If you fast charge, Tesla will permanently throttle charging

[whitex]

Yes Tesla handled it very poorly, but I think all EVs are going to have the same problem.

Not necessarily. Tesla product is a bleeding edge product, and if you live on the edge, you're more likely to fall. Other manufacturers don't let you charge to the true 100% SoC for example (which is known to potentially have adverse effects on battery life), even though that would mean they could claim more rated miles (like Tesla does). They also design their charging with similarly conservative approach, trying to ensure that the car will maintain its design specs for some expected lifespan duration. Each solution has it's advantages and disadvantages, so customers should pick the approach they prefer.

The main thing I'm pissed about is that I didn't know I was heading for a cap on SC rate. If I knew then (when I bought the car) what I know now, I'd have almost never used a SC, whereas I used it almost exclusively back then.

Most likely Tesla didn't know it back then either. Unlike say Porsche/VW, who spent a few years and few million miles testing their designs, Tesla pushes the envelope and learns on customers. "Ship it now, fix it later via OTA, fix hardware only if we have to" is Elon's famous MO.

The best way to think about it, when you're driving a Tesla, you're driving an experimental test vehicle. You get to experience the latest and greatest features before they are fully baked, so sometimes a few years before the other guys who prefer to thoroughly test before shipping, but you take the risk that it might not be good for your car's longevity.

 

[Duke-U]

I have been capped a year ago and stil use the SUC a lot, say 75% of my charge is SuC. But my battery degradation is very limited, better than some who never fast charge. I also read somewhere that fast charging is actually good for the battery, provided that temperature and charge curve is handled well.

Also I find that my time at the SuC is not significantly longer than before the cap. I typically charge to 70-75% to get home at 55-60% and charge it to 85% during the night. This way I can always immediately go if called for an emergency just after arriving home without first having to charge.

I wonder what happens to the new batteries with this ultrahigh charging power, will in 2 years from now the story repeat but now for 1000’s of M3s?

 

[David29]

I have been capped a year ago and stil use the SUC a lot, say 75% of my charge is SuC. But my battery degradation is very limited, better than some who never fast charge. I also read somewhere that fast charging is actually good for the battery, provided that temperature and charge curve is handled well.

Also I find that my time at the SuC is not significantly longer than before the cap. I typically charge to 70-75% to get home at 55-60% and charge it to 85% during the night. This way I can always immediately go if called for an emergency just after arriving home without first having to charge.

I wonder what happens to the new batteries with this ultrahigh charging power, will in 2 years from now the story repeat but now for 1000’s of M3s?

I rarely use a Supercharger as I don't travel much (especially now with the virus, of course). And I am careful never to charge over 80% at home for routine daily use, and no more than 70% most days. So, I was very surprised when I picked up my car from the service center after the MCU1 replacement about 10 days ago, and found they had charged my car to 100% on the supercharger. Why would Tesla do this, when they themselves tell you not to go above 90% unless you are on a trip? I did not recall this happening before, and I had had very nice service otherwise, so I did not say anything. But I tried to knock a few % off right away, even though I had no particular place to go. Still, the car had 90+% charge for most of 24 hours, unfortunately. (I suppose i should have taken the opportunity to do a range/degradation check, and driven it down to 5 or 10%, as a test, but that did not occur to me in time.)

 

[seenhear]

Still, the car had 90+% charge for most of 24 hours, unfortunately. (I suppose i should have taken the opportunity to do a range/degradation check, and driven it down to 5 or 10%, as a test, but that did not occur to me in time.)

I keep seeing people mention stuff like this range/degradation check. Is there a wiki or primer or something on how to DIY these battery tests?

 

[seenhear]

Other manufacturers don't let you charge to the true 100% SoC for example (which is known to potentially have adverse effects on battery life), even though that would mean they could claim more rated miles (like Tesla does). They also design their charging with similarly conservative approach, trying to ensure that the car will maintain its design specs for some expected lifespan duration. Each solution has it's advantages and disadvantages, so customers should pick the approach they prefer.

