Using TM-Spy to see Model S data.

[Larsen]

Thanks for the answer.
I wanted to share the information about the battery status from TM-Spy and listen to the comments from people on the forum with pleasure, if of course they want to share their thoughts on this matter.
Here are two screenshots of the status of battery cells, daylight is the state of the car's rest, the night car is charging. I noticed that a few "bricks" stand out for their condition - when the car is under stress on them below all and when charging, their voltage is higher than the rest. I have not worked very much with lithium batteries and I do not understand their behavior very well, but I did a lot of lead acid. In acidic, this behavior of individual cells indicates their failure (sulfation).
Tell me please your opinion, there is a failure of individual cells of the "brick" or perhaps this is an unbalance of the block of "bricks", which you can try to fix.
Thank you!

 

[speedy]

5mv difference on discharge doesn't seem like much to worry about...

 

[JohnnyG]

@Larsen , Lower voltage at steady-state would indicate either degraded cells (not able to hold an equivalent charge) or possibly pack imbalance. Higher voltage during charging would indicate a higher internal cell resistance; which would point back to cell degradation.

At a glance, it would appear that those two groups are not taking a good charge, and as such, do not have as much capacity to deliver during use.

 

[Larsen]

Thank you very much for answering my question!
I'll try to perform a charge-discharge cycle and will continue to observe the battery status. If there is interesting information, I will write for everyone.
Special thanks to the developers of the program TM-Spy, thanks to which you can identify these features of the battery. This is surely useful to many people when choosing to buy a car not under warranty.

 

[Benjamin Brooks]

Time to add the EDR recovery feature to this app?

Tesla releases new tool for people to retrieve ‘blackbox data’ after a crash

 

[boaterva]

Time to add the EDR recovery feature to this app?

Tesla releases new tool for people to retrieve ‘blackbox data’ after a crash

I *was* wondering how to not buy more hardware.....

 

[S100D]

I *was* wondering how to not buy more hardware.....

Don't worry, I don't think you will buy the hardware kit. It costs $995 but if you pre-order you can get a $200 discount.

I was expecting a $50 to $75 cable, NOT!!!!

I wonder if the cable and Bluetooth adapter could be used to talk to the EDR software?

 

[hiroshiy]

Maybe somebody already reported but since I didn't know I will report it here as well. I purchased OBDLink LX in addition to LELink and it worked with TM-Spy as well.
LELink works over BLE, and OBDLink LX works over Bluetooth (paired in advance). So people who want to use both TM-Spy and Scan My Tesla the best bet for the adapter (AFAIK) seems to be OBDLink LX.

 

[Go4IT]

Thanks for the answer.
I wanted to share the information about the battery status from TM-Spy and listen to the comments from people on the forum with pleasure, if of course they want to share their thoughts on this matter.
Here are two screenshots of the status of battery cells, daylight is the state of the car's rest, the night car is charging. I noticed that a few "bricks" stand out for their condition - when the car is under stress on them below all and when charging, their voltage is higher than the rest. I have not worked very much with lithium batteries and I do not understand their behavior very well, but I did a lot of lead acid. In acidic, this behavior of individual cells indicates their failure (sulfation).
Tell me please your opinion, there is a failure of individual cells of the "brick" or perhaps this is an unbalance of the block of "bricks", which you can try to fix.
Thank you!View attachment 282353

Looking at the groups in question (54-59 and 72-77), it looks like you have 2 entire modules that show faster signs of degradation than the other 14 modules. I have observed the same with my 2013 P85 (300.000km on the ODO) on one module (covering groups 24-29). While the deltaV you see is not so high (in mV), one must remember this is at very low DC current. DeltaV will increase as current increases (in charging or discharging). This clearly points at a degradation of the internal resistance for each of these groups (to be noted : this happens in ALL groups of a module).

Besides the reduced capacity of the faulty modules, this impacts the capacity of the entire battery pack as the BMS sees max and min voltage (resp. 4.1V and 2.9V) reached faster on these modules. This will either stop the charge (at Vmax) or stop the car (at Vmin). Balancing above 93% cannot compensate this as the current drained during balancing would not be enough to compensate.

I guess Tesla will be facing more and more of these issue and will have to address it...

 

[Larsen]

Thanks for your reply!

I managed to take a screenshot of TM Spy a few seconds before the car told Charging Complete, and I watched the charging process for a long time from 95 to 100%. As it turned out, the voltage on the cells of these suspicious blocks rose to 4.2V and then did not grow, the rest of the blocks continued to be charged. BMS does its job well, but it seems to balance the cells only inside the block and there is no balancing between the blocks.

 

[Go4IT]

Thanks for your reply!

I managed to take a screenshot of TM Spy a few seconds before the car told Charging Complete, and I watched the charging process for a long time from 95 to 100%. As it turned out, the voltage on the cells of these suspicious blocks rose to 4.2V and then did not grow, the rest of the blocks continued to be charged. BMS does its job well, but it seems to balance the cells only inside the block and there is no balancing between the blocks.
View attachment 298283

There was a typo in my previous post. The charge ends at 4.2V (not 4.1V). Sorry.

Reacting to your last post, balancing takes place in each group individually. There is no interaction between groups ; not even between groups within the same module. Balancing actually starts at 93% charge by draining current locally in any group showing higher voltage (than the others groups) using bleeding resistances. These resistances are welded right on the module's BMS PCB. Since a module houses 6 groups of cells, each BMS PCB has 6 groups of bleeding resistances (four 158 Ohm resistors for each group connected in parallel giving 39,5 Ohm total value). Hence with a voltage of 4.2 V on the group when approaching full charge, the balancing current is very limited (less than 0.4W) ; surely not enough to keep voltage of the faulty group significantly below 4.2 V (such as groups 54-59 and 72-77 in your pack, and potentially 18-23) so inevitably, as some stage, the BSM decides to stop the charge, even if other groups are still charging. In other words, the faulty groups are so of limits that even balancing cannot compensate any more.

