Phantom Drain is Gone!?

[jbcarioca]

I installed 2019.32.12.2 about two weeks ago. Eight days ago I departed for a long trip. In the past my Model 3 has lost about 0.5% per day ( my prior Model S was typically about 1%). This time after eight days I have precisely the same charge I had when I left. I did plug in the car but I set the lever at quite a bit lower than the car had already, so there has been no charge since I left.

this seems like a BIG deal to me. Have other people been having similar experience. If this continues I will save a fair amount of money I have paid to support The Phantom during the last five years!

 

[Ostrichsak]

Phantom drain is a physical/chemical fact of lithium-ion & nickel metal hydride batteries. No amount of software updates is going to "fix" it as you mentioned. The headlines make it seem like this is a "Tesla" problem when in reality it's a battery chemistry problem. Charge your phone fully, power it off & leave it on your dresser for a week. When you come back it won't be at 100%. Same with your toothbrush, headphones... Anything that takes rechargable batteries. Even lead acid discharges which is why normal 12v car batteries shouldn't be left sitting for prolonged periods. Your range indicator isn't an exact figure & what you're experiencing is something within the range of accuracy that's considered acceptable & has nothing to do with your update. Software updates can fix lots of things but basic chemistry & physics are difficult to overcome.

 

[jbcarioca]

Phantom drain is a physical/chemical fact of lithium-ion & nickel metal hydride batteries...

Not only those, lead-acid also. The point as I stated it is that the Tesla Phantom Drain seems to be gone. That was driven by other processes than simple battery loss over time, which we all know happens, just not anywhere approaching the 0.5% or 1% per day that we have been accustomed to. Obviously by the time I return to Miami six weeks from now there will be some loss. I plan to measure it over that period to see just how much it may have been. In any event The Phantom was/is a different issue.

 

[Ostrichsak]

Not only those, lead-acid also. The point as I stated it is that the Tesla Phantom Drain seems to be gone. That was driven by other processes than simple battery loss over time, which we all know happens, just not anywhere approaching the 0.5% or 1% per day that we have been accustomed to. Obviously by the time I return to Miami six weeks from now there will be some loss. I plan to measure it over that period to see just how much it may have been. In any event The Phantom was/is a different issue.

I addressed the lead acid similarity later on in the same post. It would be nice if it was something they could address via a software update but it's not going to happen as it's just a side effect of power storage. Even non-rechargeable batteries have the same issue it's just not as rapid. Hence expiration dates on Duracell & Energizer batteries. This is why they're working so hard on new battery technology because even incremental improvements mean massive advancements in the end product.

 

[mociaf9]

I installed 2019.32.12.2 about two weeks ago. Eight days ago I departed for a long trip. In the past my Model 3 has lost about 0.5% per day ( my prior Model S was typically about 1%). This time after eight days I have precisely the same charge I had when I left. I did plug in the car but I set the lever at quite a bit lower than the car had already, so there has been no charge since I left.

this seems like a BIG deal to me. Have other people been having similar experience. If this continues I will save a fair amount of money I have paid to support The Phantom during the last five years!

Since you left it plugged in, I wonder if the car will just automatically pull shore power for some things regardless of the fact that it's not actively set to be "charging"? To tell for sure, you would probably need to have data logging on either the car or on your home electricity circuit.

 

[dgatwood]

Phantom drain is a physical/chemical fact of lithium-ion & nickel metal hydride batteries.

No, it isn't. The self-discharge rate of a typical lithium ion battery is only about 0.5–1% per month. If you lose more than 1–3 miles of range per month, the problem is not self-discharge.

Most folks seem to see a discharge rate of 0.5–1% per day. That's almost two orders of magnitude higher than you would expect from the battery chemistry itself. Heck, that's twice the self-discharge rate of even lead-acid batteries, which are about as bad as batteries get, self-discharge-wise.

A loss of two or three miles per day (not uncommon) is equivalent to about 40 watts of continuous consumption. That's almost like leaving your headlights on, and your car is using that much power while turned off. This is really quite bad. Tesla should be a lot more aggressive at reducing power consumption while the car is turned off. It isn't, so the MCU and AP computers use a lot of power. I'm guessing their kernel isn't dropping into the lowest C-states (if it ever even makes it into C-states).

All of that should be fixable through software improvements.

 

[jbcarioca]

...A loss of two or three miles per day (not uncommon) is equivalent to about 40 watts of continuous consumption. That's almost like leaving your headlights on, and your car is using that much power while turned off. This is really quite bad. Tesla should be a lot more aggressive at reducing power consumption while the car is turned off. It isn't, so the MCU and AP computers use a lot of power. I'm guessing their kernel isn't dropping into the lowest C-states (if it ever even makes it into C-states).

All of that should be fixable through software improvements.

The entire point of this thread is that Tesla seems to have done just that. As of this morning, nine days after I left my car, it has lost 1%. Prior to the most recent firmware update it was losing -0.5% per day. The only explanation I can find is that Tesla has eliminated nearly all the Phantom drain.

 

[jbcarioca]

Since you left it plugged in, I wonder if the car will just automatically pull shore power for some things regardless of the fact that it's not actively set to be "charging"? To tell for sure, you would probably need to have data logging on either the car or on your home electricity circuit.

I have logging of all electricity used by my charger, including each individual draw duration and kW. Zero logging has occurred single the last eight days. Historically power has been consumed only when charge level descends below the set target. This time I set the target 20% below the indicated level at which I left the car, because I suspected there had been a major improvement. While in daily use I could not be certain that drain had reduced significantly.

 

[dgatwood]

The entire point of this thread is that Tesla seems to have done just that. As of this morning, nine days after I left my car, it has lost 1%. Prior to the most recent firmware update it was losing -0.5% per day. The only explanation I can find is that Tesla has eliminated nearly all the Phantom drain.

The only problem is that you left out two words:

The entire point of this thread is that Tesla seems to have done just that, for now.

It's amazingly easy to introduce bugs in code, either in user space or in the kernel, that result in the CPU not being idled consistently. Keeping that from happening requires constant vigilance, and given that these problems keep disappearing and then coming back for people, I'm not convinced that Tesla's software engineering team has the discipline to prevent it from coming back again, nor that they have adequate test suites in place or adequate telemetry to recognize it in a timely manner and prevent releases with those sorts of bugs from getting deployed broadly, much to my chagrin.