IMO locking down the battery capacity via software is pretty lame.
Does anyone know exactly how this is done? does the battery just never charge up all the way? does it claim it's empty when there's actually more charge left? Are some of the cells locked down and just don't charge? How does it impact the health of the battery (since not being able to charge or discharge completely would presumably affect the health of the cells)
Without these software limits, lithium ion batteries are practically useless.
Lithium-ion cells are remarkably finicky and fragile. There's no clear cut-off that demarcates "over-charged" or "over-discharged", you're just increasing the stress on the cell. A highly stressed cell has a much shorter cycle life, so there's a tradeoff between durability and capacity. The 0% and 100% you see on the charging icon are essentially arbitrary lines, chosen by the product designer to best exploit this tradeoff.
Drone racers will often charge their cells beyond 4.3v and discharge them below 3.2v to eke out every last joule of energy. These cells will lose 20% of their capacity within a few dozen charge cycles and might fail catastrophically due to dendrite formation. A perfectly reasonable tradeoff in this application, but a disaster in an electric car.
For obvious reasons, electric car manufacturers lean strongly on the side of durability. Nobody will care about an extra 20 miles of range if their battery fails after 40,000 miles. Consumer electronics manufacturers started using more conservative battery management profiles when non-replaceable batteries became the norm - by sacrificing a bit of capacity, they could double or triple the cycle life of the battery.
Note: this explanation is grossly over-simplified in the interests of brevity. Battery chemistry is ludicrously complex.
Seems like the issue you're describing is more about repeated "overcharging", so it would seem that it'd be perfectly acceptable to allow the user to override it when they deem it necessary. However, as the article states, users can permanently unlock the extra charge for $5000, so I doubt your concerns are why the limit is in place.
My understanding is that it does not charge all the way.
There are some advantages in having a s/w locked down battery rather than a physically limited battery. Since the battery charges slower when it is above 80% capacity, you can recharge to "100%" much faster with a software limited battery. Similarly, charging to 100% is supposed to cause more battery wear, so in a real 75 you don't want to do that very often. But in a 60, its no problem, since 100% is really 80% of the physical capacity.
My understanding is that it does not charge all the way.
So doesnt that make this update a little late and kind of useless?
The evacuation order was days ago. People have already left their homes--where they would have charged their cars. They already made it to where they were going on the 60kw charge they had or they didn't. The hurricane has already reached FL. http://media.miamiherald.com/static/testdir/weather/storm-ra...
hah just replied with exactly the same thing. I'd say it was a PR move, but advertising that they lock down features doesn't strike me as wise.. but then again, I'm not an ad man.
Good on Tesla for doing what's good for these specific customers than worry about their corporate image. If they continue to act on the users' behalf these events which may tarnish other companies will have no negative impact.
The evacuation orders were only for low level areas. Half of the state is pretty much out of harms way and could make a run for it at this point, if they choose.
interesting. In that case, I wonder how useful this update is, since it would presumably just allow the consumers to charge their cars for a longer period of time, but they should probably be getting the hell out of the storm path instead.
That's a good way to get a lithium fire. Lithium batteries need careful monitoring to charge correctly. To give you an idea, 4.2V is 80% charged for a chemistry I worked with. 100% is 4.23V, so you don't have much margin for error.
>>IMO locking down the battery capacity via software is pretty lame.
This is a thing in OBD2 ICE cars as well and has been for years, particularly the VW series of cars in the 2000s. The GTI had most of the same parts as the Golf but the GTI had an ECU "tuned" to maximize performance.
Hondas were not that much different, though not as egregious. The rise of Hondata and scavenging JDM parts overseas is a testament to that.
"Software" just feels different. But it really isn't, whether it's firmware (ECU) or a few extra holes/different tolerances in the machinery.
I think to me, the difference is that in ICE vehicles, you're often tuning them up (requires additional research, testing, warranties) rather than tuning them down. I also assume there's some additional hardware to support the increased stress from higher performance.
In the case of the tesla, they're taking something good and just tuning it down.
I wonder how people would react if they did this with the next iphone - i.e. it only charges to 80% unless you pay for an upgrade.
