Calibrating BMS with a software locked battery (SR)

[lateulade]

My SR+ (voluntarily downgraded to SR 5 months ago) is only showing 199 rated miles at 100% (29k miles on the clock). At the time of downgrade (at 15k miles) it was still showing 238 rated miles at 100%, and immediately went to only 208 miles on the date of downgrade.

I’ve been reading a lot about BMS and everything I’m seeing seems to indicate that BMS is calibrated when modules are full. I’ve drained to 1-2% and back up to 100% several times with no improvement, but it just dawned on me that since my battery is top-locked, I’m really only charging to ~92% capacity. Can BMS calibration really be achieved on top-locked batteries?

 

[ewoodrick]

Don't drain it low. Take it to 100% for a few days.

Most importantly, especially if you are on a capped battery, STOP WORRYING ABOUT IT.

And NEVER worry about max range in the winter.

 

[lateulade]

I’m in South Florida, so not sure what this “winter” you speak of is

I drive a ton (just hit 29k miles this week, and I took delivery on March 25th), and there are several drives I take regularly where 20 miles means the difference between having to supercharge to get back home or not — if the battery is degraded I can live with that, but if I have 20 miles of range “hiding”, I definitely want to know

 

[ewoodrick]

Sorry, missed the FL part.

The BMS should know how to work, even with capped batteries.
On those long drives, watch the energy graphs and just adjust your speed slightly. Your car will probably get 400 miles if driven optimally. the LR will go over 600.

While people speculate on top-locked batteries, I'm not really sure if we know that for sure, or if they are bottom locked, or mid range locked. It wouldn't surprise me if Tesla didn't mid-range lock them, as that should increase the battery life over the other options.

You know, winter, that's when the locals put on their parkas but the tourist have shorts on. Max temps down to the low 80's.
It was funny, we were diving in the Keys and the crew had parkas on, while we were in the water with skins on.

 

[AlanSubie4Life]

means the difference between having to supercharge to get back home or not — if the battery is degraded I can live with that, but if I have 20 miles of range “hiding”, I definitely want to know

I suggested this elsewhere but you are an excellent test case due to your use patterns:

Next time you have an opportunity to charge your car at home at a very low SoC (say 5%) up to 100%, monitor your loaded (meaning what it reads after current is at max) voltage and current in the car (will be reasonably constant once the charge starts but you can check a couple times over several hours, just in case the utility supply droops), and carefully measure the charging event time (set a reminder so you can catch the precise end of the charge event). Measure it to the nearest minute! Preferably charge at the max 7.7kW allowed by the SR. Turn off HVAC prior to beginning the charge, to avoid excessive accessory drain. Also turn off AC in your house during the charge to prevent excessive droop on your house feed (if you see it change the supply voltage to the car).

Report back here. We can use the published charging efficiency (~89%) to tell you within a couple % how much energy your battery must then contain. And compare to the “correct” (original) value.

This will resolve your mystery and answer your question about range loss and whether you have any. Can’t create energy out of thin air.

I expect 44.4kWh for you from 5% (10 rmi) to 100% (199rmi) when charging at 7.7kWh (meaning 39.5kWh available from 100% to 5%). Assuming 236V @ 32A that would be a 5 hour 53 minute charge. We’ll see! Not 100% sure about these numbers because the consumption constant for the SR is very slightly in doubt.

For precision, rather than using %, let’s switch to distance display, and log the starting rated miles and the final rated miles at 100%. % could add an additional 1% error and makes it ambiguous what your 100% charge energy is.

 

[Teedub21]

I’m in South Florida, so not sure what this “winter” you speak of is

I drive a ton (just hit 29k miles this week, and I took delivery on March 25th), and there are several drives I take regularly where 20 miles means the difference between having to supercharge to get back home or not — if the battery is degraded I can live with that, but if I have 20 miles of range “hiding”, I definitely want to know

seems downgrading to SR might not have been the best idea for someone that drives as much as you do.

 

[lateulade]

@AlanSubie4Life sadly, I’m charging on a 10-30 at the moment, so I’m at 24A. Will this affect things at all, or just make the experiment take longer?

