Pages: [1]
Author Topic: Using EDC17C46 "standalone"  (Read 4937 times)
ruan
Jr. Member
**

Karma: +2/-9
Offline Offline

Posts: 33


« on: October 27, 2021, 08:40:31 AM »

This is a pretty nooby question, but I've looked and I can't seem to figure out whether this is realistic. I'm not a VAG guy really, so bear with me - I'm not exactly down with the architecture of their cars.

Basically I'm looking at using EDC17C46 (likely from an Ibiza 6J PQ25) in a non-VAG vehicle - ideally without the cluster/gateway/BCM/ABS module. Anything aftermarket for Diesel is either shite or ludicrously overpriced, not to mention the amount of time spent on the calibration afterwards is unfeasible to make it drive anything like factory.

There's the obvious Immo related issues which are pretty dull - but is this a case of "disabling" it within software? There's so much noise out there, frankly I've not a clue. Do you _require_ the cluster at this point once it's 'disabled'? Are there any caveats I'm not considering here? I'd also probably need to pick/feed data through CAN to make this happen (send vehicle speed, receive RPM, EML etc.) is it likely to require the gateway/BCM at this point with various signals being sent to achieve this, or is it likely easier to send the frames required to make things work?

How is the start procedure handle? I believe the ECU itself controls the start procedure in tandem with other modules to crank the engine - how is this handled by most people using ECUs from other vehicles?

Pretty basic I know, but any help towards getting my head around how best to make this happen. Thanks.
Logged
prj
Hero Member
*****

Karma: +1111/-544
Offline Offline

Posts: 6196


« Reply #1 on: October 27, 2021, 09:57:53 AM »

Just disable immo.
Everything will work.

You can use my tool for logging and live tuning, so you're really not missing out much compared to a standalone and gaining a huge amount by having the correct calibration.
Reason there's so little out there for Diesel is because you can't just calibrate the injection timing and pre-injection in a shed. Serious knowledge is required so it does not run like shit.
Logged

PM's will not be answered, so don't even try.
Log your car properly - WinOLS database - Tools/patches
jarda52005
Newbie
*

Karma: +0/-0
Offline Offline

Posts: 1


« Reply #2 on: October 27, 2021, 10:14:19 AM »

There's the obvious Immo related issues which are pretty dull - but is this a case of "disabling" it within software? There's so much noise out there, frankly I've not a clue. Do you _require_ the cluster at this point once it's 'disabled'? Are there any caveats I'm not considering here? I'd also probably need to pick/feed data through CAN to make this happen (send vehicle speed, receive RPM, EML etc.) is it likely to require the gateway/BCM at this point with various signals being sent to achieve this, or is it likely easier to send the frames required to make things work?

I will try to answer this question. Immobilizer on some VAG ECUs can be permanently disabled in DFlash (EEPROM) for WFS4 and early WFS5, which should be your case as well as I believe EDC17C46 is TC1767. Later WFS5 (with SHE hardware, like TC1793 - EDC17C74) are using different structure of DFlash and I did not saw any permanent solution, but it is still possible to patch those in PFlash (Program) to ignore immo state. It will create DTCs in memory, but ECU will start the engine nevertheless.

The cluster should contain master keys for immo (Sometimes it is BCM, sometimes other units) and you should be able to see encrypted communication between cluster and ECU on 0x10 / 0x11 CANIDs or when you have only ECU, then you will see only 0x10. So technically you don't need cluster, if you are able to disable or suppress the immobilizer in ECU.

Gateway is necessary, because there are several CAN networks, and cluster is on different CAN than ECU, so GW is routing messages from one CAN network to another.

Communication on CAN is described in so called DBC files. You can find some VW PQ files on https://github.com/commaai/opendbc and you can open those in CANdb++ (tool distributed with Vector CANape or Vector CANoe, you can get demo from Vector for free). Also you can use Busmaster https://rbei-etas.github.io/busmaster/ which can import DBC file and translate you CAN messages realtime into signals. You will need PeakCAN hardware for that (or Vector hardware if money are not the problem).

You can also use DBC files to emulate signals to ECU and get it working without real ECUs. When you know which signals are necessary and which are not for running of EDC17C46, then you can build some CAN signal emulator based on some MCU, which will keep EDC17C46 satisfied and running without other ECUs.
Logged
ruan
Jr. Member
**

Karma: +2/-9
Offline Offline

Posts: 33


« Reply #3 on: October 27, 2021, 10:46:17 AM »

I will try to answer this question. Immobilizer on some VAG ECUs can be permanently disabled in DFlash (EEPROM) for WFS4 and early WFS5, which should be your case as well as I believe EDC17C46 is TC1767. Later WFS5 (with SHE hardware, like TC1793 - EDC17C74) are using different structure of DFlash and I did not saw any permanent solution, but it is still possible to patch those in PFlash (Program) to ignore immo state. It will create DTCs in memory, but ECU will start the engine nevertheless.

The cluster should contain master keys for immo (Sometimes it is BCM, sometimes other units) and you should be able to see encrypted communication between cluster and ECU on 0x10 / 0x11 CANIDs or when you have only ECU, then you will see only 0x10. So technically you don't need cluster, if you are able to disable or suppress the immobilizer in ECU.

Gateway is necessary, because there are several CAN networks, and cluster is on different CAN than ECU, so GW is routing messages from one CAN network to another.