The Teslas will suffer longer battery charges (much more than 20% longer) and significantly reduced battery life if on this new firmware. It would take improvements in hardware (or drastically reduced durability) to make it available for everything.
I just wish I could pick these modes for my phone, especially if it's a non-repairable battery. My Lenovo laptop actually has something like this - it puts the laptop to sleep at 20% and only charges to 80%, unless I choose otherwise.
>>I think to me, the difference is that in ICE vehicles, you're often tuning them up (requires additional research, testing, warranties) rather than tuning them down. I also assume there's some additional hardware to support the increased stress from higher performance.
In the case of the Golf/GTI, there wasn't. The hardware is virtually the same in both, the only difference was ECU tuning. This is the same as the Tesla, it's not tuned up or tuned down... it's identical, just rolls off the line with a different body, accessories, and ECU.
>>In the case of the tesla, they're taking something good and just tuning it down.
That's exactly what VW and car makers have been doing for decades. And Intel, AMD, Western Digital....
>>I wonder how people would react if they did this with the next iphone - i.e. it only charges to 80% unless you pay for an upgrade.
Probably about as poorly as this is being received, even though it's been going on forever.
For the same reason that CPUs and solar panels (just two examples off the cuff) that are not up to spec are sold at cheaper prices: price discrimination.
Tesla gets to sell a car it otherwise wouldn't get to sell, and a consumer gets a vehicle they otherwise couldn't afford. Why is that lame? That's capitalism.
On the one hand I understand your argument. IMO it's a strange time to live in. You own the car, but not the ability to use it to its full capacity. On top of that, from my understanding, the software is owned by Tesla, so you couldn't mod it even if you wanted to (legally).
That's a fundamental problem we've been having for a long time, then - you own the device, they own what you can do with it.
It's the very same issue a lot of us have fought tooth and nail for, for the last decade and a half.
Edit: someone replied - then deleted - that effecitvely this is a free market issue. I agree with that, and I by no means think that government needs to get invloved. Still, as a consumer of such devices/cars, I still feel uneasy (to say the least) about this scheme/business practices. It's "lame" for the reasons I outlined.
>Tesla gets to sell a car it otherwise wouldn't get to sell, and a consumer gets a vehicle they otherwise couldn't afford. Why is that lame? That's capitalism.
It's lame because the value change is one-sided.
The value is being adjusted by taking away options from a good which has the same intrinsic material value as another to change the market value of the good.
It's capitalism, sure, but it's not one of the better parts.
So you're suggesting only selling the 75kwh version, which means some folks who want it would otherwise not be able to afford it. How is that good for anyone? And if you buy a 60kwh version, would you be happier if it actually had fewer cells? Comments above indicate that means it would charge more slowly and charging would wear it out faster, but that appears to be the solution you prefer.
It's not clear to me that the batteries in the 60kwh version aren't 'reject' batteries that weren't quite up to snuff to support the full 75. So just because Tesla pushed out this update doesn't mean that every car is actually capable of getting the full 75. Where-as the cars sold as 75 are guaranteed to reach that.
I don't know the details of Tesla's batteries exactly, but I'm pretty sure they're just use a ton of 18650s. If too many of them won't hold a charge, then instead of fixing it they can just call it a 60 rather then a 75.
That's a bit silly. I suppose in someways it is nice for the consumer that they have the 60kwh version if they're basically selling it close to a loss with the hope that you'll pay for the battery upgrade, but obviously it begs the question why the 75 model can't just be the same price.
> but obviously it begs the question why the 75 model can't just be the same price.
Because Tesla would make less off of each 75, which is no good when you need cash to expand operations. Have to squeeze every dollar out of each sale, manufacturing and R&D costs money!
The 60kwh has been discontinued after tepid demand. The 75kwh is the cheapest Model S one can buy today, but that wasn't true for a few months 60kwh was up for sale.
Puhleese, if you're going to make a statement like that you need to at least attempt some sort of citation.
The vast majority (in the US at least) of price discrimination instances are cases where anyone is free to choose the cheaper option. Rich people can clip coupons too, they just decide it's not worth it for them. Heck, the whole reason for a Tesla Model 3 and Model S is price discrimination. I think it's a great thing I'll be able to get a Model 3 while richer consumers can choose a Model S if they care to.
