[noonespecial]

Could this really not be prevented by a $10 microcontroller and a big-ass relay to just disconnect the battery if it reaches a certain discharge state?

Even my iPhone can turn turtle to protect its battery when it gets too low.

Also how much power does it take to keep the damn thing plugged in? The article indicates that a 100 foot extension cord isn't enough just to break even and the car discharges even when plugged in. A 100 foot cord of cheap 16 gauge wire can still supply almost 900 watts before the voltage drops below 100v. So it uses a kilowatt just to stand-by without even charging? That's one hell of a power vampire.

I think maybe they should spend a bit more time on the fundamentals and a bit less time on fancy bird-wing doors if they want to have a real product. The real car of the future is just a big dumb tray full of indestructible nickle-iron batteries and 4 wheel-hub motors bolted to the corners.

[bradleyland]

The account of the Tesla discharging while plugged in to a 100-foot extension cord strikes me as suspect. It's anecdotal at best. Have a look at the Tesla Roadster "Charging" page:

http://www.teslamotors.com/goelectric/charging

Using 120V @ 15A, you get a charge rate of +5 miles (of range) per hour. Obviously not the optimal solution, but not a net negative either. Adding a 100 foot extension cord isn't going to diminish the available current enough to result in a net negative, so either the owner had additional electrical issues, or simply failed to plug it in and is making up excuses.

[noonespecial]

120v @ 15A is 1800 watts. As I said, I was able to measure the wattage being delivered at the end of my 100 foot 16 gauge "home depot" orange extension cord connected to a space heater and found less than 900 watts being delivered to a load that normally pulls 1750. The voltage had also dropped below 100 volts.

Its possible that the Tesla simply won't charge at all if the current or voltage on the line becomes too low. In this case, it would be "plugged in" but not charging.

Edit: Lets figure it out: Range=244mi. Battery capacity=58kWh. So: 5 miles takes 1.086kWh. 10% is lost in chemical conversion so we really need 1.207kWh for those miles. We pull at a rate of 1.8kWh/h from the plug. So 600 watts or so is lost elsewhere. We know the "always on" battery cooling system alone takes around 150 watts. Its not that far fetched.

Edit: I deleted my comment below where I calculated the cost of keeping the car on standby because I don't think I made clear enough (and didn't want to type it all twice) the difference between the actual discharge rate of the battery and the amount of power required at the charger to stop it. It just seems to take a lot of power at the charge port before any gets to the battery. The battery seems to discharge at an average rate of just 30 watts, but it seems to take a much, much greater amount of power input to prevent this and failing to provide this power has dire consequences. Why?

[jws]

The pack is 53kWh according to wikipedia and discharges in 11 weeks from full according to the article. Unless I botched a decimal place, that is a 29 watt idle load. (53000 watt hours / (11 weeks * 7 days/week * 24 hours/day) ). In perspective, that is about the same as 4 night lights.

Clearly there is about 20 times that much power available at the end of a 100 foot extension cord in a standard household outlet.

Failure modes available:

• Was not actually plugged in.

• Was unplugged then plugged back in much later, possibly unbeknownst to owner. (Someone in my household unplugs my DVR to use an outlet. Grr.)

• Was on a switched outlet. (I once bought a new drill over that. Who knew one of the duplex outlets was switched and the other not? But the new drill wasn't $40k.)

• The Tesla charger could have a cutoff where it stops charging if unable to pull X amps where X is somewhere between 7 and 15. This could be either intentional or unintentional. Brown out protection circuitry, if needed by the design, could do this. They also might fear they are setting your house on fire by dropping 800+ watts somewhere in a wall and shut off.

• Broken charger.

• Broken extension cord.

[nash]

You forgot "Owner made up story"

[Vixter]

A roadster that is plugged in, is not really "off". It will provide power for heating or cooling of the battery pack if the temps get too high or too cold. This would be much more than a small idle load.

[bradleyland]

That's because a 16 gauge extension cord isn't supposed to supply 120v @ 15A. A 16 gauge cord is only good for around 10 amps.

Which means you have a point. If the owner in question plugged the Tesla supplied 120v cord in to an additional 100-foot, 16 gauge cord, they might have a net loss situation on their hands.

[HeyLaughingBoy]

A little OT, but I've found that it's better to make your own extension cables. I have a 1kW tank heater that is about 100' from the closest outlet. The cost of 100' of 12gauge house wiring and a plug, socket and outdoor outlet was much less than even a cheap 100' extension cord and has much less voltage drop.

[noonespecial]

That depends on your application. Extension cords are more expensive because they are made of stranded wire that remains flexible and resists breaking. Household wire is solid and not meant to be rolled and unrolled repeatedly. It will get kinks and weak spots. But if your application is for a relatively static run, left in place for a long time, the household wire is better.

[tlb]

The charge circuitry probably turns on at a high charge rate, senses the line voltage drop below 100v, and shuts down for a few seconds. Stuck in this loop, it might never deliver much charge.

