Wiki Sudden Loss Of Range With 2019.16.x Software

[glhs272]

Regarding Coolant pumps running while car is idle. So, as some of us have noticed, the coolant pumps run a lot more than they used to. This seems to be related to State of Charge (SOC) and other factors. It also happens after charging even when I have a relatively cool battery. I also notice the coolant pumps run at a higher %, i.e. full blast when driving the car and the battery is relatively cool.

My uninformed wild guess is that it's not doing it to cool the battery at all. Rather it is just to confirm cell temp uniformity. The BMS on the older batteries like mine only has temperature sensors on two cells per module. One glued to a cell near the coolant inlet and near the coolant outlet. So the BMS doesn't have a picture of cell temps in the rest of the module, say the middle of the module. My guess is that Tesla found a situation/condition where a cell, (a bad cell perhaps) is getting hot and the BMS doesn't know it. This condition would be possible if the coolant pumps were idle, thus the hot cell would not dissipate heat to the rest of the module very well. So to mitigate they run the pumps continuously where there is any possibility of this happening. This allows the BMS to have a good idea of the whole module temperature, as long as the pumps are running. Those conditions are high state of charge, charging (even AC charging with a relatively cold battery), and driving. Perhaps other conditions as well.

 

[Ferrycraigs]

But surely they don't behave the same, as at some point some individual cells become so different from the others that they fail, blow their individual fuse and drop out of the brick.

So what is really being said is that to achieve balance, there must be a (long) period of low / reducing current at least current through individual bricks to allow energy absorbtion to reach equilibrium between bricks. The longer and more gentle the balancing phase, the more even the energy load across all cells will be.

For me that pretty much describes a Supercharging, or indeed any Charging to a high SoC. I have monitored my battery several times now over the past 6 months. TMSpy shows each individual brick. Whilst the spread can be around 60 mV below 50% SoC, it’s normally reduced to around 30mV by 80% and less than 20 mV by 90% and often in single digits above 95% SoC. Of course I have examples that are better, and worse, these figures are just the general trend. When people say Tesla's BMS balances the battery every time it charges, so no need to specifically balance them, I tend to agree. But there will always be exceptions. The balancing (between bricks) seems pretty progressive, but I have never seen it achieve close balancing if the charge level is set below 100%, ie if I set 85% I don’t expect it to be fully balanced by the time it stops charging at 85%. I do expect it to be pretty balanced above 95%.

 

[tga]

Model S/X do preheat the battery en route to superchargers. The Tesla app shows the battery heating icon.

I have absolutely seen bettery being heated (evidenced by available regen increasing from 20kw to 50kw +) in the 15 minutes prior to arriving at Supercharger. (compared with identical trip, same day same conditions) without Supercharging where regen remained at 20kw or less.

Raven S LR

I've never seen my car turn on the battery heater (or do anything to reduce cooling) when setting a Supercharger as an explicit destination or as a stopover on a longer trip. I've watched the data in Scan my Tesla, and nothing related to temp management changes, ever. I never use range mode.

 

[Droschke]

I have to look and see. I thought you were referring to the snowflake icon on the battery charge display on the app.

The snowflake and the battery heating icons are entirely two different things. I kept saying the "battery heating" icon on the Tesla mobile app under Climate, when en route and navigating to a Supercharger.

You need to be driving to see this icon on the mobile app. Just to be in the Park position and set the navigation to a supercharger is not going to trigger the battery heating in my experience.

 

[tga]

and has everything to do with the "Shuttling balancing" we are seeing with newer firmware revisions.

(Older?) 85kWh packs don't do this. The only balancing hardware are the bleeder resistors to drain charge from higher cells and waste it as heat. There is no hardware to "shuttle" it into weaker cells. Software can't fix this.

Isn't this "shuttling" only valid for newer cars and not the older cars?

Correct.

but I haven't managed to find real battery tear-downs showing actual components

Pics/Info: Inside the battery pack

 

[Droschke]

Regarding Coolant pumps running while car is idle. So, as some of us have noticed, the coolant pumps run a lot more than they used to. This seems to be related to State of Charge (SOC) and other factors. It also happens after charging even when I have a relatively cool battery. I also notice the coolant pumps run at a higher %, i.e. full blast when driving the car and the battery is relatively cool.

My uninformed wild guess is that it's not doing it to cool the battery at all. Rather it is just to confirm cell temp uniformity. The BMS on the older batteries like mine only has temperature sensors on two cells per module. One glued to a cell near the coolant inlet and near the coolant outlet. So the BMS doesn't have a picture of cell temps in the rest of the module, say the middle of the module. My guess is that Tesla found a situation/condition where a cell, (a bad cell perhaps) is getting hot and the BMS doesn't know it. This condition would be possible if the coolant pumps were idle, thus the hot cell would not dissipate heat to the rest of the module very well. So to mitigate they run the pumps continuously where there is any possibility of this happening. This allows the BMS to have a good idea of the whole module temperature, as long as the pumps are running. Those conditions are high state of charge, charging (even AC charging with a relatively cold battery), and driving. Perhaps other conditions as well.

Interesting theory.

BTW, if I supercharge to 90% my pumps run. To 80% they do not. So, I've been charging to 80% these days (which is about 70% for my capped car).

 

[Battpower]

Pics/Info: Inside the battery pack

Awesome! Thanks very much. That's exactly what I hoped to see. Another thread to slog through! So far, really interesting that on that battery it looks like minimal if any cell / brick balancing circuitry.

I won't comment until I'm done reading!

Anyone wanting to comment on charging behavior, battery capping, battery 'use and failure characteristics' would do well to look at these pics and some of the dialogue if you haven't seen it already. Not because it's new (thread is ancient) but it is well presented and comprehensive.