Who are "they" in this instance, these "other manufacturers" who are way more conservative with their battery estimates, placing software limits on range and charging speed preemptively to make the battery last longer? Just Porsche? You couldn't be referring to Nissan, that's for sure! LOL. To me, the Leaf was the "experimental" car; advertised at 100mi range new, which quickly fell by 10+ miles every 15-20k miles of driving. Other cars (Bolt, e-Tron, Jaguar) haven't really been on the road long enough for issues like this to crop up, so how do we know they are so conservative? In the 2015-2017 era, when I was researching buying an EV, Tesla was considered the leader in battery research and longevity. The initial stories of the Tesloop cars backed this up back in 2017.

Granted they continue to learn, but by the time I bought my Model S75 (HW2.5, Sept 2017) these battery issues were already known (ref. the first post of this thread) and despite trying to learn as much as I could about Teslas, I had no idea about this issue. The only warning I knew was "don't charge to 100% except rarely for roadtrips" which I heeded; I also would have happily heeded a warning about "don't supercharge except rarely for roadtrips" too, if they'd given me one. Instead what I got was "congrats you got one of the last cars with unlimited supercharging! Have at it!"

Despite the company explanations that the OP of this thread received in May of 2017, my in-car digital owners manual did not have any warnings about supercharging when delivered in October 2017; although I can't prove that now, as it has since been updated. I now have 56kmi on the odo, and about half of those miles were charged at Superchargers. My car now (for the past 6 months or so) seems to be capped at a max SC rate of 81kW.

 

[whitex]

Who are "they" in this instance, these "other manufacturers" who are way more conservative with their battery estimates, placing software limits on range and charging speed preemptively to make the battery last longer? Just Porsche? You couldn't be referring to Nissan, that's for sure!

"They" are Porsche & Audi so far (similar premium vehicle market segment as Tesla) - from the world's largest automotive manufacturer (VW). Nissan not exactly in the price range (they were selling $32K EV's when Tesla's least expensive EV was $80K).

Despite the company explanations that the OP of this thread received in May of 2017, my in-car digital owners manual did not have any warnings about supercharging when delivered in October 2017; although I can't prove that now, as it has since been updated. I now have 56kmi on the odo, and about half of those miles were charged at Superchargers. My car now (for the past 6 months or so) seems to be capped at a max SC rate of 81kW.

Because Tesla didn't know for sure, and they wanted customers to experiment. How else are you going to be able to say - out of 50,000 cars, so many of them supercharged often, and here is how their longevity compared to those who didn't supercharge at all, and to those who supercharged sometimes. This is very valuable data (to Tesla). I stand by my original point - with Tesla you get to drive the latest tech car few years ahead of the rest, but at a cost of the car being an experimental learning tool for the manufacturer. I wouldn't be surprised if besides different battery versions, Tesla also deployed batches of different firmware to test out different battery management strategies. That would make complete sense from Tesla point of view, allow them to learn which battery management strategy works best (at a cost of few cars which were the subjects for the less optimal strategies). Over the air updates allow them to mitigate issue they find in the field, if a particular battery management strategy starts hurting the battery too much, they can move that car out of the aggressive experiment group and/or limit the battery for safety. It's a brilliant strategy.

 

[cduzz]

So, I'm curious, if you were given a new battery right now, knowing what you now know about SC limitations and battery life limitations, would you be happy with that? If so, why is a new Tesla not an option for you? If not, then I think probably EVs are out altogether, right? I mean, isn't this problem largely a Li-Ion battery issue? Yes Tesla handled it very poorly, but I think all EVs are going to have the same problem.

The main thing I'm pissed about is that I didn't know I was heading for a cap on SC rate. If I knew then (when I bought the car) what I know now, I'd have almost never used a SC, whereas I used it almost exclusively back then.

If its true that they cap SC rate at around 30kmi of DCFCharging, I think I could make 30kmi last a long time if I only did it during long road trips, and the occasional "I'm in a pinch and need 50 miles of range to get home" situations.

I'm curious about this dcfc cap -- the battery is worn as much by regen as by dc fast

"They" are Porsche & Audi so far (similar premium vehicle market segment as Tesla) - from the world's largest automotive manufacturer (VW). Nissan not exactly in the price range (they were selling $32K EV's when Tesla's least expensive EV was $80K).


