Battery Management System - What I Learned At Tesla Service Center

[Gasaraki]

This is not a personal jab, but what you just described is the definition of range-anxiety. I think perhaps you’re over thinking this whole thing in situations where it’s really not warranted.

No, overthinking is looking at the number next to the battery and it's not saying 322 but 318. Then driving and noticing that that number is not matching how far you are driving.

Using the map to get state of charge is not anxiety. It's how you NEED to travel if you need to travel long distances and use superchargers. Superchargers are not at every exit.

 

[camalaio]

I've written, rewritten, deleted, written again, and again deleted this post a lot. But one last try, in an uncharacteristically short summary for me:

Regarding balancing, we'd probably be best to forget about the doc we've seen. By my measuring, it is either incorrect or misleading. We don't know which parts are still (or ever were) true. Even if the doc is partly correct, imbalance is so incredibly unlikely that the imbalance scenarios I've tried outlining are plainly asinine.

If you are reading this, your battery is almost certainly properly balanced. If it is not, you're likely in talks with Tesla, not this forum.

This is not a personal jab, but what you just described is the definition of range-anxiety. I think perhaps you’re over thinking this whole thing in situations where it’s really not warranted.

Personally, I think they highlighted a very realistic approach to non-anxious driving. Planning and reacting is still absolutely necessary with EVs (especially in poor weather conditions, which you can't always plan for). Considering the data you need to make a trip successfully with limited "refueling" options is absolutely not over-thinking, but required (especially where EV charging options are still thin, e.g. most places outside southern California).

errr - in most cases it is demonstrably NOT the case.
Most ICE cars after 10 years do not get anything like the same MPG compared to new, same for power output.
An ICE is degrading from the moment it is first run.

It seems like a lot of people on here haven't overly measured their ICE vehicles. Properly taken care of (and this is not hard), ICE vehicles can still get great mileage after 10 years. Our 12 year old Honda Fit still gets 5.3L/100km on the highway (worst case) and about 5.8L/100km (worst case) in the city. It's EPA rated for 7.1 L/100 km city, 5.5 highway. (lower is better for L/100km).

ICE vehicles are at their worst efficiency beginning of life (my Crosstour got something like 30% worse at new than when it was broken in!), get better after break-in (yepp, this is real), then have generally fixable (certainly preventable) degradations to efficiency later in life. A lot of engineering has gone into modern ICE, and they give incredible reliability for all that's going on.

The worst degraded mileage I've witnessed is a buddy's old 1992 Mazda MX-3 V6. It was rated for 20mpg in city, but got more like 15. It ran terribly and had multiple issues, but the actual efficiency hadn't degraded as much as you'd expect. It was just a 1992 vehicle that didn't have great efficiency to begin with, which is maybe why people think that cars old in years get so much worse? A 25 year old Tesla isn't going to be any better if it's even alive.

EVs literally get worse every day, and this is permanent and unrestorable (short of replacing the cells, which isn't "fixing" to me but replacing).

Dude I know. I have the same. 8.5 of the 9 months we own our Model 3 SR+ it’s lacking -11 to -14% capacity.

When I bring this up to Tesla they won’t help you because no error. They tell you rated range is variable, bms needs calibrations, lecture about outside temps, tire pressure, car doesn’t measure heater energy, used energy isn’t accurate, drive even slower then speedlimit...

All bullshit and excuses in three different Tesla service centers in NL. I am also pretty sure you have the right to know battery health.

And yes I have frequently driven from 90, 100% to 5% and charged back up. I never charge small narrow charges.

Want to test for yourself how rediculous it is? Ask what State of Health is exactly next time when you are at service, Tesla will tell you it’s secret. (Even though I/you own the car!)

Funny thing is they tell you it actually would cost you money to get it checked! Even when they can see true capacity, but they won’t tell you capacity..

Our SR+ is usually around 208 /240 rated miles. Been like this when we got it, effectively never has been a SR+

ScanMyTesla shows I have 44.8 kWh available capacity at 100% charge. With a 0.5kWh heatloss that means max 39.7 kWh useable with a 90% charge. Buffer is 2.1kWh

(new full pack is 54.45kWh, ours is 46.x at best)

I really don’t want to as a shareholder, but after my “cooling down” (furious) I might explore legal options..

If that doesn’t work you’ll see my personally made Documentary (I built my own E-moped, programmed it, know how BMS work, ex- electric-cartechnician and I can edit videos pretty well also!)

Tesla might disagree with your lack of narrow cycles (these would be better for the battery, but potential worse for calibration?). But don't go by the new full pack value; as far as I've seen, this value is the same for all vehicles of the same size battery and may not be representative of what your battery was at new.

