VAGCOM: Troubleshooting Dual K-Line Issues

Hello.

Some online retailers are offering VAGCOM adapters for vehicles with a dual K-line (Auto-Intern ).

Does anyone know what vehicle-side circuitry is involved, whether our adapters are compatible with it, or how they might need to be adjusted?

Hi Ulf,

As far as I understand the explanations, the dual K-line simply uses a different pair of wires (since the connector has enough free pins) to implement a second diagnostic line. In essence, you would only need to switch the adapter to use these other pins.

I suspect (but this is just a guess, not based on any knowledge!), that one of the common adapters can simply be switched to the other pin (whatever that pin may be) using a switch, and then it would reach one half of the control units, while in the normal setting it would reach the other half.

That adapter from Auto-Intern might be able to achieve this by using a control signal so that VAGCOM can precisely control an electronic switch. Alternatively, there could be TWO adapters internally, one for each bus, and the devices would then be accessed in parallel on both buses. Since only a response comes from one bus (a device is either on bus 1 or bus 2), the responses can simply be combined using an OR gate. However, this is just a guess.

According to my theory, that's overkill. A simple mechanical switch and the knowledge of which device is connected to which bus should be sufficient (you can always test which device responds where).

Best regards,

Jan.

Jan6K wrote:

As far as I understand the explanations, the double K-line is simply the use of a different pair of wires (since the connector has enough free pins) to implement a second diagnostic line.

Hi Jan,

Why are you writing about a different pair of arteries?

If we assume that the + and ground connections always remain connected, then switching to a different signal line with a "single-pole" switch would be sufficient with a second K-line .

Another idea:
Could it be that the K-signal is now being transmitted symmetrically, meaning that in addition to the original line, there is a second, synchronized line that is "voltage-inverted" and has a vehicle-side pull-down resistor?

In the analog (HiFi) realm, this type of technology is used in (or between) high-end devices to reduce noise.
Would this have any advantages in the digital realm, such as higher data transfer rates?

Best regards, Ulf.

Hi Ulf,

Sorry, I obviously only meant "oder" (or), not "Adernpaar" (pair of arteries). That happens when you write too many texts on too many different topics one after another...

I don't believe that's about a symmetrical signal, because according to internal documentation (or other sources, I read a lot about it back then), you can use a "normal" adapter in such a car to access "half" of the data in a standard way.

And elsewhere, in one of the catalogs (online, but I no longer remember the URL), there's a K -> K1/K2 switch that can be connected to a regular adapter, which further supports the described approach.

It seems this K1/K2 part is also needed for the 1551 model to drive a car like that.

Best regards,

Jan.

Jan6K wrote:

. . . "normal" adapters that query only half of the "normal" data. And elsewhere, in one of the catalogs (online, but I no longer remember the URL), there's a K->K1/K2 switch that can be connected to a regular adapter, which further supports the described approach. This K1/K2 part is probably also needed for the 1551, in order to drive such a car.

Hi Jan,

Then it should be clear: a single 1x switch on the car output is sufficient as a minimal solution.
To do this, you would only need to know which control unit is connected to which pin on the vehicle's side. Alternatively, you could painstakingly determine this by trial and error: you would need to access all control units and record which ones are accessible with each switch position.

Owners of brand new cars should be able to identify which pin is the second K-line by comparing it to the standard connector layout. The second K-line is located where there is a connector pin that shouldn't be there.
Assuming that the diagnostic ports only connect to the poles that are actually connected, like in my tractor...

What I don't understand is: Why is the tree being split? Can a K-line tree only accommodate a certain number of STGs?

Hi Ulf,

The latter. It seems that only a certain number of participants can be connected to each K-Bus. Without knowing the specifics of this particular K-Bus, but based on general knowledge of such bus systems, I would say that the limitation is either due to the bandwidth allocation (maximum bandwidth divided by the number of participants) or, more likely, the addressing scheme. If the header field for the address only has a certain number of bits (probably a few in this case), then there will eventually be a limit at 2^n.

Overall, this is probably just a temporary solution. The K-bus is there for backward compatibility, but the actual diagnostics are supposed to be done via CAN. Michael II can probably give you the exact model year for when this change occurred, but that was at least his statement here in the forum.

And via the CAN-K bus gateway, all of this arrives at the interface. Supposedly, there's also a CAN connection somewhere on the connector, so it might even be possible to perform direct CAN diagnostics...

That would make DIY solutions extremely expensive. A CAN card for a PC probably costs just under 500 euros, if I remember correctly (I once wanted to have the university buy one), and a PCMCIA card for a laptop would cost even more.

Since all the workshops have the K-bus equipment, it will probably remain with the gateway for now, and that gateway will likely also convert whatever it receives via CAN into two K-bus signals.

But, as I said, there's a lot of speculation involved.

Here are a few superficial things about CAN: http://www.informatik.hu-berlin.de/~richling/emes/13-bus.pdf

That's at least what I tell my students... of course, the details go much deeper. The CAN standard alone is about 80 pages long (you can find it online, it's called can2spec.pdf).

Best regards,

Jan.

The division into two K-lines has electrical reasons.

The K-interface requires that all participants actively pull the K-line to VCC.
The master costs 1k, and each slave costs 30k.
So, if too many components are connected to the line, the resulting pull-up resistance becomes too low, and no driver can pull the line down to ground anymore.

On the connector, pin 15 is used for the second K-line. Previously, the L-line was connected there, but it hasn't been used for years.

The distribution across both K-lines is very simple. When writing data, the test device uses both lines. When reading data, it uses an OR connection.
That's all there is to it, as both lines can never respond simultaneously. After all, I'm only addressing ONE control unit via its address, and only that unit is allowed to respond.

'Of course, a switch would probably be even cheaper than those few gates, but people: I've already had customers on the phone who were simply too stupid to plug in the power cord... and you all know about the typical workshop intelligence .'

info is available at http://www.lin-subbus.org

I would also prefer it to be done via CAN, but as mentioned above, PCMCIA cards cost around 800 Euros.

What I don't quite understand is that the car now has a K<->CAN gateway through which both the comfort and the powertrain CAN buses run. And yet, the connection speed to the transmission is still very slow... It doesn't seem like a very elegant solution to me.

CU Gremlin.

Hello.

As a theoretical minimal solution for an "automatic" adapter extension to 2 K-lines, I arrived at a configuration of 4 diodes arranged in a bridge rectifier configuration.
"With this setup, it's my opinion that you can isolate two K-ports so that they don't interfere with each other, while still allowing them to be accessed simultaneously from the PC. Furthermore, every response from the car (regardless of which K-port it arrives through) can be securely transmitted to the PC."

Referring to the adapter circuit diagram 2.2, you must "disconnect" the collector connection of T2 and reconnect the interruption using the specified bridge rectifier. The old K-line must also be disconnected.
The bridge rectifier is connected as follows: Its positive "output" connects to terminal T2 (collector), the negative side connects to resistors R9/R12, and the two new K-lines connect to the AC voltage pins.

This will reduce the voltage peaks on both the positive and negative sides of the K-Lines circuit by 0.6 volts, which should not significantly disrupt communication, especially if the trimmers are adjusted accordingly.

What do you think of the circuit?