.com/Electricity/ConfigDashboard.aspx?param1=MTMyOA==&param2=MTAwNw==&param3=ZW5lcmF0aWJkOTM5NDhlZjg2MGVhMGMzZTYyZTczMjE=

 

[boredcleaner]

Good luck! I'm approaching the 5K service quickly, so I'm very interested in your test results.

 

[Dsinned]

My RAV4 EV is parked and locked in my driveway and I just plugged in after making sure I have the center console set for a scheduled charge ending at 7am tomorrow morning. If all goes well, the charge should initiate automatically sometime after midnight - probably about 2 or 3am - completing by 7am. I parked with about 20 miles left on the range gauge, so completion of a normal charge should take about 4 or 5 hours.

When the charge starts you will see my graph (at the link below) go up to about 7.7kW and hold steady for that many hours. Right now the car is "asleep", so the graph shows "0" at the baseline. With this Enerati monitoring capability, we'll be able to see any unusal "events" happening at the Leviton charging station as far as how much power is being consumed from the charging cable.

https://www.enerati.com/Electricity/ConfigDashboard.aspx?param1=MTMyOA==&param2=MTAwNw==&param3=ZW5lcmF0aWJkOTM5NDhlZjg2MGVhMGMzZTYyZTczMjE=

 

[0GASLOL]

Wish they just add a start time. I use my phone to start charging at 12am, but sometimes the phone's schedule alarm doesnt wake me up.

 

[cwerdna]

Interesting. I wonder if this is what Tony got and if so, whether you'll have the same problems he ended up having.

 

[Dsinned]

The charge just ended about 35 minutes ago at 7:07am . . . I unplugged shortly thereafter and plugged in my VOLT for a manual charge as well. As can be seen from the graph, there are 3 charging events. The first and second relate to my RAV4 EV. The third is for the VOLT, which is still in progress as I am writing this post.

The RAV4 EV charge is no different than I have observed many times before; always the same pattern of events. First a long charging period followed by a "rest period" of several additional hours, then a cell balancing charge at a much lower charging rate for about 30 minutes. The active charging "window" resulting from "scheduling", is still the same even after my car's firmware was updated yesterday at the dealer. Sadly, that update had NO AFFECT on the problem. I could of bet money on it!!! :lol:

It's clear to me that Toyota simply does NOT understand the actual problem with scheduled charging.

The root cause of the problem as far as I'm concerned is the apparent "rest period" which ALWAYS occurs after the bulk charge. It simply waits too long! Last night, as can be seen from the Enerati graph, the rest period took ~3 hours before the 30 minutes of cell balancing occurred. Thus, the overall charging process started too early at 11:30pm, and ended too late at 7:07am, if for no other reason because of those 3 lost hours. If this charge had been done on a weekday, I would have paid more for the electricity, because the charge started early and ended later than midnight and 7am respectively. My electricity rate is lowest between midnight to 7am during the week. This is why I try to do most of my bulk (long duration) charging on weekends.

If I had moved out the scheduled completion time to say 8am, charging most likely would have started an hour later, thus after midnight when the rate would have been lowest. But then it would have pushed out even later the ending time on the other end, which would have result in paying even more for the electricity.

The bottom line . . . Toyota does not understand the problem and therefore does not understand how to fix the problem. And no matter how much real world data we try to provide to help understand, Toyota seems unaware or simply does not care.

 

[fooljoe]

The root cause of the problem as far as I'm concerned is the apparent "rest period" which ALWAYS occurs after the bulk charge. It simply waits too long! Last night, as can be seen from the Enerati graph, the rest period took ~3 hours before the 30 minutes of cell balancing occurred. Thus, the overall charging process started too early at 11:30pm, and ended too late at 7:07am, if for no other reason because of those 3 lost hours. If this charge had been done on a weekday, I would have paid more for the electricity, because the charge started early and ended later than midnight and 7am respectively. My electricity rate is lowest between midnight to 7am during the week. This is why I try to do most of my bulk (long duration) charging on weekends.

I disagree - I think the root cause is that the Rav doesn't close/open the relays upon plug-in to sample the voltage - without this step it's impossible to accurately gauge the time to charge, and the Rav apparently assumes 200-208 volts for a >20amp pilot and 120v for 20 amps or less. This seems to match actual vs planned charging time data we've seen posted.