 

[Battpower]

So, as some of us have noticed, the coolant pumps run a lot more than they used to.

My recent 2000 mile trip in UK and France had me sitting at several busy Supercharging stations. The noise from most S and X cars was really loud as they all sat whirring and whining away. If it really is to dissipate heat, then that would suggest a pretty inefficient charging process. Not sure what else the purpose would be, but ambient temps were typically 5 celcius so you would think warming the pack would make more sense if you actually want to charge at high speed...... but of course the evidence is that charge rates are being pushed down, not up. Taken any way, sounds like bad news for end-to-end efficiency.

My Raven S LR never obviously ran pumps or fans during Supercharging irrespective of SoC range and charge rate and the same seem to be the case for model 3's.

 

[JRP3]

Well, you'd hope so, but nothing is perfect and surely it is those minor differences between bricks / cells that necessitates balancing and to a large extent puts bigger stress on certain parts of the battery and can lead to premature failure of some type.

Sure if a cell has higher internal resistance it will heat up more and lose capacity faster, and if it gets bad enough it could short out and disconnect from the pack. I'm just saying there is no possible way to balance cells in parallel, any more than you could balance internal cell layers.

 

[tga]

Awesome! Thanks very much

You're welcome. He's got a lot of interesting info out there, both on TMC and his own site/Twitter feed.

 

[Battpower]

Sure if a cell has higher internal resistance it will heat up more and lose capacity faster, and if it gets bad enough it could short out and disconnect from the pack. I'm just saying there is no possible way to balance cells in parallel, any more than you could balance internal cell layers.

From a different thread:

Yes. All parallel-connected cells are balanced as a group. Nor is there any reason for a more individual charge, as they work together, the ones with more ready ions will provide a little more current, than some of them with higher internal resistance, but in the big picture, they are balanced in terms of the chemical state of the cells.

In which it is clearly explained how parallel cells within a brick function as one high capacity cell.

What I am focusing on now is to better understand early and current balancing processes.

At the moment, this:

Pics/Info: Inside the battery pack

post suggests minimal or zero high tech balancing and that the battery just depends on very well matched cells for its performance. I haven't yet found images of the bleed resistors in earlier batteries or charge shuttling circuits for newer batteries.

 

[glhs272]

From a different thread:

I haven't yet found images of the bleed resistors in earlier batteries or charge shuttling circuits for newer batteries.

Bleed resistors, yes. Charge shuttling circuits, no. Or at least not on the early packs.

The bleed resistors are definitely there on the BMB boards attached to each module. I have some of these BMB boards I took off of my modules on my off-grid solar pack. You should find it in the Pack teardown thread.

At least on early BMB boards, I can find no evidence of charge shuttling capability.

 

[Kim.T]

Interesting theory.

BTW, if I supercharge to 90% my pumps run. To 80% they do not. So, I've been charging to 80% these days (which is about 70% for my capped car).

On my 70D the pumps runs for 3 hours. Even if I only charge with 11kW for 1 hour and don't have a high SoC

 

[Kim.T]

My recent 2000 mile trip in UK and France had me sitting at several busy Supercharging stations. The noise from most S and X cars was really loud as they all sat whirring and whining away. If it really is to dissipate heat, then that would suggest a pretty inefficient charging process. Not sure what else the purpose would be, but ambient temps were typically 5 celcius so you would think warming the pack would make more sense if you actually want to charge at high speed...... but of course the evidence is that charge rates are being pushed down, not up. Taken any way, sounds like bad news for end-to-end efficiency.

My Raven S LR never obviously ran pumps or fans during Supercharging irrespective of SoC range and charge rate and the same seem to be the case for model 3's.

I have recorded a SuC session where SoC went from 25%-62%. Outside temp. 1C. Cell temp went from 31C-37C, I could not hear the fan.
It took 31min to get 24,71kW ! Shitty charging curve, but we are not bothered by the fan
We where the only one at the SuC - a 150kW....

 

[Sumpi]

Have you seen Joseph Lam comments on Facebook. He was given a new warranty battery in a 2013 p85 and charging it to 265 miles rated.

 

[lightningltd]

Have you seen Joseph Lam comments on Facebook. He was given a new warranty battery in a 2013 p85 and charging it to 265 miles rated.

Link shows not found.

 

[Chilam]

Link shows not found.

It's 289 mi at 100% rated.

 

[JRP3]

I haven't yet found images of the bleed resistors in earlier batteries or charge shuttling circuits for newer batteries.

I believe you can see the resistors on the closeup of the BMS board, they are small. Also:

A note on the balancing function of the module BMBs, They don’t actually preform balancing automatically, it’s totally up to external software to control turning on/off the bleed resistor on each cell, in the case above, the Arduino DUE software would request ADC measurements from the BMB, and then decide based on a voltage threshold if each cell should have it’s bleed resistor turned on or off. The command is then sent back to the BMB which sets the bleed resistors as requested.
Collin, Tom and myself had a lot of back and forth messaging to work out all of the module functionality.

Tesla Battery Victory Lap - EVTV Motor Verks

 

[Zextraterrestrial]

I think many of you are missing the major point of 3ry-reset's picture. Some cells/bricks being charged to 4.200V and others being at 4.183V is reasonably normal. However, the fact that ANY cells are anywhere near 4.200V when SOC=92% is a huge problem.

I would think so ...& I would expect mine to be near 4.2V at 100% ...but?

..but this is a later date...

 

[ran349]

I would think so ...& I would expect mine to be near 4.2V at 100% ...but?

..but this is a later date...

You can't compare the two bar charts because you're still charging in the second one; note the negative current. You have to look at the voltages when you are not charging.
Your battery seems pretty healthy to me. Your voltages are closer to 4.2 volts than the 4.1 where most people see the capping.