It is common economic knowledge/theory. My citation is any economic textbook on price discrimination.
Let me give you an extreme example.
Imagine that a company perfectly knew the exact maximum amount of dollars that you personally are willing to pay for their product. Then, they offer you that product for 1$ less than your maximum.
This is called perfect 1st degree price discrimination, and results in ~zero consumer welfare, because consumer welfare is the difference between actual price paid, and max willingness to pay price, which in this case are both ~equal to one another.
> Alright, then there is no 60 kWh version of the car.
That's implicit short-term thinking which I find strange given the speed at which such technology improves nowadays.
Even assuming the government made a law against artificial software lockdown punishable by death, this luxury vehicle would still come down in price and improve in quality well within the lifetime of most of the people who are currently drooling over it.
That's not an argument for making such a law, just a refutation against what is probably a false dichotomy between car and no car.
Essentially, yes. There are physical limitations to how much power an ICE can safely produce, but no sane manufacturer would tune the engine in a passenger vehicle to run at those limits. It'd make for a gas guzzler that required constant maintenance. (Which, unsurprisingly, is a perfect description of a "performance" vehicle, i.e. a racecar.)
Does it really matter if they create the artificial limitation via software or by drilling a hole in the extra cylinder? I mean, it makes a difference from the perspective of reversibility, but it shouldn't affect your view of the morality of doing the price discrimination.
but I don't know of any ICE models where the difference is solely software that locks features down (or, as you said, drilling a hole somewhere to limit functionality). They may share parts like bodies or engine blocks (not sure, I'm not a mechanic) - but there's a lot of additional hardware that makes it safe for them to beef things up via software.
not sure though - can you think of any two ICE models where difference between the two is solely locked down via software? or the only difference is a hole drilled in a cylinder, etc etc. It may exist and I'm just not aware of it.
BMW did something with the 116d, 118d and 120d. All had the same 2 litre diesel engine - and there was considerable sharing of all power related components, including the brakes. Not 100% identical across the board but pretty darn close. The 116d and 118d were detuned via ECU.
a) In the context of Tesla not using full hardware limit of batteries - this is standard on all modern low to mid personal vehicle diesels, they are very software limited on charge pressure/injected fuel from pretty early rpms. Otherwise, the high pressure turbo could charge the engine off the engine compartment. Apart from obvious legal (tax levels), or marketing-profit (price bins) reasons, it is done for the same reason Tesla does it. Running the engine in such "unlimited" mode, it would have about the same longevity as a laptop battery charging always to (chemical) 100.
b) I'm on mobile, cannot find exact models now, but yes, there were two engines having the same engine block&head, acessories, turbo, powertrain and were only tuned differently in ECU. Not sure now if it was for the same model of the car itself, though, maybe it was used across different model series. And a pretty common practice in European diesel cars with average power, loading ECU maps from higher versions - although often, the car would have smaller brakes or less Nm rated powertrain.
Yeah, some of the CUT are like this 10, 20 hp difference only in software because they don't make enough to justify different engine components.
Diesel trucks tend to be like this too. If I recall they put the same 7.2L engine in a bunch of F-250 to F-550(although there your also paying for better built frame/suspension) models.
> IMO locking down the battery capacity via software is pretty lame
This is done all the time in other industries - aircraft engines are derated, SSDs are over-provisioned, all to preserve longevity. It's also done with CPUs, perfectly good cores disabled in lower performance models.
Many kinds of batteries last longer if you don't use their full capacity. Maybe that's true here?
It's annoying, but on the other hand it would be kind of neat to avoid paying for the upgrade until the first day you find you need it. Maybe you never do?
Not really that lame or uncommon. It keeps costs down, makes replacement easier and faster, less inventory, etc. And happens all the time in lots of industries - lots of examples here about same engines indifferent states of tune, same engines and parts with different ECU / fuel injection maps, happens in electronics with chip binning, and lots of devices that have special keys to unlock more functionality or better performance.
What’s wrong with blurring the lines between hardware and software? We do in app purchases all the time, and if standardising hardware across all product levels makes things more efficient anyway, why not?