Because Tesla didn't know for sure, and they wanted customers to experiment. How else are you going to be able to say - out of 50,000 cars, so many of them supercharged often, and here is how their longevity compared to those who didn't supercharge at all, and to those who supercharged sometimes. This is very valuable data (to Tesla). I stand by my original point - with Tesla you get to drive the latest tech car few years ahead of the rest, but at a cost of the car being an experimental learning tool for the manufacturer. I wouldn't be surprised if besides different battery versions, Tesla also deployed batches of different firmware to test out different battery management strategies. That would make complete sense from Tesla point of view, allow them to learn which battery management strategy works best (at a cost of few cars which were the subjects for the less optimal strategies). Over the air updates allow them to mitigate issue they find in the field, if a particular battery management strategy starts hurting the battery too much, they can move that car out of the aggressive experiment group and/or limit the battery for safety. It's a brilliant strategy.

Yep, the S/X are the beta testers for the 3/Y family. I imagine they're constantly running experiments on clusters of 100-300 cars to see how chemistry / BMS strategies / etc affect fleet health. By the time a strategy hits the S/X it's pretty well proven but the world is full of edge cases. They can afford some moderate level of warranty support on that fleet while a catastrophic mistake in the 3/Y would crush them in warranty claims.

What's the current list price and warranty of a *new* 85kwh class battery pack for an S from tesla?

As far as throttling the battery based on DCFC, wouldn't that equally apply to any high mileage car due to regen? As far as the battery is concerned, DCFC and regen are identical.

 

[MarcG]

I'm curious about this dcfc cap -- the battery is worn as much by regen as by dc fast


Yep, the S/X are the beta testers for the 3/Y family. I imagine they're constantly running experiments on clusters of 100-300 cars to see how chemistry / BMS strategies / etc affect fleet health. By the time a strategy hits the S/X it's pretty well proven but the world is full of edge cases. They can afford some moderate level of warranty support on that fleet while a catastrophic mistake in the 3/Y would crush them in warranty claims.

What's the current list price and warranty of a *new* 85kwh class battery pack for an S from tesla?

As far as throttling the battery based on DCFC, wouldn't that equally apply to any high mileage car due to regen? As far as the battery is concerned, DCFC and regen are identical.

Regen puts energy back at ~60-70kW at most and only for a few seconds at a time, so I think it's pretty different than pumping 150-250kW of DC charging for several minutes if not half-hours at a time.

 

[whitex]

As far as throttling the battery based on DCFC, wouldn't that equally apply to any high mileage car due to regen? As far as the battery is concerned, DCFC and regen are identical.

Not exactly. Max charge rate is 120KW, mag regen ~50KW. Also, regen power varies all the time and is never continuous. Even if you regen 50KW for 3s every 30s (racing around the track scenario, where most people turn off regen anyways), that is a much different hit to the battery, as it has time to cool off between cycles, maybe distribute the charge or the heated up electrolyte, or whatever. An analog to this might be - DC charging is like heating something in the microwave oven on full power, vs. regen is like heating it 10 times longer on 10% power (which most microwave ovens just implement by turning full power on for 10% of the time) - you can heat up things on 10% power which would burn if you heated them on 100% power 10x faster. An actual electrical example - LED's, the most common way to change the LED brightness is to modulate 100% current ON/OFF, with the ON duty cycle being the desired brightness. Some LED's out there will burn out if you supply them with a stead 100% current, but will live for years if you turn full current ON/OFF many times per second, as long as it's not ON more than some % of the time and the ON pulse width is not longer than some spec.

Even if you found a mountain where you could somehow sustain max regen for a couple of hours, you'll still have additional cooling from the fact that the car is moving through the air, so still not exactly the same as supercharging. The only way I can see regen being the same as DC charging is if you put it on a dyno for a couple of hours, then it would be equivalent to a ChaDeMo charging session.