 

[TimothyHW3]

Referring to the BMS as “balanced” is confusing and inaccurate. Your method seems to be focused on calibrating the BMS.


This might work, but the Model 3 Service Manual, quoted in a previous post, indicates a .

Balanced calibrated, healthy, just wording - result is the same and the meaning understood.

As for charging - nobody really knows. I really doubt even Tesla has a perfect understanding how the BMS should calculate the charge.

The fact is that after all these reports it shows that charging small increments messes up the range estimate and charging from low to high all the time fixes that(numerous reports) Furthermore this manual says that they only balance when there is cell imbalance, which is rare and this manual also falsely claimed that above 4.0V is around 80 or 85% SOC? In reality it is closer to 70%.

So as someone pointed out this could be a different draft.

 

[TimothyHW3]

If I do suddenly get more buffer, I know that I can go a little faster. Now I'm not saying I drive slow to begin with. I drive at 75 and adjust up or down from there.

This of course kind of works, but if you really want to understand how your car works and have a good understanding of the range, you should get into the energy graph and just follow the consumption from the last period(the shortest one in the graph) and do your math from there - backwards from your capacity. But for this you need the km range or at least to understand the whole buffer below 0%.

This will give you both more confidence and more accuracy and you wouldn't be that mad man on the highway who first drove at 100mph only to creep at 60 later on and then do 80...

 

[lanbo]

The fact is that after all these reports it shows that charging small increments messes up the range estimate and charging from low to high all the time fixes that(numerous reports)

In the telegram group of users from Spain that have installed the OBD2 we've observed the contrary in a few cases. People charging 40% to 60% have increased their estimates of Nominal Full Pack. Once they charge from low to high SoC then they get worse charging estimates again.

That may mean that charging in small increments can unbalance the BMS in both ways.

 

[TimothyHW3]

In the telegram group of users from Spain that have installed the OBD2 we've observed the contrary in a few cases. People charging 40% to 60% have increased their estimates of Nominal Full Pack. Once they charge from low to high SoC then they get worse charging estimates again.

That may mean that charging in small increments can unbalance the BMS in both ways.

No, it actually means that charging short increments doesn't give enough room for the BMS to calculate the real range.

If you just charge 20% the margin of error is small. Once you charge big increments the car sees - oh, they charged 90% and we only added 63kWh this means we have 70kWh at 100%.

This is the same as any Battery health app on your phone. Tesla does have a few extra measurement points for sure, but this is the gist of it.

So if you simply drive 40%-60% the car will not readjust. So when for some reason the BMS was off by 5% by the time you started doing that, it will remain at that 5% error the whole time.
20% on LR is around 15kWh on LR and even less in MR or SR. Even if you had 5% error that is still less than half a kwh on each charge. The car can hardly make any adjustments based on that.
The only way to retest it and give it more calculating space is to charge it to nearly full and drive down to nearly 0 on one go. You will see that the used kWh will be closer to what the BMS predicted when you were charging 10-90.

Ask your Telegram friends to charge full and drive on the highway at around 100-120 until close to 0% and see what comes out of it.

 

[lanbo]

No, it actually means that charging short increments doesn't give enough room for the BMS to calculate the real range.

No to what? This is exactly what I said.

That may mean that charging in small increments can unbalance the BMS in both ways.

 

[KenC]

If the BMS can't handle calculating recharging in short increments, how does it handle regen?

 

[TimothyHW3]

What is there to handle. The regen is so minimal, 80kW in 5 seconds is about 100Wh. The car sees 100Wh coming in and adds it to the calculation.
Imagine you are in a swiming pool and only swim 5 meters, back and forth. How will you know how long the pool is if you have never swemt it fully from one end to the other?!

Now imagine the same scenario, but you know the pool was roughly 75 meters at the begining. But you swim only 20% of it somewhere on an unknown point inside and can't really be sure wether it is 18% or 22%. And you also don't know if someone didn't cut one end or the other and made the pool shorter (ie some cells keeping lower charge).

That is roughly the scenario the BMS needs to figure out.

 

[mrgoogle]

What is there to handle. The regen is so minimal, 80kW in 5 seconds is about 100Wh. The car sees 100Wh coming in and adds it to the calculation.
Imagine you are in a swiming pool and only swim 5 meters, back and forth. How will you know how long the pool is if you have never swemt it fully from one end to the other?!

Now imagine the same scenario, but you know the pool was roughly 75 meters at the begining. But you swim only 20% of it somewhere on an unknown point inside and can't really be sure wether it is 18% or 22%. And you also don't know if someone didn't cut one end or the other and made the pool shorter (ie some cells keeping lower charge).

