Hello,
I also purchased a e-golf pack a month ago and have now teared it down into bits and pieces. I think I now have got a good clue now how its built up and how I can interact with the CMCs.
I'm writing here with no questions at the moment, just want to spread some insights I learned along the way so that the next one that tries this has a smoother journey. A lot of what I'm wiring is already in this thread, so this will aim to sum that up a bit. Basically, this post is what I would like to have had from the start.
The pack I got is from a 2020 e-Golf with 50000 km on it. Got it of a junkyard in Norway for about $2500. My plan is to mount at least 72 of the 88 cells into a 1969 Austin Mini, pair it up with a Mitsubishi Outlander rear motor, rear inverter and dcdc/obc. I was planning to buy a separate BMS, but after I learned how to interact with the VW CMC:s, I have decided to program it myself and save a lot of bucks.
--------
OVERVIEW
--------
The pack consists of:
8 master modules (4s3p)
9 full slave modules (4s3p)
10 half slave modules (2s3p)
Total pack is 88s3p. Each cell is 37 Ah, each cell cluster 37*3=111 Ah. The pack interconnections is of 35 mm^2 copper busbars, both stiff and flexible mixed.
Every master module communicates over CAN (500 kbs) and the most important bits has been backwards engineered by the champs in this thread (kudos to them!). Every master module consumes about 30 mA when awake and powered from 12V.
This is what can be read and controlled over CAN:
- Read every cell voltage, accuracy 1 mV
- Read temperature of every module, 1 sensor per module. Accuracy 0.5 degC
- Read balancing status for every cell
- Control balancing on/off for every cell
The master modules seems to be very resistant against malconnections. I have not broken any masters, even though I messed up the connections many times. At some times, I have felt heat purge through the cover of the PCB when I messed up some blue connectors, but they seem to not take damage. So don't worry too much
.
------
WIRING
------
The modules are connected with busbars. These are probably of least interest as these might be hard to reuse later given their unique shapes.
In the pack you'll find two main cables harnesses, one orange and one black.
Black = 12V cabling, U30, Ignition, CAN H, CAN L etc. Connects all master modules
Orange = Balancing leads between slaves and masters + temperature leads of the slaves
The orange cabling is actually 8 separate harnesses. First thing is to unwrap most of the orange tape to find the 8 separate smaller harnesses.
Each harness has:
- One black connector = Temperature leads to the master
- One red connector = Balancing leads to master
- Two, three or four blue connectors = Balancing leads and temperature leads from slaves. More on this later...
-------------
CONFIGURATION
-------------
Every master modules is uniquely configured for a set of either full of half slave connections.
CMC1 = 2 + 2 + 2 + 2
CMC2 = 4 + 2 + 2
CMC3 = 2 + 2
CMC4 = 4 + 2
CMC5 = 4 + 4
CMC6 = 4 + 4
CMC7 = 4 + 4
CMC8 = 4 + 2
Note: CMC 5/6/7 and CMC 4/8 share the same configuration, therefore, the orange cable harness is interchangeable between those.
The master is always the most negative. From the positive pole of the master, the first slave's negative connects, and so on.
(-) Master (+)(-) Slave #1 (+)(-) Slave #2 (+)(-) Slave #3 (+)(-) Slave #4 (+)
The orange cable harness with 4 blue connectors goes to CMC1.
- The only CMC with 4 slaves
- Slave #1 blue connector with one of the cables yellow
- Slave #2 blue connector with one of the cables red
- Slave #3 blue connector with one of the cables purple
- Slave #4 blue connector with one of the cables green
The orange cable harness with 3 blue connectors goes to CMC2.
- The only CMC with 3 slaves
- Slave #1 blue connector with one of the cables yellow
- Slave #2 blue connector with one of the cables purple
- Slave #3 blue connector with one of the cables green
For the rest of the cable harnesses with only 2 blue connectors, it's a bit more tricky. As it happens, slave #1 will always have the blue connector with the yellow cable (among others, but let's focus on the yellow). If the blue connector with the yellow cable has:
- 4 cables in total, the first slave is a half module. This means the cable harness is for CMC3.
- 6 cables in total, the first slave is a full module This means CMC 4/5/6/7/8
If the other connector than the one with a yellow cable has:
- 4 cables, then its for CMC 4 and 8
- 6 cables, then its for CMC 5, 6 and 7
----------
CONNECTION
----------
In order to make the master modules behave "nicely" and send and respond on CAN as expected, they need all prerequisites fulfilled:
- The right orange cable to the right amount of slaves and the right type of slave (as per above).
- All brown wires from the black connector to GND
- All green and red, green and yellow wires from the black connector to +12V
(TO BE CONFIRMED)
- Master (+) terminal connected to Slave 1 (-)
- Slave 1 (+) connected to Slave 2 (-) and so on...
The cell temperature and balancing status is reported continuously.
The cell voltages are only transmitted as response to a message:
- Identified CAN standard: 0xBA
- Byte 0: 0x45
- Byte 1: 0x1
- Byte 2: 0x28
- Byte 3: 0x0
- Byte 4: 0x0
- Byte 5: 0x0
- Byte 6: 0x0
- Byte 7: 0x30
For testing, I've been using CANKing. Works well enough and has the functionality needed for this.
All CAN communications are documented in this DBC file. This file is based on the works of Tom-evnut, but complemented and cleaned up by me.
https://github.com/haand22/VW-e-Golf.git
--------------------
CELL CHARACTERISTICS
--------------------
I've done a full cycle test, documenting voltages and mAh:s with the equipment I had (IMAX B6). I got a total capacity of 121,2 Ah, which is too much of course... However, I will be using the data and just scale down the Ah:s a bit. The overall SOC vs. Voltage curve showcase the characteristics of the cell quite a lot:
https://github.com/haand22/VW-e-Golf/bl ... C_vs_U.png