The whole cell-balancing thing is an "extra" step which only needs to be done occasionally and should not be factored in to any algorithm deciding when to start charging given a set departure time.

 

[jspearman]

I'm guessing you are right, Dsinned, that Toyota doesn't care too much. It's not likely they will build another Tesla-powered car and are probably more interested in perfecting their own EV system than spending more dollars on this money pit. Not enough people connect the RAV4 to Tesla, so Tesla probably doesn't care much either. They have bigger cars to fry.

 

[Dsinned]

One thing is for sure and rather surprising since for decades Japanese auto manufacturers have prided themselves on the highest possible standards, but where is that same level of "QA" on these EVs with respect to something as fundamental and essential as battery charging???

 

[cwerdna]

^^^
You talking about charging or the timers being buggy?

As I mentioned in other threads, Tesla shipped the Model S without ANY timers and continued to do so for many months until a software updated added them.

 

[TeCKis300]

Dsinned, do you have any info on the software update and what it specifically addresses? I wouldn't count on it fixing all bugs across the board, but it would be interesting to know which bugs Toyota has addressed.

I think the root cause is that the Rav doesn't close/open the relays upon plug-in to sample the voltage - without this step it's impossible to accurately gauge the time to charge, and the Rav apparently assumes 200-208 volts for a >20amp pilot and 120v for 20 amps or less. This seems to match actual vs planned charging time data we've seen posted.

The whole cell-balancing thing is an "extra" step which only needs to be done occasionally and should not be factored in to any algorithm deciding when to start charging given a set departure time.

I would agree with the above and that this seems to be the root issue for Dsinned's problems. I wonder if the hardware/software has any facility to do any sampling of the voltage. My house runs pretty hot at 238.8V and my charge consistently finishes early.

 

[TonyWilliams]

The RAV4 EV charge is no different than I have observed many times before; always the same pattern of events. First a long charging period followed by a "rest period" of several additional hours, then a cell balancing charge at a much lower charging rate for about 30 minutes.

Those aren't "rest periods"; it will be actively shunting cells to discharge the highest ones, and the final low rate charge is to top off the battery after balancing.

 

[Dsinned]

Okay, maybe "rest period" is an inaccurate description, but why does it take so long to do cell balancing? ~3 hours??? Is this accounted for in the overall time to do a scheduled charge??? It would certainly appear not!

 

[TonyWilliams]

Okay, maybe "rest period" is an inaccurate description, but why does it take so long to do cell balancing? ~3 hours??? Is this accounted for in the overall time to do a scheduled charge??? It would certainly appear not!

I've seen almost 6 hours in the LEAF. How much current shunting turning into heat energy do you want in your battery? We have data now on the LEAF that we can watch the individual 96 cell pairs while that are being shunted.

Like the GOM, I don't really care what "estimate" the car displays if it hasn't sampled the voltage and knows what the pilot signal is in amps.

Since I can know that information, I'll just calculate is as always.

Miles Gained per Hour Charging
Amps/Volts -- Where ---- City Drive ----- 65mph

12 / 120 ------- Any ------ 2.8 miles ------- 3.5 miles (supplied cable with car)
12 / 240 ------ Home ----- 6.8 miles ------ 8.6 miles (EVSEupgrade.com mod)
16 / 208 ------ Public ----- 7.5 miles ------- 9.5 miles
16 / 240 ------ Home ----- 8.9 miles ------ 11.2 miles
20 / 208 ------ Public ----- 9.4 miles ------- 11.9 miles
20 / 240 ------ Home ---- 11.0 miles ------ 13.9 miles (Clipper Creek LCS-25)
30 / 208 ------ Public ---- 14.8 miles ----- 18.7 miles (typical)
30 / 240 ------ Home ---- 17.1 miles ------ 21.6 miles
40 / 208 ------ Public ---- 18.5 miles ----- 23.4 miles (just the Tesla branded ones)
40 / 240 ------ Home ---- 22.7 miles ----- 28.5 miles

 

[Dsinned]

So then, this "cell balancing" occurs over a lengthy period of several hours in addition to the normal bulk charging operation? Why is this not documented by Toyota? Is it important to be done periodically? If so, how often? And if so, why do we have to "second guess" the time it actually takes to truly complete a recharge cycle?