That is the issue you and I have. You with 3 groups and me with only one.

One more thing... While the balancing mechanism STARTS at around 93% charge, it is MAINTAINED even after the BMS has decided to stop the charge. It can actually continue for days, possibly until the next charge at 93% with the same bleeding current as calculated for each group individually by the BMS. So, if a pack is not recharged, the balancing mechanism could potentially accelerate the discharge of faulty groups and therefore bring the faulty group faster to Vmin on the long run. Maybe there is a logic in the BMS to address this and stop balancing under some conditions, I don't know.

 

[n2mb_racing]

There was a typo in my previous post. The charge ends at 4.2V (not 4.1V). Sorry.

Reacting to your last post, balancing takes place in each group individually. There is no interaction between groups ; not even between groups within the same module. Balancing actually starts at 93% charge by draining current locally in any group showing higher voltage (than the others groups) using bleeding resistances. These resistances are welded right on the module's BMS PCB. Since a module houses 6 groups of cells, each BMS PCB has 6 groups of bleeding resistances (four 158 Ohm resistors for each group connected in parallel giving 39,5 Ohm total value). Hence with a voltage of 4.2 V on the group when approaching full charge, the balancing current is very limited (less than 0.4W) ; surely not enough to keep voltage of the faulty group significantly below 4.2 V (such as groups 54-59 and 72-77 in your pack, and potentially 18-23) so inevitably, as some stage, the BSM decides to stop the charge, even if other groups are still charging. In other words, the faulty groups are so of limits that even balancing cannot compensate any more.

That is the issue you and I have. You with 3 groups and me with only one.

One more thing... While the balancing mechanism STARTS at around 93% charge, it is MAINTAINED even after the BMS has decided to stop the charge. It can actually continue for days, possibly until the next charge at 93% with the same bleeding current as calculated for each group individually by the BMS. So, if a pack is not recharged, the balancing mechanism could potentially accelerate the discharge of faulty groups and therefore bring the faulty group faster to Vmin on the long run. Maybe there is a logic in the BMS to address this and stop balancing under some conditions, I don't know.

So, how does it balance if you only charge to 90% each day?

 

[Go4IT]

So, how does it balance if you only charge to 90% each day?

Easy answer. No balancing happens if pack is charged below 93%.

Note: there is a reason for this 93% SoC. 93% SoC is the charge level where the BMS switches from Constant Current (CC) mode to Constant Voltage (CV). Indeed, since balancing depends on voltage, it cannot happen in CC mode where voltage fluctuates depending on charging conditions.

 

[n2mb_racing]

Easy answer. No balancing happens if pack is charged below 93%.

Note: there is a reason for this 93% SoC. 93% SoC is the charge level where the BMS switches from Constant Current (CC) mode to Constant Voltage (CV). Indeed, since balancing depends on voltage, it cannot happen in CC mode where voltage fluctuates depending on charging conditions.

Interesting... I guess I occasionally take long trips, so it does get to balance. Do you happen to know if the threshold for balancing was the same on the Roadster?

 

[Go4IT]

Interesting... I guess I occasionally take long trips, so it does get to balance. Do you happen to know if the threshold for balancing was the same on the Roadster?

Actually, I should re-phrase a bit. No NEW balancing computation will happen until pack is charged to 93%. Prior balancing computation may still be in effect before that. There may be a time-out (at least several days) for the prior balancing process to be stopped but I am not sure as I regularly charge about 93%.

No idea about the Roadster.

 

[SongjinDK]

I took a delivery of an 2018 Model X last week, I am wondering it is the same way to access the diagnostic port, i.e. pull the cubby underneath the screen. I tried it but am afraid of breaking it. Do I need lots of force? Thanks.

 

[hiroshiy]

Yes lots of force.

 

[SongjinDK]

Yes lots of force.

Thanks, got it working!

 

[SongjinDK]

I finally got Tm-Spy to work with my 85D!
Coincidentally, it didn't work with an adapter that looks just like the one Speedy posted above (see my post #409), but is working now with the one:
View attachment 274324

Three questions regarding the data I'm getting:
View attachment 274327

1) What do the numbers 81.7 and 83.2 mean? (circled in red)

2) Interrelated questions: what's the difference between Total Discharge and Total Charge (what are the "losses" between the 2?), and where is Total Charge measured? Before or after the onboard charger?

3) Maybe this as been widely discussed, but correct me if I'm wrong: for the car's Trip Meter to represent the effective consumption&cost (like on ICE cars), the consumption should be calculated as "Total Charge" / "KM".
In my war, that would be 22624/66241*1000 = 341 Wh/KM, which is a lot higher than the ~230 Wh/KM that I'm typical shown by the Trip Meter that, as far as I know, doesn't take into consideration (i) conversion losses, (ii) discharge losses, (iii) vampire drains (iv) consumption of some onboard electrical equipment (?) and who knows what else.

81.7 (circled in red)
is the SOC, it seems to be always the same as the SOC at the lower left corner.

Any one knows what 83.2 (circled in red) is? It seems to be always a little higher than SOC, except when SOC is 100%, it is like 98.

 

[SongjinDK]

83.2 may be the "true SOC", including the reserved power of about 3 to 4 kwh when SOC reches 0%.