You are not aware that the exact same thing is done for petrol and diesel cars (i believe a bmw 318d is the exact same engine as a 323d, a part from software/firmware). Cpus, gpus etc. etc.
There are two basic versions of that engine both have four cylinders, a 1.8L and a 2.0L, that's only a difference in sleeves and pistons but it is not just a software change.
There is no 2.3L version of that engine and I'm not aware of a 323d.
Exactly. Many years ago my father drove a Rover 75 in the UK which used an older version of BMW's 318d engine, but of course hobbled in software so as not to deliver as much power/torque as the same engine in a BMW.
For a couple hundred quid you could "chip" the engine to match or exceed the rated power of the BMW version. They even installed it in such a way that the chip could be removed before you took it in for a service so as to avoid warranty concerns etc.
>>it just seems perverse in this situation though. vehicles have more of a 'life and death' component than a digital camera
As pointed out multiple times in these comments, Tesla is not the first company to do this. VW, Honda, and many others have been doing this for decades.
This was a good move on Tesla's part, but the dark side of this is that they could also "flip a switch" to reduce the range of cars. Presumably they could also disable the car entirely, disable the brakes or rev the motor at an unsafe time, or cause the battery to catch fire.
All of this seems very implausible for a company like Tesla, but cars can last a very long time and companies and society change over those time scales. It's not hard to imagine an authoritarian government forcing an auto manufacturer to mess with the cars of protesters, journalists, or anyone else they don't like and to use the implicit threat to encourage compliance from everyone else.
I'm not saying this is a likely outcome for anyone who buys an electric car, but it is new failure mode that didn't exist in older, non-internet-connected cars.
It's kind of frustrating that there doesn't seem to be a middle ground between giving an enormous amount of power to supposedly-benevolent-for-now product manufactures or finding and disabling the car's antenna and not getting any updates again ever. (Hopefully the car doesn't just refuse to function if it's not able to phone home?)
You can remove the SIM card from a Tesla if you don't want it to phone home, and only connectivity functions are broken (obviously). This isn't a particularly satisfying solution, of course, but it's more recourse than I have against, say, Google shutting me down.
You probably actually have implied (or express) warranty protection against the manufacturer screwing with your car. Certainly for the first four years, possibly even after that. I feel like software updates ought to come with their own implied warranty, though that's far from standard understanding right now. Beyond that, it's likely that anyone deliberately degrading your car, even the manufacturer, is a tort.
Taking out the SIM card sounds like a good idea if it's physically possible (i.e. the "sim card" isn't soldered down or integrated into some other non-removable device).
Maybe the sensible thing for most IoT-enabled car owners is to get all the software updates for the first few years as the manufacturer works out all the major bugs, then disable the network when the manufacturer's attention has shifted to the next new thing and drive it until the wheels fall off.
Someone breaking into your house and stealing all your possessions will be prosecuted. So you have an implied warranty from the government for the security of your possessions.
I bet you don't leave home without locking the door though.
So I'm REALLY curious to know now - in all those road tests where people testing the car gave the 60kWh cars such glowing reviews, was Tesla flipping the switch for more range, the way ISPs flip the switch when you run speed tests??
Consumer Reports buys cars anonymously to avoid just this class of issues. I don’t know which other reviewers do this – it’s obviously more effort and expense than testing a car that the manufacturer gifted or lent you, or even that they knew the intended use of.
The 60 watt model s has the same battery as the 75 watt model s. It's well known that you can pay tesla $200 to upgrade your 60 to a 75. If the road testers were testing the 60 watt model, they'd be testing it with the software lock on.
considering that the ONLY thing it changes is the range.. and the ranges are public info... wouldn't that be pretty obvious to the person reviewing it?
I would normally hope so, but Tesla cars have definitely had a bit of "fanboy blindness" happening where they don't really think too closely about things.
This also could apply to the battery's overall lifespan - by releasing more capacity with software as the battery degrades, they can make their "60kWh" battery seem to last far longer than any other competing battery. While this is nice for the consumer, it is very disingenuous if they do it and claim "we have better batteries"
Edit: To clarify I'm not claiming they do this, but just bringing up another possible use of this "software switch" that they could use maliciously. We just can't possibly know because there's zero way TO know...