 

[cduzz]

Not exactly. Max charge rate is 120KW, mag regen ~50KW. Also, regen power varies all the time and is never continuous. Even if you regen 50KW for 3s every 30s (racing around the track scenario, where most people turn off regen anyways), that is a much different hit to the battery, as it has time to cool off between cycles, maybe distribute the charge or the heated up electrolyte, or whatever. An analog to this might be - DC charging is like heating something in the microwave oven on full power, vs. regen is like heating it 10 times longer on 10% power (which most microwave ovens just implement by turning full power on for 10% of the time) - you can heat up things on 10% power which would burn if you heated them on 100% power 10x faster. An actual electrical example - LED's, the most common way to change the LED brightness is to modulate 100% current ON/OFF, with the ON duty cycle being the desired brightness. Some LED's out there will burn out if you supply them with a stead 100% current, but will live for years if you turn full current ON/OFF many times per second, as long as it's not ON more than some % of the time and the ON pulse width is not longer than some spec.

Even if you found a mountain where you could somehow sustain max regen for a couple of hours, you'll still have additional cooling from the fact that the car is moving through the air, so still not exactly the same as supercharging. The only way I can see regen being the same as DC charging is if you put it on a dyno for a couple of hours, then it would be equivalent to a ChaDeMo charging session.

It's my understanding that the DCFC charging counter actually counts hours under ChaDeMo or under Gen 3 Supercharger in the same way. But I could also be wildly wrong.

 

[whitex]

It's my understanding that the DCFC charging counter actually counts hours under ChaDeMo or under Gen 3 Supercharger in the same way. But I could also be wildly wrong.

I heard the same thing, however even for that the regen is supposedly counted separately. While this is purely my personal speculation, I would be surprised if Tesla was not collecting much, much more detailed logs of charging and discharging, including the curves of voltage, current, time, temperatures (battery, ambient, other), age of the battery. The counters we hear of are just production level simplifications to which the service techs have access and possibly on which the throttling decisions are made. The actual analysis is likely happening on a much more detailed level at Tesla mothership from uploaded logs. Based on those findings, Tesla decides what simplified counters to implement and/or use for software throttling and diagnostics. I bet they can per car initialize new counters they decide to implement too - for example say they find out that time spent charging at some high DC rate at some high ambient temperature is particularly damaging and they decided to start counting only that time, they can implement a new counter, then initialize it with a number the mothership computer calculated from the logs collected over the years for each car. Again, I have no specific information on what Tesla does, but this is precisely how I would have designed it. For all those who though free connectivity was purely good will from Elon, you'd be wrong, it's so Tesla can continue to learn from their fleets. This is why even cars which don't pay for premium connectivity still have Tesla paid LTE for log uploads. Premium connectivity (web browsing, streaming) was a bonus for early adopters though, nothing to gain for Tesla there (hence the MCU1 browser was abandoned completely).

 

[n2mb_racing]

This is why even cars which don't pay for premium connectivity still have Tesla paid LTE for log uploads. Premium connectivity (web browsing, streaming) was a bonus for early adopters though, nothing to gain for Tesla there (hence the MCU1 browser was abandoned completely).

True, and this is why MCU1 cars have failing eMMC memory. They log everything like crazy and didn't bother thinking about how much they were logging and what effect that would have on the flash.

 

[whitex]

True, and this is why MCU1 cars have failing eMMC memory. They log everything like crazy and didn't bother thinking about how much they were logging and what effect that would have on the flash.

Or they did think about it and chose to deal with it when it becomes a problem (or hoping most of it breaks after the warranty, which would be consistent with recent halving of MCU warranty too). Remember too that MCU1 was designed before AutoPilot which added a ton of logging as part of Tesla fleet learning (at least that is what JB claimed in his presentation on AP1), so it might have been one of those decision:

1. design a new, updated MCU and retrofit all cars for free ASAP
2. design a new, updated MCU but don't worry about older ones until they fail, which would be months or years later, likely after warranty so customer will pay to fix it
3. don't collect AP data at all, until new MCU is ready

Guess which option makes most sense for Tesla economically.

 

[David29]

I keep seeing people mention stuff like this range/degradation check. Is there a wiki or primer or something on how to DIY these battery tests?