The answers to these question are baffling at best, but even more so to me personally, because, why doesn't the Chevy VOLT exhibit any such similar behavior?

My VOLT does one continuous recharge from start to finish without any evidence of belated cell balancing. Although for the last several minutes, it tapers off its charging rate, which I assumed has something to do with cell balancing. Also, I have read that GM also does in fact do cell balancing, consisting of shunting cells and topping them off, as deemed necessary during an actual normal charging operation. The VOLT's BMS does have an intricate, compartmentalized, voltage regulator control scheme that in fact measures each Li-ion cell's voltage drop, under load, and very accurately in millivolts, just for this purpose. This measurement scheme must somehow be incorporated into the actual "bulk" charging mode, but seemingly unnoticeably done, perhaps in the final phase of charging. However, it does not do it as a lengthy follow up secondary charging operation, taking many hours later to complete, like in the RAV4.

Again, I am using exactly the same Leviton L2 EVSE to charge either of my cars. It is very apparent, however, that each car uses a totally different charging scheme. The VOLT's seems very unified and effective (and remember, it has a 10yr-150k mile battery pack warranty in CA). It ALWAYS completes charging a few minutes before the pre-scheduled delayed departure time, and never starts a scheduled charge any earlier than necessary to complete by that departure time. The RAV4 simply does not do the same thing nearly as well, and because of its seemingly elongated overall charging method, it is nearly impossible to fit a full charge cycle, including typically 3+ hours allocated for cell balancing, within the desired "window" of an off-peak rate period (except perhaps on weekends).

Could it be that a relatively large HV capacity traction motor battery like in the RAV4 - all 845lbs of it - needs a lot longer time period to recharge even at a charging rate of 7.7kW (~25% under its full built-in capability) to avoid overheating itself, with cell balancing (i.e. "shunting") included??? If so, why does it have a TMS to cool itself, which I don't think I have ever seen (or heard) operating while plugged in and charging my RAV4, (albeit, I live in the SF Bay Area in a fairly moderate climate and park it outside every night as well as only charge my cars at night when its generally much cooler than in the daytime)???

It really bothers me how we all must continue to speculate, post theory after theory, trying to figure out how this battery should be properly charged. Imho, this is the worst thing about owning an EV. There is so much "secret sauce" in their power supplies that we can never truly understand how they really work or WHY!

 

[TonyWilliams]

why doesn't the Chevy VOLT exhibit any such similar behavior?

My VOLT does one continuous recharge from start to finish without any evidence of belated cell balancing.

Well, the VOLT really is a different beast. Yes, it uses the same chemistry as Nissan in the LEAF, but GM only allows the battery to operate between about 20% and 80% SOC. Balancing the cells isn't nearly so important.

In a car like the LEAF, which uses 2% - 95% of the battery, it is a big deal.

 

[Dsinned]

It's a big deal in the VOLT as well. There is plenty of "extra" electronics in the VOLT dedicated to this very function. As I said, among all the various electronic modules in the VOLT, and there are many, there is at least a couple that sense the voltage drop in "millivolts" across each cell or cluster, that can activate shunts to aid in controlling the cell balancing process. GM uses only 10.5kW of the total available storage capacity of 16kW in the VOLT. That's only about 65% of useable capacity; very conservative. Reserve charge is kept at the bottom at all times and a "full" charge never gets even remotely close to a 100%. I have seen no evidence of battery degradation whatsoever in nearly 9 months of commuting to work every day. However, I only need to recharge about twice a week, and fortunately, there are no temperature extremes to speak of while driving almost anywhere in N. CA. Imho, GM did a phenomenal job of designing the VOLT. With no help from TESLA, I might add, and with a 10 year/150,000 mi warranty, they must know a thing or two about Li-ion battery packs and how to maximize their operational longevity.

