Cim: Difference between revisions
Jump to navigation
Jump to search
(Created page with "# SAAB 9-3 CIM Dorking MCU: M306NAFGTFP 10k RAM 256K ROM ## Starting it in bench ____________________________________________ | +B PH PL +15 | | 17 18 19 20 21 22 23 24 25 | | | | 1 2 3 4 5 6 7 8 9 11 12 14 15 | |+30 GND GND | |____________________________________________| PH P-bus CAN High PH P-bus CAN Lo...") |
No edit summary |
||
(34 intermediate revisions by the same user not shown) | |||
Line 1: | Line 1: | ||
<h1>Saab 9-3 CIM Column Integrated Module</h1> | |||
[[file: cim.jpeg|450px|]] | |||
10k | <b>MCU</b>: M306NAFGTFP<br> | ||
<b>RAM:</b> 10k<br> | |||
<b>ROM:</b> 256k<br> | |||
[[file: Cim mcu.jpeg|450px]] | |||
<h2>Starting it in bench</h2> | |||
____________________________________________ | ____________________________________________ | ||
| +B PH PL +15 | | | +B PH PL +15 | | ||
Line 16: | Line 18: | ||
|____________________________________________| | |____________________________________________| | ||
PH P-bus CAN High | PH P-bus CAN High<br> | ||
PL P-bus CAN Low<br> | |||
<h3>Power</h3> | |||
12v to pin 1, 18 and 23, ground on pin 3 & 7 | 12v to pin 1, 18 and 23, ground on pin 3 & 7 | ||
pin 18 is the "wakeup" signal to CIM from ISM, it's a switch that engages when a key is inserted | pin 18 is the "wakeup" signal to CIM from ISM, it's a switch that engages when a key is inserted<br> | ||
pin 23 is the "key in on position" signal from ISM | pin 23 is the "key in on position" signal from ISM<br> | ||
<br> | |||
if 18 and 23 is not powered the CIM will not start broadcasting on the p-bus. unknown if it's listening with only +30 yet<br> | |||
<h2>I-bus</h2> | |||
All 3 I-bus pinns are connected to each other(!??) | All 3 I-bus pinns are connected to each other(!??) | ||
<h2>P-bus</h2> | |||
<h3>Broadcasted Message id's</h3> | |||
0x0C1 | 0x0C1 | ||
Line 41: | Line 43: | ||
0x381 | 0x381 | ||
<h2>GlobalTIS binarys</h2> | |||
Remove 20 first bytes from 176kb file to make all vector tables etc line up in IDA | Remove 20 first bytes from 176kb file to make all vector tables etc line up in IDA | ||
<h3>Header</h3> | |||
5F 6E | 00 01 21 | 00 C3 7A E6 | 41 41 | 01 00 0D 00 00 | 00 02 BE 00 | | 5F 6E | 00 01 21 | 00 C3 7A E6 | 41 41 | 01 00 0D 00 00 | 00 02 BE 00 | | ||
???? | same in | 12810982 | AA | write dest? same | 179712 size after 20b header | | ???? | same in | 12810982 | AA | write dest? same | 179712 size after 20b header | | ||
crc16?| all bins | module name | rev | in all bins | same in all bins | | crc16?| all bins | module name | rev | in all bins | same in all bins | | ||
<h2>CAN</h2> | |||
CIM listens to 0x245 on p and i-bus and replies on 0x645 | |||
<h2>Memory layout</h2> | |||
Dumping memory from CIM using $23 over GMLAN is possible for the following ranges | Dumping memory from CIM using $23 over GMLAN is possible for the following ranges | ||
Line 60: | Line 65: | ||
0x8000A0 contains a power on counter | 0x8000A0 contains a power on counter | ||
<h2>Serial dumping</h2> | |||
MCU is ID locked, ID is currently unknown | MCU is ID locked, ID is currently unknown | ||
<h2>EEPROM</h2> | |||
'''Type:''' 93LC66A-I<br> | |||
'''Packaging:''' SOP8<br> | |||
[[file: Cim eeprom.jpeg|450px|]] | |||
CS connected to P50 CS0 | |||
DI to P49 CS1 | |||
DO to P39 RDY/CLKOUT | |||
SK/CLK to P51 | |||
DC to GND | |||
== UART == | |||
USART 0 -> "K-Line 1" < SCL (Steering Column Lock) > | |||
USART 1 -> < SAS (Steering Angle Sensor) > (blue connector) | |||
USART 2 -> "K-Line 2" < ISM (Ignition Switch Module) > | |||
<h2>Other</h2> | |||
<pre> | |||
bojer:Random note: CIM eeprom addresses 0x65 to 0x72 must be bit map of dtc errors presen | |||
bojer: bits represent errors and if you count 1s you get number of dtcs present | |||
bojer: as in flipping bit 29 will get you B3598 04 | |||
Only variable is sas - it has additional data section at the end of eeprom file | |||
bojer: Side note on transponder - for cim to think it is valid key it needs to have isk locked AND protect write user pages (PWUP) flag on | |||
bojer: as in: page 3 first byte needs to be 0x96 | |||
chriva: Found what I wanted to know. If 0F_BE03 is not set to 0x8f it's simply not going to jump to the main firmware. | |||
As for how it knows where to enter the main firmware: That's a tribyte-address stored at 0F_BDED | |||
chriva: The utility program is to be stored at EXACTLY address 0x614 and up. All those weird bytes finally made sense | |||
