Chevy Bolt EV Forum banner

1181 - 1200 of 1234 Posts

·
Registered
Joined
·
402 Posts
Discussion Starter · #1,181 ·
@EV Engineering - I forget, are you working with a 2020 or an earlier Bolt?

If a 2020, you mentioned you've got official diagnostic tools - by any chance can you sniff the CAN buses on the diagnostic connector when performing diagnostics on one of the telematics units?

It appears in the 2020 service manual that the 2020 firewalling setup implemented with the Serial Data Gateway Module (SDGM) is a bit fancier than 2017-2019 - the OBD connector is wired only to the SDGM. The SM implies that there's a way for diagnostic tools to ask the SDGM to talk with the isolated/dirty buses instead of the clean/nonisolated buses, but obviously not how. The answer to "how" would hopefully be answered by a passive bus sniff of an attempt to talk to the telematics module since that's on the isolated buses.
It's much simpler than that. There are two networks in the car - LS-CAN and MS-CAN. Cheap ELM327 based devices can only access the MS-CAN bus. Some lower priority things, including body control module and likely telematics, are on the LS-CAN bus. The diagnostics MDI adapter can talk on both busses.

While it's possible they implemented some advanced firewalling, I would suspect it's more likely that if there is a difference, it's only that they moved something from MS-CAN to LS-CAN to free up bus space on MS-CAN.

What specifically are you looking for? I know that some telematics are accessible from the MS-CAN bus in the 2017 at least.
 

·
Registered
Joined
·
351 Posts
@EV Engineering - I forget, are you working with a 2020 or an earlier Bolt?

If a 2020, you mentioned you've got official diagnostic tools - by any chance can you sniff the CAN buses on the diagnostic connector when performing diagnostics on one of the telematics units?

It appears in the 2020 service manual that the 2020 firewalling setup implemented with the Serial Data Gateway Module (SDGM) is a bit fancier than 2017-2019 - the OBD connector is wired only to the SDGM. The SM implies that there's a way for diagnostic tools to ask the SDGM to talk with the isolated/dirty buses instead of the clean/nonisolated buses, but obviously not how. The answer to "how" would hopefully be answered by a passive bus sniff of an attempt to talk to the telematics module since that's on the isolated buses.
I can't promise when I will have time to play with but I will let you know when I have time to see what is hidden behind it....
 

·
Registered
Joined
·
351 Posts
It's much simpler than that. There are two networks in the car - LS-CAN and MS-CAN. Cheap ELM327 based devices can only access the MS-CAN bus. Some lower priority things, including body control module and likely telematics, are on the LS-CAN bus. The diagnostics MDI adapter can talk on both busses.

While it's possible they implemented some advanced firewalling, I would suspect it's more likely that if there is a difference, it's only that they moved something from MS-CAN to LS-CAN to free up bus space on MS-CAN.

What specifically are you looking for? I know that some telematics are accessible from the MS-CAN bus in the 2017 at least.
I would be all ears if you point me in right direction for more advanced dongle...2020 are no go there is some type of firewall or splitting between bus networks...
 

·
Registered
Joined
·
357 Posts
It's much simpler than that. There are two networks in the car - LS-CAN and MS-CAN. Cheap ELM327 based devices can only access the MS-CAN bus. Some lower priority things, including body control module and likely telematics, are on the LS-CAN bus. The diagnostics MDI adapter can talk on both busses.

While it's possible they implemented some advanced firewalling, I would suspect it's more likely that if there is a difference, it's only that they moved something from MS-CAN to LS-CAN to free up bus space on MS-CAN.

What specifically are you looking for? I know that some telematics are accessible from the MS-CAN bus in the 2017 at least.
No, it's not simpler than that on a 2020. The OBD connector is literally wired to the SDGM and nowhere else according to the service manual and the schematics. Many messages are automatically bridged through, but if you fire off a precondition command from the OBD connector over the SWCAN (aka LS-CAN) bus, it gets ignored since it's coming from the wrong place.

The 2020 service manual and 2017-2019 service manuals differ significantly in this regard. The 2020 is quite explicit in stating the diagnostic connector only goes to the SDGM, 2017-2019 the SDGM only firewalls between the main MS/LS buses and the "isolated" ones and the diagnostic connector is on the "main" buses. The passenger footwell secondary diagnostic connector present on 2017-2019 is missing on 2020.

Although even with the diagnostics feature of somehow commanding the SDGM to route messages from the OBD connector to the firewalled/isolated buses instead of main, firing a precondition command from the diagnostic connector might still be ignored.

