Hacker News Reader

Comment by dmurray | Hacker News Reader

> The automaker told Bolt owners to limit the charge to 90 percent, plug in more frequently and avoid depleting the battery to below about 70 miles of remaining range.

Tesla's approach here would be to issue an OTA update that told your car to treat 90% as 100% and 25% as 0. I can't decide if I like that better or worse.

View on HN

I think there's the 5 second cool factor but after that novelty you're probably just as disgruntled as the fact that you lost 35% of the range for your car.

I hope the Chevy bolt is a rare case and stories like this don't pop up with the other upcoming electric vehicles. The new lightning F150 looks very attractive.

I heard really good things about the EV mustang from Ford. I wouldn't take this to be representative of EVs in general, but rather a remark on GM.

Is it GM or LG for making bunk batteries? TFA clearly states that GM believes they will be recouping this money from LG.

Both really. The problem is supposed to have been caused by misaligned robots at LG's factories:

https://arstechnica.com/cars/2021/08/misaligned-factory-robo...

But the fault of both GM and LG is that they weren't able to detect the problem sooner.

So better regression testing on new battery deliveries and better unit tests on the batteries themselves. That's a good item to add to the after action report. Now, what do you do to the battery to make the unit test? Shake it? Heat it? For how long?

I guess I'm just not seeing fault at GM here. Trying to get there though.

In this case you can x-ray it to see the problem. They were x-raying some, but not enough.

If you're going to outsource manufacture of key components -- like the battery for an electric car, which is the single most important and expensive component in the entire vehicle -- then the buck stops with you for quality control. People inside GM may want to blame LG, but everyone outside GM will blame GM.

The Bolt has been around since 2017, and the Mach-E arrived this year. It is way too early to be deciding that maybe Ford will have better luck.

Neither company is actually making their own batteries, in any case.

I think it would be good. If you drive your car or charge it out of those ranges it can catch on fire. Removing that area of "ordinary things you can do with your car and it might make it burn up" is exactly why that is important to consider. There must be hundreds of thousands of bolts and at least some of those customers have not heard about this. The other side of it is hey you took something away from my car and I don't like that. GM has to deal with this by fixing people's cars and giving them monetary compensation.

Cool, but if i bought a low cost electric car specifically because my commute range allowed for it, for example, until your "bug fixes and changes" OTA was quietly force-pushed and i find myself on the highway with a dead battery, i'm likely to set the fkn thing on fire myself...

Obviously the manufacturer should compensate its customers, it's just better to have an EV with worse range than an unsafe EV or even a pile of scrap (if they all got recalled and probablt not fixed).

> There must be hundreds of thousands of bolts

99,356 in USA & Canada, per this number: https://www.goodcarbadcar.net/chevrolet-bolt-sales-figures-u...

I was curious about this so I looked it up, you might think it is interesting too.

Thanks for actual data to correct my guess!

I don't see how electric cars are any different than mobile phones. So basically this means a 3 year life span in terms of battery performance. Regular combustion cars though also suffer reduced mpg so I am not saying this is an apples to apples comparison.

The reduced mpg is not all that much though, maybe 2-3 mpg nothing nearly as significant as driving behavior on mpg. A small displacement honda is going to have good mpg all its life unless something has really fouled up.

I've noticed a 5mpg decrease with every car I've owned in pretty similar driving conditions over 5-10 years and immaculate maintenance.

That can’t be right, it would leave your cars immobile after like 6 years, and getting half the range out of a tank after 3 years.

Cars are good for 20 years still going roughly as far on a tank as they did. I’ll believe 10% loss of mpg over a lifetime, I don’t track that closely, but 10-20% range loss on a tank per year, nahhh.

Sorry I meant a 5mpg decrease overall not every year. That would fit with the 10% you describe.

I'm driving a 2013 F-150; it's got ~20mpg since I got it in 2014. The 2004 Corvette still gets approx. 30 (mostly highway), although it desperately needs new fuel injectors.

That is not normal. Every car I have owned has maintained its gas mileage unless something was wrong with it. My current vehicle is over 20 years old and is actually getting better gas mileage than it was when I bought it 3 years ago. I do keep up on maintenance more than the average person does tho.

Honestly in my experience, small displacement cars have typically gotten OVER epa estimate mpg. I had a 25 year old acura that got 35mpg easily each tank, epa estimate when it was new was like 30mpg iirc.

Common misconception and big reason why many avoid EVs. They are both Li-ion but battery degradation is very much mitigated in an EV because of very close attention to thermal management. Plus warranties on batteries give you something like 70% at ten years I think

While nobody will come out and say it, I’m convinced based on some extensive early career battery system engineering and various evidence from tesla that what they do is put in much more raw capacity than they sell you and then limit with software the capacity of the battery.

