GM Can info & knowledge share

[kidturbo]

I just checked that list for first time in a couple years. Glad to see some of my old shares from long ago still alive..

However the majority of the GM stuff up there is all low speed bus files, hacked together by reverse engineering raw data. Nothing of much value on the High Speed side has every been shared to that repository. Some complete HS GMLAN Global A versions are out there. These were written using the Vector tools, and tend to be held pretty close to the chest by anyone who has em. All built using data readily found in GMW3119 files, and pretty much line for line.. But includes the added node matrix in a detailed level designed to work with tools like Vehicle Spy. Would have taken weeks to write one by hand, and is the perfect base file to just add another field for tracking OS #s. Or maybe just place them in the Comments section. Either way, 99.9% of the CANbus DBC work is already done. It's just matching up the known parameter names, CAN ID's, and byte length to each known OS version. Using bin map details, or CAN sniffing one of every related platform in a running config. That would intern build the Perfect GMLAN DBC / CANbus hardware map..

But the big question is, who's willing to share, and or go dig up the latest GMW files needed to update an existing PPEI 3.6 or newer DBC, with those missing ID's and fields added over past last several years? Or better yet, a recently compiled version of said DBC? A small exert of a complete PPEI version makes the old GMLAN Bible look like sand-script. All the parameter names, bit map, and scaling are mapped out, quite correctly actually. However, my personal experience with sharing a decades worth of CANbus hacking knowledge hasn't proven very profitable lately. To me at least. But I know the perfect starter kit already exists, it's just who's willing to share it for this good cause?

[kippdipp]

Does anyone know which GMW document covers the transmission related signals such as 0x199? GMW8777 (2006) does not and neither does 8772. I would assume GMW8766 might??

[hjtrbo]

0x199 is torque request to the ecm. The docs I have on hand (same as yours except they're version 2010) do not document the transmission fully and I cannot find any reference to 0x199 in 8762, 8777 or 8766.

[kippdipp]

@kidturbo
Can you share what you have identified for the following Can IDs? (these are not covered in GMW docs I've seen): $189,$197,$199,$19D,$1A6,$1AF,$320,$520

[gmtech825]

Can anyone point me in the direction to locate the can routines/ID's in 2014+ v850 bcms?

[Tre-Cool]

@kidturbo
Can you share what you have identified for the following Can IDs? (these are not covered in GMW docs I've seen): $189,$197,$199,$19D,$1A6,$1AF,$320,$520

look in the GMW3119-2005.pdf document it has the 320/520 definitions.

[series8217]

I'm working on decoding the T43 TCM <-> E39 ECM torque messages.

Here is where I'm at with $189. I'm trying to RE this using CAN logs from my 2014 Impala with the LFX and 6T70. Next I would like to start looking at the T43 disassembly to better understand the unknown signals.

Code: Select all

BO_ 393 unknownTCM_Torque_189: 8 TCM
 SG_ MaxTorqueRequest : 3|12@0+ (0.5,-848) [0|255] "Nm" XXX
 SG_ MaxTorqueRequestActive : 4|1@0+ (1,0) [0|1] "" XXX
 SG_ Counter : 7|2@0+ (1,0) [0|3] "" XXX
 SG_ MinTorqueRequestMaybe : 19|12@0+ (0.5,-848) [0|4095] "Nm" XXX
 SG_ MinTorqueRequestActive : 20|1@0+ (1,0) [0|1] "" XXX
 SG_ Checksum : 39|16@0+ (1,0) [0|8191] "" XXX
 SG_ UNKNOWN_SIGNAL1 : 55|8@0+ (1,0) [0|65535] "" XXX
 SG_ UNKNOWN_SIGNAL2 : 63|8@0+ (1,0) [0|255] "" XXX

Message $199 also originates from the TCM and has a similar shape to $189. I'm not sure what the difference is yet.

I also have identified the transmission input shaft RPM signal in $19D.