[D_Alex]

Er, no, this is probably wrong. The voltage drop is due to resistive losses in the long (and skinny) cable, and the resistive losses depend on current flowing. So, unless there are some weird effects due to the car's electronics, you cannot stop the charging, only slow it down.

[marshray]

So the charger sees the 120 V, decides to begin charging, current flows, and the voltage drops. If the voltage drops out of range, the charger aborts because it's out of design spec and/or there's a significant amount of power being dissipated, probably by resistive heating (i.e., risk of fire).

This is probably necessary to achieve UL rating.

[nash]

I would agree. The whole document has a few little niggles that make me think it's BS, or at least severely overblown.

Additionally they quote it's happened 5 times... by an un-named service manager.

[geori]

Totally Agree. There are no names, only anecdotes and the science is suspect. This article is 100% B.S. and I wouldn't be surprised if it was planted by a competitor.

[david927]

The bigger an industry is, the harder they fight for their existence. Elon Musk is not just taking on the second biggest industry, the auto industry. He's also taking on the biggest industry: oil.

It looks like they switched from ignore/mock to attack. He's getting somewhere.

First they ignore you, then they laugh at you, then they fight you, then you win. --Gandhi

[URSpider94]

Could this really not be prevented by a $10 microcontroller and a big-ass relay to just disconnect the battery if it reaches a certain discharge state?

Not necessarily. This might help, but the battery will also have self-discharge, meaning that it can lose charge without being connected to any load at all. If you get the charge down low enough, and then wait long enough, then even a full disconnect from the system won't help.

[wisty]

You could add a $100 mobile phone, which auto-calls a service center. If 1% of cars get "bricked", that's $10,000 per car saved, not including intervention costs. Even with some false positives, it would be worth it. Of course, that cost is carried by Tesla, not the customer.

"The Tesla manager called me to warn that my car was in trouble" is a much better customer experience than "my Tesla broke down, and cost $40k to fix".

Cars used to have the same problems with oil changes. This got better, as customers were educated (at the cost of a lot of cooked engines), and cars were able to warn their owners, and run for longer without oil.

[bri3d]

That's exactly what Tesla do, even once having gone so far as tracking the car down via GPS and charging it themselves when the owner was unresponsive.

[mbreese]

According to the article they used to do that, but don't any longer.

[thebooktocome]

Certainly that's going to peg some people's creepy meters.

[toomuchtodo]

As a Model S reservation holder, it pegs my pocketbook meter. If I'm about to lose $40K on my car because of a battery, I have no problem with them showing up in my driveway to charge the vehicle if necessary.

Creepy? I'm more worried about the app-of-the-day sucking down my iPhone contacts. THAT I find to be more personal than my vehicle location.

[shinratdr]

IMO they're both creepy, and using one to justify the other isn't helpful. Apple has already come out and admitted that was a mistake and they will be rectifying it, so it doesn't really help your case much.

In both cases though it's not the action so much as it is the fact that you weren't able to opt out of it, or in many cases you aren't even aware of the possibility of it happening.

It's the same violation, and in both cases it's equally creepy. Apple got called out and did something about it. What will Tesla do?

[leoedin]

So set the cutoff higher?

Small portable gadgets using LiPoly batteries (which I assume the Tesla is using) can be left for months or years without charging, and then nursed back to life. If sufficient charge is left in them, and the circuitry is smart enough to recognise that the battery's about to die, the time before complete discharge could be extended to a much larger period of time.

The solution could even be a one-time use fuse, isolating the battery completely. Given the option, I think most owners would rather have to replace a fuse than a $40k battery.

[mjwalshe]

Yes you woudl have though that this is not exactly rocket science. Or have a mechanical switch to isolate the batery coudl this be conected to the Parking Brake.

Jermy Clarkson must be pissing himself laughing over this

[aidenn0]

LiIon cells have virtually zero self-discharge. It is the battery management system that contributes nearly all of the self-discharge to a LiIon battery. You could have such a relay, but it would require some additional tools to recover from such a state (needs to get the BMS in a happy state, or the BMS itself must have the ability to charge the battery from an unknown state).

This all would seem to make sense to install though, it's far better to have a pack that needs a reset in the shop then a brick, esp. as the battery pack in the model S will make up nearly all of the resale value of the vehicle.

[CamperBob]

Perhaps the most straightforward fix would be to simply refrain from drawing any current at all from the vehicle's main battery while parked. The alarm and any other standby processes should run from a 12-volt motorcycle battery or something, which is recharged during driving, and can be jumped like any other car battery if allowed to go dead.

That wouldn't fix the Li-ion self-discharge problem completely, but it could provide as much as a few months' worth of margin.

In any case it's batshit insane to leave early adopters stuck with a $40,000 bill for your lack of engineering foresight. That was the real surprise in the article.