 

[TonyWilliams]

It's a big deal in the VOLT as well. There is plenty of "extra" electronics in the VOLT dedicated to this very function. As I said, among all the various electronic modules in the VOLT, and there are many, there is at least a couple that sense the voltage drop in "millivolts" across each cell or cluster, that can activate shunts to aid in controlling the cell balancing process. GM uses only 10.5kW of the total available storage capacity of 16kW in the VOLT. That's only about 65% of useable capacity; very conservative. Reserve charge is kept at the bottom at all times and a "full" charge never gets even remotely close to a 100%.

You're reinforcing what I said. All of the EV manufacturers measure voltage across cells in millivolts (as shown on the Nissan LEAF graphic I posted above). All of them have dedicated "BMS" for balancing cells. All have cell shunting (again, in the above graphic).

Again, in a battery that uses such a small amount of its total capacity, should a bottom voltage cell be significantly out of balance with another, you'd barely notice with a 20% lower battery charge buffer.

I have seen no evidence of battery degradation whatsoever in nearly 9 months of commuting to work every day. However, I only need to recharge about twice a week, and fortunately, there are no temperature extremes to speak of while driving almost anywhere in N. CA. Imho, GM did a phenomenal job of designing the VOLT.

That's part of the whole thing with the Volt. Because it is well insulated, and the battery only uses a small portion of its capacity, you won't notice degradation for a long time (of course, it is degrading, as all batteries do). Yes, I think GM did do a good job, but on a cost per usable kWh or weight per usable kWh, I suspect they are a bit high compared to the competition.

With no help from TESLA, I might add, and with a 10 year/150,000 mi warranty, they must know a thing or two about Li-ion battery packs and how to maximize their operational longevity.

Uh, that warranty is REQUIRED because it's a smog certified car. The pure battery electrics only require 8yr/100,000 miles. None of them offer (currently) a battery capacity warranty... again, because it's not required EXCEPT for smog certified cars like the Volt.

Nissan will likely adopt a 66.25% capacity warranty soon.

 

[Dsinned]

Tony, where did you get that display of a Nissan Leaf battery graph? That is quite an interesting set of data. The reading in the upper right hand corner (est. capacity remaining) I suppose is the one of most interest to Leaf owners. What I am curious to know is what do the cell pairs with the vertical red lines represent? There are several, but the cell voltages are all at random levels. If I were to venture a guess, the color red must mean something significant related to rate of degradation of those corresponding cells.

 

[TonyWilliams]

Tony, where did you get that display of a Nissan Leaf battery graph? That is quite an interesting set of data. The reading in the upper right hand corner (est. capacity remaining) I suppose is the one of most interest to Leaf owners. What I am curious to know is what do the cell pairs with the vertical red lines represent? There are several, but the cell voltages are all at random levels. If I were to venture a guess, the color red must mean something significant related to rate of degradation of those corresponding cells.

The picture above is an app being developed for the LEAF. The vertical lines represent the voltage of each of the 96 cell pairs. The values to the left are voltages in millivolts. There is a note at the bottom reflecting highest, lowest, and average cell millivolt, and in parenthesis is the difference between max and min.

The cells are always random, which is why these multicell packs need BMS. In the quoted pic, the red cells are incorrect, as they should be the highest cell pairs being shunted. It has since been corrected.

There's also reading from all four battery temperature probes, etc. I believe the max deviation is displayed, or perhaps peak temp.

Some of the date is processed, but most is raw data from the LEAF's EV-CAN bus or Car-CAN bus. Any reference currently to Gids is estimated, hence the ( ' ) in front. A Gid is a community derived name (from the name of the founder of the value on the EV-CAN bus) that represent 80 Wh unit. Therefore, a 300 Gid would equal 80 * 300 = 24kWh battery.

The LEAF, when fully charged, will show 281 Gid, or 22.48kWh stored energy. About 0.48kWh is left unused at the bottom of the charge. When losses are incurred from extracting the energy from the battery, about 21kWh usable are available to power the car for a fully charged, new condition, warm battery.

The CAP(acity)% is derived from a formula with ampHours (or maybe it is SOC%; don't remember). Sorry, don't have that off the top of my head, but I also don't believe it is currently very accurate. I believe the SOC% is an indexed number.

The 12.8 volts is the actual aux battery's voltage read from the Car-CAN bus. Currently, it is manually indexed by measuring the actual voltage and inputting a correction factor.