Code: Select all

BO_ 413 TransmissionInputSpeed: 8 TCM
 SG_ UNKNOWN_COUNTER : 7|2@0+ (1,0) [0|3] "" XXX
 SG_ InputShaftRPM : 46|15@0+ (0.25,0) [0|65535] "" XXX
 SG_ TCSlipMaybe : 56|8@1+ (1,0) [0|255] "" XXX
 

[hjtrbo]

$19D
I think it's this???
BO_ 413 TransmissionInputSpeed: 8 TCM
SG_MessageHeartBeat : 6|2@0+ (1,0) [0|255] "" XXX
SG_ TurbineAngularVelocity : 48|16@0+ (0.25,0) [0|65535] "" XXX
SG_ GearRatio : 56|8@1+ (0.03125,0) [0|255] "" XXX

[series8217]

$19D
I think it's this???
BO_ 413 TransmissionInputSpeed: 8 TCM
SG_MessageHeartBeat : 6|2@0+ (1,0) [0|255] "" XXX
SG_ TurbineAngularVelocity : 48|16@0+ (0.25,0) [0|65535] "" XXX
SG_ GearRatio : 56|8@1+ (0.03125,0) [0|255] "" XXX

I don't think the "GearRatio" signal is strictly the gear tooth ratio, because the signal value varies continuously instead of only in discrete steps. I think it may be the ratio between engine RPM and transmission output shaft RPM, with torque converter slip causing the ratio to change slightly depending on conditions.
You're definitely right that it's gear ratio related though. I see the step changes when the estimated gear changes

[series8217]

Some progress on $189 and $199.
There are 4 separate torque signals spread across the two frames. Also, only the 2nd torque signal of each frame is protected by a twos-complement of the sum of the signal and rolling count.

I did verify the protection value agrees with the spec in GMW8772-1.

The protected values are likely requests to increase torque as that is much more dangerous when malformed compared to a torque reduction request.

I'm still trying to figure out what the purpose is of each of the different torque requests. I don't yet understand why more than two signals are needed. I expected to see only a max torque / torque reduction signal, and a min torque / torque increase signal. So why 4?

What I've observed so far:
- full throttle upshift from 1->2 and 2->3, TorqueReq1&3 are used and follow each other exactly.
- partial throttle upshift from 5->6, TorqueReq1&2 are used and follow each other exactly. This might be a MINIMUM torque request. The engine was reporting negative torque until these two signals became active at 18 Nm, then the engine torque gradually increased. However, the accelerator pedal was increasing slightly at the same time, so I can't be certain.
- partial-throttle upshift from 3->4, TorqueReq1&2&4 are used and follow each other exactly. The engine was reporting lower torque than the request, so this seems to be a minimum torque request / request to increase torque. However, the engine reported torque never increased to match the request. It kept following the driver requested torque.
- partial-throttle upshift from 2->3, 3->4, 4->5, TorqueReq1&2&4 are used and follow each other exactly. Again seeing these signals be higher than the reported engine torque yet the ECM doesn't seem to care.

I haven't been able to get ghidra to work yet, but hoping to understand more of what's going on when I see the code that fills in these values.

Code: Select all

BO_ 393 unknownTCM_Torque_189: 8 TCM
 SG_ TCMTorqueReq1 : 3|12@0+ (0.5,-848) [0|255] "Nm" XXX
 SG_ TCMTorqueReq1Active : 4|1@0+ (1,0) [0|1] "" XXX
 SG_ RollingCounter : 7|2@0+ (1,0) [0|3] "" XXX
 SG_ TCMTorqueReq2 : 19|12@0+ (0.5,-848) [0|4095] "Nm" XXX
 SG_ TCMTorqueReq2Active : 20|1@0+ (1,0) [0|1] "" XXX
 SG_ TCMTorque2Protection : 36|13@0+ (1,0) [0|8191] "" XXX
 SG_ NEW_SIGNAL_1 : 55|8@0+ (1,0) [0|65535] "" XXX
 SG_ MaxTorqueIncreaseRateMaybe : 63|8@0+ (1,0) [0|255] "" XXX
 
 BO_ 409 unknownTCM_Torque_199: 8 TCM
 SG_ TCMtorqueReq3 : 3|12@0+ (0.5,-848) [0|255] "Nm" XXX
 SG_ TCMTorqueReq3Active : 4|1@0+ (1,0) [0|1] "" XXX
 SG_ RollingCounter : 6|2@1+ (1,0) [0|3] "" XXX
 SG_ TCMTorqueReq4 : 19|12@0+ (0.5,-848) [0|4095] "Nm" XXX
 SG_ TCMTorqueReq4Active : 22|3@0+ (1,0) [0|1] "" XXX
 SG_ TCMTorque4Protection : 36|13@0+ (1,0) [0|4095] "" XXX
 SG_ MaxTorqueIncreaseRateMaybe : 63|8@0+ (1,0) [0|255] "" XXX