ALL POSTS about Charge Timer Failure

[TeCKis300]

I had a nice long talk with a representative at Toyota headquarters in Torrance a couple days ago.

I was able to confirm the definite existence of one Charge Timer bug related to the end of month, on the long months ending in 31, as evidenced by the 3 people above. If you search for my history in this thread, you'll note that I also posted failures on Jan 31st and Mar 31st. The bug is being worked by engineers at the moment with a fix pending.

I inquired about existing updates out in the field. The only thing I was able to glean was that those were to address compatibility issues with 3rd party chargers (read non-Leviton 40amp), and chargers with timers on the EVSE - so none of those updates would help me.

He could not confirm whether there were any bugs relating to charge timer failure on a 240V EVSE after charging on a 120V charger earlier in the day. Though he did want to get information so that he could pass it along to the engineering team for review.

I would be pretty darn happy if I could get the end of month fix as that seems to be the predominant issue I'm experiencing.

 

[fooljoe]

@Teck - If you talk to this Toyota rep again, please advise him that any upcoming software fix really needs to implement a voltage check by closing/opening the EVSE relays upon plug in!

 

[dark]

Wait a minute, mine also fail randomly, on May 23rd and other dates.

Please don't come to any conclusions of the bug.

 

[TeCKis300]

Being a software/systems engineer, you guys are a tough crowd.

These systems are very complex under the hood. In order to be productive, it is important to characterize and isolate the various failure conditions. While it's certainly plausible some issues are related, it is just as likely there are several distinct root causes.

Then there are those of you guys are that are asking for more features and more system sophistication to support a perceived as desired behavior i.e. start time, cell balancing integral to the charge cycle, etc.

I'm in no way defending Toyota, but IMO, the only real failure on their part is communicating and providing timely fixes for critical bugs like charge timer reliability.

 

[fooljoe]

I'm a bit of a software developer myself (a reliability engineer by trade), and I'm absolutely certain that implementing an algorithm to check the line voltage, read the pilot, calculate the expected time-to-charge, and subtract it from the scheduled departure time - plus some margin - to set a start time should be incredibly simple. How Toyota/Tesla got this wrong I have no idea, but the one most obvious flaw is skipping the step to check the line voltage. And of course all that could be circumvented if they would just offer us the option of setting our own start time.

I guess there's a possible second issue of the car not starting a charge event even if an appropriate start time is scheduled, and not getting something as basic as that right just speaks to how piss-poor their software engineering and QA is.

Anyway, it seems there there are so many problems related to charging the best path forward would to just scrap the system and start development from scratch. If only we could "jailbreak" our cars' computers and do that ourselves...

 

[TonyWilliams]

If only we could "jailbreak" our cars' computers and do that ourselves...

It sounds like there are several with the talent to do this. There are many guys now who have broken virtually all the code for the LEAF.

 

[cwerdna]

Being a software/systems engineer, you guys are a tough crowd.

These systems are very complex under the hood. In order to be productive, it is important to characterize and isolate the various failure conditions. While it's certainly plausible some issues are related, it is just as likely there are several distinct root causes.

Yep. And some of them can result in what looks like only one or a few symptoms.

 

[SeaMonster]

I think we can all agree that the software task of programming a simple timer shouldn't be that hard. Sadly the issues here transcend technical and spill over into the realm of business/political. The overall system design is unfortunately overly complicated by Tesla controlling the charger and Toyota controlling the UI. (It's more than that, but that is a ood high level separation.) These are separate computers and they must communicate. That complicates things, but add to that the fact that neither company wants to share their technical "jewels" so they provide each other as little as possible and assume that it should be enough. (BTW: ""enough" is a business decision and not a technical one.) There is no way that this issue should take more than a couple of weeks for a competent engineer to resolve IF they had all the system details. Tesla seems to be in the driver's seat (bad pun and ironic because their computer is in the back) so they toss software over the wall on their schedule and Toyota has to work with whatever they get.

 

[dark]

Guys, this is what I don't understand, if this issue is a software issue. Why only only few of us are having the timer failure? And not happen to all the rav4 EV?

I started to think this could relates to hardware issue where the computer can't detect the charge.

Also, some of you complain the timer charging early. Consider if your car does not charge at all when you leave in the morning, I rather have my car to start charging early than no an empty battery in the morning.

 

[SeaMonster]

Guys, this is what I don't understand, if this issue is a software issue. Why only only few of us are having the timer failure? And not happen to all the rav4 EV?