Communication on CAN is described in so called DBC files. You can find some VW PQ files on https://github.com/commaai/opendbc and you can open those in CANdb++ (tool distributed with Vector CANape or Vector CANoe, you can get demo from Vector for free). Also you can use Busmaster https://rbei-etas.github.io/busmaster/ which can import DBC file and translate you CAN messages realtime into signals. You will need PeakCAN hardware for that (or Vector hardware if money are not the problem).

You can also use DBC files to emulate signals to ECU and get it working without real ECUs. When you know which signals are necessary and which are not for running of EDC17C46, then you can build some CAN signal emulator based on some MCU, which will keep EDC17C46 satisfied and running without other ECUs.

Thanks for the response.

Reasonably familiar with trawling through DBCs and doing it without expensive tools, using SocketCAN (gs_usb) devices and the pretty generic linux tools that go with.

The real question is whether I _need_ the gateway - if whatever signals are required I can dump onto PT-CAN (not sure of the name in VAG world) then surely the gateway is redundant at that point, unless the gateway has some say in communicating with other modules for starting, although I suspect putting 12v to the starter solenoid will cause the engine to run at that point and it's like push-starting it. It's things like the clutch switch, vehicle speed and others which are required as they will drive like shit without them, with a plan in future to send the right frames for cruise control to operate.

Just disable immo.
Everything will work.

You can use my tool for logging and live tuning, so you're really not missing out much compared to a standalone and gaining a huge amount by having the correct calibration.
Reason there's so little out there for Diesel is because you can't just calibrate the injection timing and pre-injection in a shed. Serious knowledge is required so it does not run like shit.

Thanks for that - yes, your tool is the only reason that this is remotely possible. I could run other ECUs, but with VehiCAL, it's worth it to use something VAG which has full A2Ls, FRs, high-speed logging and live tuning for a fraction of the cost with significantly greater functionality and the vast majority of the work done by the OEM. It's preferable to do it this method rather than any standalone, the final hurdle for me is using those ECUs out of the vehicles they were intended for.

I've experienced the pain of calibrating diesels from (essentially) scratch by building engine combinations that were not available from the factory thus no calibration to use and going through the process by having to flash every version, particularly getting the cold start calibration right. Torn out my hair for many hours even with combustion pressure monitoring.
Logged
prj
Hero Member
*****

Karma: +1111/-544
Offline Offline

Posts: 6196


« Reply #4 on: October 27, 2021, 10:48:25 AM »

I will try to answer this question. Immobilizer on some VAG ECUs can be permanently disabled in DFlash (EEPROM) for WFS4 and early WFS5, which should be your case as well as I believe EDC17C46 is TC1767. Later WFS5 (with SHE hardware, like TC1793 - EDC17C74) are using different structure of DFlash and I did not saw any permanent solution, but it is still possible to patch those in PFlash (Program) to ignore immo state. It will create DTCs in memory, but ECU will start the engine nevertheless.
It is possible to patch both Immo4 and Immo5 so there are zero DTC's, without having to change anything in calibration.
Logged

PM's will not be answered, so don't even try.
Log your car properly - WinOLS database - Tools/patches
jcsbanks
Full Member
***

Karma: +22/-13
Offline Offline

Posts: 172


« Reply #5 on: October 28, 2021, 10:43:15 AM »

The real question is whether I _need_ the gateway - if whatever signals are required I can dump onto PT-CAN (not sure of the name in VAG world) then surely the gateway is redundant at that point

On the bench EDC17C74 does not communicate on CAN UDS directly like EDC17C64 does (without a gateway just by connecting 12V, switched 12V, ground, CAN H, CAN L with termination). The wiring diagram shows a flexray pair going to the gateway, but even with those connected and the CAN pair too, UDS comms do not get a reply from the EDC17C74 but I can talk to the gateway itself, which has a fault code for flexray failing to initialize. Adding 100ohm termination didn't help, neither did changing the VIN in the EDC17C74 EEPROM to match that in the gateway. There is more to flexray than I understand so far. I would avoid it by going for something that doesn't use it for a swap.
Logged
prj
Hero Member
*****

Karma: +1111/-544
Offline Offline

Posts: 6196


« Reply #6 on: October 28, 2021, 10:50:50 AM »

On the bench EDC17C74 does not communicate on CAN UDS directly like EDC17C64 does (without a gateway just by connecting 12V, switched 12V, ground, CAN H, CAN L with termination). The wiring diagram shows a flexray pair going to the gateway, but even with those connected and the CAN pair too, UDS comms do not get a reply from the EDC17C74 but I can talk to the gateway itself, which has a fault code for flexray failing to initialize. Adding 100ohm termination didn't help, neither did changing the VIN in the EDC17C74 EEPROM to match that in the gateway. There is more to flexray than I understand so far. I would avoid it by going for something that doesn't use it for a swap.

Except we are talking about EDC17C46 here and that is not flexray... just normal CAN, you don't need a GW or any other modules it will run the engine with the immo off, and will talk to diags.
Logged

PM's will not be answered, so don't even try.
Log your car properly - WinOLS database - Tools/patches
Pages: [1]
  Print  
 
Jump to:  

Powered by SMF 1.1.21 | SMF © 2015, Simple Machines Page created in 0.017 seconds with 14 queries. (Pretty URLs adds 0s, 0q)