That is roughly the scenario the BMS needs to figure out.

This is a common example but a BMS is never “blind”

The BMS can also use voltage levels pretty accurately in combination with learning imbalance from each block. 4.2v per cell is full while 2.5-3v is empty/near empty. (Cells have tesla tunes characteristics)

When driving normally the car isn't “blindly” loosing charge, it also monitors voltages and imbalance.

My theory is that our Model 3 SR+ figured out after 1-3 charges (we got it new) that it was a “short end of the stick” batterypack with -13% capacity.

I have seen near 90mV imbalance when driving long trips at slow highway speeds to test capacity. The more imbalances tou have the less of the capacity gou can use, because you can’t let the lowest brick go below 2.5volt (damage)

 

[TimothyHW3]

True about them roughly knowing the voltage, but since the cells are connected in parallel (no idea really how the do it or measure it) - if some of them are on the lower end of the block that can scew the results. Like in your case with the imbalance. Plus they don't really know the exact SOC% on each V level

I have seen only one guy having such high imbalance and charging nearly full to low for about 3 months straight kind of fixed his imbalance. Might be also that batman and Robin thing

At the low end the imbalance increases, but even on the low end near 0% mine was about 18-20mV. Normal is about 2-4-6.

What is yours at 90%?

 

[CertLive]

This video is quite interesting if you want to see some BMS results.

It actually shows lowest cell and highest cell values. Module overall voltage. It can actually see cells? Or does it class a cell as a pack?

 

[CertLive]

 

[CertLive]

If you want to go deep down the rabbit hole

 

[mrgoogle]

True about them roughly knowing the voltage, but since the cells are connected in parallel (no idea really how the do it or measure it) - if some of them are on the lower end of the block that can scew the results. Like in your case with the imbalance. Plus they don't really know the exact SOC% on each V level

I have seen only one guy having such high imbalance and charging nearly full to low for about 3 months straight kind of fixed his imbalance. Might be also that batman and Robin thing

At the low end the imbalance increases, but even on the low end near 0% mine was about 18-20mV. Normal is about 2-4-6.

What is yours at 90%?

At 90% it’s common to see only 4mV but when I drive from 90 to 10% my imbalance is 80mV or higher, after charging it is back to 4mV quickly.

i will try to go even lower next natural moment to test it. Curious if it gets even worse imbalance

(indicates a bad pack imo, and BMS learns how to counter that?)

 

[jmichaels]

Again, thanks for the helpful posts. One point I'd like to clarify. The last sentence on the prior post states:

The best way to balance the Model 3 pack is to set charge limit to 90% or higher and let the vehicle sit idle for hours (plugged in or not). 24 hours of balancing can reduce imbalance by 1mV.

For my purposes does that statement mean that if I periodically charge to 90-95%, and let the car stay at that level for as much as 24 hours, that I'll effectively rebalance the battery pack to obtain maximum range (given the state of my battery pack)? If so, shouldn't I then just charge from 30-80% as "standard", and then charge to the higher level "occasionally"? That seems to suggest that I'll have maximum battery life (staying in a 30-80% range), and maximum range (rebalancing the battery pack to offset low cell voltage and other issues?

Lencap, read the posts, thanks for asking the questions I am dealing with as well. Anything more beyond the post above? Is this your final plan of attack?

 

[buttershrimp]

Finally a post that describes an important fact about stated range and gives a small clue to the mystery 8 mile knee capping that occurred to most Teslas late last year. The stated range is important for psychological purposes and despite a lot of smart people on this thread, this is the first post I’ve found that describes a plausible explanation of what is happening with how the stated range is affected by something new in BMS algo. Bjorn did some good videos on this a while back and found a key assumption was changed in the algo and was in effect hiding range somewhere in the pack. It was confusing as hell why both of my Teslas, a model S and Raven X lost 10 miles of stated range over two updates late last year. I’ve yet to get the psychological miles back but my test of actual ability to drive super 311 miles without stopping. Let me go on record as saying how dumb it is that tesla did this, creating a perception of lost range is just dumb. Super super dumb even if it is perfectly logical.

Greetings!

Like many others I've noticed a drop in range on my 2018 LR RWD Model 3. The loss is about 8-10% over the last few months. I assumed that it was due to the new Version updates, vampire loss or a combination of both. When I brought my car to the Tesla service center to rotate my tires I mentioned it to the tech. He offered to check my battery, which he did, and his comments surprised me.