It's pretty clear that personal behavior will effect how likely you are to see this.

- If you don't plug in every night then you are more likely to NOT be plugged in on the 31st of the month and not see that issue.

- If you don't use the 110 charger for manual charges between 240V timed charges, then you won't see that issue.

- If you don't run your "tank" low, then you will not see issues related to longer charge times.

So it makes complete sense to me that some might never or rarely see issues.

 

[fooljoe]

I'll add to SeaMonster's list that if you don't have a 20 amp or less 240v EVSE you won't run into the problem of the car assuming any such EVSE is 120v, making the already early start time ridiculously early (like yesterday).

 

[Dsinned]

Another explanation is that perhaps the majority of RAV4 EVs are ONLY being "manually" charged. Not all EV households are on "TOU" electric rate plans; so those people are probably paying the same fixed rate per kWh for their power consumption 24/7 (except as may change from baseline to "tiered" rates). In this case, it does not matter when you charge the RAV4 EV. It make more sense to just charge it manually every time. However, even some of these people surely must see "refusals to initiate charge" issues as has been reported here from time to time. But they are usually present to witness such behavior and just retry till it starts charging. This is a "minor" nuissance compared to a complete failure to charge at all, or way outside the scheduled "window" based on a timed event, while the car is plugged in for an overnight period. This is when the car owner is usually not able to actually monitor what is suppose to be happening thought the night and/or early morning. Also, TOU rate payers are probably still in the minority.

The main reason this is such a big problem, is because owners like me, are trying to charge their cars when their "TOU" electric rate are lowest, usually from midnight to 7am on weekdays, and this by its very nature is far more conducive to an automated or "scheduled" charging event. However, this is when I am least likely to be able to insure and verify that the charging event actually occurs during those particular hours.

Probably the most frustrating thing about this is that I am doing exactly the same overnight "scheduled" charging (using the same Leviton charger) to my Chevy VOLT. This works flawlessly each and every time, and I have NEVER had any problems!

 

[fooljoe]

Also, TOU rate payers are probably still in the minority.

You think? I would imagine the majority of EV drivers do switch to TOU.

 

[dark]

WRONG statement, I plug in every night, and my timer failed during random dates. The timer failure has nothing to do with the calendar date.- If you don't plug in every night then you are more likely to NOT be plugged in on the 31st of the month and not see that issue.

WRONG statement, I do not use 110 charger, it is stored in the storage compartment, and I use a 240v 30amp charger- If you don't use the 110 charger for manual charges between 240V timed charges, then you won't see that issue.

WRONG statement, My daily commute is around 50 miles, and I always have 1/2 remaining battery when I use the timer.- If you don't run your "tank" low, then you will not see issues related to longer charge times.

Also, I set the timer not because of a cheaper rate, the battery would least longer if you charge it and use it right away. Instead if I don't use the timer, I come home at 6, charge it right away would takes 2.5 hours, assuming it finishes at 9pm, then the battery would be sitting for over 9 hours at 80% overnight. I feel that is not very good for the battery life.

 

[TonyWilliams]

... then the battery would be sitting for over 9 hours at 80% overnight. I feel that is not very good for the battery life.

I wouldn't sweat that at all.

 

[fooljoe]

WRONG statement, I plug in every night, and my timer failed during random dates.

Here's one thing I can tell you for certain, these failures are NOT random. Computers are incapable of true randomness - they do what they're programmed to do, and they do it faithfully. Whatever the root cause(s), it's clearly due to human programming error.

 

[SeaMonster]

WRONG statement, I plug in every night, and my timer failed during random dates. The timer failure has nothing to do with the calendar date.- If you don't plug in every night then you are more likely to NOT be plugged in on the 31st of the month and not see that issue.

WRONG statement, I do not use 110 charger, it is stored in the storage compartment, and I use a 240v 30amp charger- If you don't use the 110 charger for manual charges between 240V timed charges, then you won't see that issue.

WRONG statement, My daily commute is around 50 miles, and I always have 1/2 remaining battery when I use the timer.- If you don't run your "tank" low, then you will not see issues related to longer charge times.

Also, I set the timer not because of a cheaper rate, the battery would least longer if you charge it and use it right away. Instead if I don't use the timer, I come home at 6, charge it right away would takes 2.5 hours, assuming it finishes at 9pm, then the battery would be sitting for over 9 hours at 80% overnight. I feel that is not very good for the battery life.