Early in the life of the battery this leads to slowed degredation and as the battery agrees the software uses more and more of the raw capacity to make it appear like the capacity is being lost much slower than it is… then at some point capacity will fall off a cliff when the excess is depleted.

Comparable to how SSDs have capacity reserved for bad blocks.

It is probably good engineering but it would be nice for the mechanics to be sold truthfully and some level of control to be given to the owner.

I mean, you don't have to guess. There are plenty of battery teardown videos where they count the cells. IIRC it's on the order of 5-10% kWh reserved.

Wouldn’t this be extremely easy to find out with a battery tear down/test?

I mean the Chevy Volt uses only 50% of it's battery capacity but it can more easily do that thanks to having a built-in gasoline engine that operates more like a generator for the electric engine. When I found out that they limited the battery for maximum longevity I was less stressed out about the battery life. They still degrade over time and until they become more modular like the Ultium system GM is putting into their next-gen EVs it's still a big costly single point of failure.

It's not as high as 50%.

GM also does this with the Bolt and Volt. I was under the impression that Tesla did this less so.

This is clearly the case as every time there is a mass evacuation from somewhere they issue an OTA update to release more battery capacity.

I'm pretty sure every product with lithium-ion batteries does this, but it's not so that they can use the extra capacity later, it's because letting these kind of batteries either fully charge or discharge is itself harmful to the batteries, so to get a useful lifespan you have to have some amount of extra capacity.

With phones and regular usage you are likely doing full charge/discharge cycles every day or two, which is hard on the battery. That's not really the case with electric cars. When you have 200-300 miles of range, and you only regularly drive, say, 20 or 30 miles per day, it's pretty easy to keep the battery around 70%-80%, which should significantly extend the life span.

Obviously the batteries still degrade, but they should be usable long past 3 years assuming regular driving habits.

At least with a car you might be able to replace the battery. Maybe. Of course the auto companies have the same incentives as the mobile phone companies to want you to replace the whole thing.

It's pretty easy to replace phone batteries, I wish people would stop spreading this myth. Popular models can nearly al be done by somebody that could change their own oil in a car. And there are phone repair shops in nearly every town for those who want to spend $20-$50 to have a pro do it.

Please don't discard a phone merely because of a bad battery, unless you've tried to get a replacement!

So, just like a phone then? Those batteries are also replaceable, even if they aren’t user-replaceable.

https://support.apple.com/iphone/repair/service/battery-powe...

Electric cars have been around sufficiently long that we already know this isn't the case. Even with these early technology EVs. The only EV with any kind of problem that is sort-of similar to that is the original Leaf.

Most phones use LiPo Batteries not LiIon, so they're not directly comparable. Afaik LiPo is known to have a lower lifespan.

A friend with a 6yo P85D says his battery performance is still in the high 80%s vs brand new. What to do with large numbers of EVs at the end of their useful battery life is still an unsolved problem though.

>The new lightning F150 looks very attractive

you will be happy to know Ford is using SK batteries after switching away from LG. SK just recently settled for $2billions a case that revealed they stole LG battery technology. Bolt uses LG batteries.

LG had a chain of costly recalls and replacements for all of their clients

"LG Energy Solutions, the company that makes the battery for the Bolt and Kona EV, has not had a good year. First, they agreed to replace the 82,000 batteries sold to Hyundai for the Kona EV, Ioniq, and Elec City buses. Although the initial rumors were from a faulty battery separator, Hyundai later said that the problem was badly folded tabs. GM emphatically pointed out that they use a different separator, and a different factory. Thus neither of those problems should apply to the Bolt fires.

Porsche recently initiated a recall on a loss of power in its Taycan LG batteries, and Ford also moved from LG in its Mustang Mach-E to SK in its Ford F-150 Lightning."

> I can't decide if I like that better or worse.

It's a really good question. Let's think about it.

As you've outlined it, Tesla's approach prioritizes the user's confidence in the product, rather than the user's knowledge about the product.

Naturally, when a product is faulty, the user's confidence in the product is inversely proportional to their knowledge about the fault; the less I know about the fault, the more unshaken is my confidence in the product.

So I guess it comes down to this: when a product fails, would you prefer to react with "wow, how unexpected" or "figures"?

Apple went with the Tesla model and had a PR mess as well as lawsuits

Apple's solution was akin to taking a Tesla going 80 mph on a freeway, suddenly turning on "limp" mode and locking the car to 10 mph, all without informing the user anything abnormal was going on.

EV owners are conscious about range. Tesla would have said that they were limiting range until vehicles are repaired.

Apple didn't tell anyone and a slower device is harder to quantify in normal use than not being able to drive the same distance between charges.

The EV equivalent would be reducing acceleration slowly over the lifetime of the vehicle with the helpful (to the company) side-effect that new models seem a lot more responsive.

Isn't this basically what Apple should have done to avoid their throttling lawsuit? Told their customers to change their behaviors?