Edit: To clarify, by "ignored" I mean "precondition did not start", however firing off HV wakeup frames over SWCAN with the PHY configured to HV wakeup mode definitely causes nodes on the bus to wake up and start transmitting for a bit.

I can't promise when I will have time to play with but I will let you know when I have time to see what is hidden behind it....
Thanks!

I would be all ears if you point me in right direction for more advanced dongle...2020 are no go there is some type of firewall or splitting between bus networks...
P1 (Under-the-Dash) - low-level access to two standard differential CAN buses, and also to the SWCAN bus, with complete control over the SWCAN PHY via GPIO.
 

·
Registered
Joined
·
8,205 Posts
@Telek,

You have said you have never seen the Bolt balance cells. Yesterday, I decided to charge to 100%, and do a long drive.


I went for a walk in the woods, and came back when I figured it would have about gone from hilltop to full. As is often the case, the green light on the dash was steady, and the car said full, but the EVSE indicated it was still running, and I could hear the electronics loop pump. I pulled out my phone and checked Torque Pro. It ran like this for about 20 minutes. The accessory battery charger was using 250-300 watts. Other than the two onboard chargers, and their pump, nothing else appeared to be happening, yet the DC side of the charger was showing 2-3 kW. The cell delta has never changed from 0.20-0.26 V. If it was balancing, it wasn't like any BMS I have ever seen.

not balancing 1.jpg not balancing 2.jpg not balancing 3.jpg not balancing 4.jpg
 

·
Registered
Joined
·
402 Posts
Discussion Starter · #1,186 · (Edited)
@Telek,

You have said you have never seen the Bolt balance cells. Yesterday, I decided to charge to 100%, and do a long drive.


I went for a walk in the woods, and came back when I figured it would have about gone from hilltop to full. As is often the case, the green light on the dash was steady, and the car said full, but the EVSE indicated it was still running, and I could hear the electronics loop pump. I pulled out my phone and checked Torque Pro. It ran like this for about 20 minutes. The accessory battery charger was using 250-300 watts. Other than the two onboard chargers, and their pump, nothing else appeared to be happening, yet the DC side of the charger was showing 2-3 kW. The cell delta has never changed from 0.20-0.26 V. If it was balancing, it wasn't like any BMS I have ever seen.

View attachment 34095 View attachment 34096 View attachment 34097 View attachment 34098
Yeah definitely not balancing, but hard to tell what's going on. Could just be topping up the charge. Without the unbroken soc raw it's hard to see.

I am convinced that GM can update the parameters OTA without letting us know. I've seen different behaviors recently. Especially cold weather charging - it is way slower than it used to be until the battery is quite warm. I'm pretty sure this changed after November, despite not being in for the recall update.

So normally speaking when charging to full, my SOC raw usually stops between 96-97. Maybe occasionally it'll go to like 97.5 but that's very rare. As soon as the charge completes, it resets to 96.001%.

A couple weeks ago I was at a 24kW charger and the SoC raw went all the way to 98.9% - pulling about 3-4kW (just like you saw) all the way up. I have NEVER seen it do this before, and I pay close attention. As soon as it completed it reset back to 96.001% again. What's odd is that my cell voltages were normal - 4.175V max, which isn't unusual. 399.5V sum total.

So looks like exactly what you saw. But that's not balancing, it's just pushing how high it charges.

The balancing in the BMS can only very slowly drain the high cells. So this definitely isn't that.
 

·
Registered
2019 Chevy Bolt LT, Cajun Red
Joined
·
575 Posts
quick question for those, like myself, who has a 2019 (or later maybe?), non-recalled Bolt.

I noticed since last summer/fall that my battery capacity has reduced about 5% from estimates performed then. I currently have 20,000 miles on my Bolt.
Has anyone else noticed this quick (to me) reduction in capacity? (mine went from estimated 57 kwh in summer, to 55 kwh in winter to currently 53 kwh based on telek’s formula, i dont have the ah in front of me, but its down to 166 ah or something and 53.5 kwh est.)

Call me paranoid, but could Chevy have updated our software over the air secretly? to reduce available battery usage 5% just like the recall fix on Korean built battery packs?
It just seems odd to me. I’m keeping watch to see if my capacity goes back up in torquepro as the weather warms and I do some deep discharges to see if my ah number increases.
 

·
Registered
Joined
·
402 Posts
Discussion Starter · #1,188 ·
quick question for those, like myself, who has a 2019 (or later maybe?), non-recalled Bolt.