@Teedub21 I’m not unhappy with the downgrade decision, I just wish I wasn’t “losing” 20% of my battery vs the promised 10%. I’m never far from a supercharger, so it’s not a huge deal, but I’d prefer not to unnecessarily stop. Given unlimited funds, I’d have a P3D though

 

[AlanSubie4Life]

New @AlanSubie4Life sadly, I’m charging on a 10-30 at the moment, so I’m at 24A. Will this affect things at all, or just make the experiment take longer?

It will just reduce the efficiency a little but I think we can back calculate what it should be. I calculate instead of 89% it will be 88% efficient or so at 240V*24A.

0.893 = (eta * 240V* 32A - 250W) / 240V*32A

This is the actual overall efficiency from the EPA document. 250W is from basic estimates others have done. About 60W of this or so goes to pack related operations during the charge but it does not really matter. We can then calculate eta, the raw DC-DC efficiency:

eta = 0.926 (raw AC-DC efficiency)

So at 24A/236V (TBD)

(24A*236V*0.926-250W)/(236V*24A)

= 0.882 overall efficiency

So:
(199rmi-10rmi)*209Wh/rmi/0.882 = 44.8kWh (energy from wall)

Again, assuming 24A @ 236V:

44.8kWh/24A/236V = 7 hours 55 minutes.

That is what I expect if you have lost the energy.

If you actually had no energy loss (had the EPA test energy of ~48.2kWh available), I would expect this charge to the (incorrectly labeled) 199rmi from 5% to take:

48.2kWh*0.955 * (1 - 0.05) = 43.7kWh (energy added to battery)

43.7/0.882 = 49.6kWh (energy from wall)

49.6kWh/24A/236V = 8 hours 45 minutes

So the difference would be VERY noticeable.

Anyway, we will see and we can double check those calculations later. You just need to do:
1) Log your miles at start of charge, and then charge to 100%, and log your miles at 100%. Obviously you do not have to start at 10rmi, we can adjust the calculations, but the lower the better, otherwise people might say “you’ll recover the missing energy at the bottom,” or something like that...
2) Pay close attention to loaded voltage, and keep it constant during the charge, for accuracy.
3) Measure charge time precisely. Also confirm it is still charging at 32A at the end of the charge...should not be any taper for you, but we want to be sure the experiment is valid...

 

[AlanSubie4Life]

Anyway, we will see and we can double check those calculations later. You just need to do:
1) Log your miles at start of charge, and then charge to 100%, and log your miles at 100%. Obviously you do not have to start at 10rmi, we can adjust the calculations, but the lower the better, otherwise people might say “you’ll recover the missing energy at the bottom,” or something like that...
2) Pay close attention to loaded voltage, and keep it constant during the charge, for accuracy.
3) Measure charge time precisely. Also confirm it is still charging at 32A at the end of the charge...should not be any taper for you, but we want to be sure the experiment is valid...

@lateulade as I said earlier, other than quick checks (using app) at beginning and near the end of charge, we want to keep accessory use minimized, so:

1) Close doors and leave car alone while charging. Again, a couple very quick checks to monitor voltage from the app during charge are ok. Using the app probably adds no additional energy use during charge.

2) Be sure to turn off HVAC before beginning charge. That will minimize energy use that occurs during charge. Use the app to check voltage rather than opening car.

3) Double check that your house AC does not cause voltage sag, before starting charge, and if so, turn it off for the 8 hour test.

Data to capture:
Starting rated miles at ~5%:
Final rated miles at 100%:
VoltageStart (with 24A current flow, not open circuit):
Voltage Middle (optional):
VoltageNearEnd:
Current: 24A
Charge Duration: (set a reminder so you can capture the exact completion of the charge...though perhaps app notification time stamp will be enough?)

And the hope is that all the voltage measurements are exactly the same or very close at least.

 

[vickh]

Sorry, missed the FL part.

The BMS should know how to work, even with capped batteries.
On those long drives, watch the energy graphs and just adjust your speed slightly. Your car will probably get 400 miles if driven optimally. the LR will go over 600.

While people speculate on top-locked batteries, I'm not really sure if we know that for sure, or if they are bottom locked, or mid range locked. It wouldn't surprise me if Tesla didn't mid-range lock them, as that should increase the battery life over the other options.
.

I'm at about 205 miles @100% on a SR too. only 8k miles.