Mechanically, the battery is better yes. However since it's becoming rather obvious that the general population had no idea how this battery limit worked, it means that Tesla could say "our battery is better" and be lauded for making the better "60kWh" battery. They don't have the "better" battery, because you're comparing an apple to a bigger apple and saying the bigger one is better at being an apple.
Sounds like a perfectly valid strategy to me. You can create a longer lasting battery by investing in R&D or spending more on the battery itself.
No consumer is buying batteries by looking at what technology was used to create it. They're buying it for X range, X recharge cycles, for X dollars. Those are all of the variables. So giving them more recharge cycles for the same money IS better.
This is exactly the way the SSD market works btw.. in fact its worse, because the largest manufacturers are actively investing money to make shorter life (but denser MB/area) cells (SLC > MLC > TLC), while building more spare capacity into each drive to meet their reliability targets.
If a 75kwh battery downrated in software to act as a 60kwh battery performs better, isn't that still "better" than a cheaper "real" 60kwh battery? Is that really different from using a more expensive cell design in the "better" battery?
Range depends a lot on driving conditions. Someone test-driving a car is unlikely to get anywhere near the published range, as they'll be pushing it a lot harder than typical conditions.
You still have to plug it in and charge the battery to 75kw.. if they were driving it, and they upgraded it, it wouldn't change anything. So when they started the test on a full battery, it either had the estimated range of a 60kw battery or a 75kw battery.
I understand the value to managing the hardware and software of your deployed devices centrally and with protection from users breaking things or becoming a road hazard, and the significant difficulties in developing for an automotive environment. I still wish we could buy Tesla's hardware and run something open source on it like we can with Apple.
Why would you buy hardware for double the going rate for a given set of specs and remove the OS that you paid so much for? Macbooks are OK, nothing against them, but I never understood who would buy one to run linux on. (I bought a Dell, and it actually included support out of the box)
Of course you can... first of all, the OS itself is (at least partially) open-source. Second, there's tons of apps in the App Store that are open-source. And third, you can obviously also compile open-source software yourself and install it too.
lol I own 2 macbook pros and a mac pro - wiped apple software off an installed linux on all of them. never looked back.
They all work great (a little bit of extra work setting up) - but you are right, I wouldn't have bought them if I knew I was going to ditch apple software ahead of time.
Tesla had preorders for lower capacities and chose to honor the deal and charge for the extra capacity instead of build another smaller battery pack. They could have told people who preordered the smaller one to pucker up. They also probably have over provisioning so they could in theory enable more capacity in an emergency.
Or they could have given those people a free upgrade. It would have cost them nothing except the opportunity to try and wring more money out of customers.
A find it humorous that this is a real world equivalent to mobile game speed boosts, where you pay to save time. It's interesting to ponder the monetization model this could open up. Imagine paying extra to unlock another 10 MPH on the highway (or unlock the speed limit governor). Before going on a long trip you pay for extra distance. Level 3, 4 and 5 Autonomous AI can be toggled via in-app purchase on your mega-screen.
One step further; the car is free and self driving, but you're bathed in a 360 degree never ending stream of advertisements, targeted to you based on where you've been, and where you're going (not just your digital/google footprint but real world).
Perhaps it's like what Sony did with the PS3 a long time ago: they manufactured it with an eight-core processor but only allowed access to seven of them. The manufacturing yields were really low at launch so they manufactured an eight-core and played the odds that they'd be able to fix the yields later.
The PS3 had one PPE core, and eight SPE cores. One SPE was disabled for yield purposes and of the remaining seven only six were available for usage by end developers.
intel did this back in the days when they sold 80486 CPU with DX and SX designation, the latter lacking math coprocessor a.k.a FPU.
In assembly line, all chips would start off as DX with math FPU. In order to make SX variant, they would use just disable FPU function, though they eventually made different chip to cut out FPU altogether.
Does anyone know exactly how this is done? does the battery just never charge up all the way? does it claim it's empty when there's actually more charge left? Are some of the cells locked down and just don't charge? How does it impact the health of the battery (since not being able to charge or discharge completely would presumably affect the health of the cells)