I noticed since last summer/fall that my battery capacity has reduced about 5% from estimates performed then. I currently have 20,000 miles on my Bolt.
Has anyone else noticed this quick (to me) reduction in capacity? (mine went from estimated 57 kwh in summer, to 55 kwh in winter to currently 53 kwh based on telek’s formula, i dont have the ah in front of me, but its down to 166 ah or something and 53.5 kwh est.)

Call me paranoid, but could Chevy have updated my software over the air secretly? to reduce my range 5% just like the recall fix on Korean built battery packs?
It just seems odd to me. I’m keeping watch to see if my capacity goes back up in torquepro as the weather warms and I do some deep discharges to see if my ah number increases.
Do you know what the Ah rating was before? Keep in mind that my December update changed the formula, and as a result it's a little lower than before (by a few %).
 

·
Registered
2019 Chevy Bolt LT, Cajun Red
Joined
·
575 Posts
Do you know what the Ah rating was before? Keep in mind that my December update changed the formula, and as a result it's a little lower than before (by a few %).
I do. I will double check and post the exact reading I had if this is wrong. all my torquepro readings are after your update in December. My prior estimates of 57kWh were done from driving and comparing usage from 100% to 0%. but your estimates since december matched what I see usable in my Bolt since on my trips and viewing the Bolts energy used screen and SOC.

After your December update it was 172 ah.
hoping this new 166 figure is just a one off as others have said they have seen. im not at home, so trying to find my notes.
 

·
Registered
Joined
·
402 Posts
Discussion Starter · #1,191 ·
I do. I will double check and post the exact reading I had if this is wrong. all my torquepro readings are after your update in December. My prior estimates of 57kWh were done from driving and comparing usage from 100% to 0%. but your estimates since december matched what I see usable in my Bolt since on my trips and viewing the Bolts energy used screen and SOC.

After your December update it was 172 ah.
hoping this new 166 figure is just a one off as others have said they have seen. im not at home, so trying to find my notes.
The Ah reading is raw - so that won't change from before my December update to after. But I don't think I had a reading for the raw Ah value before.
 

·
Registered
Joined
·
234 Posts
quick question for those, like myself, who has a 2019 (or later maybe?), non-recalled Bolt.

I noticed since last summer/fall that my battery capacity has reduced about 5% from estimates performed then. I currently have 20,000 miles on my Bolt.
Has anyone else noticed this quick (to me) reduction in capacity? (mine went from estimated 57 kwh in summer, to 55 kwh in winter to currently 53 kwh based on telek’s formula, i dont have the ah in front of me, but its down to 166 ah or something and 53.5 kwh est.)

Call me paranoid, but could Chevy have updated our software over the air secretly? to reduce available battery usage 5% just like the recall fix on Korean built battery packs?
It just seems odd to me. I’m keeping watch to see if my capacity goes back up in torquepro as the weather warms and I do some deep discharges to see if my ah number increases.
I've wondered this. Late last year (whenever Telek released the update that provided an Ah estimate for 2019s), I recorded 188.9 Ah. As I checked heading into winter, the number started dropping...which I expected due to the falling temps. 187.2, 184.4, 180.5. I didn't check for a while and the other day it read 168.8! I really hope that estimate rebounds as we head into warm weather as that kind of degradation over a few months is unlikely (especially given how I drive). I did wonder if Chevy pulled something. I haven't had the battery update or any other service work done. We'll see.
 

·
Registered
Joined
·
8,205 Posts
I really hope that estimate rebounds as we head into warm weather
Don't count on it!

 

·
Registered
Joined
·
8,205 Posts
It will be interesting to see if part of the final fire software patch, involves hiding the PIDs for Ah, SoC Variation, etc.
 

·
Registered
Joined
·
234 Posts
Don't count on it!

Do you think it's realistic that I lost 20 Ah in a few months? The pack will be dead in no time if the trend continues. I have wondered if only charging to 75% and never to a full charge has affected the reading in some way.
 

·
Registered
Joined
·
234 Posts
It will be interesting to see if part of the final fire software patch, involves hiding the PIDs for Ah, SoC Variation, etc.
It will be interesting. I'm planning on taking a reading before and after the "fix" (next month, right?).
 

·
Registered
2020 Chevrolet Bolt
Joined
·
433 Posts

·
Registered
Joined
·
8,205 Posts
Do you think it's realistic that I lost 20 Ah in a few months? The pack will be dead in no time if the trend continues. I have wondered if only charging to 75% and never to a full charge has affected the reading in some way.
Not possible. I hadn't looked back at your earlier posts. What is your SoC Var PID say? What is you min/max cell voltage delta?
 
1181 - 1200 of 1234 Posts
Top