Yes, that's what I mean.This strongly suggests "unidirectional" - I think (if I understand Uwe's use of the term correctly)
But it's nice to have empirical confirmation (thanks kamold) to what is claimed in the documentation.
-Uwe-
Yes, that's what I mean.This strongly suggests "unidirectional" - I think (if I understand Uwe's use of the term correctly)
That would be a very interesting experiment. I can tell you the VIN gets announced on the bus continuously, in a multiplex message repeating every 800ms. And for MQB, there's some signature-looking stuff embedded in the message I had previously thought was CP related. PQ has a similar cyclic mux message but with a different ID and there's no signature in it.The other question that you haven't asked is: is it possible avoid CP errors when swapping constellation modules? The answer is YES (well....... it's a type of YES in theory). I've not done this, but theoretically, if it's possible to hot-swap any of the constellation modules after a successful CP challenge/response process has been completed (i.e. post ignition switch-on) - it should be possible to avoid CP errors. Of course this would only apply for the current ignition cycle - once the ignition is switched-off and if the replaced module remains installed, a CP error will occur at the next ignition switch-on
Do you know which module is sending those messages?That would be a very interesting experiment. I can tell you the VIN gets announced on the bus continuously, in a multiplex message repeating every 800ms.
I'm 99% sure it's the gateway, but I'd have to isolate every bus or test a lone bench gateway to be sure. With my current rig I can only tell you it isn't coming from anything on Extended CAN.Do you know which module is sending those messages?
-Uwe-
199.5 05D2 8 01 5A 5A 5A 41 41 5A 44
399.6 05D2 8 02 44 30 33 37 35 31 38
600.0 05D2 8 00 00 00 00 00 57 56 57
^ms
I did not have any particular use in mind. It was more an academic question that would confirm whether CP actually works the way it is described, i.e. "unidirectionally".Then the question, what would be the use case of two gateways matched to the same car?
I'm 99% sure it's the gateway, but I'd have to isolate every bus or test a lone bench gateway to be sure. With my current rig I can only tell you it isn't coming from anything on Extended CAN.
On MQB, the VIN is muxed in a cyclic message 0x6B4 and it has the funny prefix. On PQ, the VIN is muxed in 0x5D2 the exact same way, but the prefix is all zeros on the only test car I've looked at so far. Unfortunately the test car is, uh, not in factory condition... it's the R36 Passat in kamold's signature. We are working with him to port openpilot to the PQ35/PQ46/NMS family. So I can't say what a factory B6 or B7 Passat would do, or exactly when CP might have been introduced or whether it exists in any form on those vehicles.
An example scenario:
Turn car to accessories.
Unplug instrument cluster.
Plug in foreign instrument cluster (used) and you WON'T get a CP message immediately.
Newly installed cluster will power up and display as normal. ........................
Kamold:............. except that it appears (I don't know why) that in the case of a used cluster at least, a successful CP interrogation process must have been completed before "hot swapping" the module.
After reading these posts, I had thought that I had a solution to dealing with CP error on the hex17 module on my MQB test bench: simply energize the cluster module after activating my simulated ignition switch - allowing sufficient time for the CP master to complete its task?
Alas not successful - same CP error occurs!! Admittedly the test-bench is a very unique set of conditions - but the experiment does add further insight into the mysterious CP process. All I have to do now is understand why - hopefully not as difficult as my current other quest: finding the answer to meaning of life!!
I suspect that once the CP master records an error on a module - all subsequent attempts to hot-swap further modules are infected with the same complaint (in the newly introduced module). Kind-of similar to jyoung's observations about the transmission of repeated VIN messages on the CAN bus - perhaps?
Don
Even though I 'only' have an Mk5.1, this thread is interesting.
One thing comes to mind though, what about any "right to repair" laws like some places have for cell phones and electronics?
Couldn't the whole right-to-repair process allow us to 'defeat' the whole CP issue? ("Defeat" may not be the right word - i mean legally bypass/work around it)
Another solution is to downgrade mk7 modules to mk6
If the CP would be doable, it would’ve been done already. CP scheme has been around for several years already.
And who would like to downgrade to mk6 modules? And that would not be even possible in most cases.
.............But the whole CP scheme is to battle theft and illegal trade of components........