One source is the YouTube videos by Bjorn Nyland of Norway. He has done extensive and regular battery testing on his own cars (mostly Teslas, including Models S, X, 3) as well as other brands he is testing, for several years. You can find his videos at https://www.youtube.com/channel/UCG1QcV31eoSaX4rE8avQL4A

 

[tga]

It's my understanding that the DCFC charging counter actually counts hours under ChaDeMo or under Gen 3 Supercharger in the same way. But I could also be wildly wrong.

I heard the same thing, however even for that the regen is supposedly counted separately.

FWIW, Scan My Tesla reports:

• Total kWh for AC charging
• Total kWh for DC fast charging (Chademo and Superchargers combined)
• Total kWh for regen
  
• Number of charge cycles (not differentiating AC or DCFC)
• Number of discharge cycles

 

[aerodyne]

Here is an example:

 

[brucet999]

Here is an example:View attachment 572763

If I were to run such a scan on my car, what, exactly would those various readings tell me?
400V system 0.03kW tells me what?
battery power 0.19kW
DC-DC current and voltage is presumable about the 12V battery?
12V systems 160W?
DC-DC output power 122W?
BMS max discharge 333 kW What does this tell me? I'm not sure that I have ever drawn that much. Most times I see less than 200 on the dash display.
BMS max charge 96.9 kW I'm guessing that this is the supercharging limit imposed by software?

 

[aerodyne]

If I were to run such a scan on my car, what, exactly would those various readings tell me?
400V system 0.03kW tells me what?
battery power 0.19kW
DC-DC current and voltage is presumable about the 12V battery?
12V systems 160W?
DC-DC output power 122W?
BMS max discharge 333 kW What does this tell me? I'm not sure that I have ever drawn that much. Most times I see less than 200 on the dash display.
BMS max charge 96.9 kW I'm guessing that this is the supercharging limit imposed by software?

If I were to run such a scan on my car, what, exactly would those various readings tell me?
400V system 0.03kW tells me what?
battery power 0.19kW
DC-DC current and voltage is presumable about the 12V battery?
12V systems 160W?
DC-DC output power 122W?
BMS max discharge 333 kW What does this tell me? I'm not sure that I have ever drawn that much. Most times I see less than 200 on the dash display.
BMS max charge 96.9 kW I'm guessing that this is the supercharging limit imposed by software?

Answer to most of your questions is yes. I have a dual motor car, so 333 Kw is close to advertised.

There is a LOT more. Screen capture just shows the common stuff I put into a special tab. Knowing your battery capacity and cell imbalance is priceless.

For more, see this excellent YT vid by Dave:

Also, the SMT thread, there is an excel files somewhere attached that shows all 100+ parameters:

Vendor - Scan My Tesla, a CANBUS reader for Android

 

[srk67]

Received official confirmation for my model x 75d

Per your request, our remote diagnostics team looked into this further and found:

Autodiagnostic has detected that the peak fast charge current is affected by accumulated wear from DC charging (Supercharging), the current max DC charge current in ideal condition is 244.0 Amps. The limit was 470.0 Amps when the HV battery was new and the limit will go as low as 244.0 Amps when the limit will reach its lowest level. The vehicle has been affected by the amount of Supercharging this vehicle has had.

We recommend incorporating a slower rate of charge whenever possible to prevent further Supercharging rate loss.

As the vehicle ages, the HV battery pack and cells also age and accumulate some expected fatigue through repeated charge/drive cycles. Wear over time makes the battery cells unable to accept as much fast-charging power as they once did.

 

[VikH]

Received official confirmation for my model x 75d

Per your request, our remote diagnostics team looked into this further and found:

Autodiagnostic has detected that the peak fast charge current is affected by accumulated wear from DC charging (Supercharging), the current max DC charge current in ideal condition is 244.0 Amps. The limit was 470.0 Amps when the HV battery was new and the limit will go as low as 244.0 Amps when the limit will reach its lowest level. The vehicle has been affected by the amount of Supercharging this vehicle has had.

We recommend incorporating a slower rate of charge whenever possible to prevent further Supercharging rate loss.

As the vehicle ages, the HV battery pack and cells also age and accumulate some expected fatigue through repeated charge/drive cycles. Wear over time makes the battery cells unable to accept as much fast-charging power as they once did.

What does that mean for your max kw? How much DC charging have you done to reach this level?