uart2 goes to pin 17 of CIM (comm with ISM) | |||
UART0 - SCL communication | |||
UART1 - SAS + bootloader | |||
UART2 - ISM communication | |||
0 and 2 are multiplexed | |||
I think in normal conditions all of them run at 9600 | |||
12V tx and rx on same line | |||
I suspect 0xc3 might be CIM options | |||
add/remove sensonic steering wheel controls changes here c2 87 to e2 87 | |||
and I can confirm that in those same two bytes is also stored info about cruise control | |||
PM]bojer: I think it will be 8080 | |||
[10:57 PM]bojer: Or at least 8000 | |||
[10:57 PM]bojer: 0x0080 might be steering buttons | |||
[8:25 PM]bojer: Note to self and others: cim is picky about what you have stored in P3 of key transponder (one with config bits) - there are 3 bytes left that MUST match | |||
[8:29 PM]bojer: In working added key it should be: 96AA4854, no problem. my key for some reason (bit flip?) had 96AA4855 and CIM refused to accept key | |||
[8:35 PM]bojer: new key should have 06AA4854 | |||
</pre> |
Latest revision as of 20:35, 5 February 2023
Saab 9-3 CIM Column Integrated Module
MCU: M306NAFGTFP
RAM: 10k
ROM: 256k
Starting it in bench
____________________________________________ | +B PH PL +15 | | 17 18 19 20 21 22 23 24 25 | | | | 1 2 3 4 5 6 7 8 9 11 12 14 15 | |+30 GND GND | |____________________________________________|
PH P-bus CAN High
PL P-bus CAN Low
Power
12v to pin 1, 18 and 23, ground on pin 3 & 7
pin 18 is the "wakeup" signal to CIM from ISM, it's a switch that engages when a key is inserted
pin 23 is the "key in on position" signal from ISM
if 18 and 23 is not powered the CIM will not start broadcasting on the p-bus. unknown if it's listening with only +30 yet
I-bus
All 3 I-bus pinns are connected to each other(!??)
P-bus
Broadcasted Message id's
0x0C1 0x0C5 0x180 0x1F5 0x380 audio RDS status? 0x381
GlobalTIS binarys
Remove 20 first bytes from 176kb file to make all vector tables etc line up in IDA
Header
5F 6E | 00 01 21 | 00 C3 7A E6 | 41 41 | 01 00 0D 00 00 | 00 02 BE 00 | ???? | same in | 12810982 | AA | write dest? same | 179712 size after 20b header | crc16?| all bins | module name | rev | in all bins | same in all bins |
CAN
CIM listens to 0x245 on p and i-bus and replies on 0x645
Memory layout
Dumping memory from CIM using $23 over GMLAN is possible for the following ranges
0x100000 - 0x1003FF 0x100400 - 0x1017FF 0x200000 - 0x20001A 0x800000 - 0x800110
0x8000A0 contains a power on counter
Serial dumping
MCU is ID locked, ID is currently unknown
EEPROM
Type: 93LC66A-I
Packaging: SOP8
CS connected to P50 CS0 DI to P49 CS1 DO to P39 RDY/CLKOUT SK/CLK to P51 DC to GND
UART
USART 0 -> "K-Line 1" < SCL (Steering Column Lock) > USART 1 -> < SAS (Steering Angle Sensor) > (blue connector) USART 2 -> "K-Line 2" < ISM (Ignition Switch Module) >
Other
bojer:Random note: CIM eeprom addresses 0x65 to 0x72 must be bit map of dtc errors presen bojer: bits represent errors and if you count 1s you get number of dtcs present bojer: as in flipping bit 29 will get you B3598 04 Only variable is sas - it has additional data section at the end of eeprom file bojer: Side note on transponder - for cim to think it is valid key it needs to have isk locked AND protect write user pages (PWUP) flag on bojer: as in: page 3 first byte needs to be 0x96 chriva: Found what I wanted to know. If 0F_BE03 is not set to 0x8f it's simply not going to jump to the main firmware. As for how it knows where to enter the main firmware: That's a tribyte-address stored at 0F_BDED chriva: The utility program is to be stored at EXACTLY address 0x614 and up. All those weird bytes finally made sense uart2 goes to pin 17 of CIM (comm with ISM) UART0 - SCL communication UART1 - SAS + bootloader UART2 - ISM communication 0 and 2 are multiplexed I think in normal conditions all of them run at 9600 12V tx and rx on same line I suspect 0xc3 might be CIM options add/remove sensonic steering wheel controls changes here c2 87 to e2 87 and I can confirm that in those same two bytes is also stored info about cruise control PM]bojer: I think it will be 8080 [10:57 PM]bojer: Or at least 8000 [10:57 PM]bojer: 0x0080 might be steering buttons [8:25 PM]bojer: Note to self and others: cim is picky about what you have stored in P3 of key transponder (one with config bits) - there are 3 bytes left that MUST match [8:29 PM]bojer: In working added key it should be: 96AA4854, no problem. my key for some reason (bit flip?) had 96AA4855 and CIM refused to accept key [8:35 PM]bojer: new key should have 06AA4854