Dark, before you go claiming that what I said is WRONG, please go try reading what I wrote. Then I'd recommend reading it again. My statements are all correct. Those are known issues that increase the likelihood that you will see a timed charge failure. I am in no way claiming that this is a complete list. It is simply the three most identifiable ways that I am aware of to "usually" cause the issue. And the point was to answer your question of why some folks may not have seen issues. If you don't do the things in that list, then you are less likely to bump into those causes of timed charge failures. This isn't one bug.

For me, the first two are 100%. The third is hit and miss. I think it is telling that for number 2, you yell WRONG, but then you state that you never use the 110 charger. If you don't use it, how would you know if it would or wouldn't increase the chances that you'll see a timed charge failure? Maybe you could try listening to someone that does use it instead of yelling that I am wrong.

Again, keep in mind that I didn't claim that anything was 100%. I simply answered why some people's behaviors might be such that they would be less likely to see a timed charge failure. In fact some folks might never see one. I believe every one of those statements is correct.

 

[dark]

Fooljoe and Seamonster, with respect I do not understand why you guys think you know the cause of the timer failure bug.

It has no time, date and remaining charge pattern. You all talking like you know the pattern but you don't and you are misleading everyone here and Toyota.

The timer failure has nothing to do with time and date. It is not as simple as you think. It does not have any pattern based on how we operate, that's why I said it fail randomly.

Again, you are misleading based on your assumptions that occurred on your case.

 

[SeaMonster]

With respect: I do not know the deep down cause of the timer charge failures, but I DO know how to reproduce them. I have a spreadsheet that has every single charge for 8 months. It also charts many external factors to enable it to show correlations (or lack thereof). I have talked directly with Toyota techs. (I spoke with them again last night, twice. Not the dealer, Toyota.) Toyota has confirmed some of those behaviors that I listed and you claimed were false. I have pursued many theories (some mine and some Toyota's) to see if they correlated. Some did, some did not. Toyota instrumented my car with recording devices (talking to both the Toyota and Tesla OBD connectors) and had me go through my regular routine for several weeks recording diagnostic data and using my externally observed data to correlate that with.

Additionally, I have now had a Rav4 EV rental for more than 30 days (while mine was in the shop) and I have reproduced many/most of these failure behaviors with that car as well.

So tell me. Why do you say there is no pattern? Is that based on a well documented analysis? Or is that just your opinion? I can show you patterns. Real actual patterns. Yes, I have no idea where a broken line of code is or which subroutine(s) or components are in need of repair. But I can show you behavior that will reproduce a timer failure 100%. And I can show other behaviors that will reproduce it less than 100%, but statistically more often than just chance.

So let's take this down a notch. The path forward is to find useful data and provide that to Toyota. I encourage everyone to take careful notes. Theories are great for forums, but not for notes. Record facts only and as much as you can. Use your theories to help you decide what raw data to collect. As others have noted, software is deterministic and given the same inputs will always produce the same outputs. (Yes, i know there are exceptions to that, but basically it is true.) The issue here is that there are many inputs and these need to be observed and recorded and isolated as best possible. I will be talking to Toyota again today and my message is that they need to engage the community to help them, and they need to communicate more. (Of course Toyota is huge and I am just talking to one guy. But it's a start.)

Peace!

 

[fooljoe]

@dark -

SeaMonster laid it out quite well. There's no "timer failure bug" - there are lots of issues here. I don't pretend to know why some charges never start, which of course is the most serious problem. But it's very interesting how many of us had a failure on the 31st, so that's information that could definitely be of value in trying to figure out what is causing these problems. And it's very interesting how others have had failures after 120v charging - but I haven't tried to reproduce this myself.

Here's what I DO know for sure: The car doesn't command the EVSE to close the relays momentarily upon plug-in, something that other EVs (at least the Leaf and Volt) do, and something that is REQUIRED to obtain an accurate estimate of the time required to charge. If you're plugged into a 240 volt / 20 amp EVSE and the car thinks you're on 120, your charge start time is going to be WAY off, potentially so early that it's before the time you're plugging in - which can cause the charge to never happen (which is apparently yet another bug).

That the car omits this voltage check is just a hugely amateurish oversight, IMO - but I'd agree, as others have mentioned - that the issue of scheduled charges not happening at all is worse than them happening early. It seems to me that there are just so many flaws with the timer system that it needs a complete rewrite, and in that case I'd want to make sure the voltage check gets in there as well as fixes for other issues.

So let's take this down a notch...

+1!