I remember reading in another SR thread that the batteries were top-locked, given the charging curve and regen at 100% (although I do get the dots at 100%)

 

[AlanSubie4Life]

I'm at about 205 miles @100% on a SR too. only 8k miles.

Would be interesting to see your charging results as well! We'd have something to compare to! Your charging time from ~5% to 100% should be slightly longer than @lateulade. At 24A charging it should take about 15 minutes more if these rated miles are accurate representations of energy available. Probably to eliminate a variable, should stick with setting the current limit, to emulate 24A charging.

 

[smashcz]

Calibrating won't fix it. The only thing that can fix it is Tesla release an update and give us what we're supposed to have, but it looks like it will never happen. (I'm at 202 @ 100%)

 

[AlanSubie4Life]

Calibrating won't fix it. The only thing that can fix it is Tesla release an update and give us what we're supposed to have, but it looks like it will never happen. (I'm at 202 @ 100%)

If you frequently draw your car down to low SoC, you should do the measurements above too! Lots of data points then.

Ideally we’d have an owner who still has 235 or more rated miles with a 2019 do the test as well.

That would make it pretty definitive and we could finally resolve the debate about whether this rated mile loss is “real.” (I believe it is based on other evidence, but still has to be proven in a very straightforward manner that is clear to all!)

 

[RLEE]

My sr maxes out around 206 to 212 at 100pct. Lower when it's cold out. Btw if you want the numbers you should do a search for teslamate. I run it on a raspberry pi and have all the stats on mu car 24/7 without giving my information out to a 3rd party.

 

[lateulade]

If you frequently draw your car down to low SoC, you should do the measurements above too! Lots of data points then.

Ideally we’d have an owner who still has 235 or more rated miles with a 2019 do the test as well.

That would make it pretty definitive and we could finally resolve the debate about whether this rated mile loss is “real.” (I believe it is based on other evidence, but still has to be proven in a very straightforward manner that is clear to all!)

Thanks! Waiting for a day this week when I can do this while everyone is out of the house (so I can kill the AC).

Curious why you think the range loss is real, particularly in light of the fact that my SR+ still showed 238rmi the day before I downgraded, at which point it immediately dropped to 209rmi. Why wasn’t that ~10% degradation reflected in the SR+ rmi?

Again, super thankful for you sharing all your knowledge with us!

 

[pdx_m3s]

For SRs, it would be neat if Tesla held rated range constant at 220 until degradation was truly below 220 from full pack capacity... essentially “winding down” the top-lock as degradation increased (could probably stay at 220 for years). I’m guessing it’s an optics thing and Tesla doesn’t want to give people the impression that the SR is immune to degradation.

 

[vickh]

If you frequently draw your car down to low SoC, you should do the measurements above too! Lots of data points then.

Ideally we’d have an owner who still has 235 or more rated miles with a 2019 do the test as well.
!)

you mean 220 for an SR, right?

 

[vickh]

For SRs, it would be neat if Tesla held rated range constant at 220 until degradation was truly below 220 from full pack capacity... essentially “winding down” the top-lock as degradation increased (could probably stay at 220 for years). I’m guessing it’s an optics thing and Tesla doesn’t want to give people the impression that the SR is immune to degradation.

I agree and expect that to be a selling point that next owner can unlock protected battery, if the top lock is truly immune to degradation.

 

[AlanSubie4Life]

Curious why you think the range loss is real, particularly in light of the fact that my SR+ still showed 238rmi the day before I downgraded, at which point it immediately dropped to 209rmi. Why wasn’t that ~10% degradation reflected in the SR+ rmi?

I could well be wrong. It may be that the energy content is higher per rated mile than the Energy Screen suggests. This would be really easy to determine though - just on your next long drive, log beginning and end rated miles, and take a picture of the trip meter stats. And don’t spend any time in park.
I’ve never seen this done definitively for the SR. If there is more than 209Wh/rmi then it would all make sense. But that would not align with the Energy screen pictures we have seen (and that would be an exception vs. all other vehicles).

 

[AlanSubie4Life]

you mean 220 for an SR, right?

Sure. But there just are not many SRs out there. Easier to just take a 2019 SR+ showing close to full range, and do the same experiment but then just charge to 92% or so.