He said that Telsa has a Battery Management System (BMS) in all of its cars, and that the purpose of that system is to balance the battery load. It does this by various readings, but the key point is that if you don't discharge your battery below 20% capacity the BMS system is dormant. That means that it begins to sense that your capacity is being limited so it gradually decreases your range (I may be explaining this incorrectly). The point is that his instructions to restore full range were very different from what I thought was proper battery management.

I typically keep the battery charged from 30-80%, rarely going below 20% or above 90%. He said that will maximize battery life, but not maximize range. To do that you have to "cycle" the battery to use most of its range. He suggested that I NOT plug in the charger whenever the car is in the garage. Instead run the battery down to 10% or less then charge it up to 90% or more. Repeat this process for several cycles. The BMS will sense the changes in the battery usage and gradually restore the full range.

He noted several things:
1) My battery capacity hasn't been permanently lost. The BMS is curtailing range and following his procedure will help restore full range (he assured me that the Tesla battery test on my car shows that I have at least 8% more capacity than is being made available in normal use).
2) If the battery charger is plugged in the BMS doesn't work! You MUST keep the car disconnected from any charger to engage the BMS system.
3) Using the Supercharger after the BMS is reset does not decay the battery capacity or range on the M3 in any way. The car is designed for this type of use.

This is very different from what I thought was best practice, especially the part about not charging when I'm not driving the car. I'm not saying that he's right and everyone else is wrong, but I'm certainly going to try what he suggested and I'll report back after several cycles to let you know what I find. Frankly, it makes sense to a degree. There isn't any other way to explain my range loss.

It may also be that many of us who have blamed software or other losses on decreased range can regain the lost range by actively engaging the BMS program through the process I just described.

Hope it work for all of us!

 

[Gasaraki]

Finally a post that describes an important fact about stated range and gives a small clue to the mystery 8 mile knee capping that occurred to most Teslas late last year. The stated range is important for psychological purposes and despite a lot of smart people on this thread, this is the first post I’ve found that describes a plausible explanation of what is happening with how the stated range is affected by something new in BMS algo. Bjorn did some good videos on this a while back and found a key assumption was changed in the algo and was in effect hiding range somewhere in the pack. It was confusing as hell why both of my Teslas, a model S and Raven X lost 10 miles of stated range over two updates late last year. I’ve yet to get the psychological miles back but my test of actual ability to drive super 311 miles without stopping. Let me go on record as saying how dumb it is that tesla did this, creating a perception of lost range is just dumb. Super super dumb even if it is perfectly logical.

You want to under estimate than over estimate. Do you want the car to instantly die when it hits 0% or a safety buffer under 0%? I prefer a buffer under 0%.

 

[buttershrimp]

You want to under estimate than over estimate. Do you want the car to instantly die when it hits 0% or a safety buffer under 0%? I prefer a buffer under 0%.

Neither actually, as that summary misses that I’m intending to say. I wish I could describe it better to make the point.

 

[buttershrimp]

Neither actually, as that summary misses that I’m intending to say. I wish I could describe it better to make the point.

You want to under estimate than over estimate. Do you want the car to instantly die when it hits 0% or a safety buffer under 0%? I prefer a buffer under 0%.

ok here goes... let’s just look at the raven X with 5,000 miles even though I think it’s an important detail that I also own a model S.... with 40,000 because anecdotal scenarios aren’t all that helpful, BUT Bjorn’s videos and my experience at the same time means I am seeing something... and yes.... it’s super stupid by tesla either way.

If you own a 5000 mile Raven model x and have 325 mile range at 100% soc, then two updates and one month later, you can’t go higher than 312 EVER for the next 10 months.... that is super stupid lack of foresight on Tesla’s side. Particularly because it only affected perceived max range, which is really only useful for psychological purposes when the car can run similar stats on 311 mile trip pre and post knee capping. Ask yourself how many owners call and bug the service center claiming loss of range.... not understanding the issue as you (or I) do? That’s dumb.... perception of full range absolutely matters, just as suddenly capping your newish car at
13 miles fewer range overnight is really dumb. At least give the owner a heads up on why and how... they can live with the answer. Same thing with my model S happened... used to have 318 mile range max then over 40,000 it goes down to 311 max range before knee capping update, one day later whammy! 311 to no more than 294!!!! It’s never gone back to normal, which is why I appreciate the sheer brilliance of the OP at least asking the questions and getting an answer that is not totally clueless from service center.
I did the super long trip anyway and verified what Bjorn found, that the range is being hidden somehow based on algo. Though it s possible perhaps likely that it’s not at all on purpose, it was still made with minimal thought into ripple effect of owners thinking their tesla battery loses chunk of range early. It’s not smart decision to keep reasoning quiet but I guess that’s what happens when people do stupid class action lawsuits on autopilot technology not being delivered as hoped....