Nah, they should have just been transparent about the throttling.

They still throttle now after the lawsuit, you can just stop it if you want to. Also an important difference with Apple is that the iPhone throttling wasn’t a safety concern.

The throttling was a reaction to poor battery performance at the technical endish-of-life part of the product. After X cycles the battery recharge ability was limited, couldn't be fully charged. It also had increased internal resistance, meaning an increased power draw (eg CPU/GPU load increase) can't be sustained. If too much current is drawn, the voltage drops, and the result may be a brown-out event (voltage too low -> shutdown/reboot). Since the battery couldn't deliver, Apple limited use of the battery so it doesn't run into those brown-outs.

This is all sound and well and well-known, which is why you design in a safety factor so that you have plenty of margin even at end of life.

Why didn't they? Since no internal Apple discussions surface, I can just speculate it's either,

  * incompetence, they didn't anticipate this
  * cost-driven decision, despite Apple products being top-dollar devices and even low-range devices work better in this aspect
  * design-driven decision, larger battery would mean a larger device

I'm betting the last point. Someone (Jobs? Ive?) had a fixed upper size limit on the device, and at the same time a fixed minimum CPU/GPU performance limit to not feel sluggish.

Around that time there were a ton of battery recalls for phones nowhere near the end of their life - mine was one.

I think they just got a bad batch of batteries or were too aggressive cost optimizing and this was the fix. 5&6 era iPhones never had this problem and their batteries degraded just the same.

They've tried that before, it turns out that "you're holding it wrong" isn't the best customer experience.

I still can’t believe they went for that

Not just that they went for that - but that they ended up in that situation to begin with. IIUC, the outer rim was two (or more) metal pieces, whereof one acted as antenna. The signal degradation occurs when the user holds the phone and bridges the antenna-piece with a non-antenna piece. This affects the antenna tuning and attenuates the signal.

Just missing out on such a basic thing is a bit boggling to me. My iphone 4 was working pretty good though, but I was also mostly in the inner city where connectivity is good to begin with.

> Tesla's approach prioritizes the user's confidence in the product

More like Tesla prioritizes removing user error

Tesla's approach would limit significant danger to human life & property. I'm going to guess that many Bolt owner may never end up knowing about this issue.

I would vote for Tesla's approach, followed by a partial refund for loss of advertised functionality (or a class action)

As long as it included messaging and an admission of fault, that seems significantly safer for the general populace.

To be fair this is the norm for pretty much any rechargeable electronic device. Lithium chemistries in particular are sensitive to damage from over discharging. "5% battery remaining on the display" is pretty much never "5% of available energy available"

Yes, but raw cells basically stop being rechargeable when you discharge them past a certain cell voltage. They either won’t recharge or are rather likely to burn when you do.

I think practicality is governing each company’s choice; there’s no way GM is capable of pushing an OTA update to change all those cars’ firmware. If they could push an update over cellular, I would bet you GM would do it in a heartbeat.

They could easily achieve the same 90% == 100% charging result with “myChevrolet” (onstar).

Car sends status update that charge has hit >=90%: stop charge command issued server side. Last I checked, the car sends a vehicle status update every percentage tick on charging so this wouldn’t be perfect but it’d be better than nothing.

Both GM and Ford have OTA updates on their modern cars. They just don't do it very often. The upside of Tesla pushing an update every week is they're good at it.

Tesla did pretty much exactly this to some Model S owners. And got sued over it. I don't know if that has been fully resolved yet. Some people got screwed out of a lot of battery capacity and performance.

Tesla's approach is safer for everyone. Now, if you want to discuss ethics, that's a totally different thing.

Yeah, this kind of shit just isn't going to work. If I have to be told how how to maintain my fuel source, then these cars with this battery technology aren't ready for mass adoption. They just aren't.

GM's first fix was to go in for a software update that limited max charge to 90%. This of course took scheduling and sitting in a dealer for a little over an hour.

I'd much rather have the OTA option.

As a Bolt owner and advocate for right-to-repair/etc - I'd prefer the OTA option for sure.

They should have limited the charge from the beginning. No way the engineers didn’t realize deep cycling is bad for the battery.

It's not that they were deep-cycling the cells before. Discharging lithium cells to 0 volts is very bad for them; the cars definitely restrict the cells from being discharged below a certain point. Rather, the potential for danger due to this rare, hard-to-detect manufacturing flaw is great enough that Chevy is asking Bolt owners to self-impose even tighter restrictions on how low the batteries will be discharged.

> They should have limited the charge from the beginning. No way the engineers didn’t realize deep cycling is bad for the battery.

Have you even read up on the manufacturing faults behind this? It's a cell manufacturing robot alignment problem that went unnoticed, not a deep cycling problem.

Thanks for the correction.

Reply on news.ycombinator.com